Rural Long Tail Public Service and the Correction Mechanism: Evidence from China 981164022X, 9789811640223

This book firstly analyzes the status and characteristics of rural long tail public service and its unbalance in detail.

123 60 4MB

English Pages 278 [274] Year 2021

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Acknowledgements
Contents
About the Author
1 Introduction
1.1 Background
1.2 Research Ideas
1.3 Main Innovation
2 Theory and Concept
2.1 Theory
2.1.1 Long Tail Theory
2.1.2 Mechanism Design Theory
2.1.3 Resource Dependence Theory
2.2 Concept
2.2.1 Rural Long Tail Public Service
2.2.2 Imbalance of Rural Long Tail Public Service
2.2.3 Correction Mechanism of Rural Long Tail Public Service
References
3 Attributes of Rural Long Tail Public Service
3.1 The Demand Side Attribute of Rural Long Tail Public Service
3.2 The Supply Side Attributes of Rural Long Tail Public Service
3.2.1 From the Perspective of Service Type
3.2.2 The Financial Expense Perspective
3.2.3 From the NGO Perspective
3.3 Imbalance Between Supply and Demand of Rural Long Tail Public Service
3.3.1 Spatial Distribution
3.3.2 Time Sequence Distribution
3.3.3 Satisfaction Cost
References
4 The Influencing Factors of the Imbalance of Rural Long Tail Public Services
4.1 The Demand Side Factors of Rural Long Tail Public Services
4.2 The Supply Side Factors of Rural Long Tail Public Services
4.2.1 Government Financial Constraints
4.2.2 The Deviation of the Government’s Rational Choice
4.2.3 Immature NGOs
4.2.4 Lack of a Supply “Market”
4.3 The Factors of Imbalance: Special Education
4.4 Factors of the Imbalance: Special Health
4.5 Factors of Imbalance: Elderly Care
4.6 The Factors of Imbalance: Special Finance
4.7 Summary
References
5 The Measurement of the Imbalance of Rural Long Tail Public Services
5.1 Measurement Method
5.2 Measurement of Imbalance Index of Rural Long Tail Public Services: Rural Special Education
5.3 Measurement of Imbalance Index of Rural Long Tail Public Services: Rural Special Health
5.4 Measurement of Imbalance Index of Rural Long Tail Public Services: Rural Special Elderly Care
5.5 Measurement of Imbalance Index of Rural Long Tail Public Services: Rural Special Finance
5.6 Summary
References
6 Correction Mechanism of the Imbalance of Rural Long Tail Public Services
6.1 Basic Mechanism Design
6.1.1 Basic Setting
6.1.2 Commitment Mechanism
6.1.3 Government Regulation
6.2 Incentive Compatibility and Information Efficiency Mechanism
6.3 Interval Design of Imbalance Correction Mechanism
6.4 Groves-Clark Correction Mechanism
6.5 Nash Equilibrium Correction Mechanism
6.6 Dynamic Adjustment Mechanism
6.7 Accurate Matching Mechanism
References
7 The Application of the Correction Mechanism: Internet + NGO
7.1 Network Externality of Imbalance Correction Mechanism
7.2 Internet + NGO
7.2.1 Characteristics of Network Economy
7.2.2 Practical Mechanism
7.3 Case Study: JD.com Public Welfare Foundation Platform
References
Conclusion: Efficiency-Fairness Complementation
Appendix
Recommend Papers

Rural Long Tail Public Service and the Correction Mechanism: Evidence from China
 981164022X, 9789811640223

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Ji Luo

Rural Long Tail Public Service and the Correction Mechanism Evidence from China

Rural Long Tail Public Service and the Correction Mechanism

Ji Luo

Rural Long Tail Public Service and the Correction Mechanism Evidence from China

Ji Luo The School of Public Policy and Management Tsinghua University Beijing, China

ISBN 978-981-16-4022-3 ISBN 978-981-16-4023-0 (eBook) https://doi.org/10.1007/978-981-16-4023-0 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

To my beloved grandma in the heaven, who raised me up.

Acknowledgements

This book is based on my doctoral thesis. Parts of the contents have been published on PLOS ONE (Latent but not absent: The ‘long tail’ nature of rural special education and its dynamic correction mechanism. PLOS ONE, 2021, 16(3): e0242023). In the process of writing this book, I cannot help but think of all kinds of difficulties and tribulations I experienced during my doctoral education, including the monotonous life of sticking to the academic field for more than three years, the academic loneliness of sitting on the bench and the invisible social, family and emotional pressure of nearly 30 years old. In particular, when the papers rejected, the research results failed, and the academic views denied, the feeling of depression and frustration that I could not get paid back once occupied my life and mind for a long time. But I always firmly believe that every step I take counts. I believe that all my efforts today can be seen in the sky. One day, it will prove worth it. Thanks to all the teachers who have come all the way. My doctoral supervisor, Prof. Xie Shun, is the guide to lead me into the research field of NGOs participating in public service. Professor Xie has given me great care and help not only academically but also in my life. In my heart, Prof. Xie is like a lighthouse of my academic career, guiding me forward. In addition, I would like to thank Zhang Zhibin from Flinders University, Kathleen Donnelly from Birmingham City University, Zhang Yi from University of International Business and Economics for their valuable comments on my manuscript, which have benefited me a lot. Thank my parents for their support and care for my life! They are my strong backing, but also let me understand that blood is thicker than water and never ask for selfless love. With gratitude, I will try to live up to your expectations and be a proud son. Thanks to my friends in my academic career! In particular, Wang Tianwei has always spared no effort to help me when I encounter academic difficulties, especially in the empirical research of econometrics. The reason why my econometrics level has made a qualitative leap from zero is related to the guidance of him. I want to thank my love Daisy Wang. We have known each other in Guizhou University. It has been six years since we graduated from Guizhou University, and we have experienced nearly four years of long-distance love. Only the wearer knows where the shoes pain. Thank you for always supporting me, not abandoning, not vii

viii

Acknowledgements

giving up, giving me warmth and strength. This year we are about to enter the palace of marriage, and I also hope that we can always be happy! Finally, I want to say to my grandmother in the heaven. You have gone to a paradise without pain. I hope you can be happy and healthy in that world. I will always miss you! We will meet again one day!

Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Research Ideas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 Main Innovation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 2 4 8

2 Theory and Concept . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.1 Long Tail Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.2 Mechanism Design Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.3 Resource Dependence Theory . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Concept . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1 Rural Long Tail Public Service . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2 Imbalance of Rural Long Tail Public Service . . . . . . . . . . . . . 2.2.3 Correction Mechanism of Rural Long Tail Public Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11 11 11 13 16 17 17 22 28 34

3 Attributes of Rural Long Tail Public Service . . . . . . . . . . . . . . . . . . . . . . 3.1 The Demand Side Attribute of Rural Long Tail Public Service . . . . 3.2 The Supply Side Attributes of Rural Long Tail Public Service . . . . . 3.2.1 From the Perspective of Service Type . . . . . . . . . . . . . . . . . . . 3.2.2 The Financial Expense Perspective . . . . . . . . . . . . . . . . . . . . . 3.2.3 From the NGO Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Imbalance Between Supply and Demand of Rural Long Tail Public Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.1 Spatial Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.2 Time Sequence Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.3 Satisfaction Cost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

37 37 38 40 43 47 51 51 64 74 84

4 The Influencing Factors of the Imbalance of Rural Long Tail Public Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 The Demand Side Factors of Rural Long Tail Public Services . . . . .

89 90 ix

x

Contents

4.2 The Supply Side Factors of Rural Long Tail Public Services . . . . . . 4.2.1 Government Financial Constraints . . . . . . . . . . . . . . . . . . . . . . 4.2.2 The Deviation of the Government’s Rational Choice . . . . . . 4.2.3 Immature NGOs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.4 Lack of a Supply “Market” . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 The Factors of Imbalance: Special Education . . . . . . . . . . . . . . . . . . . 4.4 Factors of the Imbalance: Special Health . . . . . . . . . . . . . . . . . . . . . . . 4.5 Factors of Imbalance: Elderly Care . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6 The Factors of Imbalance: Special Finance . . . . . . . . . . . . . . . . . . . . . 4.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 The Measurement of the Imbalance of Rural Long Tail Public Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 Measurement Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Measurement of Imbalance Index of Rural Long Tail Public Services: Rural Special Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Measurement of Imbalance Index of Rural Long Tail Public Services: Rural Special Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4 Measurement of Imbalance Index of Rural Long Tail Public Services: Rural Special Elderly Care . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5 Measurement of Imbalance Index of Rural Long Tail Public Services: Rural Special Finance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

90 91 92 93 94 95 137 145 153 156 159 163 164 165 171 182 193 198 203

6 Correction Mechanism of the Imbalance of Rural Long Tail Public Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1 Basic Mechanism Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.1 Basic Setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.2 Commitment Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.3 Government Regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Incentive Compatibility and Information Efficiency Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3 Interval Design of Imbalance Correction Mechanism . . . . . . . . . . . . 6.4 Groves-Clark Correction Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5 Nash Equilibrium Correction Mechanism . . . . . . . . . . . . . . . . . . . . . . 6.6 Dynamic Adjustment Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.7 Accurate Matching Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

215 220 224 227 230 233 235

7 The Application of the Correction Mechanism: Internet + NGO . . . . 7.1 Network Externality of Imbalance Correction Mechanism . . . . . . . . 7.2 Internet + NGO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.1 Characteristics of Network Economy . . . . . . . . . . . . . . . . . . . 7.2.2 Practical Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

237 238 240 240 242

205 205 205 212 214

Contents

xi

7.3 Case Study: JD.com Public Welfare Foundation Platform . . . . . . . . 246 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249 Conclusion: Efficiency-Fairness Complementation . . . . . . . . . . . . . . . . . . . . 251 Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253

About the Author

Dr. Ji Luo born in 1990, is Postdoc Researcher in the School of Public Policy and Management, Tsinghua University, Beijing, China. His major is public economics. Now his research field focus on Big Data and Machine Learning, Health and Green Economics, Education Economics, and Urban Economics. Dr. Ji Luo got his Bachelor degree of English in Hunan University of Science and Technology, China (2009–2013). He got his Master degree of Public Management in Guizhou University, China (2013–2016). He has been Visiting Student in the Media School, Birmingham City University, Britain (2015). He got his Ph.D. of Economics in Guangxi University, China (2016–2019). Dr. Ji Luo has a lot publication papers, which include in International Journal of Environmental Research and Public Health, Neural Computing and Applications, PLOS ONE, Journal of Supercomputing, Environmental Science and Pollution Research International, and Korean Journal of Policy Studies. As for academic practice, Dr. Ji Luo is Editorial Board Member and Guest Editor of Biological Environment and Pollution, The Open Public Health Journal, Journal of Economics and Management Sciences. He is also Reviewer of Decision Support Systems, Journal of Supercomputing, VOLUNTAS, China Nonprofit Review. Dr. Ji Luo has also working background overseas. He has been PR Consultant for British real estate and public policy at Morale Consultant Ltd, Nottingham, UK, in 2015. Dr. Ji Luo host and participate in some projects, which includes Research on the Pattern of NGOs Participating in Social Governance; Research on the Localization Path of the Development of Rural NGOs; Research on High-quality Development of Chengdu-Chongqing Twin City Economic Circle; Construction and Evaluation of the Dual Attributes of Social Enterprise Charity Behavior from the Perspective of Globalization; Research on the Development of Green Funeral Industry in Guangxi Autonomous Region.

xiii

Chapter 1

Introduction

Abstract This chapter is the introduction to this book, which analyzes the heterogeneity of rural public service and its objective requirements for different types of suppliers. By dividing rural public demand into head demand and long tail demand, a theoretical model is constructed according to the characteristics of the different types of demand. Different types of suppliers are matched, in order to realize the long tail coupling of rural public service supply and demand. Many scholars have analyzed the matching degree of supply and demand of rural public services based on the neoclassical price theory model. However, the logic of public choice behavior behind public goods is difficult. And the theoretical analysis to escape the shackles of the effective disclosure of demand information and incentive compatibility is also not easy. This book intends to provide a reasonable explanation for the necessity, rationality and comparative advantages of different types of suppliers in response to the long tail demand for rural public services. This will be achieved through mathematical proof, extension analysis, theoretical expansion and empirical research. Based on the current situation and characteristics of rural long tail public service, this book analyzes the factors, mechanism and development trends that affect the imbalance of supply and demand in practice. By introducing lag term and proxy variables into the model, this book constructs a dynamic adjustment and self-correction mechanism of the imbalance between supply and demand. This illustrates that the response subject, as represented by local government, provides the minimum guarantee for the degree of imbalance, based on political assessment, reputation effect and other factors. However, the impact of the governance level of other social sectors on the degree of imbalance may be based on the organizational form of diversification and differentiation, ultimately leading to no significant impact. Combined with the change of the macro-environment and the heterogeneous development trend of public demand, this book puts forward an information efficiency and incentive-compatible mechanism to meet the public demand of rural residents. This is achieved by constructing a correction mechanism design model. Combined with the experience and lessons in the practice of rural construction, and on the premise of the revelation principle, this book explains the application of this correction model in a specific institutional environment and with regard to economic resource allocation.

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 J. Luo, Rural Long Tail Public Service and the Correction Mechanism, https://doi.org/10.1007/978-981-16-4023-0_1

1

2

1 Introduction

1.1 Background In recent years, with the continuous improvement in people’s living standards, the basic public services (in terms of social security) have been gradually met. Increasingly, the people’s demand for public services (products) is putting forward new requirements in terms of product type, quantity, quality and degree of free choice. Rural areas have the characteristics of vast territory and the scattered distribution of resources and population. These areas are especially affected by terrain, climate and the natural environment, and they have a relatively backward development level. These issues lead to the fragmentation, minoritization and atomization of the demand for public services. To a large extent, this poses new challenges to the supply of these rural public services, including from the aspects of technical level, economic efficiency and social welfare. This in turn leads to relatively less demand for large-scale centralized public services and more scattered demand. As such achieving scale effect becomes difficult. These characteristics (of rural public service) mean there are many inherent defects in rural public service, such as low utilization rate, high investment cost, small scale effect, and social benefit being superior to economic benefit. With the concept of shared development deeply rooted in the hearts of the people, the government, as the main body of morality and the entity responsible for public service, should better respond to social needs. However, when the growth rate of fiscal revenue slows and funds become limited, the government still needs to perform the functions of resource allocation. They must stabilize the economy and redistribute income, and reasonably allocating all kinds of fiscal funds according to the existing economic and social development needs. It is difficult to be like the “omnipotent government” of the past, the “housekeeper type” that could supply everything, regardless of cost. Due to the inherent defects of the government itself and the government’s multi-objective consideration in the supply of public services, especially from the perspective of public finance and supply efficiency, the spectre of “government failure” is of increasing concern. This is especially serious when the government supplies other types of rural public services. The limited financial resources available to the government make it difficult to match the increasingly diversified, heterogeneous and personalized public needs of rural residents. In addition, the financial pressure on the government is increasing. In the background of the effective restriction of the scale of government public expenditure, the demand for rural public services continues to grow. The government does not consider the economic cost of providing unlimited public services; doing so is unsustainable. The contradiction between supply and demand is gradually forcing the government to realize that some rural public services that can be better supplied by society should be transferred to the society for supply. This approach would replace the “all-inclusive” supply approach used in the past. This involves the division of labor and the efficiency of the government and society in the supply of rural public services.

1.1 Background

3

Rural public service—and its imbalance between supply and demand—has been the focus of academic attention in recent years. The qualitative and quantitative analysis of rural public services is gradually increasing. However, there has been very little analysis of the imbalance between supply and demand and its correction mechanism from the perspective of changes in demand preference. With regard to the goal of matching the supply and demand of rural public services, existing literature mainly analyzes the research of different response subjects based on their own efficiency and scope of action. Focusing on the formation, change and influence of the abovementioned imbalance, the overall conclusions tend to focus on the static perspective of economic paradigm research. These scholars take the classical economic theory as the premise to carry out the Pareto optimal improvement of the matching of supply and demand. However, there is a lack of a “trading market” of public goods, as well as the irrational motivation of the behavior subject. This makes it difficult to align the gap between the classical economic price model and the polycentric governance theory of public management. The studies lack logical and practical thinking about the rationality of the imbalance between supply and demand under various constraints. The conclusions fall into a certain dilemma between “should be reasonable” and “how to be reasonable”. Generally speaking, existing research has the following deficiencies: (1)

(2)

Most existing studies analyze the supply status and problems of rural public services from a separate perspective. Few studies selectively analyze the adaptability and comparative advantages of different suppliers according to the different characteristics and types of rural public demand. Although some scholars have classified rural public services, it is difficult to distinguish the complexity of diversification and heterogeneity of rural public demand, simply by the degree of publicity. Existing research lacks a better theoretical system; the studies are unable to explain the supply classification based on the complexity and heterogeneity of demand subjects. As such, this research remains fixed on analyzing the issues of pure public goods, quasi-public goods, and club goods by comparing the publicity of different types of rural public demand. A simple analysis of the government, market and NGOs is carried out in these studies, while a more detailed analysis of the heterogeneity preferences of the supply main body is lacking. Most existing researches generally analyze the problems of rural public services from a static perspective. Rarely is the dynamic perspective used to analyze the correction mechanism of this imbalance. Discussing the imbalance of rural public service only through qualitative description means the research is incomplete. Rural public demand is in a state of constant change, especially given the great improvements in the material conditions of rural residents. The volatility of demand and the relative stability of supply cause the matching degree, type and duration of both sides to be unstable and dynamic. Generally speaking, the supply of rural public services is based on demand changes, with a certain time lag.

4

(3)

(4)

1 Introduction

Determining how to selectively refine and customize the supply for different types of response subjects is an integration and reasonable extension of the above two points. In the horizontal dimension, rural public services may have different demand disclosure and supply mechanisms. In the vertical dimension, both the supply and demand sides will actively or passively adjust their own behavior during the interaction, causing the imbalance to present autocorrelation and volatility in time sequence. Do the efficiency and governance level of different suppliers have impacts on the dynamic adjustment of the supply and demand sides, in such a way as to change the steady-state of equilibrium? Existing literature rarely quantitatively analyzes the setting of the correction mechanism from the perspective of mechanism design and the significant impact of that mechanism on the equilibrium effect. Although previous research suggests that the construction of this mechanism should include accurate identification, research on the revelation principle and implementation theory applied to rural public service is still scarce. Existing researches mostly remain at the level of the macro-analysis of problems and simple advocacy of supply and demand matching. However, these studies do not explain how to construct the precise matching of rural public service demand and supply in theory. Simply establishing a reasonable matching relationship between demand and supply at the operational level is not adequate. The real question is whether, why and how an effective response can be made to the complex and special issues of rural public demand.

In short, there is no clear distinction between the rural public demands that have been effectively responded to and those that have not. If the effective response to rural public demand is to accurately match the supply and demand of rural public service, the research objectives of this book cannot be satisfied by simply describing the complexity and diversity of rural public demand, and generally advocating the diversification of the rural public service supply mode. We should clearly point out that the most complex and difficult demands in rural public services are the minority demands that relate to discrete distribution. We should focus on clarifying the special attributes of those demands that have not yet received an effective response. What conditions should be met for an effective response to such demands? Can we build an effective response mechanism for such demands?

1.2 Research Ideas Focusing on how to better meet the diversified and minority public demand in rural areas, we should first classify them according to the characteristics of rural public services. Next, we divide the classifications according to the comparative advantages of different response subjects. Then, the imbalance between the supply and demand of rural public services is analyzed, and we explore how to establish a complete and perfect correction system of rural public services.

1.2 Research Ideas

5

First, how should rural public services be classified? On the one hand, the classifications are related to the characteristics of the different quantities and qualities of rural public service. There is a sequence of satisfaction between the two types. Specifically, after the security type of public service is satisfied, the demand for the development type of public service will be stimulated. On the other hand, we can classify public services based on their publicity (noncompetitive and non-exclusive) and whether the supply has scale effect. One is the public services with larger publicity, and the centralized supply has scale effect. The other, with smaller publicity, will have more difficulty realizing centralized supply. How to integrate and summarize the different types of public services according to the inherent characteristics of rural areas is the first problem to be solved in this book. Second, we examine whether and how to divide the supply of rural public services. The supply of rural public services has been an issue of wide concerned for the government and the community. This book proposes measures to promote the development of public services in rural areas, narrowing the service gap between urban and rural areas and improving regional service equalization. These measures are designed to ensure that rural residents enjoy the same level of public services as urban residents. Many scholars have proposed that the comparative advantages of government and NGOs should be brought into play; the multi-center supply of public services should also be realized on the basis of partnership and complementation. However, few existing studies examine the division among different rural public service providers based on the classification of the characteristics of different rural public services. Thus, it is difficult to avoid falling into the pattern of separate analysis or general discussion of different providers; realizing the real transformation from “why” to “how” is also difficult. How to divide responsibility for services among different suppliers, so as to give full play to the overall synergy, is the second major problem to be solved in this book. Third, is there a correction mechanism for the imbalance between the supply and demand of rural public services? The imbalance between the supply and demand of public services refers to the fact that supply and demand cannot match and cannot be rationalized in terms of quantity, quality and level. In essence, the demand for rural public services is spontaneous, endogenous, and “embedded” in the economic and social environment of rural areas. However, the supply of rural public services is more a response to demand, and the response lags behind in time. As such, the objective demand cannot be met through the adjustment of supply degree, thus affecting the stimulation and disclosure of demand. Some scholars have studied the imbalance between the supply and demand of rural public services, emphasizing the importance of knowing residents’ individual demand preferences for the balanced supply of public goods. However, these studies generally analyze the supply–demand mismatch between urban and rural areas from a static perspective, finding that the mismatch is due to insufficient supply, diversified demand, and institutional defects. Rarely do such studies analyze the causes of this imbalance from a dynamic perspective, based on the classification of different rural public services. How to analyze the impact of dynamic adjustment on the degree

6

1 Introduction

of imbalance correction according to the efficiency and governance of different suppliers is the third question to be answered in this book. The supply and demand equilibrium mechanism of rural public service is fundamentally the interaction and compatibility between demand and the response subject. With the continuous development of economy and society in rural areas, local governments are finding it difficult to meet the diversification and heterogeneity of demand. “Government failure” is also threatening to appear under the background of asymmetric finance. Most existing theories use the neoclassical price influence mechanism to analyze the imbalance of supply and demand equilibrium in time, space, satisfaction, quantity, quality, variety and so on. However, these theories ignore the failure of the “rational economic man” hypothesis in the theoretical model itself, which leads to other responding subjects (such as NGOs) to enter the “missing market” to supply services. It is this theoretical refinement defect of diversified public demand that causes the “marginal adjustment” paradigm of classical economics to lack explanatory power (public choice) in the imbalance between supply and demand, especially in the self-correction mechanism. In recent years, more and more literatures have begun to face up to this problem; researchers have started to study the response mechanism of public demand heterogeneity. Therefore, a new research paradigm (especially a theoretical model as well as an empirical test) is urgently needed. In response, this book creatively applies the “Long Tail theory” to rural public service. We classify and distinguish the head and long tail of the demand with different characteristics and nature in rural public services. This book provides mathematical proof and an empirical analysis for different types of demand characteristics, so as to expand the application frontier of the power-law distribution of the Long Tail theory to a certain extent. A new theoretical perspective for the heterogeneous classification of rural public services is also provided. On this basis, we use the aggregation of long tail demand to analyze the feasibility of scale economy and scope economy of rural public service responders. Based on the diversified preference of long tail demand, the welfare level of responders and the spillover of different public goods, we build a comparative static and dynamic equilibrium model of “long tail equilibrium” and “head equilibrium”. Then, on the basis of the supply and demand equilibrium model of rural long tail public service, this book selects education, health, elderly care and finance for special empirical research. The economic and social factors that affect the dynamic imbalance of rural long tail public service are analyzed, and we test whether a selfcorrection mechanism exists in time and space. In addition, considering the mutual influence of long tail public demand and head public demand, this book constructs a seemingly unrelated model (SUR) to deeply analyze the interaction and coupling between head and long tail, so as to better improve the theoretical system of long tail demand of rural public services. In addition, by comparing the characteristics and advantages of the government and NGOs in terms of responding to the demand for rural public services, this book analyzes the long tail division mechanism based on heterogeneous preference for rural public services. In addition, a correction mechanism is constructed that meets the

1.2 Research Ideas

7

requirements for information efficiency and incentive compatibility. Based on mechanism design theory, this book makes an empirical study with NGOs as the research subject. The interactive sequence and mechanism of the book’s dual attributes in terms of the supply of public services put forward new theoretical perspectives and solutions for the role of NGOs in new rural construction. Generally speaking, based on the neoclassical economic model and combined with the theory of public choice and mechanism design, this book creatively puts forward a theoretical system of long tail demand for rural public services. The heterogeneous preference of public demand is taken as the analytical framework, and the existence and rationality of this power-law distribution is proven through mathematical derivation and empirical analysis. On this basis, this book constructs a correction of the imbalance between supply and demand. The mechanism expounds the comparative advantages and functions of different subjects (especially NGOs) in the supply of public services, based on their own inherent attributes, so as to construct a complete and sound division of systems. This method not only effectively meets the head public demand with scale effect, but also reasonably supplements and strengthens the long tail public demand. Different from the traditional research perspective, this book combines microeconomics, public economics, public management, sociology and other interdisciplinary theoretical perspectives. The research presented here is more based on the constraints of supply and demand balance, internal structure and correction mechanism, which can enhance the practical explanatory power and applicability of this theory. By studying the imbalance between the supply of and demand for rural public services, especially the long tail demand, this book depicts the changing characteristics, influencing factors and development trends of heterogeneous demand preference. The feasible ways of categorizing different types of public demand in terms of mode, quantity, quality, type and development mechanism of dynamic imbalance between the matching of supply and demand is also revealed. This book shows the factors that restrict the equilibrium and correction of different subjects, especially for NGOs. An imbalance correction mechanism is put forward that conforms to the dual attributes of long tail public service and meets the rural public demand preference, thus realizing the accurate identification and effective supply of multiple subjects. In detail, this book includes the following three aspects: (1)

The heterogeneity of rural public service and its objective requirements for different types of suppliers is analyzed. By dividing rural public demand into head demand and long tail demand, a theoretical model is constructed according to the characteristics of different types of demand. Different types of suppliers are matched, in order to realize the long tail coupling of rural public service supply and demand. Many scholars have analyzed the matching degree of the supply and demand of rural public services based on the neoclassical price theory model. However, the logic of public choice behavior behind public goods makes it difficult for this theoretical analysis to escape the shackles of effective disclosure of demand information and incentive compatibility. This book intends to provide a reasonable explanation for the necessity, rationality and

8

(2)

1 Introduction

comparative advantages of different types of suppliers in terms of responding to the long tail demand for rural public services. This explanation is supported by mathematical proof, extension analysis, theoretical expansion and empirical research. Based on the current situation and characteristics of rural long tail public service, this book analyzes the factors, mechanism and development trends that affect the imbalance of supply and demand in practice. By introducing lag term and proxy variables into the model, this book constructs a mechanism that dynamically adjusts and self-corrects the imbalance between supply and demand. Our work illustrates that the response subject represented by local government provides the minimum guarantee for the degree of imbalance, based on political assessment, reputation effect and other factors. However, the impact of the governance level of other social sectors on the degree of imbalance may be based on the organizational form of diversification and differentiation, leading to no significant impact.

Through the proposed dynamic panel model, this book intends to prove that the governance efficiency of different response subjects in the different spatial (provincial level) supply of public services will indeed affect the dynamic balance in terms of the long tail demand for rural public services. According to the local government’s financial expenditure, economic development level, socio-geographical environment, population density, education level and other factors, this book analyzes the basic ways and means to respond to the public demand behind the imbalance correction mechanism. In addition, the driving force and resistance that affect the demand response are explained. By explaining how the dynamic development process of the supply efficiency of governance subjects affects the balance of rural public services, we can reveal the “black box” of the imbalance that the neoclassical economic model fails to do, to some extent. This helps to tap the potential advantages of different response subjects and provides a reference for further improving the performance of public service supply. (3)

Combined with the changes in the macro-environment and the heterogeneous development trend of public demand, this book puts forward an information efficiency and incentive-compatible mechanism to meet the public demand of rural residents by constructing a correction mechanism design model. Combined with the experience and lessons learned in the practice of rural construction, and on the premise of the revelation principle, this book explains the application of this correction model in specific institutional, environment, and economic resource allocation.

1.3 Main Innovation Theoretically, based on the differences in public demand concentration and universality, this book creatively uses the Long Tail theory to divide rural public service

1.3 Main Innovation

9

into head and long tail. The existence and rationality of this power-law distribution of public service is proven through mathematical deduction, thereby expanding the understanding of traditional economics and public management with regard to the distribution and connotation of rural public service. Based on the division of long tail and head of rural public service, this book reveals that a dynamic imbalance mechanism of self-correction exists between the supply and demand of rural public services. In addition, the governance efficiency of different subjects in the supply will have a significant impact on the equilibrium results. Existing researches generally analyze the problems of rural public services from a static perspective. This includes the mismatch between supply and demand between urban and rural areas that is caused by insufficient supply, poor expression of demand, institutional defects and other reasons. However, these studies are rarely based on the classification of the different characteristics of rural public services. This book analyzes the adjustment mechanism of this imbalance from a dynamic perspective. The author argues that, due to the volatility of demand and the relative stability of supply, the matching degree, type and time of both sides are unstable and in a constant state of dynamic change. Generally speaking, the supply of rural public services is based on changes in demand, with a certain time lag. Based on other political factors (such as the punishment mechanism and the minimum limit of demand mismatch by the supplier), as well as the concern regarding the public choice of the response subject about the reputation effect, the degree of mismatch caused by the time lag is more of a dynamic equilibrium, with a time-series self-correction mechanism. In addition, this book creatively puts forward a correction mechanism design that addresses the imbalance of rural long tail public services, based on the division of different supply subjects under the established economic and social environment. Different subjects are faced with different natural endowments, economic levels, population status, social cultures, fiscal and taxation systems and other environmental conditions. Accordingly, this book analyzes the impact of the proposed correction mechanism on the dynamic imbalance of rural public service long tail demand. The comparative advantages of NGOs in information efficiency and incentive compatibility are illustrated, and the future development trend of the correction mechanism is put forward. From the perspective of mechanism design, this book analyzes whether the setting of a rural public service construction correction mechanism is reasonable and examines the mechanism’s significant impact on the balance effect. The revelation principle and implementation theory, which are based on information efficiency and incentive compatibility, are applied to the research of rural public services. This book also makes a contribution to the existing theoretical literature. In addition, the objective fact of imbalance of rural long tail public services is determined by both supply and demand. In particular, its discrete and fragmented characteristics, to a large extent, restrict different suppliers’ ability to meet the rural long tail public demand. This book, for the first time, starts with the degree of discretization and fragmentation of rural long tail public services in different fields. Then, the severity of this imbalance between supply and demand is empirically measured, and an imbalance index system that can be used to quantify the imbalance level is constructed. This work makes up for the lack of existing descriptive research

10

1 Introduction

in this field, to a certain extent. Through the construction of an imbalance index system, the overall contradiction of the imbalance phenomenon can be analyzed, as well as the direction of various factors on the overall change results. In addition, the future trends of imbalance and the correction mechanism can be predicted.

Chapter 2

Theory and Concept

Abstract This chapter presents the theory and concept of this book. Related theories include the Long Tail theory, Mechanism Design theory and Path Dependence theory. The related concepts include rural long tail public service, the imbalance of rural long tail public service, and the correction mechanism of rural long tail public service. Based on the Long Tail theory, rural public services can be divided into two categories. One is comprised of the head public services represented by infrastructure construction, the social security system, basic education and health. This kind of rural public service has strong universality, wide demand, high homogeneity, concentrated distribution, scale effect and scope economy in supply. These characteristics are in line with the spillover characteristics of public goods. The other category is the rural long tail public demand, which has the characteristics of individuality, dispersion, heterogeneity and concealment. These characteristics make it difficult to meet the scale effect in the supply; rural long tail public demand also has information disadvantages in terms of demand identification.

2.1 Theory 2.1.1 Long Tail Theory The Long Tail theory was first put forward by Anderson (2007), and was initially applied to private goods and the market. Anderson believed that most of the private goods and market demand is based on power-law distribution. Popular products are in the middle, that is, at the head of demand. The supply of this market has scale effect and scope economy, creating the potential for huge profits. At the two ends of the market demand are long tail niche products with decreasing demand. The demand for these products is relatively small, scattered and fragmented, and does not have economies of scale. However, the long tail demand will always extend to both ends and will not disappear to zero. With the development of supply technology and the entry of large-scale amateur producers into the supply chain, the marginal cost to increase supply reduces to zero. This is what makes the supply of long tail niche products profitable (Fig. 2.1). © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 J. Luo, Rural Long Tail Public Service and the Correction Mechanism, https://doi.org/10.1007/978-981-16-4023-0_2

11

12

2 Theory and Concept

Fig. 2.1 Map of the long tail theory

The aggregation effect of the aggregate creation of personalized demand can even be compared with the first popular products. Anderson believed that, by encouraging new producers to enter and improve the efficiency brought about by technological innovation, a long tail of “niche products” can be created. In addition, the market share of popular products can be reduced. There are many controversies surrounding the theory and empirical research of the long tail effect. Brynjolfsson et al. (2006) and Goldmanis et al. (2009) believed that the development of e-commerce has caused the search cost faced by customers to be reduced. The former maintained that this development causes the long tail effect; the latter stated that the reduction of search costs produces the “superstar effect” of supply (Rosen, 1981). Further, Rosen maintained that the development of early communication technologies, such as radio and television, has increased the economies of scale of high-quality suppliers. The continuous development of a network has a profound impact on the supply quality and diversification of products in different industries. Anderson believed that individual micro niche suppliers in different industries create a long tail, while at the same time reducing the relative importance of mainstream products. The increase in access mechanisms, especially with regard to the transformation from a small number of “hit” products with monopoly power to the decentralized and fragmented production of many independent “niche” products, has brought about an important standard effect. This is not only because niche products can satisfy a wider range of diverse preferences, but also because niche products increase uncertainty in the

2.1 Theory

13

market, encouraging talented individuals to enter the market and bring along technological innovation. The long tail effect is also a component of the welfare debate about competitive policy, cultural fragmentation and network neutrality. There are still many controversies about the definition and connotation of the Long Tail theory. Anderson believed that the diversity of products available increases consumer welfare. The Long Tail theory can be applied to the Internet, finance, the library industry, news media publishing industry, entertainment and consumption, logistics, education, insurance and other industries. For example, from the perspective of how Internet finance companies create value for users and obtain income based on that value, some scholars have analyzed the impact of the number of users, transaction willingness, transaction risk and big data application on the income of Internet finance companies. The problems existing in the resource integration and service of libraries have been analyzed by scholars such as Genoni (2007). The study concluded that a long tail phenomenon exists in the utilization rate of digital resources in libraries. Generally speaking, although the Long Tail theory has been widely used and practiced in all walks of life, few scholars have applied the theory to the field of public services and products. As such, the theory has not been used to analyze the characteristics and distribution of different public needs, which is the innovation of this book.

2.1.2 Mechanism Design Theory It is generally believed that the Mechanism Design theory originated from Hurwicz (1960). Hurwicz regarded a mechanism as an information exchange system. In this system, different participants pass information to each other (using the information center), and release the corresponding results according to the previously established rules. The Mechanism Design theory holds that different systems (such as market economies and planned economies) can be compared (Reiter, 1973). Groves (1973) further studied the importance of incentive mechanisms. He believed that, under the condition of incentive compatibility, a rational economic man will tend to have private information and may also take hidden actions. Subsequently, the further development of the revelation principle greatly promoted the accurate expression and mathematical simplification of mechanism design. Under the condition of the revelation principle, only the direct mechanism needs to be considered to find the optimal solution of the relevant configuration selection problem. Because the direct mechanism has a simplified mathematical structure, the optimal allocation problem can potentially be solved internally (Myerson, 1979, 1982, 1986). However, the defect of the revelation principle is that solving the problem of equilibrium multiplicity is difficult. This means that, with the revelation principle, only the unique optimal solution can be obtained in equilibrium; other suboptimal solutions cannot be obtained. The introduction of an execution mechanism successfully solved the mechanism design of multiple equilibrium suboptimal solution (Maskin, 1999). Next, the main content of the mechanism design that will be used in this book will be reviewed.

14

(1)

2 Theory and Concept

Incentive compatibility and the revelation principle

Hurwicz (1986) believed that the first step in the optimal mechanism design represented by the realization of social goals is setting an appropriate feasible mechanism set and equilibrium standard to predict the participants (Fudenberg & Tirole, 1991). Under the appropriate mechanism design, different behavior participants will choose to disclose private information spontaneously, in line with the individual interests of each participant. The principle of the direct mechanism is that, by analyzing the real information pertaining to different behavior participants, one-by-one, they will correspond to a specific result. Incentive compatibility (IC) is defined as follows: If each participant in the mechanism has an incentive to report his or her own behavior type truthfully, and if this choice satisfies his or her dominant strategy, then the mechanism conforms to incentive compatibility. In a Nash environment and standard exchange economy, there is no incentive compatible mechanism to satisfy the participation constraints (the welfare of all actors in the participation mechanism will not get worse), in order to achieve Pareto improvement. The revelation principle is that there is a direct incentive mechanism for the equilibrium result of any mechanism (Aumann & Hart, 2006). Myerson believed that the revelation principle should not only be effective among participants with different private information. He maintained that the principle should also be able to find the optimal direct incentive mechanism, even in a multi-stage dynamic game when participants take unobservable actions. Therefore, the revelation principle holds that this optimal mechanism is a subset of the general mechanism set. Also, in a Bayesian environment, due to the possession of private information, the Pareto optimal improvement may fail. (2)

The dominant strategy mechanism of rural public service

In this book, the Mechanism Design theory can be used to analyze the correction mechanism of rural public service. Those with different types of demands for rural public services (long tail or head demand) are willing to disclose private information as part of the payment for different types of services. However, they also have incentives to hide their real demand information, so as to reduce the sharing cost (Olson, 2009). Groves pointed out that, in a situation where the income effect of public services is ignored, a mechanism can be designed that encourages different demanders to reveal the dominant strategic equilibrium of their true willingness to pay. This will help achieve a balanced allocation of public services. The key to this mechanism design is to design appropriate taxes or subsidies that will effectively internalize the external cost (EC) of each consumer. The Vickrey Clarke Groves (VCG) mechanism can achieve this public service allocation efficiency under equilibrium. However, VCG cannot meet the strict budget balance and moderate transfer payment requirements (Vickrey, 1961). Therefore, based on the equilibrium of possession strategy, the design of a rural public service correction mechanism must achieve full efficiency by saving net payment surplus (Mas-Colell et al., 1995).

2.1 Theory

(3)

15

Bayesian mechanism of rural public service supply

Bayesian theory holds that those who demand different public services expect to maximize utility. Under the Bayesian Nash equilibrium, Myerson and Satterthwaite (1983) assumed that the preferences of different public service participants are quasilinear and that they satisfy the single crossover. In addition, d’Aspremont and GérardVaret (1979) held that, in a dominant strategy mechanism, incentive compatibility constraints can be satisfied. As long as each supply and demand participant discloses his/her own type of strategy, the overall utility can be maximized. Based on the Bayesian Mechanism Design theory, this book holds that those who participate in rural public service supply and demand are also expected to maximize the utility. Because the incentive compatibility is easier to meet in the Bayesian mechanism, the correction mechanism for the rural public service supply and demand imbalance can also achieve Pareto effective incentive, under the Bayesian mechanism. In the supply of rural public services, if there are enough economic actors participating in decision-making and production, the probability of each individual economic actor affecting the equilibrium result will be averaged and diluted. However, each independent economic actor can still determine the cost expenditure of his/her own initiative. Also, the mechanism is designed to ensure that the cost expenditure is optimal for each individual (Fudenberg & Tirole, 1991; Olson, 2009). (4)

The correction mechanism of imbalance between the supply and demand of rural public services

The direct mechanism of the revelation principle holds that, although incentive compatibility can ensure that every economic participant has a dominant strategic equilibrium, multiple equilibrium suboptimal solutions may exist. The correction mechanism of the imbalance between the supply and demand of rural public services shows that, although the specific equilibrium under different game situations can correspond on a one-to-one basis with the real demand of the direct mechanism, the multiple equilibrium of the game may lead to supply failure (Leininger et al., 1989). Therefore, the implementation mechanism determines that the general and complex mechanism design should be able to ensure that the multiple equilibrium results are optimal for the supply and demand equilibrium (Maskin, 2008). Maskin also believed that the executive mechanism needs to satisfy the general theorem of social choice rules, which mainly include Maskin’s monotonicity and no-veto condition. In the supply of rural public services, the conditions without veto power may be the most difficult to meet. However, under these conditions, the social choice rules conform to Nash’s executable mechanism to correct the imbalance between supply and demand. In the Nash equilibrium of public service supply and demand under a complete information game, social choice rules are easier to implement (Maskin & Sjöström, 2002). To expand or refine the supply and demand correction execution mechanism from the two aspects of information dynamics and incompleteness, Maskin’s theorem can be used to achieve Bayesian equilibrium of the correction mechanism under conditions of incomplete information. Since the implementation mechanism of social choice rules is effective under the perfect equilibrium of a sub-game, this book

16

2 Theory and Concept

applies those conditions to the supply and demand correction mechanism of rural public services. In addition, the maximum three-stage dynamic game can achieve the goal of maximizing the social welfare of all stakeholders (Moore, 1992; Moore & Repullo, 1988). Generally speaking, this book applies the Mechanism Design theory to the correction mechanism of the imbalance between the supply and demand of rural public services. On the basis of exogenous information, different economic participants have common priori knowledge. In addition, based on the assumption that the supply and demand participants are completely rational, an effective mechanism design can reveal the information with regard to complete imbalance. These assumptions can be applied to the design of network supply (“Internet+”) and computer systems. They can also be used to deal with mechanism design based on limited information disclosure and incomplete rationality. However, the difficulty with correcting the imbalance lies in overemphasizing the balance of dominant strategies and the Nash implementation mechanism after the event. This situation may not be satisfied and realized in rural areas.

2.1.3 Resource Dependence Theory The main point of the Resource Dependence theory (Pfeffer & Salancik, 2003) is that any organization needs resources to survive. Because of the limited capacity of a single organization, many resources cannot be produced directly; they need to be obtained from the outside. Therefore, an interactive relationship exists between the organization and its external environment, because of the dependence on resources. This also includes other types of organizations in the environment. As the external environment is based on organizational control, different organizations rely on their environment to obtain various external resources. Therefore, in turn, these resources actually affect the organization. According to the Resource Dependence theory, on the one hand, the resources upon which an organization relies include personnel, capital, social legitimacy, customers, technology and material input. On the other hand, in the same external environment, the degree of dependence between different organizations depends on the relationship between resources and organizational survival, the degree of internal access to resources, and the availability of alternative resources. When a resource is highly exclusive and the resource’s substitutability is small, the organization requiring that resource will be highly dependent on other organizations (Burt, 1983). According to the Resource Dependence theory, different organizations depend on each other. This interdependence can be measured by different dimensions, especially as the interdependence relates to unequal power among organizations (Clegg et al., 2006). Interdependence can be divided into competitive and symbiotic interdependence; different types of organizations can also deal with interdependence through different strategies. Burt (1983) believed that social networks create conditions and facilities for interdependence. Due to the different positions occupied by economic

2.1 Theory

17

actors in a network, non-competitive positions have more advantages in terms of interdependence than relatively competitive positions. This book uses the social network theory of resource dependence and the theory’s further development to study the rural public service correction mechanism from the network perspective. The realm of social network theory includes Granovetter’s (1985) “Embedding theory” and Burt’s (1992) “Structural Hole” theory. The latter theory is mainly used to analyze the imbalance of supply and demand from the perspective of social network relationships or structures, as well as the interaction of various stakeholders. Burt’s theory also puts forward the construction of network organization and platform on the micro basis of structuralism. Generally speaking, the Resource Dependence theory and Social Network theory reveal the interdependence and selection ability shared between different organizations (Galaskiewicz & Wasserman,1989 ). This book seeks alternative correction mechanism resources through this kind of dependence, so that different suppliers can better “embed” the environment. This book also focuses on the supply decisions taken by different types of organizations (governments or other NGOs) in order to adapt to the interdependence of other organizations in their environment. The differentiated action constraints in correction are emphasized, while focusing on the trade-off and symbiosis between the independent supply of different types of suppliers. Although the Resource Dependence theory/Social Network theory emphasizes the individual organization as the basic unit of analysis, the theory can still be applied to the network supply platform formed by different organization alliances and sharing. Therefore, this theory has the advantages of resource dependent groups and cost.

2.2 Concept 2.2.1 Rural Long Tail Public Service Rural long tail public service is aimed at rural head public services. According to the Long Tail theory, rural public services can be divided into two categories. The first category is made up of the head public services represented by infrastructure construction, social security systems, basic education and health. This category of rural public service has strong universality, wide demand, high homogeneity, concentrated distribution, scale effect and scope economy in supply, in line with the spillover characteristics of public goods. The second category is the rural long tail public demand at a higher level. This type of demand has the characteristics of individuality, dispersion, heterogeneity and concealment, and also has the characteristics of private goods. With this type of demand, meeting the scale effect in the supply is difficult. Rural long tail public demand also has an information disadvantage in terms of identification, which leads to high search and supply costs. The definition of rural long tail public service can be elaborated from the following aspects:

18

2 Theory and Concept

From the perspective of service distribution, rural public service also reflects the long tail curve distribution characteristics of large quantity, concentrated distribution and strong homogeneity of basic service demand. With an increase in demand heterogeneity, the level of demand gradually decreases and tends to be small. For example, the level of basic public needs represented by basic education is high. Compared with basic education, the level of demand for elderly care services in rural areas is obviously much lower. Due to the traditional culture of “family support for the aged” in China’s rural areas, the public demand for “collective support for the aged” has the characteristics of minority, heterogeneity and personalization. At the same time, even within the high homogeneity of basic education demand, the number of high schools in rural areas only accounts for 0.51% of the number of primary schools in China. Compared with primary education, rural residents’ demand for high school education is significantly different. On the one hand, this is because more rural junior high school graduates choose to go to the town to study in senior high school. On the other hand, the demand levels are also different because the individual differences of rural residents’ demand for senior high school education are significant; there is obviously less urgency than the demands for primary schools. This differentiation reflects the compromise effect of the depth and breadth of various needs in rural public life, and is a comprehensive coordination of horizontal and vertical differences. With regard to the level of service, the rural long tail public service is at a higher level. According to Maslow’s hierarchy of needs (1943), rural head public services, such as basic medical care, basic education, infrastructure construction and so on, are at a lower level of demand than universal basic public services. However, demand for the long tail public services, which have limited audience groups (with common personalized niche preferences) or which are difficult to supply centrally due to decentralized demand distribution, is at a higher level. From an individual point of view, the demand for rural long tail public service is quite small. However, the overall social welfare (social surplus) is large enough and certainly affects the quality of life, happiness and sense of acquisition of rural residents. Therefore, satisfying the different levels of rural long tail public service is the true embodiment of the completeness and equalization of public service. From the perspective of service satisfaction cost, it is the unbalanced distribution of rural human resources, geographical environment and systems that leads to the regional imbalance of demand. This imbalance makes the supply cost of meeting the demand for different rural public services and the search cost of obtaining information very high (information asymmetry). According to the theory of supply cost in the Long Tail theory, with the development of technology, more and more amateur producers have advantages in information acquisition, mobile cost, and exclusivity for specific needs. As such, technology is causing the marginal cost of supplying this minority demand to continue to decline, or even to reach zero. Unbalanced rural long tail public service can achieve customized small-scale supply for specific needs. In addition, supply monopolies can be broken, supply costs can be reduced, and scale economy can be formed within a specific range. From the perspective of service supply and demand matching, the information filtering mechanism in the Long Tail theory holds that, due to the inherent drawbacks

2.2 Concept

19

of the market mechanism, the supply quality of private products is bound to be uneven. When it is possible that the marginal cost of long tail products could be zero, the market filtering mechanism is particularly important. Anderson (2007) believed that a recommendation and rating mechanism can help the market accurately match demand and supply. In the supply process of rural long tail public services, the government, as a single supplier, has inherent defects in the incentive preference disclosure mechanism. The government’s own information disadvantage makes it difficult to accurately identify and meet the demand for rural long tail public services with wide distribution, variety and strong heterogeneity. The level of efficiency of this kind of “noise-filled” supply is not high. However, by customizing services for different individual needs, the realization of an accurate niche supply can be achieved, with better screening and filtering of information, and ultimately realizing the accurate matching of supply and demand. From another perspective, the Long Tail theory holds that the driving force of the long tail of a private market is to realize the connection between demand and supply through the democratization of production tools and distribution tools. The term “democratization of production tools” mainly refers to the participation of mass grassroots forces, such as crowdsourcing and crowdfunding in production supply, which help to realize the “decentralization” of supply. This theory is also in line with rural long tail public service. By purchasing services, the government outsources its own limited and unsatisfied supply services to other social suppliers, thus realizing the “decentralization” of rural long tail public services. Because of the spillover, charity function and sociality of rural public service, even a minority public service can drive other social suppliers to actively supply services, thus producing certain social benefits. This positive externality is also the internal driving force behind the development of many grassroots organizations, especially charitable organizations. Generally speaking, the long tail public service in rural areas refers to small, heterogeneous and personalized public services. These characteristics are due to the scattered and fragmented distribution of resources and people in rural areas. Although the demand for each type of rural long tail public service is small in quantity and weak in intensity, there are many kinds of these services, creating a strong agglomeration effect. The long tail public service in rural areas is related to the “incentive factors” proposed by Herzberg, who stated that, although the lack of supply will not necessarily affect the most basic life security, if a service can significantly enhance the satisfaction and sense of acquisition of some rural residents, those public services will have a strong incentive effect. Rural long tail public service has dual attributes, namely the degree of spillover as public goods and the degree of preference difference as private goods. This book examines the definition, supply efficiency, structure and technology incentive effect of rural long tail public service, through a unified theoretical framework. Due to the development of modern network technology, the diversified supply of public goods creates a long tail effect and also leads to an imbalance between supply and demand. In particular, technological change leads to the continuous change of spillover and a dispersion of rural public services. In addition, the dynamic substitution of fixed cost in the early stage and the marginal variable cost in the later stage leads to a long

20

2 Theory and Concept

tail effect in the diversified level of rural public services. The cost of rural long tail public service supply can be divided into two parts. The first one is the internal cost (IC) without spillover effect, based on the long tail attribute. The other is the EC with spillover effect based on the public attribute. Compared with the IC, a reduction of EC will lead to a monopoly effect of marginal supply. Compared with the EC, a reduction of IC will increase the competitiveness of public service supply through the potential access mechanism. Therefore, the development of network technology (“Internet+”) has different effects on rural public services with different long tail attributes. Based on the relevant theories of Blundell (1987), the long tail and head public services in rural areas can be defined as:   gi = g/ q(N /S)LT L T = log(N /S) [g/(qgi )] Here, LT is the long tail attribute of rural public services; N is the number of people with a certain type of public demand in the region, and S is the area of the region. Also, gi is the per capita consumption of a certain type of public services; G is the total consumption of this type of public services, and Q is the total number of types of public demand. When a certain kind of rural public service has universality and homogeneity, q → 1, gi → g; the long tail attribute L T → 0. When a certain kind of rural public service is fragmented and personalized, q → ∞, gi 0, EC  (N ) > 0. In order to emphasize the impact of cost structure changes, we assume that both the government and NGOs can improve supply technology. In supply quantity N, N H is the quantity of head public service, and N L T = N − N H is the quantity of long tail public service. The proportion of long tail and head is λ = N L T /N H < 1. Each consumer has a specific unit demand for rural public services (each consumer consumes one unit of head or long tail services). The welfare benefits of N H to consumers with head preference and long tail preference are VH and VL T respectively (due to spillover, N H also produces welfare benefits for long tail consumers without actual consumption). Similarly, the welfare benefits of N L T to consumers with head preference and long tail preference are W H and W L T , respectively.

2.2 Concept

21

The assumption is made that the evaluation of head consumers is higher than that of long tail consumers (VH > VL T ). Therefore, most consumers (rural residents) are more inclined to prefer rural head public services. The utility difference of the same type of consumers (head preference or long tail preference) for the same type of rural public services (head service or long tail service) is VH − W H and VL T − W L T . This utility difference can be used to express the vertical difference between head public service and long tail public service. The utility difference of the same type of rural public service (head service or long tail service) to different types of consumers (head preference or long tail preference) is VH −VL T and W H −W L T . This utility difference can be used to express the horizontal difference between head public service and long tail public service. Different consumers can change their preferences, which will influence the horizontal and vertical differences of different types of rural public services. This phenomenon is very important in terms of understanding the long tail of rural public services. The “prices” of different types of rural public services are assumed to be PH and PL T ; these “prices” are reflected in the tax revenue of the government and the service fee or social donation fee collected by other NGOs. For the head preference of consumers, their own benefit maximization requires VH ≥ PH and W H ≥ PH . Similarly, for consumers with long tail preference, the maximization of their own benefits requires V L T ≥ PL T and W L T ≥ PL T . This book assumes that each supplier (government and NGO) can only produce one kind of rural public service (head or long tail) and realize cost compensation (neither the government nor NGOs aim to make a profit). Let IC and EC account for θ I C and θ EC , respectively. Therefore, the total social welfare utility of the different types of rural public services meets the following requirements: Ni Vi ≥ θ I C I C(Ni ) + θ EC EC(Ni ) Ni Wi ≥ θ I C I C(Ni ) + θ EC EC(Ni ) i = H, L T Based on the above, first of all, the assumption is made that the government and NGOs, in the supply of rural public services, are participating in a static game. When the “price” of public services is selected, each consumer makes a choice, based on their own utility maximization. This book focuses on the potential supply of rural long tail public services, because rural head public services have always existed. In a competitive environment with elastic demand and total cost structure, there is no pure Nash equilibrium among different suppliers. However, consumers with different preferences have path dependence when they enjoy public services; they can always enjoy any level, quantity and quality of public service that they previously obtained. Meanwhile, however, different suppliers will choose to continue to supply only when the welfare effect increases. In this static equilibrium, none of the suppliers have an incentive to reduce the quantity and quality of supply, thus reducing externality. This unique pure strategic equilibrium is a natural extension of the standard Nash Bertrand equilibrium (Baye & Morgan, 1999).

22

2 Theory and Concept

Therefore, this book can define the different attributes of rural long tail and head public services, which are reflected in the respective services’ different equilibrium prices. In the following, PH∗ and PL∗T are, respectively, the equilibrium prices of rural head and long tail public services:  H (PH , PL∗T ) ≤  H (PH∗ , PL∗T ) PH < PH∗  L T (PH∗ , PL T ) ≤  L T (PH∗ , PL∗T ) PL T < PL∗T Here, H , LT are the overall social welfare of rural head and long tail public services, respectively. When all suppliers achieve the equilibrium price of supply, no supplier is willing to reduce the quality and quantity of supply; this is in line with the traditional Nash Bertrand equilibrium. Set ψi (P−i ) as the sum of the equilibrium price formed by setting the quantity and quality of supply for other suppliers (−I) faced by supplier I. Then, the static equilibrium of rural long tail public service can be expressed as follows: H (PH∗ , PL∗T ) =  L T (PH∗ , PL∗T ) =

max

 H (PH , PL∗T )

max

 L T (PH∗ , PL T )

PH ∈ H (PL∗T )

PL T ∈ L T (PH∗ )

2.2.2 Imbalance of Rural Long Tail Public Service (1)

Rural long tail public service balance

Considering the coexistence of rural head and long tail public services, Anderson (2007) believed that the entry of long tail products has a crowding out effect on head products. This book assumes that consumers with head preference will spontaneously choose head public service; conversely, consumers with long tail preference will spontaneously choose long tail public service. The necessary conditions for the long tail equilibrium of rural public service are as follows: VH − W H > V L T − W L T The head preference consumers will only spontaneously choose the head public service when they obtain more consumer surplus than they would if choosing the long tail public service: P H ≤ PL T + (VH − W H )

(2.1)

2.2 Concept

23

Similarly, the long tail consumers will only spontaneously prefer to choose long tail public services when they obtain more consumer surplus than if they choose head public services: P H ≥ PL T + (VL T − W L T )

(2.2)

In this static equilibrium, the strong spillover effect of head public services (such as pure public goods) will affect not only the head preference consumers, but also the long tail preference consumers. The government has no incentive to reduce the supply price (tax), because that would attract additional ineffective consumption on the part of long tail consumers, which means that: P H N H − EC(N H ) − θ I C I C(N H ) ≥ [PL T + (VL T − W L T )]N − θ EC EC(N ) − θ I C I C(N H ) The above formula can be transformed into: PH ≥

1 {[PL T + (VL T − W L T )]N − θ EC [EC(N ) − EC(N H )]} NH

(2.3)

Similarly, long tail suppliers (NGOs) only supply to long tail consumers with their specific preferences: P H ≤ (W H − VH ) +

1 {[PL T N L T + θ EC [EC(N ) − EC(N L T )]} N

(2.4)

The price that satisfies (2.3) and (2.4) is the equilibrium price, under which the government and NGOs have no intention to reduce the price and thereby attract potential consumers. In this equilibrium, increases in the differentiation (diversity) and spillover EC of head and long tail public service levels can increase the overall welfare of different suppliers (government and NGOs). However, the increase of IC caused by the marginal decrease of scale effect will reduce the overall welfare of different suppliers. In the long tail equilibrium of rural public service, the welfare level of the head and long tail suppliers is:  LH = PH N H − θ EC C(N H ) − θ I C I C(N H )

(2.5)

LL T = PL T N L T − θ EC C(N L T ) − θ I C I C(N L T )

(2.6)

Here, C is the average marginal cost of supply, including EC and IC. Because of the convexity of the cost function, the supply quantity, quality and price competition among suppliers will decrease with the increase in the spillover cost. Therefore, for both the government and NGOs, increasing the supply of public services to attract consumers from the other side is costly. This also means that the reduction of total

24

2 Theory and Concept

supply cost (internal and spillover cost) via technological reform will depend on the change of the cost structure. For example, for the head public services, a relative reduction of spillover costs will make the proportion of IC higher, thus reducing the diseconomy of supply scale and increasing supply through fiercer competition. For long tail public services, a relative reduction of IC of supply will increase the overall welfare level of supply. (2)

Rural head public service balance

Now consider that all long tail preference suppliers do not enter the rural public services “market”; the head equilibrium is only formed by the supply from the head supplier (government). Rosen (1981) believed that, in this equilibrium, the government can achieve a monopolistic overall welfare level. In order to meet the needs of long tail consumers, (2.2) is transformed into: PLT ≤ P H − (VL T − W L T ) The welfare level of the long tail preference supplier should meet: B L T ≤ [PL T − (VL T − W L T )]N L T − θ EC EC(N L T ) − θ I C I C(N L T ) In the head equilibrium of rural public service monopolized by the government, the equilibrium price (tax) of public service set by the government should be satisfied at the same time: PH ≤ (VL T − W L T ) + θ EC C(N L T ) +

1 + λ θ I C I C(N L T ) λ N

P H ≤ (VH − W H ) + θ EC C(N H ) +

θ I C I C(N H ) N

The welfare level of the head preference supplier (government) in this equilibrium is:   θ I C I C(N L T ) H H = (VL T − W L T ) − θ EC [C(N ) − C(N L T )] N + λ

(2.7)

or HH = (VH − W H )N In the head equilibrium of rural public services, the vertical differentiation of head public services relative to long tail public services will significantly improve the welfare level in the equilibrium. This occurs because the increase of internal fixed cost (IC) can reduce the entry competition threat of other long tail preference suppliers, which in turn leads to an increase in the overall welfare level. Therefore, just like the long tail equilibrium of rural public services, the impact of total cost reduction on the welfare level in the head equilibrium depends on the reduction of EC relative to IC.

2.2 Concept

25

In real life, the long tail balance and head balance of rural public service are not fixed; rather, each will transform the other during the interaction. Three situations exist that will lead to the transition from head equilibrium to long tail equilibrium, namely (1) an increase in horizontal differentiation (diversification); (2) an increase in long tail consumers’ preferences, and (3) an increase in the IC. By comparing Eqs. (2.5) and (2.7), one can find that, for the head supplier, only if LL T ≥ 0 The welfare level in the long tail equilibrium is higher than that in the head equilibrium. This means that, in the long tail equilibrium, there is a mutual benefit effect between the head supplier and the long tail supplier. When the long tail supplier can create a producer surplus in equilibrium, the head supplier tends to choose an equilibrium price that will not “crowd out” the long tail supplier, in order to ensure the long tail equilibrium remains stable. On the other hand, when the long tail supplier creates a negative producer surplus in the equilibrium price, that supplier will choose to withdraw from the supply “market”. This will create the formation of head equilibrium with only head suppliers. Therefore, the key to the transition between the two equilibria lies in the welfare level of long tail suppliers under the equilibrium price. In addition, when the supply technology level improves, the welfare level of long tail suppliers will also improve. This book sets the share of different supply costs, θ EC and θ I C , as the two ends of the welfare curve coordinate axis of the long tail supplier (θ EC as the X axis, θ I C as the Y axis). Curve I is the long tail welfare indifference curve (concave to the origin). One can find that the zero-welfare horizontal line LL T = 0 divides the coordinate plane into two parts, namely the long tail equilibrium and the head equilibrium. Above LL T = 0, long tail suppliers can create a positive producer surplus in long tail equilibrium, and any cost proportion in this region can achieve long tail equilibrium. Below LL T = 0, the long tail suppliers will choose to withdraw from the supply “market” because of the creation of a negative producer surplus. This leaves only the head suppliers in place and maintains the head equilibrium (Fig. 2.2). The most important characteristic of a long tail welfare indifference curve is that its slope is always positive. This is because the welfare level of long tail suppliers increases in line with the increase of spillover cost and decreases in line with the decrease of IC. Assuming that the cost structure of technology that is initially available is point ‘a’ in the figure, because point ‘a’ is above LL T = 0, rural public services can achieve long tail equilibrium. When the IC of supply is significantly reduced compared with the EC, the relative cost share of EC and IC moves from point ‘a’ to point ‘b’, and the supply of the single supplier can achieve scale effect. At this cost level, on the one hand, the producer surplus of long tail suppliers is negative in equilibrium, in which case they will choose to withdraw from the supply “market”. This leads to the transformation from long tail equilibrium to head equilibrium. On the other hand, when the EC of supply is significantly lower than the IC (externality is reduced and privacy is increased), the relative cost share of EC and IC

26

2 Theory and Concept

Fig. 2.2 The equilibrium of long tail and head in rural public service

moves from point ‘b’ to point ‘c’. The long tail suppliers will then choose to enter the supply “market” and re-realize the long tail equilibrium. From the perspective of demand, the welfare level of long tail suppliers will decrease in line with the increase of VH − VL T or W L T − W H . Therefore, the rural long tail public service with significant level differentiation (diversification) has a lower zero-welfare level line, and long tail equilibrium is more likely to occur. When λ → 0, the slope of the zero-welfare level line M L T → ∞. Therefore, consumers with larger long tail preferences can serve more personalized needs. Generally speaking, this dynamic process will continue to occur with the change of supply technology. Rural public services will continue to alternate in the long tail equilibrium and head equilibrium. The welfare level of niche providers increases in line with the decrease of IC relative to spillover cost, and decreases in line with the decrease of spillover cost relative to IC. In the process of the mutual transformation of the two equilibria, due to the lagging of supply relative to demand, there will be a mismatch and an imbalance between supply and demand. (3)

Imbalance between the supply and demand of rural long tail public services

The imbalance of the supply and demand of rural long tail public services has the following situations: When the spillover cost EC and IC are very high, the monopoly effect of head suppliers tends to lead to an insufficient supply of long tail public services. When the spillover cost EC and IC are very low, the long tail effect of long tail suppliers tends to lead to the excessive entry of long tail public services. In the long tail equilibrium, both head suppliers and long tail suppliers tend to engage in an excessive diversification of supply, which leads to imbalance. In the head equilibrium of rural public service, social surplus S H is: SH =

1 λ VH N + VL T N − θ EC C(N )N − θ I C I C 1+λ 1+λ

2.2 Concept

27

In the long tail equilibrium of rural public service, social surplus S L is:  1 1 λ VH N + W L T N − θ EC S = C(N H ) 1+λ 1+λ 1+λ  λ C(N L T ) N − 2θ I C I C + 1+λ L

When S H = S L , we can define an effective supply line on the plane θ I C − θ EC , so that different marginal suppliers of rural public services enter the supply “market”; this is no different from social planners maximizing social surplus θEC = M ∗ θ I C + B ∗ Here, M ∗ and B ∗ are the slope and intercept of the effective supply line, respectively. When the cost and welfare level are above the effective supply line, social planners tend to achieve long tail equilibrium. Under an effective supply line, social planners tend to maintain the head balance. As M ∗ > 0, the effective supply line is upward sloping. As mentioned above, when the EC is relatively high, head equilibrium is more effective; when the IC is relatively high, long tail equilibrium is more effective. However, the effective supply line and zero welfare level line are not consistent; they overlap, and the deviation between them may make the supply ineffective. As shown in Fig. 2.3, due to the inconsistency between the effective supply line and the zero-welfare level line, the plane is divided by the suppliers into four parts: a, b, c and d. When λ is small, the zero-welfare level line of long tail suppliers is steeper than the effective supply line. When the scale of consumers is given as N, the supply of long tail suppliers will transfer some head preference consumers and

Fig. 2.3 The disequilibrium of rural long tail public service

28

2 Theory and Concept

reduce the spillover of supply. The imbalance between supply and demand caused by this substitution effect is reflected by either the excessive supply or insufficient supply of long tail suppliers (as shown in areas ‘c’ and ‘d’ in Fig. 2.3). When the spillover cost EC is relatively high, the exit of long tail suppliers reduces the competition between suppliers. The spillover of this substitution effect is due to the fact that the incumbent supplier has no ability to fully digest the consumer surplus of long tail consumers. Therefore, on the one hand, when the spillover cost EC is high and the IC is small, the new long tail suppliers can save the supply cost and transfer the producer surplus to the long tail consumers. On the other hand, when the spillover cost EC and IC are both high, the head supplier thinks that the monopoly supply of public goods is in line with the optimal total social welfare. Therefore, under the differentiation of rural public service levels, a change in different supply costs will lead to either insufficient or excessive supply, which will then lead in turn to imbalance between supply and demand.

2.2.3 Correction Mechanism of Rural Long Tail Public Service Previous analysis has shown that achieving the balance of rural long tail public service depends on the relative size of spillover cost and IC. In addition, welfare maximization may change in line with changes in the cost structure. The development of supply technology will create the long tail effect of rural public service supply and increase the diversity of products. However, when the initial IC and spillover cost are high, there will be either insufficient supply or excessive supply, both of which will lead to demand imbalance. Let SL T be the supply level and D L T the demand level of rural long tail public services in period t:

SL T

SL T = D L T = E G L T (Q 1 ) + E SL T (Q 2 ) − I ne f f L T [E xeG + E xe S ]

DL T =

N 

U1L T (Q 2 ) + U2L T (Q 1 ) − TL T (Q 1 ) − Fe L T (Q 2 )

(2.8)

(2.9)

i=1

Here, E G L T and E SL T , respectively, represent the financial expenditure and social expenditure of head suppliers and long tail suppliers in providing rural long tail public services in t period. Also, Q 1 and Q 2 represent the number of rural long tail public services provided by head suppliers and long tail suppliers, respectively. Next, I ne f f L T represents the inefficiency in supply; this is used to measure the governance level of supply; E xeG and E xe S represent the administrative expenses of the head supplier and the long tail supplier, respectively; U1 is the utility of the long tail attribute of rural long tail public service to consumers, which has the base

2.2 Concept

29

summability. Then, U2 refers to the utility of the public attribute part of the rural long tail public service to consumers. This utility is indivisible and exclusive, and the amount of consumption for each consumer is the same; TL T is the price (the tax paid by consumers) of providing long tail public services for head suppliers in t, and Fe L T is the fee charged for the long tail part of providing long tail public services for long tail suppliers in t. All variables meet the following requirements: ∂EGLT /∂Q1 > 0, ∂ES L T /∂ Q 2 > 0, ∂ I ne f f L T /∂(E xeG + E xe S ) > 0, ∂U1L T /∂ Q 2 > 0, ∂U2L T /∂ Q 1 > 0, ∂ TL T /∂ Q 1 > 0, ∂ Fe L T /∂ Q 2 > 0 However, this static equilibrium correction mechanism ignores the dynamic and sequential relationship between supply and demand, to some extent. Because the long tail public service is more personalized and volatile, its supply is more based on demand satisfaction. Therefore, the degree and quantity of supply has a certain time lag, compared with the generation of demand. When there is an imbalance between the supply and demand of public services in the long end of t, there may be the following two situations: When D L T > SL T , based on social welfare, public opinion, performance appraisal and other purposes, different suppliers had incentives to increase supply in the t + 1 period (that is, SL T +1 > SL T ), to make up for the inefficiency and imbalance caused by the ineffective demand in t, to a certain extent. The demand subject may inhibit the full expression of its demand in t + 1, as the demand in t has not been effectively met (that is, DT +1 < DT ), in order to improve the efficiency and imbalance. When SL T > D L T , the excessive supply will increase the burden of suppliers and consumers and can cause a waste of resources. In t + 1, different suppliers have incentives to reduce the supply, based on cost saving (that is, SL T +1 < SL T ). The excessive supply of public services in t may stimulate the generation and effective expression of demand in t + 1, so as to promote the development of public services (that is, D L T +1 > D L T ). To sum up these two situations, the supply and demand dynamic equilibrium correction mechanism of rural long tail public services meets the following conditions: ∂(D L T +1 − SL T +1 )/∂(D L T − SL T ) < 0 S L T +1 = E G L T +1 (Q 1 ) + E SL T +1 (Q 2 ) − I ne f f L T +1 [E xeG (Q 1 ) + E xe S (Q 2 )] + f (D L T − SL T , A T ) D L T +1 =

N 

(2.10)

U1L T +1 (Q 2 ) + U2L T +1 (Q 1 ) − TL T +1 (Q 1 )

i=1

− Fe L T +1 (Q 2 ) + g(D L T − SL T , BT )   (D L T − SL T , A T ) > 0, g(D (D L T − SL T , A T ) < 0 f (D L T −S L T ) L T −S L T )

(2.11)

30

2 Theory and Concept

f (0, A T ) = 0, g(0, BT ) = 0 where A T and BT , respectively, represent the supply capacity and demand characteristics of different subjects, which are endogenous. When ST +1 = DT +1  ∂Q1 : EG L T +1 (Q 1 ) − I ne f f L T +1 [E xeG (Q 1 )] + TL T +1 (Q 1 ) = U2L T +1 (Q 1 )

∂ Q 2 : E SL T +1 (Q 2 )

− I ne f

f L T +1 [E xe S (Q 2 )]

+

FeL T +1 (Q 2 )

=

N 

 U1L T +1 (Q 2 )

i=1

The optimal supply level for the head supplier in t + 1 depends on the indivisible utility, tax revenue and self-governance level brought about by the public attribute part of the rural long tail public service. The optimal supply level for the long tail suppliers in t + 1 depends on the superimposable utility of the long tail attribute part of the rural long tail public service, consumption fees and self-governance level. Therefore, the imbalance of long tail public services has a certain degree of selfcorrection mechanism in a time sequence. As shown in Fig. 2.4, when the supply and demand level of public services in period t falls in interval I, that is DT > ST , the supply and demand level of period t + 1 tends to be satisfied. When the supply and demand level falls in interval II, that is, when DT < ST , the supply and demand level of period t + 1 tends to satisfy ST +1 < DT +1 . When the supply and demand level falls to DT = ST in period T, that is, on the 45° line, the supply and demand level of period t + 1 tends to satisfy ST +1 = DT +1 . This imbalance is related to the governance level of different suppliers (the proportion of supply administration and other service expenditures):

Fig. 2.4 Supply–demand correction mechanism of rural long tail public service

2.2 Concept

31

    SL T +1 I ne f f L T +1 (E xeG + E xe S ) < 0, D L T +2 g(D L T +1 − SL T +1 ) < 0 → D L T +2 [I ne f f L T (E xeG + E xe S )] < 0

Under the condition of fixed expenditure, the lower the governance level of the supplier is in the t period, the lower the effective supply level will be in the t + 1 period. A low level of supply in the t + 1 period will further indirectly and negatively affect the level of the effective demand in the t + 2 period, in such a way that the real demand cannot be effectively revealed and satisfied: ∂(D L T − SL T )/∂ I ne f f L T [E xeG + E xe S ] > 0 The governance levels of different rural long tail public service providers are different, and these differences have different impacts on the correction mechanism of long tail public service. Because long tail suppliers are more fragmented and customized, and because the cost compensation point is low, the proportion of individual administrative expenditure does not necessarily have a significant impact on the overall imbalance of long tail public services. The “head effect” and governance level may have a greater impact on the overall imbalance between the supply and demand of long tail public services. In addition, an excessive supply of rural long tail public services will also cause imbalance and efficiency losses. Long tail level differentiation (diversification) will expand this impact and lead to a change in consumer welfare distribution. Even if the welfare of long tail consumers is improved in the long tail equilibrium (the satisfaction of demand diversity), the welfare level of head consumers may deteriorate (the “price” of head public services is increased). This book argues that, when the IC is lower than the spillover cost, the supply and demand of rural long tail public services will be unbalanced, due to excessive product diversification. The head suppliers tend to reduce the excessive differentiation of public services and social fragmentation by monopolizing supply and by creating a social consensus, so as to correct the imbalance. In this kind of imbalance correction mechanism, one should consider that the head and long tail suppliers use different H L , TIHC ), (TEC , TILC ) and have corresponding supply costs. supply technologies (TEC According to the previous analysis, based on technical differences, the “price” of the long tail equilibrium of rural public services (regardless of the relative share of EC and IC) is: H H EC H (N H ) ≥ [PL T + (VL T − W L T )](N ) − TEC EC H (N ) P H N H − TEC L L PLT N L T − TEC EC L T (N L T ) ≥ [PH + (VH − W H )](N ) − TEC EC L T (N )

The welfare level of head and long tail suppliers in the correction mechanism is: HL =

1 [(1 + λ)(VH − W H ) − λ(VL T − W L T )]N 1 + λ + λ2

32

2 Theory and Concept H   TEC λC H (N ) − (1 + λ)C H (N H ) N 1 + λ + λ2 L   TEC + (1 + λ)C L T (N ) − λC L T (N L T ) N − TIHC I C H 2 1+λ+λ λ = [(VL T − W L T ) − (1 + λ)(VL T − W L T )]N 1 + λ + λ2 H   λTEC + (1 + λ)C H (N ) − C H (N H ) N 1 + λ + λ2 L   λTEC + C L T (N ) − (1 + λ)C L T (N L T ) N − TILC I C L T 2 1+λ+λ

+

L LT

This expression is similar to that of social surplus; different suppliers may also not necessarily adopt “optimal”, but may choose “suboptimal”, supply technology. For the two kinds of suppliers, the marginal benefits to reduce the spillover cost are:   λC H (N ) − (1 + λ)C H (N H ) N ∂HL =− H 1 + λ + λ2 ∂(−TEC )   λC L T (N ) − (1 + λ)C L T (N L T ) N ∂nL =− L 1 + λ + λ2 ∂(−TEC ) From another perspective, the social surplus in the long tail imbalance correction mechanism is: TH 1 λ VH N + W L T N − EC C H (N H )N 1+λ 1+λ 1+λ L λTEC − C L T (N L T )N − TIHC I C H − TILCT I C L T 1+λ

SL =

Therefore, for the two kinds of suppliers, the marginal social welfare required to reduce the spillover cost is:   λC H (N ) − (1 + λ)C H (N H ) N C H (N H )N ∂ LH ∂S L > − = = H H 1+λ 1 + λ + λ2 ∂(−TEC ) ∂(−TEC )   L λC L T (N ) − (1 + λ)C L T (N L T ) N ∂S ∂ LL T λC L T (N L T )N > − = = L L 1+λ 1 + λ + λ2 ∂(−TEC ) ∂(−TEC ) The above inequality means that, when the long tail imbalance of rural public services occurs, different suppliers face “sub-optimal” entry motivation to improve the spillover cost saving technology. The corrective mechanism is efficient in terms of helping suppliers reduce ICs

2.2 Concept

33

∂iL ∂ SL = I Ci = , i = H, L T ∂(−TI C ) ∂(−TI C ) Assuming that the long tail suppliers can improve their supply technology and L reduce their ICs, the rural public service will re-realize long tail equilibrium (LT > L 0). Because of the decrease of the IC LT and the increase of the spillover EC, the long tail suppliers have an incentive to correct the imbalance by improving technology, reducing ICs and increasing spillover costs. The cost savings brought about by the improvement of supply technology means the long tail can be further extended. Moreover, with the increasing degree of differentiation of the supply level, long tail consumer preferences become more heterogeneous, which gives marginal suppliers greater supply incentives. Now, considering the supply network (Internet+), the supplier has strong motivation to increase the social welfare level by increasing the spillover costs or reducing ICs. When the welfare level of the head and long tail suppliers is positive under the long tail equilibrium price, they are especially inclined to adopt the network supply L technology. In other words, when the network supply changes the structure θ EC and θ ILC between different costs, the network supply can meet the demand: L H =  H H − H < 0

 LL T > 0 ∂HH = C H (N )N H ∂(−TEC ) ∂HS = I CH ∂(−TIHC ) Therefore, when the long tail imbalance of rural public services appears, the head suppliers lack the incentive to improve the network supply technology in order to save the spillover cost. Meanwhile, the long tail suppliers will adopt the network supply technology with low IC and high spillover cost, in order to correct the imbalance of supply and demand. This kind of network incentive effect is stronger when the differentiation of public service level (product diversification) increases. The supply network (Internet+) will also have a positive impact on the supply and demand imbalance of rural long tail public services. A networked supply platform and information sharing can enhance the cooperation and complementarity between suppliers, allowing the new long tail suppliers to correct the imbalance by changing the cost structure or creating new services. A ‘network of supply’ means that all suppliers in the network enjoy the same level of information, which in turn has a positive impact on rural long tail public services with strong time demand elasticity. In the model, the long tail preference increases the level of consumer welfare W L T . In the long tail equilibrium, the welfare level of long tail consumers decreases for the IC, but increases for the spillover cost. Therefore, the cost “flattening” brought

34

2 Theory and Concept

about by network supply can reduce the average cost of supply C and the long tail suppliers’ spillover costs. Generally speaking, the correction mechanism of the imbalance between the supply and demand of rural long tail public services depends on the supply cost structure, the differentiated motivation of different suppliers and the demand time elasticity of long tail consumers. The time elasticity of a supply network (Internet+), together with preference neutrality and cost reductions, can encourage the rural long tail public services to achieve balance. Specifically, in the supply of rural public services, a reduction of spillover costs makes achieving long tail equilibrium easier. In addition, a relative reduction of IC makes achieving head equilibrium easier. Therefore, the realization of this equilibrium depends on the change of cost structure (the relative size of IC and spillover cost) and the transformation of horizontal differentiation and vertical differentiation.

References Anderson, C. (2007). The long tail: How endless choice is creating unlimited demand. Random House. Aumann, R. J., & Hart, S. (2006). Handbook of game theory with economic applications (Vol. II). Baye, M. R., & Morgan, J. (1999). A folk theorem for one-shot Bertrand games. Economics Letters, 65(1), 59–65. Blundell, R. (1987). Econometric issues in public sector economics. Journal of Public Economic Theory, 10, 643–671. Brynjolfsson, E., Hu, Y. J., & Smith, M. D. (2006). From niches to riches: Anatomy of the long tail. Social Science Electronic Publishing, 47(4), 67–71. Burt, R. S. (1983). Corporate profits and cooptation: Networks of market constraints and directorate ties in the American economy. Academic Press. Burt, R. S. (1992). Structural holes. Harvard university press. Clegg, S. R., Courpasson, D., & Phillips, N. (2006). Power and organizations. Pine Forge Press. d’Aspremont, C., & Gérard-Varet, L. A. (1979). Incentives and incomplete information. Journal of Public Economics, 11(1), 25–45. Davis, S. J., Murphy, K. M., & Topel, R. H. (2004). Entry, pricing, and product design in an initially monopolized market. Journal of Political Economy, 112(S1), S188–S225. Fudenberg, D., & Tirole, J. (1991). Game theory. Galaskiewicz, J., & Wasserman, S. (1989). Mimetic processes within an interorganizational field: An empirical test. Administrative science quarterly, 454–479. Genoni, P. (2007). Libraries, the long tail and the future of legacy print collections. Libres, 17(1). Goldmanis, M., Hortaçsu, A., Syverson, C., & Emre, Ö. (2009). E-commerce and the market structure of retail industries. The Economic Journal, 120(545), 651–682. Granovetter, M. (1985). Economic action and social structure: The problem of embeddedness. American Journal of Sociology, 91(3), 481–510. Groves, T. (1973). Incentives in teams. Econometrica, 41(4), 617–631. Hurwicz, L. (1960). Optimality and informational efficiency in resource allocation processes. In Mathematical methods in the social sciences. Hurwicz, L. (1986). On informationally decentralized systems. In Decision and organization: A volume in honor of Jacob Marschak. Kendall, T. D., & Tsui, K. (2011). The economics of the long tail. The BE Journal of Economic Analysis & Policy, 11(1).

References

35

Leininger, W., Linhart, P. B., & Radner, R. (1989). Equilibria of the sealed-bid mechanism for bargaining with incomplete information. Journal of Economic Theory, 48(1), 63–106. Mas-Colell, A., Whinston, M. D., & Green, J. R. (1995). Microeconomic theory (Vol. 1). Oxford University Press. Maskin, E. (1999). Nash equilibrium and welfare optimality. The Review of Economic Studies, 66(1), 23–38. Maskin, E. (2008). Mechanism design: How to implement social goals. American Economic Review, 98(3), 567–576. Maskin, E., & Sjöström, T. (2002). Implementation theory. In Handbook of social choice and welfare (Vol. 1, pp. 237–288). Maslow, A. H. (1943). A theory of human motivation. Psychological Review, 50(4), 370. Moore, J. (1992). Implementation, contracts, and renegotiation in environments with complete information. In Advances in economic theory (Vol. 1, pp. 182–281). Moore, J., & Repullo, R. (1988). Subgame perfect implementation. Econometrica, 56(5), 1191– 1220. Myerson, R. B. (1979). Incentive compatibility and the bargaining problem. Econometrica, 47(1), 61–73. Myerson, R. B. (1982). Optimal coordination mechanisms in generalized principal–agent problems. Journal of Mathematical Economics, 10(1), 67–81. Myerson, R. B. (1986). Multistage games with communication. Econometrica, 54(2), 323–358. Myerson, R. B., & Satterthwaite, M. A. (1983). Efficient mechanisms for bilateral trading. Journal of Economic Theory, 29(2), 265–281. Olson, M. (2009). The logic of collective action: Public goods and the theory of groups. Harvard University Press. Pfeffer, J., & Salancik, G. R. (2003). The external control of organizations: A resource dependence perspective. Stanford University Press. Reiter, S. (1973). Informational efficiency of iterative processes and the size of message spaces. Journal of Economic Theory, 8(2), 193–205. Rosen, S. (1981). The economics of superstars. The American Economic Review, 71(5), 845–858. Vickrey, W. (1961). Counterspeculation, auctions, and competitive sealed tenders. The Journal of Finance, 16(1), 8–37.

Chapter 3

Attributes of Rural Long Tail Public Service

Abstract This chapter explains how the inherent attributes of rural long tail public services lead to a mismatch and imbalance between supply and demand. The imbalance of rural long tail public service has its own characteristics in spatial distribution, time sequence distribution, category level, satisfaction cost and so on. From the supply side of rural long tail public service, the long tail attribute belongs to the field of power-law distribution in mathematics. Compared with the standard normal distribution, the slope of the demand distribution of different types of rural public services is shown to be larger (the decline rate is faster), and drags a long “thick tail”, which approximately obeys the unilateral power-law. The slope of head is large and limited in terms of types; the slope of “tail” is small but with many types. The standard normal distribution rules hold that the samples are independent of each other, while the power-law distribution rules hold that the samples are related to each other. An interactive relationship does indeed exist between different types of rural long tail public services. Therefore, the long tail can be extended and developed. The supply side attribute of rural long tail public service can be analyzed from the aspects of service type, financial expenditure and NGOs. From the perspective of different service types, the rural long tail public service is “embedded” in the local cultural, social and economic environment. These services reflect the mutual integration and fuzzy boundary with the head public service. From the perspective of fiscal expenditure, rural long tail public service has the characteristics of concealment and unlimited extension. This book further explains the attributes of rural long tail public service from the NGO perspective.

3.1 The Demand Side Attribute of Rural Long Tail Public Service Following a study by Virkar and Clauset (2012), this book can analyze the long tail attributes of rural long tail public service by using the fitting degree between the double logarithm distribution of power-law distribution and a straight line. The mathematical expression of power law distribution is as follows:

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 J. Luo, Rural Long Tail Public Service and the Correction Mechanism, https://doi.org/10.1007/978-981-16-4023-0_3

37

38

3 Attributes of Rural Long Tail Public Service

y = cx −r This formula can be transformed into another formula by the logarithm of x and y: Lny = Lnc − r Lnx where −r is the slope of the double logarithmic distribution. From the perspective of demand preference, rural long tail public services are not completely independent; rather, they are related. For example, drinking water facilities and water conservancy facilities are generally positively correlated. Minimum living security and income increases are closely related to social welfare, while agricultural technology promotion, agricultural product supply and demand information are related to agricultural production guidance. It is the correlation between these demands that causes rural public services to have the “fat tail” attribute of power-law distribution. This book uses existing literature to sort the demand intensity of various rural public services (Fang, 2011; Liu, 2006a, 2006b; Sun & Lin, 2008; Wang, 2008; Zhang et al., 2011). By integrating the ranking indicators and classification results of existing literature, the proportion scores of 22 types of rural public service demand are calculated and shown in Table 3.1. Take the 22 types of rural public demand from high to low as 22 to 1. Then, take the natural logarithm of the x-axis (demand score) and y-axis (frequency) to get Fig. 3.1. To some extent, the double logarithm distribution of rural public demand is approximately fitted to a straight line. Also the reason the slope of the curve tail (BC segment) suddenly increases (but still straight-line fitting) is precisely because other types of demand in the tail contain many types of long tail demand. However, those types are not further refined, due to the limitation of indicators. Therefore, from the perspective of demand preference ranking, the long tail attribute and correlation of rural public services can be better reflected: in the interaction and integration of different types of rural long tail public services. As the curve extends to the tail, the demand quantity decreases sharply, while the demand intensity weakens. However, this kind of minority demand will not disappear completely. Instead, this demand will continue to form a new thick long tail. In particular, if there is more data and the demand type is more subdivided, the minority demand will fit the power-law distribution better.

3.2 The Supply Side Attributes of Rural Long Tail Public Service The supply attributes of rural long tail public service can be analyzed from the aspects of service type, financial expenditure and NGOs.

3.2 The Supply Side Attributes of Rural Long Tail Public Service

39

Table 3.1 Comprehensive weighted ranking of 22 types of rural public service Type

Percent % Wang (2008)

Medical service

21.5

Liu (2006a, 2006b)

8.5

Minimum living 20.8 security Rural policy

Fang (2011)

Zhang et al. (2011)

13.7

18

7.2

13.5

13.0

Increased income

Sun & Lin (2008)

9.1

9.6

16.4

7.1

11.7

13.6

Demand ranking

9.4

14.22

1

12.9

13.60

2

13.0

13.00

3

12.50

4

11.70

5

11.58

6

10.40

7

12.5

Elderly care

Weighted mean proportion score

Children’s education

19.1

Natural disaster relief

10.4

Rural road construction

19.2

3.4

4.8

5.3

16.8

9.90

8

Pre-job training

12.3

11.1

3.0

3.0

17.1

9.30

9

Agricultural technology

10.7

14.3

2.1

9.03

10

Water conservancy facilities

14.9

10.7

2.3

5.5

7.77

11

8.1

6.1

8.1

7.43

12

13.2

11.1

1.4

1.4

6.78

13

6.2

3.7

6.43

14

Public security Agricultural product information

6.4

5.5

Environmental protection

9.4

Cultural and sports activities

8.4

6.3

4.4

2.5

9.1

6.14

15

Communication facilities

14.1

3.4

1.9

0.7

3.8

4.78

16

3.4

3.6

2.9

5.5

7.77

17

3.20

18

3.03

19

2.40

20

1.80

21

Drinking water facilities Legal aid Rural cooperative credit

3.2 3.5

2.1

Social preferential treatment

2.4

Agricultural guidance

1.7

3.5

1.8

(continued)

40

3 Attributes of Rural Long Tail Public Service

Table 3.1 (continued) Type

Percent % Wang (2008)

Others

Liu (2006a, 2006b)

Fang (2011)

Zhang et al. (2011)

0.3

Sun & Lin (2008)

Weighted mean proportion score 0.30

Demand ranking

22

Note The sorting frequency and data are classified and integrated, and the weighted average is carried out according to the percentage of frequency. The weight is the reciprocal of the number of statistics in each line of the classification requirements. For example, if there are four statistics in the minimum living security line, then each statistics is multiplied by 1/4. The vacancy item indicates that the student does not have this indicator. In order to ensure the consistency of data, the original data takes one place after the decimal point; the final weighted proportion takes two places after the decimal point. Of the above, the original data of Wang (2008) is the demand order. Therefore, the average value is taken according to the total number of the three regions, and then the reciprocal value is taken. In the index of Liu (2006a, 2006b), the frequency proportion of rural road, water supply and communication construction are divided into rural road construction, drinking water facilities construction and communication facilities construction

7

a

Natural logarithm of frequency

6 5 4 3

b

2 1

c 0

Natural logarithm of score

Fig. 3.1 Double logarithm distribution of the demand preferences for rural public services

3.2.1 From the Perspective of Service Type From the perspective of different service types, rural long tail public service is “embedded” in the local cultural, social and economic environment. As such, rural long tail public service reflects the mutual integration and fuzzy boundary with the

3.2 The Supply Side Attributes of Rural Long Tail Public Service

41

head public service. Taking the rural elderly care service as an example, the traditional family elderly care mode still accounts for 96.2% of the head. Other types of rural long tail elderly care services only account for a small proportion (community elderly care 4.4%, institutional elderly care 1.9%, NGO elderly care 0.3%; data source: China Civil Affairs Statistical Yearbook). However, the distribution of this demand level is quite different in different rural areas, being affected by local cultural customs, family conditions and personal characteristics (Pathike et al., 2017; van Eeuwijk, 2006). This book can choose different types of rural public services to further explain the attributes of rural long tail public services (Table 3.2). From the distribution of rural domestic water services (see Fig. 3.2), this book finds that most rural residents in China still use tap water (head), while the demand for bottled water and purified water accounts for a small proportion (long tail). With the modernization and expansion of tap water infrastructure, tap water is widely used and popularized in most rural areas. Other types of natural water sources (well water, pond, river, lake and sea) have begun to retreat to the second line (Zhang & Li, 2009). However, due to the low return rate of water supply, scattered consumption and rural tradition, private enterprises selling bottled water have no incentive to enter the rural areas. Therefore, very few rural residents use bottled pure water (Liu, 2006a, 2006b). The distribution of rural housing services (Fig. 3.3) shows that most rural residents still choose to build their own houses. In fact, due to historical and traditional cultural influences, self-built housing has always been vigorous in rural China. As such, other types of housing services, such as public rental housing and commercial housing, are rarely witnessed (Garriga et al., 2017). However, due to the impact of urbanization and the transformation of ideas, more and more rural residents have begun to choose commercial housing, causing these long tail public services to extend along the curve (Chen et al., 2011). Table 3.2 Different categories of rural public service Service type

Long tail (%)

Head (%)

Domestic water

Bottled water/purified water/filtered water (0.58)

Tap water (57.39)

Energy fuel

Solar/biogas (1.09)

Firewood (42.47)

Toilet

Public restrooms (3.64)

Self-built toilets (96.36)

Garbage disposal

Corridor garbage passage (0.25)

Public dustbins (37.81)

Housing

Public rental housing/low rent housing (2.05)

Self-built houses (87.10)

Children’s education

International schools (0.16); migrant workers’ school (0.16)

Ordinary schools (49.03)

Child care

Nursery (7.27); nurse (0.04)

Family care (92.69)

Medical care

Special schools (4.06)

Village clinics/town health centers (56.87)

Data Source China Family Panel Survey, the respondents were residents in rural areas, the same as below

42

3 Attributes of Rural Long Tail Public Service

Number

Rainwater

Barrel water

River water

Cellar water

Pond water

Well water

Tap-water

Type

Fig. 3.2 Rural living water demand distribution

Number

Public rent

Low rent

Public

Partial property

Market commercial

Kin & Friend

Independent property

Type

Fig. 3.3 Rural living houses demand distribution

More extreme is the distribution of rural financial services (Fig. 3.4). Because the vast majority of rural residents have no demand for related financial services, other types of financial services are almost invisible (the number is too small to record). This phenomenon is related to the imperfection of China’s rural financial market (Feng et al., 2004), high saving rate culture (Yang et al., 2012) and low asset liquidity (Uchida et al., 2009). However, these long tail financial services have not disappeared completely; they are simply divided into more fragmented demand groups (such as rural elites and intellectuals).

3.2 The Supply Side Attributes of Rural Long Tail Public Service

43

Number

Foreign

Trust products

National debt

Futures

Fund

Stock

No demand

Type

Fig. 3.4 Rural living finance demand distribution

Therefore, based on different types of rural public services, the distribution and extension of the long tail curve will be selectively “embedded” in the local cultural, social and economic environment. Also, in the mutual integration with the head public service, rural public service “embeds” a certain kind of micro-heterogeneous group and survives. Different types of rural long tail public services have different degrees of micro spillover, but on the whole, they are consistent with their macro-social environment.

3.2.2 The Financial Expense Perspective From the fiscal expenditure perspective, rural long tail public service has the characteristics of concealment and unlimited extension. When the government (and the private sector from which the government purchases services) supplies public services to rural residents, based on the reasons of cost and operability, the government will not “package” all public services. Rather, the government will choose to supply the public services with strong universality and scale effect. For each type of personalized and heterogeneous public service, the government also undertakes the moral responsibility of supply. However, the government’s own capital constraints and high supply costs result in shortages of supply and vacancy. This can be better reflected in the proportion of different types of rural public service expenditure among governments (Table 3.3).

44

3 Attributes of Rural Long Tail Public Service

Table 3.3 Classifications of rural public services based on governmental budget Type

Head

Long tail

Public education

General education (primary, secondary, higher) 78.8% (0.2, 0.7, 99.1%) Vocational education 7.4%

Adult education 0.17% Radio and television education 0.14% Education abroad 0.20% Special education 0.31%

Public hospitals

General hospitals 78.1% Other specialized hospital 17.9%

Gynecology and obstetrics hospitals 2.3% Children’s hospitals 1.1% Psychiatric hospitals 0.4% Occupational disease prevention hospitals 0.2%

Public health

Major events and emergency Maternal and child health care handling 37% institutions 0.02% Disease prevention and control institutions 21.6% Other public health institutions 9%

Social security and employment

Social security fund 34.7% Minimum living security 8.8%

Jobseeker support 4.6% Social welfare 3.0% Disability 2.1% Natural disaster 1.0% Red cross 0.09% Extremely poor people 0.10%

Culture

Library 22.2% Art performance groups 23.1% Cultural exchange 13.9% Cultural protection 6.5%

Cultural activity 2.5% Art performance places 0.9%

Science and technology

Technical study 24.5% Application research 29.3% Fundamental research 9.2%

Science and technology services 3.8% Popularization of science and technology 2.0% Social sciences 1.3% Science and technology exchange 0.5%

Environmental protection

Pollution prevention 28.1% Energy conservation 14.9% Emission reduction 8.1% Ecological protection 7.8% Grain for green 8.1%

Natural forests 5.0% Renewable energy 3.6% Circular economy 1.5% Desert control 1.1% Watershed management 1.2% Returning grazing land to grassland 0.5% Environmental monitoring 1.3% (continued)

3.2 The Supply Side Attributes of Rural Long Tail Public Service

45

Table 3.3 (continued) Type

Head

Long tail

Agriculture

Agricultural reclamation 29.8% Rural public welfare 16.9% Production subsidy 9.2% Quality of agricultural products 7.0%

Agricultural resource 4.7% Pest control 3.6% Information services 3.2% Disaster prevention 1.0% Industrialization 0.2% Agricultural products processing 0.3%

Note The figure after each type is the percentage of the demand expenditure in the government budget of the total demand expenditure. The data sources are the budget (decision) tables of various government departments and the Financial Statistics Yearbook of China. The dividing point between head and long tail public services is 5% of the government’s public budget expenditure

As can be seen from Figs. 3.5 and 3.6, except for the pure public goods (such as national defense and security) that are completely provided by the government, when the government supplies all other types of rural public services, the public expenditure for different types of services also approximately fits the long tail powerlaw distribution. The government focuses on certain public services with strong policy pertinence, high demand, homogeneity and universality for centralized and key supply. For each type of public demand, in addition to these head public services that the government mainly supplies, there are many fragmented long tail public Number 3500

3000

2500

2000

1500

1000

500

0

Fig. 3.5 Distribution of governmental budget on all types of rural public service. Note The data are taken from the final national accounts of general public budget expenditure on the official website of the Ministry of Finance

46

3 Attributes of Rural Long Tail Public Service

Logarithm of amount 9 8 7 6 5 4 3 2 1 0 -1

Sorting logarithm

Fig. 3.6 The double logarithm distribution of governmental rural public service budget. Note The data are taken from the final national accounts of general public budget expenditure on the official website of the Ministry of Finance

services. These long tail public services account for a small proportion of each type of public demand, all of which have weak intensity. Therefore, based on the perspective of government expenditure supply, rural long tail public service has the “dis-embeddedness” of fiscal expenditure. This concealment is reflected in the fact that the government neither supplies directly nor supplies by purchasing from social forces, but selectively ignores supply. The fragmentation trend is often very serious; there may be few needs in the same rural area for some services (such as the needs of residents in the same village for their children’s piano and art education, or the treatment of some rare diseases). To some extent, this concealment is reflected in the fact that, with the continuous extension and refinement of the long tail curve, the number of requirements will continue to decrease. However, such requirements will not tend to zero and will always exist. Under the existing technical conditions and information costs, the cost of providing these services is very high. In addition, the concealment of rural long tail public services makes it difficult, in terms of financial expenditure, to accurately identify the long tail. However, due to the lack of an internal supervision mechanism, the government’s public choices with regard to supply can easily lead to waste and corruption. The government responds selectively from the perspective of its own benefit. The contradiction between supply and demand should cause the government to gradually realize that it should classify supply according to the characteristics of demand. Some of these demands should also be transferred to the society, instead of relying on the “omnipotent government” type of supply, regardless of cost. The infinite extension of rural long tail public service is also reflected by the multiobjective and “failure” of the government in the supply of services. Streamlining

3.2 The Supply Side Attributes of Rural Long Tail Public Service

47

administration, delegating power and reducing expenditure has become the “standard configuration” of “limited government”.

3.2.3 From the NGO Perspective Finally, from the NGO perspective, this book further explains the attributes of rural long tail public service. As mentioned above, considering the government’s inherent defects in the supply of rural public services, some services that are not suitable for the government to supply will be outsourced to other organizations (private enterprises or NGOs) in the form of purchases (Qi & Guo, 2017). Among them, NGOs, represented by non-profit organizations and based on their own charitable attributes, have the advantages in information cost and demand identification. As such, NGOs will actively choose specific public services to achieve supply (Table 3.4). One can find that, compared with rural head public services, the rural public services purchased by the government are more refined, decentralized and specific. Table 3.4 List of rural public services purchased by governments Type

Demand list

Education

A nutrition improvement plan for school students at the rural compulsory education stage, and the service of rural children’s preschool education itinerant teaching points At the compulsory education stage, rural boarding school accommodation, student canteens, clinical services, special education for blind children, deaf mute children, mentally handicapped children and other rural people

Employment

Rural labor transfer and employment services, rural labor skills and entrepreneurship training schemes, employment policy consulting businesses, rural entrepreneurship micro guarantee loan agencies, etc.

Housing

Rural indemnificatory housing insurance, etc.

Social security

Rural minimum living insurance information verification, rural endowment insurance agency, rural medical insurance agency, rural serious-illness medical insurance, rural-poor free basic funeral

Health

Rural disease prevention, health examination, health records management, early nutrition intervention in rural poverty-stricken areas, health examination of rural elderly over-65s, rural AIDS prevention, etc.

Culture and sport Rural residents’ cultural information resources sharing, film screening, delivering books, news books, plays and other public welfare culture, rural residents’ free fitness activities guidance, maintenance of rural cultural and sports infrastructure, etc. Disability

Practical skills training for the rural disabled, life care for the rural disabled, reading for the blind, sign language for the deaf, rehabilitation activities for the rural disabled, etc.

Source Guide catalogue of government purchasing public services from social forces

48

3 Attributes of Rural Long Tail Public Service

The relatively large number of NGOs, with their competitive advantages, can selectively supply according to their own supply characteristics and thus meet the diversified rural public service system requirements. It is the competition caused by the dispersion of demand and the division of supply that causes the supply of rural long tail public services to no longer have the characteristics of monopoly and concentration. Therefore, from the perspective of NGOs, rural long tail public service has the following attributes: First, rural long tail public service has an aggregation effect from tail to head. The “large scale amateur producers”, which are represented by a large number and variety of NGOs, are in line with the service characteristics brought about by the scattered resources and dispersed housing in rural areas. The “grassroots” nature of NGOs enables them to directly contact rural residents; they also have the advantage of information. As such, NGOs can realize the customized service of supply. As “long tail aggregators”, NGOs collect the heterogeneous public needs of different rural areas and different groups of people through information sharing. This makes it easy for NGOs to find and identify supply needs and to act as an information filtering mechanism. The information flow between NGOs creates a trend of cost fragmentation, which can be divided by outsourcing, in order to reduce the response cost. Anderson believes that this agglomeration effect plays the role of democratic distribution, making supply by NGOs possible by reducing the marginal cost. Second, rural long tail public services have the attributes of embeddedness and interaction. In practice, NGOs provide diversified rural long tail public services for the purpose of emotional expression, entertainment, and reputation dissemination, and not just for economic reasons. Many rural charity organizations provide rural relief and other charity activities in remote mountainous areas, precisely because the spread of social “positive energy” is conducive to the reputation of the organization. As such, the provision of such services can bring “hidden benefits” to the organization. According to Granovetter (1985), the development of rural NGOs is the result of interaction, identification and choice between NGOs and the social environment. Rural NGOs, as the suppliers of rural long tail public services, have continuously emerged and developed, taking public demand as the organization orientation. Based on the differentiation, diversification and minority aspects of the internal demand of rural long tail public service, rural NGOs are part of an institutional arrangement created to meet this demand. Third, the providers of rural long tail public services are directly faced with individual needs. However, realizing the leap forward to the transmission of services between individuals is difficult; more emphasis is placed on areas of human concentration. Now, NGOs are conducive to the transformation of rural long tail public service to head. According to the Long Tail theory, the tail product can transfer to the head and flatten the long tail curve. With rural public services, due to the uneven level of development, only the long tail meets the needs of the minority. The improvement in rural residents’ living standards may lead to an increase in the universality of demand, which in turn will be transformed into the head by the public. Due to the nature of non-distributive constraints, NGOs are more trusted. The profits obtained

3.2 The Supply Side Attributes of Rural Long Tail Public Service

49

by NGOs cannot be redistributed among members; they can only be used for organizational reproduction. All of the inputs of NGOs are used to provide services and improve service efficiency. Fourth, rural long tail public service has the attribute of club products. Due to the large number of NGOs, each NGO can only aim at a specific long tail, so as to realize the internal response to the long tail. The agglomeration effect of NGOs can better identify the real “long tail demand” of different residents in the same rural area (or even different rural areas), through their spontaneous choice of “voting with their feet”. Due to the information symmetry among the members of the organization, there is no motivation for any member to engage in free riding. They can better adopt exclusive charging to compensate for their costs and to improve efficiency. Fifth, rural long tail public service has the characteristics of flexibility and humanization, being good at attracting philanthropists and encouraging citizens to participate. In practice, NGOs pay more attention to individual needs, and mobilizing the enthusiasm of individual suppliers is easier. Various kinds of NGOs are created, developed and maintained by different “minority groups”, based on those groups’ own needs. This supply–demand relationship of “minority creates minority, minority responds to minority” is more in line with the characteristics of a smallscale economy. The emphasis is that, on the basis of the new division of labor, and according to the principle of the reasonable distribution and optimal allocation of resources, having decentralized organizations respond to public demand is more suitable to the characteristics of decentralized demand in rural areas. Generally speaking, when supplying rural long tail public services, the government will make more compromises on demand preferences, resulting in what Anderson (2006) called “minimum consistency”. This is a way to pursue the greatest common divisor at the expense of individual heterogeneity. As independent grassroots decentralized subjects, NGOs can better fit the attributes of rural long tail public services (Figs. 3.7 and 3.8). Based on the different attributes of rural long tail public services from the perspective of NGOs, can the distribution of rural NGOs fit the characteristics of long tail distribution? According to the classified statistics of China’s rural NGOs, the number of NGOs (arranged from high to low) also shows a concentrated head and a “thick” tail. The double logarithm distribution of this trend fits the straight line well and is highly matched with the rural long tail public service. However, the distribution of the corresponding types of indicators is not consistent. The indicators located at the head of the distribution of rural NGOs do not correspond to the head of rural public services. This inconsistency can be explained from the following aspects: firstly, the demand distribution of rural NGOs and rural long tail public services (especially with regard to the division of head and long tail) has a certain degree of “upside down”. The long tail of rural public services is the head of all types of rural NGOs. This phenomenon shows that NGOs are more inclined to provide rural long tail public services. Secondly, due to technical defects, the classification index of rural NGOs is not adequately subdivided. Many rural NGOs that provide rural long tail public services are generally classified; this method cannot adequately reflect the minority nature and heterogeneity of rural long tail public

50

3 Attributes of Rural Long Tail Public Service 1200 1000 800 600 400 200 0

Fig. 3.7 The distribution amounts of rural NGOs providing long tail service. Data source Official website of the Guoxin Cloud NGO data center, selected according to organization name and service area Number logarithm 8 7 6 5 4 3 2 1 0

Fig. 3.8 The double logarithm distribution of rural NGOs for the long tail

Sorting logarithm

3.2 The Supply Side Attributes of Rural Long Tail Public Service

51

services. Imagine, if we refine the different types of rural NGOs, this thick long tail can be extended to the right and tend to infinity. Finally, although rural NGOs should be created and should exist to supply these “long tails”, due to the limitations of resources, technology and other aspects, these rural NGOs fail to adequately achieve the goal of supply in reality. This phenomenon can be called “social disenchantment” in terms of functional objectives.

3.3 Imbalance Between Supply and Demand of Rural Long Tail Public Service 3.3.1 Spatial Distribution From the perspective of spatial distribution, due to the huge differences in economy, culture, customs, geographical environment and other aspects in different regions of China, the imbalance of rural long tail public services also presents a large spatial distribution difference. Generally speaking, the degree of imbalance in the eastern region is less than that in the central and western regions; the variance of the degree of imbalance in the central and western regions is also large, showing strong volatility. This book analyzes the spatial distribution characteristics from the following aspects: First of all, from the perspective of the elderly services in rural areas, China’s rural areas are deeply influenced by traditional Chinese culture. In addition, due to the relatively scattered and closed rural environment, as well as the reduced mobility and weak professionalism of personnel, the children and home-based elderly care modes in rural areas are still overseen by the majority of rural families. The proportion of homes that look after children and handle self-care for the elderly is about 56% (data are from China general social survey (CGSS), the same as below). The government and other professional institutions only provide approximately 10% of professional elderly care public services. Rural professional elderly care is still at the long tail end at this stage, while home-based elderly care is at the head end (Fig. 3.9). However, further analysis shows that, even for the same type of rural long tail public services, the demand of this long tail fluctuates significantly (with large variance). The proportion of rural public elderly care demand in different provinces can reach nearly 20 times that of other provinces. The areas with high demand for public elderly care in rural long tail are not the economically developed areas in the traditional sense. Rather, these areas tend to have a relatively fragmented geographical environment and low population density. The reason behind this may be that the high proportion of net outflow of population in these areas means many rural residents with elderly care needs do not have their children around. This makes it difficult for the children to support their elderly parents. In addition, due to the relatively inconvenient transportation systems in these areas, it is difficult for elderly rural residents to support themselves at home. The availability of convenient transport is

52

3 Attributes of Rural Long Tail Public Service

Fig. 3.9 Demand spatial distribution of rural care for the elderly. Source China general social survey (CGSS) A41. Who do you think should be responsible for providing care for the elderly with children? The spatial distribution of demand is based on the proportion of answers; the data pertaining to the rural areas in Shanghai, Tianjin, Guangdong and Tibet are missing

not as good as the availability of convenient transport in eastern areas. The elderly’s higher transportation cost confirms the long tail nature of the rural long tail elderly care demand. At the same time, this situation further shows that the complexity of rural long tail public services cannot be explained solely by the level of economic development. This book can approximately measure the supply of rural public elderly through the number and distribution of rural elderly care institutions and the number of people enjoying elderly care services. If the supply and demand of rural elderly care public services match, the greater the demand is, the greater the supply will be. In addition, the distribution trend between the two areas should be consistent. However, the truth is that, whether from the perspective of per capita possession of rural elderly care institutions or the proportion of elderly people receiving services, the supply and demand of rural elderly care are mismatched and unbalanced. The rural public service demand in areas with a greater supply of public services (such as Beijing, Jiangsu, Sichuan, Hubei, etc.) is not high throughout the whole country. However, the supply level in areas with higher demand for rural elderly public services, such as Inner Mongolia, Qinghai, Guangxi, is relatively low throughout the whole country.

3.3 Imbalance Between Supply and Demand of Rural Long Tail …

53

Fig. 3.10 Demand–supply spatial distribution of rural care for the elderly (I). Note The public elderly care service supply is measured by the number of per capita elderly care institutions in local rural areas (per 10,000 people). The demand for public pensions is the same as above

This shows that, as a kind of long tail service, an imbalance exists between the supply and demand of rural elderly care public services in different regions. In addition, this imbalance presents two levels of differentiation in different regions. Some regions have surplus supply (beyond the local demand level), while some regions have an insufficient supply to meet the local demand level (Figs. 3.10 and 3.11). According to the previous theoretical analysis, the more serious the imbalance between the supply and demand of rural long tail public services is, the lower the satisfaction of local residents for this service should be. One can see that the areas with higher comprehensive scores for rural elderly public service supply are mainly distributed in the provinces with higher populations in the central and western regions (possibly related to the large number of samples). However, the areas with higher satisfaction scores are mainly those in western regions with less dense populations, such as Qinghai, Gansu, Ningxia and Inner Mongolia. These findings have a certain fit with the proportion of rural elderly care public demand. One can infer that a positive correlation exists between the regional demand intensity and the proportion of demand satisfaction. However, no strict correlation exists between the regional demand intensity and the comprehensive scores related to supply satisfaction (Figs. 3.12 and 3.13). Next, this book analyzes the spatial distribution of the supply and demand imbalance of rural financial long tail public services. As mentioned above, due to the immaturity of and restrictions caused by financial market perfection, marketization, credit constraints, capital flow and the high savings concept in rural areas, the demand of rural residents for relevant financial public services also conforms to the long tail attribute belonging to the long tail public services. From the spatial distribution of

54

3 Attributes of Rural Long Tail Public Service

Fig. 3.11 Demand–supply spatial distribution of rural care for the elderly (II). Note The number of public pension providers is measured by the number of local people receiving public pension services, divided by the total population. The demand for public pensions is the same as in the figure above

Fig. 3.12 Supply satisfaction spatial distribution of rural care for the elderly. Source CGSS B15. Are you satisfied with the government’s performance in the following aspects? With regard to providing appropriate living security for the elderly, the calculation method of comprehensive scores is as follows: very satisfied = 2 points, satisfied = 1 point, average = 0 points, dissatisfied = −1 point, and very dissatisfied = −2 points. In the calculation method of satisfaction proportion, very satisfied and satisfied were the majority proportions in all samples

3.3 Imbalance Between Supply and Demand of Rural Long Tail …

55

Fig. 3.13 Relationship of demand intensity and satisfaction with rural elderly care

the matching of various rural financial long tail public services, the demand of rural financial long tail public services can be seen to have large differences in spatial distribution (the highest value is 14 times of the minimum value). In addition, the regions with strong demand are mainly the regions with high population density in the central and western regions. The supply and demand of rural financial long tail public services also present a relatively consistent spatial distribution trend (except in Hunan, Jiangsu, Inner Mongolia, etc.). This finding proves that the level of imbalance of rural financial long tail public services is relatively small (Fig. 3.14). Then, this book analyzes the spatial distribution characteristics of rural long tail energy public services. As mentioned above, many rural areas in China are affected by the traditional culture of the self-sufficiency of natural economy; many such areas are also rich in terms of natural resources. A considerable proportion of rural residents use coal, firewood and other natural fuel supplies as their main living energy, while the rural energy public services with scale economy (represented by pipeline gas) are in the long tail. A significant gap exists in the demand for long tail energy public services between different regions in China, with the demand of rural areas with high population density (such as Sichuan, Shandong, Liaoning, etc.) accounting for a relatively high proportion. The reasons behind this distribution trend may be related to the various local climate, geographical environment and energy reserves, but the distribution trend is not related to the economic gap between the East and the West. In addition, a phenomenon exists whereby the supply is excessive and does not match the demand in areas such as in Hubei, Jiangsu, Ningxia, and Shanxi. From Fig. 3.15, one can further see the spatial distribution of this imbalance between supply and demand. Some areas, represented by Liaoning, Chongqing and Hunan, still have unsatisfied demand (demand is greater than supply). Conversely, areas such as Shanxi, Ningxia and Jiangsu have an oversupply (supply is greater than

56

3 Attributes of Rural Long Tail Public Service

Fig. 3.14 Cumulative distribution of rural long tail financial service. Source CGSS. The demand for financial services comes from question A67. Is your family currently engaged in the following investment activities? The supply of financial services comes from C12. Do you or your spouse have the following assets? Considering the consistency of demand and supply, only four kinds of rural financial public services are considered: bond, foreign exchange, stock and fund

demand). This finding shows that, in the spatial distribution of imbalance of rural long tail energy services in China, insufficient supply and excessive supply exist at the same time. These situations should be considered separately. Next, from the perspective of rural social governance, due to the influence of traditional culture and customs, such as “Guanxi” in China (Bian, 2019), many economic and cultural activities in rural society are “embedded” in their specific social situations. In addition, due to the influence of rural clan culture, rural residents’ demand for social governance is also relatively small and discrete. In this book, six indicators of social governance have been selected, including environmental pollution, community construction, road traffic, social security, food safety and market order. These indicators are used to build a comprehensive satisfaction degree of supply and demand. The satisfaction degree in different regions of China is also found to be different in terms of spatial distribution (Fig. 3.16). The above can be further explained from different sub-indicators. From the perspective of environmental pollution, the areas with higher satisfaction (with a higher degree of balance between supply and demand) are mainly the central and western regions. These areas are rich in natural resources. The reason behind this finding may be that the externality of natural environmental protection can easily cause market failure. The environmental protection of local natural resources needs to be achieved through certain legal rules and regulations. These legalized “public goods” are mainly provided by the government, which has a certain amount of power. Green investment in different regions can be regarded as the intensity of the local

3.3 Imbalance Between Supply and Demand of Rural Long Tail …

57

Fig. 3.15 Supply–demand spatial distribution of rural long tail energy service. Source CGSS. The demand data sources are E12.a: the most commonly used cooking equipment; E52.a: the most commonly used heating equipment. The supply data source is E12.b: the type of subsidy when purchasing the most commonly used cooking equipment; E89: energy acquisition and consumption. When considering the scope matching of demand and supply, only pipeline gas and pipeline liquefied gas public pipeline are considered as long tail energy services. The method used to calculate long tail energy public service demand is as follows: the number of respondents demanding pipeline gas/pipeline liquefied gas for heating and cooking, divided by the total number of samples. The method used to calculate the long tail energy public service supply is as follows: the sum of heating and cooking pipeline gas/pipeline liquefied gas public network supply and subsidies, divided by the total number of samples

supply of environmental governance, while the level of satisfaction with environmental governance in regions with higher investment, such as Inner Mongolia and Ningxia, is not high (Fig. 3.17). The levels of satisfaction with rural long tail market order in different regions are not consistent with the local market levels. According to a previous theoretical analysis, the higher the level of marketization is, the higher the satisfaction for market order should be. However, Zhejiang and Jiangsu, which have the highest marketization level, are only in the mid-level in terms of satisfaction scores. Meanwhile, Qinghai, which has the highest degree of satisfaction with market order, has a lower level of marketization. Therefore, this finding also illustrates the imbalance of public services in terms of market order. This mismatch is also reflected in the social governance of rural road construction, social security, and community construction. Generally speaking, from the perspective of time and space, the imbalance of rural long tail public services in central and western provinces is relatively higher than in other provinces in China (Fig. 3.18).

58

3 Attributes of Rural Long Tail Public Service

Fig. 3.16 Supply–demand satisfaction spatial distribution of rural long tail law service. Source CGSS. F13. What do you think of the level of handling affairs according to law in the following aspects of social governance? Six indicators of legalized public service, including environmental pollution, community construction, road traffic, social security, food safety and market order, were selected and scored according to the degree of satisfaction (very low equals 1 point; relatively low equals 2 points; average equals 3 points; relatively high equals 4 points and very high equals 5 points). Sum the scores and divide them by the number of effective samples to get the proportion of satisfaction scores of each index. Then, average the proportion of the satisfaction scores of the six types of services to get the comprehensive satisfaction score

This imbalance between supply and demand is also reflected in the legal governance of rural road construction, social security, community construction and other public services (Figs. 3.19, 3.20, and 3.21). Generally speaking, from the perspectives of time and space, the imbalance of rural long tail public services in central and western provinces is relatively higher. Finally, rural special education is focused upon to prove the imbalance of rural long tail public services. One type of education, specifically, special education for children with special needs, is located at the end of the curve near the long tail in the distribution of different types of education-related public services. This demand belongs to the rural long tail public services (to be proved later). Due to the concealment of rural long tail public service demand, this can be proved from the perspective of supply. Firstly, from the distribution of the number of different types of education schools, rural special education does show the characteristics of power-law distribution, while the SES of special education is distributed at the “tail” end of the long tail curve, which belongs to the long tail service in rural areas (Fig. 3.22). The above can be further proved by the double logarithm diagram of the powerlaw distribution (as shown in Fig. 3.23). This research finds that the double logarithm function of the number distribution of schools offering different types of education

3.3 Imbalance Between Supply and Demand of Rural Long Tail …

59

Fig. 3.17 Supply–demand spatial distribution of environmental governance in rural long tail law service. Note The source and calculation method used for environmental governance legalization satisfaction are the same as above. The environmental governance investment data come from China’s Environmental Statistical Yearbook, which is measured by the proportion of environmental pollution governance investment in GDP

Fig. 3.18 Supply–demand spatial distribution of market order in rural long tail law service. The source and calculation method of market order legalization satisfaction are the same as above. The marketization index is derived and calculated from Fan et al. (2003)

60

3 Attributes of Rural Long Tail Public Service

Fig. 3.19 Supply–demand spatial distribution of road construction in rural long tail law service. Source The method used to calculate road construction legalization satisfaction is the same as that used above. The highway mileage figure, as an index of road construction supply, comes from the China Statistical Yearbook, and is divided into different regions, according to the length of transportation lines

Fig. 3.20 Supply–demand spatial distribution of social security in rural long tail law service. Source The method used to calculate social security legalization satisfaction is the same as used above. The supply of public security comes from China’s Procuratorial Yearbook, and is measured by the proportion of corruption cases in the total number of public officials

3.3 Imbalance Between Supply and Demand of Rural Long Tail …

61

Fig. 3.21 Supply–demand spatial distribution of community construction in rural long tail law service. Source The method used to calculate community construction legalization satisfaction is the same as above. The data pertaining to the supply of public security come from China’s Civil Affairs Statistical Yearbook; public security is measured by the number of community service institutions in each region

Fig. 3.22 Distribution amounts of different categories of schools. Source China Education Statistical Yearbook. Here, RPS refers to ordinary primary schools, PEI refers to preschool education institutions, JSS refers to junior secondary schools, APS refers to adult primary schools, LC refers to illiterate schools, RSSs refers to ordinary high schools, VHS refers to vocational high schools, RSS refers to ordinary secondary vocational schools, SWS refers to technical schools, ASS refers to adult secondary vocational schools, SES refers to special education schools, AJS refers to adult junior secondary schools, HVC refers to a higher vocational college, HEIDP refers to an undergraduate college, NH refers to a private college, AHS refers to an adult high school, AH refers to an adult undergraduate college, RI refers to other research institutes, VJS refers to a vocational junior high school, and CWS refers to a community correction school. The same applies below

62

3 Attributes of Rural Long Tail Public Service

Fig. 3.23 Double logarithm function of distribution amounts of different categories of schools. Source China Education Statistical Yearbook

has a good degree of linear fitting, thus proving the rationality of the power-law distribution. In the same way, one can also prove that special education is at the end of the long tail curve of power-law distribution. This can be done through the distribution of the number of teachers and students in different types of education (Figs. 3.24, 3.25, 3.26 and 3.27).

Fig. 3.24 Distribution amounts of faculties in different categories of schools. Source China Education Statistical Yearbook

3.3 Imbalance Between Supply and Demand of Rural Long Tail …

63

Fig. 3.25 Double logarithm function of the distribution amounts of faculties in different categories of schools. Source China Education Statistical Yearbook

Fig. 3.26 Distribution amounts of students in different categories of schools. Source China Education Statistical Yearbook. Here, NC refers to undergraduate education, SC refers to specialist education, UAH refers to adult undergraduate education, WU refers to network undergraduate education, MD refers to a master’s degree, ADP refers to an in-service Master’s degree, and DD refers to doctoral education. Other indicators are consistent with the above indicators

64

3 Attributes of Rural Long Tail Public Service

.

Fig. 3.27 Double logarithm function of the distribution amounts of students in different categories of schools. Source China Education Statistical Yearbook

It is assumed that the educational needs of rural special needs children are reasonable and objective. The proportion of uneducated special needs children in rural areas is taken as a sign of imbalance. For example, in places such as Beijing, Tianjin, Shanghai, Zhejiang and other eastern regions, because of their own comparative advantages in economic income and marketization, the imbalance degree of rural special education is less than 1%. Conversely, the imbalance in Shaanxi, Gansu, Guizhou, Xinjiang and other central and western regions is very serious (the average is 10%, or more than ten times that of the eastern region) (Fig. 3.28). This difference in spatial distribution is consistent with the concealment of rural long tail public service demand. The cost of search and identification to meet this demand is higher, which in turn puts forward higher requirements for the supply of technology and resources. These requirements are closely related to the development level of a local economy, culture, society and system (which will be proved in the next chapter).

3.3.2 Time Sequence Distribution The time sequence distribution of the imbalance of rural long tail public services reflects the trend of expanding and deepening, with the continuous development of economic, social, cultural, and institutional constraints, market-oriented reform and other factors in different regions of China. The deepening of the imbalance in time

3.3 Imbalance Between Supply and Demand of Rural Long Tail …

65

Fig. 3.28 The supply–demand disequilibrium spatial distribution of rural special education. Source The data come from China Disabled Persons’ Federation, China Rural Statistical Yearbook and China Demographic Yearbook. The calculation method is as follows: the proportion of children with special needs not enrolled in rural areas = the number of children with special needs not enrolled in rural areas/(total number of children with special needs * proportion of rural population of children)

sequence is not only related to the long tail attribute of rural long tail public service, but is also related to the interaction, regulation, supplement and even competition effect with other head public services. Firstly, this chapter measures the diachronic change of the imbalance of rural long tail public services from the perspective of the separation of different stakeholders (community providers and rural demanders). Based on the time limits of CFPS community surveys, this chapter selects the data from 2008 and 2014 to calculate the supply–demand ratio. In 2008, the supply was measured by the village/residence questionnaire, using the following questions: 1.11 Is your village/residence now called the residents’ committee or the village committee? 1.1.4 Do you have the following public welfare leisure/exercise facilities in your village/residence? The adult questionnaire was used to measure demand, including the following questions: 1.3 What is your current household registration status? 5.3 Last year, in your spare time, how often did you engage in the following activities? 6.15.2 Is your daily life taken care of? The method used to calculate the imbalance of long tail public services is the supply/demand ratio. In 2014, the supply was measured by the village/residence questionnaire, as follows: 1.11 Is your village/residence now called the residents’ committee or the village committee? We chose those rural samples who selected the village committee. We selected the number of elderly activity places in the community and divided the number of elderly community service institutions by the total number of samples to

66

3 Attributes of Rural Long Tail Public Service

measure the proportion of elderly care supply. We selected the number of communityowned sports venues, divided by the total number of samples, to measure the proportion of sports and fitness supply. We used the adult questionnaire to measure demand as follows: 1.3 What is your current account status? We chose the rural household registration as the demand sample. Question 2.3 asked, “Do you have or have you started to receive any of the following endowment insurance items?” We chose the new rural social endowment insurance (new rural endowment insurance), rural endowment insurance (old rural endowment insurance) and basic endowment insurance as the samples of the elderly group. According to the question: In the past 12 months, who took care of you when you felt sick? We chose the proportion of those who answered there was no one to take care of them as the elderly care demand. With regard to the question, “How many times have you exercised in the past week?” we chose the proportion of answering at least once as the sports and fitness needs. The method used to calculate the imbalance of long tail public services is the supply/demand ratio. As mentioned above, on the one hand, rural areas are affected by the traditional mode of providing for the aged (home-based care); public nursing homes and activity institutions are still in the long tail service end. On the other hand, based on the traditional self-sufficient natural economy and farming mode typically found in rural areas, the sports and fitness need of rural residents (for public sports facilities and venues) are also more discrete and private. Therefore, this chapter first selects these two types as the representatives of rural long tail public services. As can be found from Fig. 3.29, the supply–demand ratio of rural long tail public services represented by different interest actors fluctuated across time. Compared with 2008, the demand side (rural residents) for rural elderly care and sports public services increased in 2014. However, the supply side (represented by rural village 6 5 4 3 2 1 0

Sport service

Elderly care 2008

2014

Fig. 3.29 Time variation of supply–demand ratio for different stakeholders in rural long tail public service. Source China Family Panel Studies (CFPS)

3.3 Imbalance Between Supply and Demand of Rural Long Tail …

67

committees) remained relatively stable, so the supply–demand ratio decreased. This finding is consistent with the development trend of heterogeneity, individuation and minority of rural residents’ demands. The supply–demand ratio has decreased because of the continuous improvement and progress of the economic development level, marketization degree, modern culture and social life style of public services in rural areas. At the same time, due to the concealment of rural long tail public service demand, there is a certain time lag in the identification and supply of that demand. Therefore, compared with rural long tail public service demand, the supply development related to long tail public service demand is lagging behind, leading to the deepening of the imbalance. In order to further analyze and understand the characteristics of the deepening of the imbalance, this chapter selects the satisfaction or realization degree of the long tail demand of rural residents across different times as the measurement index. Firstly, this chapter selects rural specialized hospitals (specialized hospitals for a certain types of specific diseases, which are more heterogeneous, private and in “niche” than general hospitals) to measure the satisfaction and service level of rural residents with the supply of such hospitals as the imbalance index. The adult questionnaire asked, “Are you mainly engaged in agriculture now?” We selected the samples who answered “Yes” as the rural residents. According to the question, “If you go to see a doctor, where do you usually go?”, we selected the sample who answered “specialized hospital” as the demand. The two questions “Are you satisfied with the overall medical conditions?” and “What do you think of the medical standard there?” were chosen as the matching index of supply and demand. The responses were scored as follows: 2 points were assigned to “very satisfied (very good)”; 1 point was assigned to “satisfied (good)”; 0 points were assigned to “soso”; −1 point was assigned to “dissatisfied (bad)”; and −2 points were assigned to “very dissatisfied (very bad)”. The average value of the comprehensive score was calculated as the supply and demand matching index. The higher the score is, the higher the degree of matching and the lower the imbalance level will be (Fig. 3.30). This chapter finds that, except for the missing sample value in 2009, the degree of supply–demand matching of rural specialized hospitals in other years was more volatile. A downward trend is shown from 2011, which means that the degree of imbalance was becoming greater. Affected by age, physical exercise and health-care knowledge, the willingness of rural residents to choose community hospitals for chronic diseases is still relatively small. However, the willingness of rural residents in areas with higher economic development levels to choose such hospitals is significantly higher than in other areas (Zhang & Han, 2018a, 2018b). From the supply side perspective, this kind of demand is reasonable. A considerable proportion of township hospitals have suggested that rural patients should choose specialist hospitals for the treatment of special diseases with unknown etiology (Liu et al., 2018b). The supply side perspective also puts forward the need to establish a rural grading treatment system, including heterogeneous hospitals, as “Internet + medical care”. Secondly, this chapter analyzes the public demand for rural cooking fuel as a measure of the imbalance of rural long tail public services. As mentioned above, coal, firewood and liquefied petroleum gas are the main cooking fuels still in demand

68

3 Attributes of Rural Long Tail Public Service 3 2.5 2 1.5 1 0.5 0

2008

2009

2010

2011

2012

2014

2016

Fig. 3.30 Time distribution of supply–demand matching for rural special hospitals. Source China Family Panel Studies (CFPS)

in rural areas of China. The demand for electricity-based cooking fuel, such as for induction cookers and microwave ovens, is fragmented. Some scholars have indicated that the proportion of induction cookers and microwave ovens in rural areas of China is about 10% of all cooking appliances (Zhang & Xu, 2011). This section estimates the demand preferences of rural residents using power supply (induction cookers, rice cookers, microwave ovens, etc.) as cooking fuel and the matching degree and supply level of local electricity. The rural long tail cooking fuel public service represented by electric cookers has been found to have an increasing matching degree. Before 2009, the quality and quantity of rural electricity supply could not meet the increasing demand for cooking electricity in many rural areas of China, resulting in a negative matching degree. However, in line with the comprehensive improvement of China’s rural power supply, the cost of electricity for residents became lower (He, 2009). Cooking with electricity has gradually achieved the partial replacement of other traditional cooking fuels, such as liquefied petroleum gas, firewood and so on (Maes & Verbist, 2012). In fact, the continuous extension of the long tail public service has helped to realize the transformation from long tail to head in some areas. In 2016, the proportion of electricity cooking demand for rural residents increased to 20.9%, which was 7% higher than in 2008. Some scholars believe that firewood and straw are still used as cooking fuel in most rural areas in China, while electric cookers only account for a small proportion of appliances (Liao et al., 2016). However, the use of rural cooking electricity in China has increased significantly, in some areas accounting for more than 50% of the fuels used (Zhao et al., 2019). This also fully proves that the rural long tail can change to the head after the comprehensive improvement of infrastructure supply quality. However, the fact that the difference in service demand is still fluctuating, especially in the central and western regions of China, must still be

3.3 Imbalance Between Supply and Demand of Rural Long Tail …

69

considered. Due to the poor economic situation, the demand for cooking electricity in many western regions remains at a low level, and the duration distribution is relatively stable (Bonjour et al., 2013). The family questionnaire asked, “How many mu of land do you manage?” The samples who chose an amount larger than 0 were taken as the samples of rural residents. Another question was, “What is the main fuel for cooking in your home?” The samples who electricity were taken as the demand sample. Yet another question was, “How about the power supply of your home (existence and quality)?” As a sample of supply, the answer “no power” was given −3 points; “frequent power failure” was given −2 points; “occasional power failure” was given −1 point, and “never have a power failure” was given 2 points. The average value was then calculated to get the matching index over the years. The higher the index is, the higher is the matching degree. Since the 2016 questionnaire does not include the power supply to homes, the 2016 data are not considered (Fig. 3.31). Then, this chapter discusses the diachronic distribution of rural education long tail public service. Due to the relatively backward level of education in rural areas, and given the influence of traditional cultural concepts, public education in rural areas exists in the form of universal and homogeneous compulsory education (Gao et al., 2016). Individual training and guidance (for things such as English tutoring) has a relatively high demand level and strong heterogeneity. Even though this type of education has become an indispensable supplement to public education in urban areas, the demand in rural areas is still relatively discrete and fragmented. In the children’s questionnaire, A4 asks what the children’s current household registration type is. The sample of rural household registration is taken as the sample of rural residents. Next, G203 asks which of the children’s household registration types have participated in or are participating in tutoring/tutoring classes? The cumulative frequency of choosing foreign languages is divided by the total number of 0.6 0.4 0.2 0 -0.2

2008

2009

2010

2011

2012

2014

-0.4 -0.6 -0.8 -1 -1.2

Fig. 3.31 Time distribution of supply–demand matching for rural long tail cooking fuels. Source China Family Panel Studies (CFPS)

70

3 Attributes of Rural Long Tail Public Service

samples as the supply degree of rural long tail education services. In the adult questionnaire, A2 asks what the current household registration status is. The sample of rural household registration is taken as the sample of rural residents. Next, D1 asks how important the respondents think it is to be able to use the following language in communication: “What do you think of your ability to use English in communication?” and G18 asks, with regard to using a foreign language, “Do you or do you need to use a foreign language in this job?” (2014, 2016). Based on the demand level (where 1–5 points are allocated to very unimportant, very important, neither unimportant nor important, very satisfied, and very dissatisfied, respectively), divide the cumulative score for English by the total number of samples to get the demand degree of rural long tail education service. Finally, the matching degree of supply and demand is calculated by dividing the score of demand degree by the score of supply degree. This type of public education emphasizes small-scale teaching and customized “tailoring” and belongs in the category of rural long tail public service. According to the time distribution of the matching degree of English tutoring among rural residents in Fig. 3.32, the matching degree shows a trend of decreasing at first and then increasing (related to the difference in the selection of indicators). Before 2012, the supply of English tutoring and tutoring in other subjects in rural areas remained stable, but the demand for such heterogeneous education services rose sharply. Since 2014, more rural areas have begun to provide and carry out personalized education services through extracurricular teaching and tutoring, which has helped alleviate this imbalance trend. Some scholars believe that there are both external and internal reasons for the insufficient supply of education, as represented by English tutoring in rural areas (Mahmud & Bray, 2017). The external reasons lie in the lack of objective conditions under which rural students can participate in extracurricular tutoring. For example, 5 4 3 2 1 0

2010

2011

2012

2014

2016

-1

Fig. 3.32 Time distribution of supply–demand matching for rural long tail education service. Source China Family Panel Studies (CFPS)

3.3 Imbalance Between Supply and Demand of Rural Long Tail …

71

there is a lack of teachers, and the ability of students’ families to pay for such tutoring is limited. This long tail education service belongs in the category of club goods and has a certain degree of profit compensation effect. The internal reason is that rural families’ demand for this supplementary education is still in the developing stage; their subjective willingness is also not strong. However, under the premise of pursuing the balance of education between urban and rural areas, the demand for extra-curricular tutoring in rural areas should also be paid due attention. Taking rural extra-curricular tutoring as the starting point, the supply of and satisfaction with rural long tail education services can effectively eliminate the negative impact of rural parents’ migrant work on their children’s education. This supplementary supply of pursuing education quality can ensure the effective improvement of rural children’s comprehensive quality and the effective accumulation of human capital (Du Plessis, 2014). Finally, this chapter discusses the diachronic distribution of the imbalance of rural special medical service as a long tail public service. According to the previous definition of long tail public service, the demand for medical service of rural residents with special diseases, such as physical disability and other infectious diseases (AIDS, rabies, etc.) also has the characteristic of heterogeneity. Basic medical insurance and basic services (such as the new rural cooperative medical scheme) are often insufficient in terms of meeting the needs of these special needs groups. As such, they are in an imbalance state of insufficient supply, to a large extent. The basic information question BC001 asked, “What is your current hukou status?” The residents who chose “rural household registration” were taken as the sample population of rural demand. In the area of health status and functioning, DA005 asked, “Do you have any of the following disabilities?” Residents with a physical disability, brain damage, mental retardation, blindness, deafness or severe stuttering were selected as the demand samples. Next, DA044 asked, “Have you been diagnosed with any of the following infectious diseases?” Rural residents suffering from at least one of tuberculosis, hepatitis B, malaria, rabies, schistosomiasis, AIDS, Japanese encephalitis, dysentery, measles, brucellosis, gonorrhea and syphilis were also taken as demand samples. These demand subjects were asked question DB022, “Who can help you most with dressing, bathing, eating, getting up, going to the toilet, controlling defecation, housework, cooking, shopping, managing money, making phone calls, taking medicine and other difficulties?” The residents who responded that they “have no one to help” were chosen as the manifestation of insufficient supply; the sample number of this option was divided by the total sample number to get the occurrence ratio of imbalance (insufficient supply). As shown in Fig. 3.33, the special medical demand groups were selected as the main body of the long tail medical public service. Also, the diachronic distribution presents a fluctuating state of declining at first and then rising. In particular, the diseases of the rural residents represented by the disabled are becoming more complex (Lezzoni et al., 2006). More rural disabled residents are having difficulty getting the care of public medical personnel; their families provide most of the care (Rasmussen et al., 2019). This situation causes this special medical service to continue to extend along the long tail, showing a trend of diachronic fragmentation.

72

3 Attributes of Rural Long Tail Public Service 0.2 0.18 0.16 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0

2008

2011

2012

2013

2015

Fig. 3.33 Time distribution of supply–demand matching for rural long tail special health service. Source CHARLS (China health and retirement longitudinal study)

As confirmed by the research of Li et al. (2015b), China’s rural areas have different levels of economic development. In recent years, the prevention and treatment of special major diseases, such as AIDS and rabies, has gradually been included in the compensation and protection benefits of the new rural cooperative medical scheme (Chen et al., 2018). Many regions in China have introduced special disease compensation policies with their own characteristics (Li et al., 2015a). Based on their own differences in health conditions, these regions use different security methods for special groups with rare diseases. However, the lack of unified norms and guidance means the supply of more long tail disease prevention and control is still absent. In addition, this book can further support the perspective from the diachronic distribution of the matching degree of rural long tail special medical insurance. The rural medical insurance system in China includes new rural cooperative medical insurance (xinnonghe), medical insurance for urban and rural residents, public medical treatment, medical assistance, commercial medical insurance, and other benefits. The rural medical insurance system helps to improve the happiness and satisfaction of rural residents (Han & Gao, 2020). However, the public service of medical insurance, which is mainly based on the overall planning for major diseases, has limited substitution scope in terms of rural family elderly care. This situation has led to a large gap in the demand for special medical insurance, based on the characteristics of rural heterogeneity. The basic information question BC001 asked, “What is your current hukou status?” The individual residents who chose rural household registration were taken as the sample population of rural demand. The health status and functioning question DA005 asked, “Do you have one of the following disabilities?” Rural residents with one or more of the following disabilities: physical disability, brain damage, mental retardation, blindness, deafness, hoarseness or severe stuttering, were chosen as the demand sample. Question DA044 asked, “Have you been diagnosed with any of the following infectious diseases?” Those rural residents who answered that they had or

3.3 Imbalance Between Supply and Demand of Rural Long Tail …

73

had previously had one or more of tuberculosis, hepatitis B, malaria, rabies, schistosomiasis, AIDS, Japanese encephalitis, dysentery, measles, brucellosis, gonorrhea, syphilis were also chosen for the demand sample. These demanders were asked EA001, with relation to their health care and insurance, “Do you currently participate in the following medical insurance schemes?” The residents who answered that they had “no insurance” are regarded as the embodiment of the insufficient supply of special medical insurance. The sample number of this option was divided by the total sample number to get the occurrence ratio of the imbalance (insufficient supply). As shown in Fig. 3.34, in the distribution of the special medical insurance service in the long tail, the degree of imbalance was lower in 2015 than in 2011. However, the distribution was still in the “stable” state of insufficient supply and continuous extension of demand. The proportion of rural special medical insurance demand groups represented by rural disabled elderly is still relatively low. In addition, significant differences exist in the matching of the supply and demand of medical insurance for different types of disability characteristics. For example, the amount of medical insurance and medical services provided for those with intellectual disabilities under new rural cooperative medical schemes is lower than the proportion of those who receive limb disability coverage. Let us sum up the above-mentioned diachronic distribution of the imbalance of various types of rural long tail public services. This section finds that rural long tail public services, based on their own heterogeneity, have significant volatility in terms of demand and supply. Some categories, such as special medical insurance and public education, have a diachronic convergence trend in the imbalance. This is related to the ongoing continuous improvement and development of the supply level and diversification of different suppliers. There is also a trend of diachronic divergence in the degree of imbalance in other categories, such as special medical services, cooking fuel services, and specialized hospital public services. This phenomenon is 0.09 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0

2011

2012

2013

2015

Fig. 3.34 Time distribution of supply–demand matching for rural long tail special Medicare. Source CHARLS (China health and retirement longitudinal study)

74

3 Attributes of Rural Long Tail Public Service

related to the continuous extension of the long tail demand of rural residents and a certain degree of concealment. Therefore, conducting a customized analysis of different types of rural long tail public services is urgently required.

3.3.3 Satisfaction Cost From the perspective of the imbalances’ satisfaction cost, compared with the head public service, there is no scale effect in the supply cost of rural long tail public service. This leads to head public service supply in the form of discretization, fragmentation and individuation, making it difficult to aggregate the demand between people and regions and thereby reduce the supply cost. Different suppliers need to pay higher search and matching costs to correct the mismatch. The government’s information cost in terms of identification and supply is high. In addition, the government does not have the advantage of mobility, which leads to a further increase in the satisfaction cost. Based on the analysis of the supply and demand characteristics of rural areas with heterogeneous demand, general differences have been found in the degree of demand for elderly care, geographical employment and public transport among rural residents represented by the elderly and family core groups. Meanwhile, the unreasonable layout and spatial distribution of public facilities further worsens the satisfaction cost of mismatch. The development characteristics of demand can be further analyzed from the perspectives of the satisfaction cost of different types of rural long tail public services. First of all, take rural long tail special healthcare as an example. For rural residents with special needs (such as rural residents with disabilities and rare diseases), the imbalance of supply is also reflected in the cost of meeting the demand. The cost to rural residents with special needs to self-finance medical care has been rising over time. However, the income from pension subsidies did not increase significantly from 2010 to 2019. The basic information question BC001 asked, “What is your current hukou status?” The individual residents who answered rural household registration were taken as the rural population demand sample. A health status and functioning question was “Do you have the following disability problems?” Rural residents with one or more of the physical disabilities of brain damage, mental retardation, blindness, deafness, and severe stuttering were selected as the demand samples. A work, retirement and pension question asked, “In the past year, how much did you get from your pensions (including basic government pension, public institutions and enterprises, supplementary pension of enterprises, pension of urban and rural residents, commercial pension, pension subsidy for the elderly, etc.)?” We take these amounts as the pension subsidy income. The question was also asked, “What is the total cost of your medical expenses in the past month?” These amounts were taken as the self-financing medical cost.

3.3 Imbalance Between Supply and Demand of Rural Long Tail …

75

250

3000 2500

200

2000 150 1500 100 1000 50

0

500

2008

2011

2012

Self-financing cost

2013

2015

0

Pension

Fig. 3.35 Satisfaction cost of rural long tail special Medicare. Source CHARLS (China health and retirement longitudinal study)

Figure 3.35 shows that the imbalance cost of rural long tail special health insurance is facing an increasing trend to a certain extent. On the one hand, the supply cost for rural residents with special needs remains stable. On the other hand, the demand cost (including opportunity cost) is rising. The heterogeneous demand of rural residents for medical insurance will affect the degree of trust in the government; social capital and the cost–benefit ratio will also be affected (Norstrand & Xu, 2012). In addition, the gradual disappearance of any family risk sharing mechanism means rural residents with special health insurance needs will be faced with greater opportunity cost. The increase in the mismatch cost also means rural residents with special needs will be faced with adverse selection in terms of medical services. Their willingness to participate is also related to the size of their pension subsidy income, reimbursement ratio and coverage. Under existing institutional arrangements, rural residents with long tail medical insurance demand have insufficient motivation to participate in insurance schemes. The new rural cooperative medical scheme is mainly based on coverage for common diseases, making it difficult to meet the diversified heterogeneous needs and thereby leading the rural residents covered by such schemes to fall into a new round of poverty trap (Liu et al., 2018a). This can be further illustrated by the characteristics of the imbalance of rural long tail health services with regards to meeting the cost. Basic information: question BC001 asked, “What is your current hukou status?” The individual residents who chose rural household registration were taken as the sample population of rural demand. A health status and functioning question asked, “Do you have the following disability problems?” Rural residents with one or more of the physical disabilities of brain damage, mental retardation, blindness, deafness, or severe stuttering were selected as demand samples. Question GB005 asked, “In the past year, what was the total income from the agricultural products and forest products produced by your

76

3 Attributes of Rural Long Tail Public Service

family?” The average amount of income over the year was taken as the total income of rural residents. For the question, “What was your family’s expenditure for health consumption in the past month?” If the respondent couldn’t remember clearly, the default was set to 0. The average household health consumption expenditure over the year, divided by the total income of rural residents, was taken as the health proportion and used to measure the satisfaction cost of the imbalance. Another question was, “How much did your family spend on taxes and miscellaneous fees (excluding income tax) handed over to relevant government departments in the past year?” The average amount of income was regarded as the “price” (average supply cost) of public health services. As shown by Fig. 3.36, the mismatch cost experienced a U-shaped fluctuation and gradually increased. The mismatch cost can be seen as a supplement to the insufficient supply of rural general public medical services. The long tail medical satisfaction cost raised by rural residents with special needs presents a relatively stable development trend. This means that, compared with the head health public services, which are mainly supported by public expenditure, the long tail special health public services still occupy a stable proportion and have not further increased in line with the increasing cost. The underlying logic behind this finding may be that, with the continuous improvement of rural residents’ income, the coverage of medical insurance has not been effectively followed up. This means rural residents will find it difficult to be reimbursed for the expenses of treating many chronic diseases. For rural residents with special health needs, the opportunity and entry costs of choosing to receive services have not been reduced. These residents will consciously take the appropriate proportion of their own income and expenditure to self-finance special health services. This sector is not sensitive to the change in the satisfaction cost of their medical insurance. Moreover, with the continuous occurrence of 60

0.8 0.7

50

0.6 40

0.5

30

0.4 0.3

20

0.2 10

0.1 0

2008

2011

Tax fee

2012

2013

2015

0

Health

Fig. 3.36 Satisfaction cost of supply–demand matching for rural long tail special health service. Source CHARLS (China health and retirement longitudinal study)

3.3 Imbalance Between Supply and Demand of Rural Long Tail …

77

rural residents moving to the city in China (Zhang et al., 2019), the mobile cost to the remaining rural residents is higher; the rural residents with special needs also account for an important proportion of those remaining residents. The supply mode of club goods (outside the self-financing medical public services) may continue to increase. Another possibility is that the increase in the self-financing of health expenditure on the part of rural residents with special needs will offset the rapid rise of medical service prices. Since the drug price elasticity of health expenditure in China is significantly negative, an increase in self-financing cost expenditure will not be effectively controlled. In fact, such an increase may further stimulate the rise of short-term medical prices, making it difficult for the medical subsidies to restrain the rising of rural health costs. Hence, the satisfaction cost of residents’ access to long tail medical services is actually increasing. Secondly, heterogeneous rural residents have a relatively small and fragmented long tail demand for housekeeping services. Although the supply to meet this kind of demand is mostly fulfilled by the market-oriented mode, the government is relied upon to provide a sound and reasonable household market access system and relevant management norms. The greater the expenditure ratio of rural residents’ housekeeping service consumption is to their agricultural income, the greater their demand for rural long tail domestic public services will be. The tax paid by rural residents (excluding income tax) can be regarded as the “price” of the public services provided by the government. The basic information question BC001 asked, “What is your current hukou status?” The individual residents who chose rural household registration were taken as the sample population of rural demand. A health status and function question asked, “Do you have the following disability problems?” Rural residents with one or more of the physical disabilities of brain damage, mental retardation, blindness, deafness, and severe stuttering were selected as the demand samples. Question GB005 asked, “In the past year, what was the total income received for agricultural products and forest products produced by your family?” The average amount of residents’ income over the year is regarded as the total income of those rural residents. Part GE009 asked interviewees to state “your family’s expenditure in the following items in the past month: the expenditure of nanny, hourly workers, servants, etc.” If the residents could not remember clearly, the default was set to 0. The residents’ average household consumption expenditure over the year, divided by the average total income of rural residents, was taken as the proportion of household services and used to measure the satisfaction cost of the imbalance. Another question was, “How much did your family spend on taxes and miscellaneous fees (excluding income tax) handed over to relevant government departments in the past year?” The average amount of income was regarded as the “price” (average supply cost) of public services. In Fig. 3.37, the proportion of the housekeeping service consumption of rural residents with heterogeneous demand (taking rural disabled residents as the sample) shows an upward trend. Meanwhile, the supply “price” (tax) of rural public service remains stable. The satisfaction cost of the rural housekeeping service mismatch is

78

3 Attributes of Rural Long Tail Public Service 0.0035

60

0.003

50

0.0025

40

0.002

30

0.0015

20

0.001

10

0.0005 0

2008

2011

Tax

2012

2013

2015

0

Housekeeping

Fig. 3.37 Satisfaction cost of supply–demand matching for rural long tail housekeeping service. Source CHARLS (China health and retirement longitudinal study)

also rising. There are several reasons behind this. On the one hand, with the continuous increase in the urbanization of rural residents, the opportunity cost of human capital (including nannies and other housekeeping service providers) in rural areas is rising. The “price” of the supply of these services is also rising (Buera & Kaboski, 2012). On the other hand, due to the decrease in the working elderly population and the rapid increase in the elderly, rural housekeeping service is faced with weak institutional supply, a lack of responsibility on the part of housekeeping enterprises, and insufficient social security (Yang, 2021). A contradiction exists between the huge demand for rural housekeeping services and the low efficiency of service supply. Therefore, this mismatch leads to a rise in service costs, as reflected in the increasing proportion of household services consumed by rural residents. Third, this section analyzes the characteristics of the satisfaction cost of the imbalance of rural long tail transportation public services. Relevant studies have pointed out that the provision of rural public transport services is limited by the fragmentation of demand, as well as by geographical separation and economic income levels (Bocarejo & Oviedo, 2012). Public transportation service often faces the trade-off between social welfare and economic benefits (Currie & Stanley, 2008). In the current situation of the continuing separation of urban and rural dual systems in China, with the continuous improvements in urban areas, many rural residents are choosing to enter urban areas. The demand on the part of rural residents for public transport to shorten the time and distance of commuting is becoming more and more intense (Partridge et al., 2010). This is reflected in the imbalance of rural long tail transportation public services. Basic information question BC001 asks, “What is your current hukou status?” The individual residents who choose rural household registration were taken as the sample population of rural demand. A health status and function question asked, “Do you have any of the following disability problems?” Rural residents who had one or

3.3 Imbalance Between Supply and Demand of Rural Long Tail …

79

more of the physical disabilities of brain damage, mental retardation, blindness, deafness, or severe stuttering were selected as demand samples. Question GB005 asked, “In the past year, what was the total income from agricultural products and forest products produced by your family?” The residents’ average amount of income over the year was taken as the average total income of rural residents. Question GE009 asked, “What was your family’s expenditure in the in the past month on transportation expenses?” If the residents could not remember clearly, the default value was set to 0. This section divides the annual average of transportation consumption expenditure by the average total income of rural residents as the proportion of transportation public services. The result is used to measure the satisfaction cost of the imbalance of long tail transportation public services. Another question asked, “How much did your family spend on taxes and miscellaneous fees (excluding income tax) handed over to relevant government departments in the past year?” The average amount of income was regarded as the “price” (average supply cost) of the public services. As shown in Fig. 3.38, the proportion of transportation expenses needed to meet the cost also shows a rising trend. On the one hand, the increase of rural residents’ transportation consumption cost reflects the continuous development of the rural economy, the continuous improvement of rural residents’ quality of life and the demand for information exchange (Starkey et al., 2002). On the other hand, rural residents’ demand for public transportation services should be supplied according to local conditions, such as the use of small electric vehicles in areas with a poor geographical environment. This approach would help meet the special living and commuting needs of rural residents. Fourth, this section analyzes the satisfaction cost of rural long tail entertainment public services. Rural long tail entertainment public services include the provision of cultural and entertainment facilities, such as cinemas and opera houses. This 0.014

60

0.012

50

0.01

40

0.008

30

0.006

20

0.004

10

0.002 0

2008

2011

Tax

2012

2013

2015

0

Transportation

Fig. 3.38 Satisfaction cost of supply–demand matching for rural long tail transportation service. Source CHARLS (China health and retirement longitudinal study)

80

3 Attributes of Rural Long Tail Public Service

kind of demand has typical characteristics of minority and heterogeneity. With the improvements in rural income and education levels, the supply of entertainment public services has an obvious increasing trend (Royo-Vela, 2009). Basic information question BC001 asked, “What is your current hukou status?” The individual residents who chose rural household registration were taken as the sample population of rural demand. A health status and function question asked, “Do you have the following disability problems?” Rural residents with one or more of the physical disabilities of brain damage, mental retardation, blindness, deafness, or severe stuttering were selected as demand samples. Question GB005 asked, “In the past year, what was the total income from agricultural products and forest products produced by your family?” The residents’ average amount of income over the year was regarded as the average total income of rural residents. Question GE009 asked, “What was your family’s expenditure on culture and entertainment (opera, drama, film, etc.) in the past month?” If the resident could not remember clearly, the default value was set to 0. The residents’ average culture and entertainment consumption expenditure over the year, divided by the average total income of rural residents, was taken as the proportion of culture and entertainment public services. The result was used to measure the satisfaction cost of the imbalance. Another question was “How much did your family spend on taxes and miscellaneous fees (excluding income tax) handed over to relevant government departments in the past year?” The average amount of expenditure was regarded as the “price” (average supply cost) of public services. As shown in Fig. 3.39, the imbalance satisfaction cost of rural long tail entertainment public services (the proportion of cultural and entertainment expenditure) increased significantly after 2011, remaining high and stable until around 2014. This development trend shows that the rural residents’ increasing demand for long tail 0.0035

60

0.003

50

0.0025

40

0.002

30

0.0015

20

0.001

10

0.0005 0

2008

2011

Tax

2012

2013

2015

0

Entertainment

Fig. 3.39 Satisfaction cost of supply–demand matching for rural long tail entertainment service. Source CHARLS (China health and retirement longitudinal study)

3.3 Imbalance Between Supply and Demand of Rural Long Tail …

81

entertainment public services is mainly met by market-oriented payment. However, the absence of government supply has greatly lifted the satisfaction cost of this imbalance. The supply of opera public service, as represented by the “new rural construction”, has caused this type of cultural entertainment communication to gradually turn into family club-type public goods. In practice, the new mode of government purchase and supply is aimed at the development strategy of cultural differentiation. However, this strategy also encounters such problems as insufficient implementation, the weak ability of NGOs, and disconnection between cultural products and community people’s needs (Rao, 2008). This results in high satisfaction cost (sunk cost and opportunity cost of early construction) and affects rural residents’ satisfaction with cultural and entertainment services. Fifthly, this section analyzes the satisfaction cost of rural long tail education public service. Rural long tail education public service refers to the demand type of heterogeneous rural residents for other decentralized education services (such as special education, rural tutoring, and art training). This kind of long tail education service is based more on the market-oriented supply of club goods. However, the government provides the corresponding institutional environment and access policies, thereby playing an effective role in the rational operation of this kind of market. From this perspective, the market cultivation of long tail education public service has a public nature. Basic information question BC001 asks, “What is your current hukou status?” The individual residents who chose rural household registration were taken as the sample population of rural demand. A health status and function question asked, “Do you have the following disability problems?” Rural residents who said they had one or more of the physical disabilities of brain damage, mental retardation, blindness, deafness or severe stuttering were selected as demand samples. Question GB005 asked, “In the past year, what was the total income from agricultural products and forest products produced by your family?” The average amount of residents’ income over the year was taken as the average total income of rural residents. Question GE009 asked, “What was your family’s expenditure on education and training, etc. in the past month?” If the residents could not remember clearly, the default value was set to 0. The residents’ average annual education consumption expenditure was divided by the average total income of rural residents, as the proportion of education services. The result was used to measure the satisfaction cost of the imbalance. Residents were also asked, “How much did your family spend on taxes and miscellaneous fees (excluding income tax) handed over to relevant government departments in the past year?” The average amount of residents’ expenditure was regarded as the “price” (average supply cost) of public services. As shown in Fig. 3.40, the imbalance of rural long tail education public service is basically stable and at a high level. That is to say, the government has not realized an effective supply to correct the privatization of satisfaction cost. The fragmentation and minority of such demand is very serious and is highly related to the income levels of rural residents and the cost of education. Hence, the satisfaction cost does not show a further increasing trend. From the perspective of the supply of special education, the existing standardized mode of many developing countries cannot meet

82

3 Attributes of Rural Long Tail Public Service 60

0.25

50

0.2

40

0.15

30 0.1

20

0.05 0

10

2008

2011

Tax

2012

2013

2015

0

Education

Fig. 3.40 Satisfaction cost of supply–demand matching for rural long tail education service. Source CHARLS (China health and retirement longitudinal study)

the personalized teaching needs of rural children with special needs (Kalyanpur, 2011). When recruiting students, special schools restrict many pupils from entering on the basis of the students’ registered residence, age, intelligence quotient and other factors. For example, many deaf children in rural areas are excluded (Garner, 2009). In addition, at the supply system level, the imbalance is deepened, due to the imperfect mechanism of “learning in a regular class”, receiving insufficient attention from society, the dislocation of public opinion and a lack of civil power. From the perspective of the demand for special education, the families of children with special needs in rural areas can easily find themselves in a state of poverty, meaning they are unable to pay for the expenses related to special education (Pasachoff, 2011). These children have certain physical and psychological defects and are too easily forgotten by society. Constrained by traditional ideas, the concept of special education is generally lacking in China’s rural areas; the level of parents’ enthusiasm for children’s special education is also not high (Osgood, 2008). These factors lead to the mismatch of public education services in the long tail of the village, and the satisfaction cost has been maintained at a high level. Finally, this section analyzes the satisfaction cost of rural long tail fitness public service. Of all the above types of rural long tail public services, the fitness needs are more at the end of the long tail curve. Rural areas in China are still dominated by agriculture-related activities (planting, breeding, etc.), which have the characteristics of human and physical intensity (Siciliano, 2012). After a long period of intense physical labor, historically, most residents normally had little demand for public fitness facilities. However, in recent years, with the progress of rural economic and social development, the supply of rural public sports facilities has also become insufficient in the “new rural construction” (Chen & Liu, 2020).

3.3 Imbalance Between Supply and Demand of Rural Long Tail …

83

Basic information question BC001 asked, “What is your current hukou status?” The individual residents who chose rural household registration were taken as the sample population of rural demand. A health status and function question asked, “Do you have the following disability problems?” Rural residents who said they had one or more of the physical disabilities of brain damage, mental retardation, blindness, deafness, or severe stuttering were selected as demand samples. Question GB005 asked, “In the past year, what was the total income from agricultural products and forest products produced by your family?” The average amount of residents’ annual income was regarded as the average total income of rural residents. Question GE009 asked, “What was your family’s expenditure on fitness in the past month?” If residents could not remember clearly, the default value was set to 0. This section divides the residents’ average annual fitness consumption expenditure by the resident’s average total annual income. The result is taken as the proportion of fitness services and is used to measures the satisfaction cost of the imbalance. Another question asked, “How much did your family spend on taxes and miscellaneous fees (excluding income tax) handed over to relevant government departments in the past year?” The average amount of resident’s expenditure was regarded as the “price” (average supply cost) of public services. As shown in Fig. 3.41, the imbalance of rural long tail fitness public service shows a downward trend, to a certain extent. Local governments have an incentive motivation to pay attention to the implementation and supply of rural sports fitness projects, and to encourage rural residents to participate in sports and fitness activities. However, although the supply of fitness infrastructure is scientifically conducive to the formation of a healthy lifestyle for rural residents, there is a spatial and geographical mismatch (Chen et al., 2020). In short, the rural fitness public service supply represented by fitness equipment is more inclined to be built in township center areas 0.014

60

0.012

50

0.01

40

0.008

30

0.006

20

0.004

10

0.002 0

2008

2011

Tax

2012

2013

2015

0

Fitness proportion

Fig. 3.41 Satisfaction cost of supply–demand matching for rural long tail fitness service. Source CHARLS (China health and retirement longitudinal study)

84

3 Attributes of Rural Long Tail Public Service

(Cui, 2012). For the widely-scattered rural residents, the transaction cost is higher. The care with regard to the humanistic sports demand of rural residents is insufficient, causing the policy implementation to be volatile and limited. Therefore, this decline in a certain proportion of the satisfaction cost does not mean that the mismatch is alleviated; rather, the mismatch is related to the lack of an expression mechanism and feedback channel. Generally speaking, although different types of rural long tail public services have different characteristics in terms of the satisfaction cost (culture, transportation, housekeeping, and medical insurance are rising, while education and health are stable, and fitness is declining), the logic behind these services remains the same. Based on the heterogeneity, minority and decentralization of rural long tail public services, there are always imbalances in time, space and subject of demand. This is reflected in the satisfaction cost with dynamic fluctuations in time series and geographical differences in space. This kind of satisfaction cost is not what rural residents should have to bear; rather, they should only have to bear the opportunity cost and transaction cost based on their own needs. The complementary effect of the basic public services is not strong; even the substitution and spillover effect based on their own income elasticity may appear. Therefore, to analyze the causes and influencing factors of the imbalance of rural long tail public services, the heterogeneous paths and mechanisms of different cost types must be analyzed from the perspective of economic input and output. This is the content of the next chapter.

References Anderson, C. (2006).The long tail: Why the future of business is selling less of more. Hachette Books. Bian, Y. (2019). Guanxi, how China works. Wiley. Bocarejo, S. J. P., & Oviedo, H. D. R. (2012). Transport accessibility and social inequities: A tool for identification of mobility needs and evaluation of transport investments. Journal of Transport Geography, 24, 142–154. Bonjour, S., Adair-Rohani, H., Wolf, J., Bruce, N. G., Mehta, S., Prüss-Ustün, A., Lahiff, M., Rehfuess, E. A., Mishra, V., & Smith, K. R. (2013). Solid fuel use for household cooking: Country and regional estimates for 1980–2010. Environmental Health Perspectives, 121(7), 784–790. Buera, F. J., & Kaboski, J. P. (2012). The rise of the service economy. American Economic Review, 102(6), 2540–2569. Chen, J., Bai, Y., Zhang, P., Qiu, J., Hu, Y., Wang, T., & Gong, P. (2020). A spatial distribution equilibrium evaluation of health service resources at community grid scale in Yichang, China. Sustainability, 12(1), 52. Chen, J., Dong, H., Yu, H., Gu, Y., & Zhang, T. (2018). Impact of new rural cooperative medical scheme on the equity of health services in rural China. BMC Health Services Research, 18(1), 1–7. Chen, J., Guo, F., & Wu, Y. (2011). One decade of urban housing reform in China: Urban housing price dynamics and the role of migration and urbanization, 1995–2005. Habitat International, 35(1), 1–8.

References

85

Chen, Q., & Liu, T. (2020). The effectiveness of community sports provision on social inclusion and public health in rural China. International Journal of Environmental Research and Public Health, 17(2), 597. Cui, R. (2012). Research status on equalization of public sports service for nationwide fitness in Hebei. Physics Procedia, 25, 2298–2303. Currie, G., & Stanley, J. (2008). Investigating links between social capital and public transport. Transport Reviews, 28(4), 529–547. Du Plessis, P. (2014). Problems and complexities in rural schools: Challenges of education and social development. Mediterranean Journal of Social Sciences, 5(20), 1109. Fang, K. (2011). Research on demand preference, structure and expression mechanism of rural public service—Based on the questionnaire survey and statistics of Eastern, Central, Western and Northeast China. Agricultural Economy and Management, 4, 46–53. (in Chinese). Fan, G., Wang, X. L., Zhang, L. W., & Zhu, H. P (2003). Report on the relative progress of marketization in various regions of China. Economic Studies, 3, 9–18+89 (in Chinese). Feng, X., Amp, H. P., & He, G. (2004). Why we need rural financial pluralizationin China: An analysis from a local knowledge paradigm perspective. China Rural Survey, 5, 001. Gao, Y., He, Q., Liu, Y., Zhang, L., Wang, H., & Cai, E. (2016). Imbalance in spatial accessibility to primary and secondary schools in china: Guidance for education sustainability. Sustainability, 8(12), 1236. Garner, P. (2009). Special educational needs: The key concepts. Routledge. Garriga, C., Hedlund, A., Tang, Y., & Wang, P. (2017). Rural-urban migration, structural transformation, and housing markets in China (No. w23819). National Bureau of Economic Research. Granovetter, M. (1985). Economic action and social structure: The problem of embeddedness. American Journal of Sociology, 91(3), 481–510. Han, H., & Gao, Q. (2020). Does welfare participation improve life satisfaction? Evidence from panel data in rural China. Journal of Happiness Studies, 21(5), 1795–1822. He, L. (2009). Financing rural renewable energy: A comparison between China and India. Renewable and Sustainable Energy Reviews, 13(5), 1096–1103. Kalyanpur, M. (2011). Paradigm and paradox: Education for All and the inclusion of children with disabilities in Cambodia. International Journal of Inclusive Education, 15(10), 1053–1071. Lezzoni, L. I., Killeen, M. B., & O’Day, B. L. (2006). Rural residents with disabilities confront substantial barriers to obtaining primary care. Health Services Research, 41(4p1), 1258–1275. Li, C., Hou, Y., Sun, M., Lu, J., Wang, Y., Li, X., & Hao, M. (2015a). An evaluation of China’s new rural cooperative medical system: Achievements and inadequacies from policy goals. BMC Public Health, 15(1), 1–9. Li, Y., Long, H., & Liu, Y. (2015b). Spatio-temporal pattern of China’s rural development: A rurality index perspective. Journal of Rural Studies, 38, 12–26. Liao, H., Tang, X., & Wei, Y. M. (2016). Solid fuel use in rural China and its health effects. Renewable and Sustainable Energy Reviews, 60, 900–908. Liu, Y. (2006a). Constructing the supply system of farmers’ demand-oriented public goods—Based on the analysis of a national rural public goods demand questionnaire. Journal of Central China Normal University, 45(2), 15–23 (in Chinese). Liu, Y. Q. (2006b). Formating presants’ demand-oriented supply system of pubic goods—Based on a surrey of public goods demand in rural areas in China. Journal of Huazhong Normal University (Humanities and Social Sciences), 2, 003 (in Chinese). Liu, Y., Guo, Y., & Zhou, Y. (2018a). Poverty alleviation in rural China: Policy changes, future challenges and policy implications. China Agricultural Economic Review. Liu, Y., Zhong, L., Yuan, S., & van de Klundert, J. (2018b). Why patients prefer high-level healthcare facilities: A qualitative study using focus groups in rural and urban China. BMJ Global Health, 3(5). Maes, W. H., & Verbist, B. (2012). Increasing the sustainability of household cooking in developing countries: Policy implications. Renewable and Sustainable Energy Reviews, 16(6), 4204–4221.

86

3 Attributes of Rural Long Tail Public Service

Mahmud, R., & Bray, M. (2017). School factors underlying demand for private supplementary tutoring in English: Urban and rural variations in Bangladesh. Asia Pacific Journal of Education, 37(3), 299–309. Norstrand, J. A., & Xu, Q. (2012). Social capital and health outcomes among older adults in China: The urban–rural dimension. The Gerontologist, 52(3), 325–334. Osgood, R. L. (2008). The history of special education: A struggle for equality in American public schools. Greenwood Publishing Group. Partridge, M. D., Ali, K., & Olfert, M. R. (2010). Rural-to-urban commuting: Three degrees of integration. Growth and Change, 41(2), 303–335. Pasachoff, E. (2011). Special education, poverty, and the limits of private enforcement. Notre Dame Law Review, 86, 1413. Pathike, W., O’Brien, A. P., & Hunter, S. (2017). Moving on from adversity: An understanding of resilience in rural Thai older people. Aging & Mental Health, 1–8. Qi, L., & Guo, J. (2017). Understanding government purchasing public services in China: Case study of Guangdong and Yunnan. American Journal of Industrial and Business Management, 7(03), 312. Rao, S. S. (2008). Social development in Indian rural communities: Adoption of telecentres. International Journal of Information Management, 28(6), 474–482. Rasmussen, J. D., Kakuhikire, B., Baguma, C., Ashaba, S., Cooper-Vince, C. E., Perkins, J. M., & Tsai, A. C. (2019). Portrayals of mental illness, treatment, and relapse and their effects on the stigma of mental illness: Population-based, randomized survey experiment in rural Uganda. PLoS Medicine, 16(9), e1002908. Royo-Vela, M. (2009). Rural-cultural excursion conceptualization: A local tourism marketing management model based on tourist destination image measurement. Tourism Management, 30(3), 419–428. Siciliano, G. (2012). Urbanization strategies, rural development and land use changes in China: A multiple-level integrated assessment. Land Use Policy, 29(1), 165–178. Starkey, P., Ellis, S., Hine, J., & Ternell, A. (2002). Improving rural mobility: Options for developing motorized and nonmotorized transport in rural areas. The World Bank. Sun, C., & Lin, W. (2008). Analysis of farmers’ demand willingness for rural public services— An empirical study based on a nationwide survey of farmers. Journal of China Agricultural University, 25(3), 134–143. (in Chinese). Van Eeuwijk, P. (2006). Old-age vulnerability, ill-health and care support in urban areas of Indonesia. Ageing & Society, 26(1), 61–80. Uchida, E., Rozelle, S., & Xu, J. (2009). Conservation payments, liquidity constraints and off-farm labor: impact of the Grain for Green program on rural households in China. In An integrated assessment of China’s ecological restoration programs (pp. 131–157). Springer, Dordrecht. Virkar, Y., & Clauset, A. (2012). Power-law distributions in binned empirical data. Siam Review, 51(4), 661–703. Wang, Q. (2008). Analysis of rural public service supply desirability and demand degree from the perspective of farmers: A case study of three counties and cities in Shandong Province. Shandong Social Sciences, 3, 152–155. (in Chinese). Yang, D. T., Zhang, J., & Zhou, S. (2012). Why are saving rates so high in China? University of Chicago Press. Yang, L. (2021). Community-based elderly care in Beijing: Status and prospects. Analysis of the Development of Beijing, 2019, 291–325. Zhang, L., Ding, Z., Qiu, L., & Li, A. (2019). Falls and risk factors of falls for urban and rural community-dwelling older adults in China. BMC Geriatrics, 19(1), 1–17. Zhang, L. R., Li, J. C., & Fan, H. L. (2011). Research on rural public service demand preference and satisfaction based on income difference. Chinese Public Administration, 10, 118–122. (in Chinese). Zhang, N., & Xu, W. (2011). Analysis of household electricity consumption based on energy self selection behavior. China Rural Economy, 7, 72–84. (in Chinese).

References

87

Zhang, R., & Li, H. (2009). Current situation analysis on China rural drinking water quality. Journal of Environment and Health, 26(1), 3–5. Zhang, M. S., & Han, J. F. (2018a). Building a new rural pension model of “charity + poverty alleviation + industry”. Academic Journal of Zhongzhou, 258(6), 68–73. Zhang, X., & Han, L. (2018b). Which factors affect farmers’ willingness for rural community remediation? A tale of three rural villages in China. Land Use Policy, 74, 195–203. Zhao, N., Zhang, Y., Li, B., Hao, J., Chen, D., Zhou, Y., & Dong, R. (2019). Natural gas and electricity: Two perspective technologies of substituting coal-burning stoves for rural heating and cooking in Hebei Province of China. Energy Science & Engineering, 7(1), 120–131.

Chapter 4

The Influencing Factors of the Imbalance of Rural Long Tail Public Services

Abstract This chapter examines the factors that influence the imbalance of rural long tail public services. The imbalance of rural long tail public services is caused by a series of complex factors, including many aspects of economy, society, culture and the institutional environment, with multi-subjectivity and dimensions. This chapter first describes the imbalance from both sides of supply and demand. Then, according to the different fields of rural long tail public services, empirical research is conducted from the perspectives of rural special education, special health, elderly care and finance. With regard to the demand side factors, the overall reason for the imbalance of rural long tail public service is that rural residents’ public demand is irrational, discrete, fragmented and atomized. The irrationality of rural residents’ long tail demand refers to the irrationality based on personal preferences and cognition when choosing the required public services. In particular, there is often an irrational side in the selection of the types of public services with information cost and human capital requirements. The indifference utility curve of farmers in terms of public service demand always tends to be extended, without considering the decision-making cost. The decision is related to the non-economic factors existing in the demand selection, which means the over-marginal cost expenditure of the supply curve is unable to measure the total cost, and this makes the demand cost deviate from the expected income. Of course, the irrational demand is related to the rational factors behind that demand, such as the decentralized management mode of farmers, as well as social system obstacles (such as the household registration system) in the marketization process. With regard to the factors that affect the supply side of rural long tail public services, this section intends to analyze the supply reasons and influencing factors of the imbalance from the following perspectives: government financial constraints, government rational choice deviation, immature NGOs, and the lack of a supply “market”. After summarizing the overall reasons for the imbalance in this section, the next three sections will show the results of an empirical regression analysis of specific rural long tail public services, including special education, special health, special elderly care, and special finance. Field-based reasoning and deduction will also be conducted with regard to the above-mentioned influencing factors.

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 J. Luo, Rural Long Tail Public Service and the Correction Mechanism, https://doi.org/10.1007/978-981-16-4023-0_4

89

90

4 The Influencing Factors of the Imbalance of Rural Long Tail …

4.1 The Demand Side Factors of Rural Long Tail Public Services The overall reason for the imbalance of rural long tail public service is that rural residents’ public demand is irrational, discrete, fragmented and atomized. The irrationality of rural residents’ long tail demand refers to the irrationality based on personal preferences and cognition when choosing the required public services. In particular, there is often an irrational side in the selection of the types of public services with information cost and human capital requirements (such as special health needs, special literature and art needs, etc.). The indifference utility curve of farmers in public service demand always tends to be extended, without considering the decision-making cost. This demand is related to the non-economic factors existing in the demand selection, which means the over-marginal cost expenditure of the supply curve is unable to measure the total cost and causes the demand cost to deviate from the expected income (Buchanan, 1978). Of course, the irrational demand is related to the rational factors behind that demand, such as the decentralized management mode of farmers, and the social system obstacles (such as the household registration system) in the marketization process. The dispersion, fragmentation and atomization of demand refer to the tendencies of scattered flow, urban–rural separation, fragmentation of residence and disintegration of traditional culture in the production and life of rural residents. This, in turn, leads to the phenomenon of minority and heterogeneity in rural residents’ demand for public services. This discrete and fragmented demand state leads to a certain “structural tearing” of rural families, especially for rural “left behind families”, which will affect the orderly progress of rural life and production. This dispersion of demand is the essential distribution feature of long tail public services and brings huge information and cost challenges to the effective and accurate identification of suppliers. On the other hand, the dispersion of demand is also related to the income levels of rural residents and the local geographical characteristics. The dispersion presents a trend of aggregation dispersion distribution, from township to peripheral rural areas. The dispersion also decreases in line with the increase of rural residents’ net income. The radiation effect of the surrounding central city economic circle will also affect the dispersion of local rural residents’ demand. Generally speaking, the irrational and discrete demand of rural residents will significantly affect the marginal cost of rural long tail public service supply, as well as the improvement of investment satisfaction and the social welfare level.

4.2 The Supply Side Factors of Rural Long Tail Public Services This section analyzes the supply reasons and influencing factors of the imbalance of rural long tail public service from the following perspectives: government financial

4.2 The Supply Side Factors of Rural Long Tail Public Services

91

constraints, government rational choice deviation, immature NGOs, and the lack of a supply “market”.

4.2.1 Government Financial Constraints Buchanan and Musgrave (1999) proposed that the supply subject of public goods depends on the subject’s own publicity, as related to the efficiency of different supply methods (collective supply or private supply). For any type of public goods or services, the expected results of public (government) supply must be compared with the expected results of using a non-collective and voluntary market supply. Under the background of an increasing fiscal deficit in China, the government having multiple objectives in responding to the “long tail demand” of rural public services may lead to “government failure”. With the transformation of China’s government from an “omnipotent government” to a “limited government”, streamlining administration, delegating power and reducing expenditure have become the standard practices of the “limited government”. From 2010 to 2019, public expenditure on agriculture was the central expenditure with the largest increase. The expenditure on agriculture, forestry and water rose from 126.8 billion yuan in 2010, to 702 billion yuan in 2019, an increase of 454%. However, responding to the increasingly diversified, heterogeneous and personalized public demand of rural residents is still difficult. The financial pressure on the government is increasing day-by-day. The scale of government public expenditure has been effectively limited. As such, it is difficult for the government to respond to the “long tail demand” without considering the economic cost. This kind of restriction on government finances in the supply of rural long tail public services reflects the contradiction between the limited supply capacity and the unlimited demand space. The limitation of supply capacity means that the supply subject represented by the government must supply (not produce) the public services suitable for the government’s own characteristics. Here, one must distinguish the difference between the supply and production of public services. As Buchanan (2014) said, the public subject represented by the government morally undertakes the “supply” of all citizens’ public services. However, the government does not necessarily undertake the “production” of public services. In fact, the government can actively attract other entities to participate in the necessary production through purchasing and outsourcing. In the context of building a service-oriented government in China, the limitation of supply capacity is in line with the objective reality. In addition a precise institutional arrangement of a government supply boundary can effectively avoid the alienation trend of a borderless supply. The infinite nature of rural long tail public service demand space means that rural residents’ perceptions and judgments of their own needs will have a certain subjectivity. Although all needs are based on certain objective realities, the needs associated with rural long tail public service may have an evolution path of infinite accumulation and diminishing marginal effect in the process of meeting. Based on the analysis

92

4 The Influencing Factors of the Imbalance of Rural Long Tail …

of the value premise between unlimited demand and limited supply, the long tail of rural long tail public service can better reflect the limited rationality of the demand subject’s cognitive breadth and depth. The unlimited extension of this demand objectively reflects the improvement of rural social material living conditions, as well as the improvement of rural residents’ public awareness.

4.2.2 The Deviation of the Government’s Rational Choice The government’s political decision-making in the supply of public services is a complex process and is different from the supply mechanism of private goods in the market (Buchanan, 2014). In the process of transferring, assembling and transforming the choice of individual public services into collective choices, the government is affected by information asymmetry and the deviation of rational choice. This is a kind of politically-motivated behavior that includes different subjects in a more diversified dimension. As an entity organization, the government also needs to consume all kinds of resources in production. According to the public choice theory, with the continuous development of the economy and the continuous improvement of people’s material lives, the public service supply undertaken by the government is prone to the inequality of financial and administrative power. Financial competition between local governments will also help ensure that governments at all levels selectively supply public services when they are faced with a financial deficit, in order to make up for the lack of fiscal revenue (Buchanan & Masgrave, 1999). The cost of providing public services includes the opportunity cost of subjective evaluation by decision-makers, and this cost cannot be measured by objective indicators. For example, this cost cannot be calculated by the collective welfare function that sums up the utility that individual demanders are willing to give up (Buchanan, 1978). In the decision-making structure of non-market choice, assuming that the real cost and benefit of providing rural public services can be measured, only those policy-makers who pursue fair distribution and efficiency and make decisions according to mechanical choice can guarantee the Pareto optimality of non-market mechanism. Such a scenario is obviously not in line with reality. In addition, the government’s rational bias is to be more inclined to supply public services with social benefits and obvious results. This bias has a distribution distortion effect on the cost of public services (tax). Therefore, the heterogeneity of demand, the goals of officials in China’s political ecosystem, and the characteristics of the financial budget (such as “emphasizing” infrastructure construction investment while “neglecting” the improvement of people’s livelihoods and welfare) can explain the lack of supply of rural long tail public services. This phenomenon can be explained from the perspectives of fiscal decentralization, local expenditure competition, promotion incentives and rational behavior.

4.2 The Supply Side Factors of Rural Long Tail Public Services

93

4.2.3 Immature NGOs As another subject of public service supply, NGOs can make up for the defects of government supply, based on their own advantages. This has become one of the hot topics of public service research in recent years. However, based on the increasing complexity and discretization of public demand, how to better resolve the contradiction between the people’s growing needs for a better life and the unbalanced and inadequate development of NGOs’ characteristics is still controversial. The development of NGOs in China is restricted by the environment, resources and other conditions, and is still in an immature state. The internal driving force of NGO demand in China is restricted by the factors and conditions of economic and social differences among provinces. From the perspective of culture and resource supply, the lack of a citizen culture and identity leads to the congenital dilemma and resource dependence of the development of NGOs in China. From the perspective of institutional structure, the development of China’s NGOs is restricted by the laws and regulations represented by the dual management system. Meanwhile, China’s unique “administrative absorption of society” leads to the development of NGOs depending on the external environment system (Thornton, 2013). Generally speaking, China’s NGOs are underdeveloped in both the social and economic aspects of providing rural long tail public services. China’s NGOs are also difficult to compare with more mature organizations based in civil societies in the West. Recently, the new mode of China’s NGOs’ public service supply, namely “social enterprise”, has been the subject of constant debate. Arguments swirl around this mode’s legal identity, method of action, and the conflict and cooperation of this mode with traditional organizations. The NGOs represented by social enterprises are engaged in the supply of charitable public services. They can supply charitable services more efficiently through the market-oriented and commercial operation mode, including personnel autonomy, responsibility clarification, resource transparency and social credibility. However, the enterprise risk and market failure faced by the “social enterprise” mode can also easily lead to the loss of social basis in the supply of rural long tail public services, and the evolution into a product of administrative dependence. Therefore, the dual attributes of NGOs are embodied in their mutual assistance and self-help in the supply of rural long tail public services. A construction relationship of sequence exists. Some organizations, when they are first established, take public welfare altruism as their purpose. Some organizations go on to transform to a business model, in order to overcome the problem of sustainable social effects in the process of developing. Likewise, some NGOs, when they are first established, take public welfare altruism as their purpose. These organizations originated from traditional private enterprises. Then, after achieving a certain scale of economic benefits and based on the needs of their own development, they began to pursue corporate social responsibility and invest in social influence. Still other NGOs set up at the beginning with charitable mutual assistance and self-economic sustainability

94

4 The Influencing Factors of the Imbalance of Rural Long Tail …

balance as their development goal. Based on the construction sequence of these dual attributes, the immaturity and imperfection of the development of China’s NGOs can hardly reflect the deep relationship between the existing theories for localization needs and heterogeneous development approaches. Especially in China’s unique economic and social environment, the “acclimatization” and self-transformation of Western imported NGOs in China are also related to the construction sequence of the dual attributes of Chinese NGOs.

4.2.4 Lack of a Supply “Market” Because of the non-competitive, non-exclusive, externality and homogeneity characteristics of consumption, it is difficult for public goods (especially pure public goods) to make profits through private supply in the market. The result for the suppliers, therefore, is “market failure”. However, rural long tail public service has the characteristics of competition, exclusiveness, separability and heterogeneity of consumption, similar to club goods. Theoretically, rural long tail public service demand can be satisfied to a certain degree through private market supply. As Anderson (2007) pointed out, with the development of supply technology and the entry of large-scale amateur producers into the supply chain, the marginal cost of increasing supply is reduced to zero, which in turn makes the supply of long tail niche products profitable. This effect of creating aggregated personalized demand can even be compared with the most popular products. The democratization of production tools has transformed consumerism into “participatory productivism”; the convergence of the roles of consumers and producers has also made the boundary between them increasingly blurred. This mode of “self-production” is embedded in the context of largescale voluntarism and amateurism, especially in the case of grassroots “amateur” producers. Although the supply level and scope are uneven, with the increase of frictionless liquidity in the long tail, even a long tail with small demand can stimulate strong resonance, thus producing the agglomeration effect and forming an “agglomeration of long tail”. The aggregation effect of small “long tail demand” suppliers is what creates the decentralized small demand (which was originally thought to have no profit space) and forms scale effect and scope economy through the integration and aggregation of the Internet. However, when applied to the long tail demand of rural public services in China, a significant gap still exists between the market agglomeration and scale supply effect brought about by the long tail aggregator and the reality of the situation. With the continuous extension of rural long tail public service demand, the role boundary of supply is also changing. This change of supply mechanism is reflected in the influence and change of public nature, technological progress, market fairness, efficiency criteria, policy orientation and private capital scale. The “market” supply mechanism of public service is a mechanism that profit-making organizations use to compensate for their expenditures by charging fees. The fees are set according to market demand. Due to the imperfection of China’s market economy, a need exists

4.2 The Supply Side Factors of Rural Long Tail Public Services

95

for deeper political system reform. The rights of citizens to democracy and freedom have not been fully reflected in the system. The unstable and uncertain factors in the internal and external environment of the market economy have also increased. Especially in recent years, with the great downward pressure on China’s economy, the pressure for further reform of the market economy has been objectively weakened. This has made it difficult for market actors to show enough initiative and enthusiasm in providing rural long tail public services. Therefore, the lack of “market” supply represented by rural long tail public services is not only related to the inherent demand characteristics and nature of rural long tail public service, but is also related to the objective market and institutional environment in rural areas. For a long time, this will be a normalization phenomenon faced by residents in rural areas in China. After summarizing the overall reasons for imbalance in this section, the next three sections will conduct an empirical regression analysis, using specific rural long tail public services (including special education, special health, special elderly care, and special finance). Field-based reasoning and deduction will also be conducted on the above-mentioned influencing factors.

4.3 The Factors of Imbalance: Special Education Rural special education refers to the type of education provided for school-age children with physical or mental disabilities in rural areas. Based on the heterogeneous needs of rural special children, rural special education has the inherent attributes of customization and privatization (Dhuey & Lipscomb, 2013). Due to the vast territory, scattered population, strong natural risk and relatively backward economic development typical of rural areas, on the one hand, rural special education also reflects the characteristics of fragmented distribution, geographical isolation and unbalanced development. Relative limitations and shortages in terms of financial and community resources must also be faced (Collins & Ludlow, 2018). On the other hand, the demand for special education in rural areas is also increasing in line with the continuous diffusion of the heterogeneity of special needs children. The above factors can easily cause the imbalance of rural special education in developing countries, and clearly put forward higher requirements for the accurate identification of the demand for this type of education. The concept of rural special education has been widely accepted by the public. However, unlike other types of education, how the unique internal attribute of rural special education is connected with its supply imbalance and how to correct this imbalance continues to puzzle mainstream academia. Few studies exist that examine the embeddedness of rural special education in terms of supply and governance structure (see Edmonds & Spradlin, 2010). From the broader perspective of general special education expenditure, one must consider the interaction and regulatory role behind this imbalance. Besides the government, other types of social actors have different incentive effects in this self-correction mechanism. Due to the heterogeneity of different providers and actors, a significant gap exists between the supply and

96

4 The Influencing Factors of the Imbalance of Rural Long Tail …

demand of rural special education (Sindelar et al., 2018). However, the accurate identification of different social actors and adequate expenditure input should help to alleviate the shortage of supply. Affected by the terrain, culture and customs of rural areas, the provision of rural special education should be consistent with the internal characteristics of rural special needs children. Although the satisfaction of rural special education will not affect the basic needs and security of most people, this type of education can improve the marginal welfare of local special needs children under the background of diversified public needs. In order to better understand the inherent attribute of rural special education in terms of the supply shortage and mismatch, we use the Long Tail theory to understand the incentive mechanism of different actors in correcting the imbalance. On the long tail curve of education public service, general basic education, as the head demand, has universality, homogeneity and scale effect. As a long tail, rural special education has the characteristics of minority and heterogeneity. Based on the discrete and fragmented population and geographical distribution of rural areas, rural special education has dual attributes. On the one hand, as a kind of public education service, rural special education is not only conducive to the improvement of individual human capital and knowledge skills (Jung & Thorbecke, 2003), but also has positive externalities that affect local social stability and economic development (Yin et al., 2017). On the other hand, rural special education also has certain characteristics of long tail private attributes: demand concealment, distribution fragmentation, utility divisibility, competitiveness and exclusiveness (Mason-Williams, 2015). In essence, rural special education is still a public service. However, compared with other head types of education, the publicity of this type of education is relatively weak. Therefore, the dual attributes of rural special education are the heterogeneous and personalized supplement under the premise of basic education. The long tail attribute of rural special education presents challenges relating to the accurate identification, positioning and immediate effective supply of different suppliers, making it easy for the suppliers to be faced with the imbalance. The suppliers have different incentives to correct the imbalance. The government, due to the political governance performance and the pressure of public opinion, has incentive motivation to improve the attendance rate of rural special education, or at least control the dropout rate in the lowest range. The inherent altruistic and charitable attributes of other social participants mean they tend to care about the needs of vulnerable groups (including rural special needs children). Even so, for different suppliers, the supply and investment of rural special education is not immediate but has a time lag. When determining the supply level, the number of suppliers is based more on the existing demand than the new demand. Therefore, the correction mechanism of rural special education has a certain lag. Set ST and DT as the supply and demand of rural special education (RSE) in period T. In the static optimal equilibrium, the following holds true: ST = D T

(4.1)

4.3 The Factors of Imbalance: Special Education

ST = E G T (IGT ) + E S T (I ST ) − I ne f f GT (IGT ) − I ne f f ST (I ST ) DT = U P T (IGT ) +

N 

U L T (I ST ) − TT (IGT ) − FeT (I ST )

97

(4.2)

(4.3)

i=1

In the supply equation, E G T and E S T are the utility functions of governments and other social sectors (such as NGOs, private enterprises and individual donations), respectively, for RSE provision in period T. Also, IGT and I ST are the inputs of expenditures from governments and other social sectors, respectively, for RSE in period T; I ne f f GT is the function of governance inefficiency for governments in RSE provision, and I ne f f ST is the function of governance inefficiency for other social sectors in RSE provision. In the demand equation, U P T is the utility function of the public goods nature of RSE for consumers; this function has public visibility and spillover effects. The authors feel that this variable is more reflected by the input and provision of governments. Based on the maximization of social welfare as a whole, on the one hand, governments are inclined to invest in and satisfy those public goods with high homogeneity and scale effects. Hence, U P T is the function of IGT . On the other hand, U L T is the utility function of the ‘long tail’ nature of RSE for consumers in T, which has superposition property due to its private separability. The authors feel that U L T is more satisfied and aroused by the input of other social sectors, due to the scattering distribution, various categories, and the large number of these organizations, as well as their customized services. Also, N is the consumption amount of RSE; TT is the ‘price’ (tax paid by consumers) of RSE provided by governments, and Fe T represents the service fees for the RSE provided by other social sectors (which could be zero for charitable provisions). All variables satisfy the following: ∂ E G T /∂ IGT > 0, ∂ E ST /∂ I ST > 0, ∂U P T /∂(IGT ) > 0, ∂U L T /∂(I ST ) > 0, ∂ TT /∂(IGT ) > 0, ∂ FeT /∂(I ST ) > 0. However, this static equilibrium neglects the dynamic fluctuation of RSE. Due to the ‘long tail’ nature of RSE (such as its preference for concealment and a scattering distribution), it is difficult for suppliers to accurately recognize, locate and effectively provide RSE in time. This issue increases the likelihood of insufficient supply or a mismatch of supply and demand (mismatch hereafter) for RSE. When this mismatch occurred in the past, different suppliers typically had the impetus and motivation to help correct this imbalance (supply insufficiency) with respect to different factors. Governments face political, governance performance, and social reputation requirements to enhance the RSE enrollment rate, or to at least limit the drop-out rate to a minimum standard. Other social sectors (especially NGOs) have the intrinsic nature of altruism and charity to help disadvantaged groups, including uneducated

98

4 The Influencing Factors of the Imbalance of Rural Long Tail …

special needs children. However, it is also difficult for them (especially governments) to invest in RSE provisions and observe or achieve instant effects. This issue indicates a time-lag effect in the impact of investment in RSE correction. In addition, when supplying RSE, suppliers are more focused on satisfying existing needs than on emerging demands. Therefore, compared with the emergence of demands, the correction mechanism of RSE provision has time lags. Hence, (4.2) and (4.3) could be adjusted into a dynamic state: ST +1 = E G T (IGT ) + E S T (I ST ) − I ne f f GT (IGT ) − I ne f f ST (I ST ) DT +1 = U P T (IGT ) +

N 

U L T (I ST ) − TT (IGT ) − FeT (I ST )

(4.4)

(4.5)

i=1

In the dynamic optimal equilibrium, the following is observed: ST +1 = DT +1

(4.6)

The first-order optimal conditions of IGT and I ST satisfy the following:  ∂ IGT : E G T (IGT ) + TT (IGT ) − I ne f f GT (IGT ) = U P T (IGT )

 ∂ I ST : E ST (I ST ) + FeT (I ST ) − I ne f f ST (I ST ) =

N 

U L T (I ST )

(4.7)

(4.8)

i=1

Based on (4.7) and (4.8), on the one hand, the optimal supply level of RSE for governments in period T + 1 depends on the governmental provision capabilities, tax income, governance inefficiency and public nature of RSE. On the other hand, the optimal supply level of RSE for other social sectors in period T + 1 depends on the social sectors’ provision capabilities, consumption fee charges, governance inefficiency and the ‘long tail’ nature of RSE. Therefore, the RSE mismatch has an intertemporal dynamic correction mechanism. When ST < DT , the RSE suppliers are inclined to enhance their recognition and provision to make ST +1 → DT +1 and thereby correct the RSE mismatch. Considering the different functions and roles of expenditures on RSE provision, the total supply expenditures of governments and other social sectors (E x GT and E x ST , separately) were divided into four categories: capital expenditure Cap T , administrative expenditure Adm T , welfare expenditure W el T and scholarship expenditure Sch T . The authors believed that E x GT and E x ST were as follows: E x GT = Cap GT + Adm GT + W el GT + Sch GT

(4.9)

E x ST = Cap ST + Adm ST + W el ST + Sch ST

(4.10)

4.3 The Factors of Imbalance: Special Education

99

Here, Cap T included the construction and capital expenses for teaching infrastructure, equipment investment (such as computers), and school building repairs. These types of expenses could enhance the quality of the supply and have relatively stronger positive externalities than other expenses, which means that their crowding-out effects would not hamper the supply efficiency and levels. Specifically,   the following holds true: IGT (CapGT ) > I ST (Cap ST ) > 0. Next, Adm T includes hospitality spending, operating expenses and other official service expenses. The authors believed that a moderate proportion of Adm T is necessary, but an excessively high proportion may crowd out other supply expenditures and cause resource waste. Hence, Adm T is used as an index to reflect the provision and governance inefficiency of RSE. Then, W el T includes basic salaries, supplementary wages, welfare payments and the bonuses paid to RSE faculties. The authors feel that W el T may have doubleedged effects on the RSE correction mechanism. From the positive perspective, W el T could, on the one hand, be added into the function of provision input as the   increase in the quantity of RSE (IGT (W el GT ) > 0, I ST (W el ST ) > 0). On the other hand, the improvements in welfare and the benefits for RSE faculty members could motivate their work enthusiasm (Mot T is the function of working motivation). Hence, W el T could also affect the quality of RSE provision and be added into the   function of motivation (MotGT (W el GT ) > 0, Mot ST (W el ST ) > 0). However, from the negative perspective, W el T may also crowd out other categories of expenses, such as Cap T , and could therefore be adverse to RSE correction, especially under the distortion and misallocation of W el T . Which of the above conditions plays the more dominant role in RSE correction remains to be verified. Next, Sch T includes the grants, scholarships and subsidies provided to RSE students and their families. Compared with other categories of expenditure, the direct beneficiaries of Sch T are the RSE students themselves. Hence, Sch T could simultaneously affect both the supply and demand equation of RSE. From the supply perspective, Sch T could enhance the degree of satisfaction and help lower the financial barriers currently faced by RSE students, as one necessary part of provision quality. Hence, Sch T should be added into the function of provision input   (IGT (Sch GT ) > I ST (Sch ST ) > 0). From the demand perspective, however, the improvement in economic treatment may arouse and motivate additional unsatisfied potential RSE demand from students’ contemporaries and families (Mos T is the func tion of demand motivation for RSE children and their families; MosGT (Sch GT ) >  0, Mos ST (Sch ST ) > 0). This mechanism could widen the gap between supply and demand. Hence, the Sch T effect that plays the more dominant role still needs to be determined. In all, in the dynamic correction mechanism of RSE based on the different categories of expenditures, the supply and demand equation can be concluded as follows: ST +1 = E G T [IGT (CapGT , W el GT , Sch GT )] + E ST [I ST ((Cap ST , W el ST , Sch ST )] − I ne f f GT [IGT (Adm GT )] − I ne f f ST [I ST (Adm ST )]

100

4 The Influencing Factors of the Imbalance of Rural Long Tail …

+ MotGT [IGT (W el GT )] + Mot ST [I ST (W el ST )]

(4.11)

N      DT +1 = U P T IGT (CapGT , W el GT , Sch GT ) + U L T I ST (Cap ST , W el ST , Sch ST )







i=1

 − TT IGT (E x GT ) − FeT I ST (E x ST )     + MosGT IGT (Sch GT ) + Mos ST I ST (Sch ST )

(4.12)

When ST +1 = DT +1 , the first-order optimal condition could be changed to the following: ∂ IGT : U P T [IGT (CapGT , W el GT , Sch GT )] = E G T [IGT (CapGT , W el GT , Sch GT )]   + TT [IGT (E x GT )] − I ne f f GT [IGT (Adm GT )] + MotGT [IGT (W el GT )]  − MosGT [IGT (Sch GT )]

∂ I ST :

N 

(4.13)

U L T [I ST (Cap ST , W el ST , Sch ST )] = E ST [I ST (Cap ST , W el ST , Sch ST )]

i=1  + FeT [I ST (E x ST )] − I ne f f ST [I ST (Adm ST )]   + Mot ST [I ST (W el ST )] − Mos ST [I ST (Sch ST )]

(4.14)

Different supply subjects (governments and other social sectors) have various impacts on the correction mechanism of RSE, based on their differentiated intrinsic characters and governance efficiency. Social sectors, such as NGOs, individual organizations and private enterprises, could supply more fragmented and exclusive products, due to their advantages in terms of diversification and flexibility, which have lower cost compensation points (Dimaggio & Anheier, 1990). However, governments, which serve as the primary undertakers of public service, both morally and with respect to duty, are inclined to satisfy more homogeneous demands and place more emphasis on the overall social welfare and economic benefit of provision. Hence, governments could apply specific influences on RSE correction by particular expenditure categories. Based on the analysis above, this book raises the following theoretical hypotheses: Hypothesis 1 Cap T (especially infrastructure construction) has significant positive effects on the dynamic correction of RSE mismatch. Hypothesis 2 Adm GT has significant negative impacts on the dynamic correction of RSE. However, Adm ST is not significant, due to the minor proportions. Hypothesis 3a W el GT has significant positive effects on the dynamic correction of RSE. However, W el ST is not significant. Hypothesis 3b W el GT has significant negative effects on the dynamic correction of RSE. However, W el ST is not significant.

4.3 The Factors of Imbalance: Special Education

101

Hypothesis 4a Sch GT has significant positive effects on the dynamic correction of RSE. Hypothesis 4b Sch GT has significant negative effects on the dynamic correction of RSE. Rural public services in China have undergone significant adjustments since 2003. Chinese central governments have elevated the strategy of ‘public finance covering the rural’ (Yang & Tian, 2009). Due to data constraints and other factors, this book uses panel data from 30 provinces in China (excluding Hong Kong, Macau, Taiwan and Tibet), from 2003 to 2014. All data in the models are from the China Education Statistics Yearbook, Statistical Yearbook of China’s Educational Expenditure, Statistical Yearbook of the Cause of Disabled People in China, China Rural Statistics Yearbook, China Civil Affairs Statistics Yearbook, China Population and Employment Statistics Yearbook, China Statistical Yearbook, the official website of China Disabled Persons’ Federation, and the Report on the Marketization Index of China’s Sub-provinces (2003–2016). As previously discussed in the former theoretical analysis, expenditure categories are divided into capital, welfare, scholarship and administrative expenses. For each category of expenditures, governments are separated from other social sectors (such as NGOs, private firms and individual donations). In addition, considering the dominant role of governments in the construction of teaching infrastructure, infrastructure construction is detached from other capital expenses serving as independent variables. Due to the ‘long tail’ of RSE, especially its demand concealment, time is required for different expenditures to play their roles in recognizing and correcting the mismatch of RSE. Hence, the first-order time lag of all kinds of expenditures is set as the variables, in order to verify the dynamic correction of RSE. Finally, all the missing values are changed to zero. To date, only the dynamic correction mechanism of RSE has been independently considered. As a necessary component of the overall education system, RSE has interaction effects with other education categories, especially basic education, which serves as the ‘head’ of the curve. Notably, the division of the ‘head’ and ‘long tail’ in public demand is neither absolute nor static. With the continuous development of the economy, society and culture, rural citizens have the potential to change the level, quantity and quality of their public demand preferences. This potential causes the demands of the ‘head’ and ‘long tail’ to interact and interchange. Hence, system estimation and correct estimation bias must be used. Though the covariant of these education categories (especially the ‘head’ and ‘long tail’) have no direct connection, their disturbance term may be related. The seemingly unrelated regression (SUR) can reasonably be used to measure the interaction effects. In addition, SUR could also help to correct heteroscedasticity, panel autocorrelation, and contemporaneous correlation (HPAC; see Blackwell, 2005). Considering the ‘long tail’ nature of RSE, with the dual features of positive externalities as normal education categories and individuality as specific education categories, it is difficult to separate these two attributes and analyze their functions independently. System estimation and SUR give us an implication that can be used to address this issue. The authors thought that, to some extent, the externalities

102

4 The Influencing Factors of the Imbalance of Rural Long Tail …

of RSE, given its own mismatch (uneducated RSE students), are comparable to the externalities of other ‘head’ education categories, especially basic education, in terms of the illiteracy rate. From this perspective, the function of different expenditures could be seen from basic education (mainly primary and middle school) as the approximate ‘proxy variables’ of the positive externalities of RSE with its ‘long tail’ nature. With regard to the relativity, the externality of education (including special and basic education) could be comparable. With regard to the exclusiveness, the effects of expenditure of rural basic education (RBE) have no direct relationship with other RSE control variables. Based on the Long Tail theory, the underlying mechanism is that heterogeneous ‘long tail’ demand could aggregate and evolve into mass ‘head’ demand. The concentrated ‘head’ demand could also extend along the curve to become individualized ‘long tail’ demand, in line with the successive satisfaction and transcendence of demands. In this way, the dual attributes of RSE could be divided by different variables. Different types of RSE expenditure are used as the variables, in order to measure the ‘long tail’ nature on the disequilibrium and correction. We use different types of basic education expenditure as the approximate proxy variables. The basic panel data model is as follows in Eq. (4.15): U nedu it = α + β1 Fi S Eex pit + β2 SoS Eex pit + β3 Fi B Eex pit + β4 SoB E exp + β5 Contr olit + μi + yeart + εit (4.15) where i represents provinces, and t represents years. Also, Unedu is the proportion of uneducated RSE children, used as dependent variables (DVs). In panel SUR, Unedu is also used to measure the illiteracy rate among children in RBE. Next, FiSEexp is the proportion of different governmental expenditures for RSE; SoSEexp is the proportion of the different expenditures for RSE in other social sectors; FiBEexp is the proportion of different governmental expenditures for basic education, and SoBEexp is the proportion of the different expenditures for basic education in other social sectors. Then, Control encompasses other control variables; μ represents the individual effects of each province (no change over time); year represents time effects, and ε is the random error. The definitions and calculation methods of all variables and their statistical descriptions are listed below (Tables 4.1 and 4.2): The time trend of Unedu in different provinces of China below shows great variances in different regions. Some regions, such as Guangdong, Heilongjiang, Jiangsu and Zhejiang, have very steady trends over time. However, others regions, such as Gansu, Qinghai, Shaanxi and Tibet, have violent fluctuations. Regions were seldom able to decrease progressively and continually in Unedu across this period. Hence, a requirement exists to control the variation of each panel as individual effects (Fig. 4.1). The density distribution of each kind of expenditure by different social sectors could also be drawn, in order to check their relative size, as shown below (Fig. 4.2).

4.3 The Factors of Imbalance: Special Education

103

Table 4.1 Checklist of all variables Sign

Name

Unedu

The ratio of uneducated RSE children The registered amount of rural children with special needs not enrolled in school/the whole rural population in the 0–14 age group

Calculation

Illi

The illiteracy rate

Visual

The number of uneducated rural blind The registered amount of rural blind children with special needs children with special needs not enrolled in school * the rural percentage of population in the 0–14 age group

Hearing

The number of uneducated rural deaf children with special needs

The registered amount of rural deaf children with special needs not enrolled in school * the rural percentage of population in the 0–14 age group

Intellectual

The number of uneducated rural intellectually-challenged children with special needs

The registered amount of rural intellectually-challenged children with special needs not enrolled in school * the rural percentage of population in the 0–14 age group 4

Physical

The number of uneducated rural children with special physical needs

The registered amount of rural children with special physical needs not enrolled in school * the rural percentage of population in the 0–14 age group

Psychiatric

The number of uneducated rural children with special psychiatric needs

The registered amount of rural children with special psychiatric needs not enrolled in school * the rural percentage of population in the 0–14 age group

Multi

The number of uneducated rural children with multiple special needs

The registered amount of rural children with multiple special needs not enrolled in school * the rural percentage of population in the 0–14 age group

fiSEcapper

The capital expenditure on special education coming from governments (except infrastructure construction)

The financial expenditure of special education on equipment investment and repairs for school buildings/the total fiscal expenditure of special education

soSEcapper

The capital expenditure on special education coming from other social sectors (except infrastructure construction)

The social expenditure of special education on equipment investment and repairs for school buildings/(total expenditure of special education—total fiscal expenditure of special education)

The whole rural illiterate population/the whole rural population beyond 6 years of age

(continued)

104

4 The Influencing Factors of the Imbalance of Rural Long Tail …

Table 4.1 (continued) Sign

Name

fiBEcapper

The capital expenditure on rural basic The financial expenditure of rural basic education coming from governments education (primary and middle school; (except infrastructure construction) the same below) on equipment investment and repairs for school buildings/total fiscal expenditure of rural basic education

Calculation

soBEcapper

The capital expenditure on rural basic education coming from other social sectors (except infrastructure construction)

The social expenditure of rural basic education on equipment investment and repairs for school buildings/(total expenditure of rural basic education—total fiscal expenditure of rural basic education)

fiVEcapper

The capital expenditure on rural vocational high schools coming from governments (except infrastructure construction)

The financial expenditure of rural vocational high schools on equipment investment and repairs for school buildings/total fiscal expenditure of rural vocational high schools

soVEcapper

The capital expenditure on rural vocational high schools coming from other social sectors (except infrastructure construction)

The social expenditure of rural vocational high schools on equipment investment and repairs for school buildings/(total expenditure of rural vocational high schools—total fiscal expenditure of rural vocational high schools)

SEconper

The infrastructure construction expenditure on special education (mainly comes from governments)

Total expenditure on teaching infrastructure construction/total expenditure of special education

BEconper

The infrastructure construction expenditure on rural basic education (mainly comes from governments)

Total expenditure on teaching infrastructure construction/total expenditure of rural basic education

VEconper

The infrastructure construction expenditure of rural vocational high schools (mainly comes from governments)

Total expenditure on teaching infrastructure construction/total expenditure of rural vocational high schools

fiSEadmper

The administrative expenditure of special education that comes from governments

The financial expenditure of special education on official receptions and operation costs/total fiscal expenditures of special education

soSEadmper

The administrative expenditure of special education from other social sectors

The social expenditure of special education on official receptions and operation costs/(total expenditure of special education—total fiscal expenditure of special education) (continued)

4.3 The Factors of Imbalance: Special Education

105

Table 4.1 (continued) Sign

Name

Calculation

fiBEadmper

The administrative expenditure of rural basic education from governments

The financial expenditure of rural basic education on official receptions and operation costs/total fiscal expenditures of special education

soBEadmper

The administrative expenditure of rural basic education from other social sectors

The social expenditure of rural basic education on official receptions and operation costs/(total expenditure of rural basic education—total fiscal expenditure of rural basic education)

fiVEadmper

The administrative expenditure of rural vocational high schools that comes from governments

The financial expenditure of rural vocational high schools on official receptions and operation costs/total fiscal expenditures of rural vocational high school

soBEadmper

The administrative expenditure of rural vocational high schools that comes from other social sectors

The social expenditure of rural vocational high schools on official receptions and operation costs/(total expenditure of rural vocational high school—total fiscal expenditure of rural vocational high school)

fiSEwelper

The welfare expenditure of special education that comes from governments

The financial expenditure of special education on basic salaries, supplementary wages, welfare payments and bonuses for teachers/total fiscal expenditure of special education

soSEwelper

The welfare expenditure of special education that comes from other social sectors

The social expenditure of special education on basic salaries, supplementary wages, welfare payments and bonuses for teachers/(total expenditure of special education—total fiscal expenditure of special education)

fiBEwelper

The welfare expenditure of rural basic education that comes from governments

The financial expenditure of rural basic education on basic salaries, supplementary wages, welfare payments and bonuses for teachers/total fiscal expenditure of rural basic education

soBEwelper

The welfare expenditure of rural basic education that comes from other social sectors

The social expenditure of rural basic education on basic salaries, supplementary wages, welfare payments and bonuses for teachers / (total expenditure of rural basic education—total fiscal expenditure of rural basic education) (continued)

106

4 The Influencing Factors of the Imbalance of Rural Long Tail …

Table 4.1 (continued) Sign

Name

Calculation

fiVEwelper

The welfare expenditure of rural vocational high schools that comes from governments

The financial expenditure of rural vocational high schools on basic salaries, supplementary wages, welfare payments and bonuses for teachers/total fiscal expenditure of rural vocational high schools

soVEwelper

The welfare expenditure of rural vocational high school that comes from other social sectors

The social expenditure of rural vocational high school on basic salaries, supplementary wages, welfare payments and bonuses for teachers/(total expenditure of rural vocational high school—total fiscal expenditure of rural vocational high school)

fiSEschper

The scholarship expenditure of special education that comes from governments

The financial expenditure of special education on scholarships and donations for special needs students and their families/total fiscal expenditure of special education

soSEschper

The scholar expenditure of special education that comes from other social sectors

The social expenditure of special education on scholarships and donations for special needs students and their families / (total expenditure of special education—total fiscal expenditure of special education)

fiBEschper

The scholarship expenditure of rural basic education that comes from governments

The financial expenditure of rural basic education on scholarships and donations for normal students and their families/total fiscal expenditure of rural basic education

soBEschper

The scholarship expenditure of rural basic education that comes from other social sectors

The social expenditure of rural basic education on scholarships and donations for normal students and their families/(total expenditure of rural basic education—total fiscal expenditure of rural basic education)

fiVEschper

The scholarship expenditure of rural vocational high schools that comes from governments

The financial expenditure of rural vocational high schools on scholarships and donations for normal students and their families/total fiscal expenditure of rural vocational high school (continued)

4.3 The Factors of Imbalance: Special Education

107

Table 4.1 (continued) Sign

Name

Calculation

soVEschper

The scholarship expenditure of rural vocational high schools that comes from other social sectors

The social expenditure of rural vocational high schools on scholarships and donations for normal students and their families/(total expenditure of rural vocational high school—total fiscal expenditure of rural vocational high school)

Income

The rural per capita income

The rural per capita income / GDP deflator index (10 million RMB per capita)

Group

The rural per capita amount of social cultural organizations

The amount of rural social cultural organizations/local rural population (10,000 per capita)

Computer

The rural computer room area devoted to special education

Extract directly from raw data (100 km2 )

SEbud

The budget expenditure of special education per RSE student

The total budget expenditure of special education/(total number of rural student numbers * 1000)

Marketization

Marketization degree

Quote and measure following Fan et al. (2003)

BEbud

The budget expenditure of basic education per rural normal student

The total budget expenditure of basic education/total population of rural primary and middle school students

Rurapop

Rural population

Extract directly from raw data (1 million)

Note Due to the limitation of raw data, all types of education expenditure in 2012 are absent. The mean values of 2013 and 2011 are set as the substitution value for all types of education expenditures in 2012

The above shows that the expense proportions of infrastructure construction on RSE and rural basic education (RBE) are not high, as one may imagine. Instead, most of each of these expenses actually only accounts for a small proportion. The largest proportion in terms of mean values is that of the financial welfare expenditures on RSE. This finding implies that Chinese governments are increasingly concerned about the maintenance and motivation of RSE faculties. In contrast, other social sectors spend less on the welfare of RSE faculty members, due to their charitable and nonprofit nature. For the main models of each category of expenditure, normal fixed effects (1) are first set as the base. Then, the interaction of RSE with RBE is considered, in order to set a one-way random effects estimation of Panel SUR (2). Then, Illi is put as both the independent variables in the first equation and the DV in the second equation (this was not reported in the table). Considering the variance components of the ‘long tail’ nature of RSE, mixed-effects maximum likelihood (MEML) (3) is used to separate

108

4 The Influencing Factors of the Imbalance of Rural Long Tail …

Table 4.2 Statistical description of all variables Variable

Obs

Mean

Std. dev

Min

Max

Unedu

357

7.10702

5.215005

0.0385852

26.56842

Illi

357

1.188133

2.166472

0.0116183

18.76768

Visual

357

512.9236

691.1267

0

4558.614

Hearing

357

572.4153

753.1509

0

4204.157

Intellectual

357

998.4957

1052.541

0.3038585

6148.554

Physical

357

927.891

982.8844

0.5064309

5504.857

Psychiatric

357

228.9013

275.1201

0

1601.605

Multi

357

512.7749

533.5723

0.1012862

2744.542

fiSEadmper

357

0.1318472

0.563777

0.102482

0.4101175

soSEadmper

357

0.2024637

1.382536

0

0.9098361 0.7673993

fiSEwelper

357

0.5090991

0.1243886

0.0567159

soSEwelper

357

0.1665162

0.1592334

0

0.7850163

fiSEschper

357

0.1377826

0.0676744

0.0159894

0.3689919

soSEschper

357

0.2246669

0.1928869

0

0.8429715

fiSEcapper

357

0.152024

0.1136325

0.0085221

0.8508184

soSEcapper

357

0.4061389

0.2151026

0

1

SEconper

357

0.0568255

0.0984161

0

0.6013585

fiBEadmpe

357

0.1132947

0.0592606

0.006082

0.2917688

soBEadmpe

357

0.1931151

0.1179527

0.017348

0.5848703

fiBEwelper

357

0.6531947

0.1520218

0.3820042

0.956934

soBEwelper

357

0.213845

0.164165

0.119222

0.8009072

ifBEschper

357

0.1063493

0.843946

0.0000249

0.3314167

soBEschper

357

0.2510808

0.2471642

0.000028

0.8578329

ifBEcapper

357

0.1005923

0.0520529

0.170487

0.291999

soBEcapper

357

0.2869179

0.1420838

0.0468524

0.8607732

BEconper

357

0.0245548

0.0257385

0

0.2307642

fiVEadmpex

248

0.1195169

0.0855835

0

0.6554511

soVEadmper

248

0.2705165

0.1880046

0

1

fiVEwelper

248

0.4187416

0.2001796

0

0.9911215

soVEwelper

248

0.184076

0.1779395

0

0.7986076

the variation of the intercept and coefficients between RSE and RBE (Marchenko, 2005). First, the results of capital expense and infrastructure construction are as follows (Table 4.3): From the regression results, the coefficients of infrastructure construction (SEconper and Beconper) were found to be significantly negative. This supports Hypothesis 1, which holds that the infrastructure investment on RSE has large

4.3 The Factors of Imbalance: Special Education

109

Fig. 4.1 Time trend of the proportion of uneducated RSE children (Unedu) in different provinces of China (2003–2014) (unit: proportion). Note Data sources are China Rural Statistics Yearbook and the official website of China Disabled Persons’ Federation. Drawn by STATA 15.0

spillover effects on the correction of RSE supply insufficiency. From the hypothesis above, both the public good and the ‘long tail’ nature of RSE infrastructure investment could help satisfy the heterogeneous RSE demand, with high effectiveness, in the system equation. In addition, most coefficients of other capital expenses (fiSEcapper, soSEcapper, fiBEcapper and soBEcapper) are not stable and significant. In other words, compared with infrastructure construction, other capital expenses, such as the renovation of school buildings and the purchase of other teaching facilities, have smaller marginal effects and may even crowd out other expenditures. The underlying reasons behind these findings may be due to the failure of signaling for RSE demands, which in turn makes trial and error with repetitive investment necessary for correction to occur. Second, the results of administrative expenses on the correction are as follows (Table 4.4): The coefficients of governmental administrative expenses for RSE (fiSEadmper) were found to be significantly positive in most models. This finding provides support for Hypothesis 2, such that Adm GT has significantly negative impacts on the dynamic correction of RSE. With the ‘long tail’ nature of RSE, the proportion of administrative expenses reflects the degree of governance efficiency of RSE. Specifically, the higher the proportion spent on official receptions and operation costs is, the lower the level of governance efficiency will be. Namely, for a particular education category, it is

110

4 The Influencing Factors of the Imbalance of Rural Long Tail …

Fig. 4.2 Density distribution of each type of expenditure by different social sectors. Note Data sources are China Education Statistics Yearbook, and the Statistical Yearbook of China’s Educational Expenditure. Drawn by STATA 15.0

necessary to maintain the smallest proportion of administrative expenses, which will serve as the lubricant of RSE implementation. However, what the standard is and how to maintain that standard remains academically debatable. Some clues can be noted from the coefficients of governmental administrative expenses for RBE (fiBEadmper). Representing the nature of public service in RSE, the financial administrative expense of RBE has positive effects on the correction of illiteracy. As a kind of pure and basic public good with overwhelming universality and homogeneity, the necessary administrative expense also has the potential to act positively on the provision. Hence, the divergent natures of RSE are double-edged and should be discussed separately, rather than being examined from only one side. In addition, the coefficients of other social sectors’ administrative expenses for RSE and RBE (soSEadmper and soBEadmper) are not stable, although some of them are significantly negative, especially that of soSEadmper. Compared with governments, other social sectors have more flexible organizational and institutional structures, which could therefore accrue lower standards of administrative expenses in provision. However, these effects may be overshadowed by the inactive and limited participation of those private organizations, due to policy constraints. Third, the results of welfare expenses on the correction are as follows (Table 4.5): The coefficients of governmental welfare expenses for RSE (fiSEwelper) were found to be significantly positive. This finding supports Hypothesis 3b, which holds

4.3 The Factors of Imbalance: Special Education Table 4.3 Results of capital expense on RSE correction

111 (1)

(2)

(3)

Fe

Panel SUR

PCSE

DV

Unedu

Unedu/Illi

Unedu

fiSEcapper

−2.187

−0.543

0.066

(−0.888)

(−0.867)

(0.038)

soSEcapper

0.132

0.019

−0.147

(0.129)

(0.056)

(−0.162)

SEconper

−6.573***

−3.535***

−4.111**

(−3.411)

(−4.774)

(−2.259)

fiBEcapper

1.145

16.255***

11.037***

(0.173)

(9.002)

(2.613)

soBEcapper

2.519

5.589***

2.518

(1.139)

(8.377)

(1.620)

BEconper

−43.028**

−13.311***

−32.183***

(−2.146)

(−4.531)

(−3.375)

Year

No

Yes

Yes

Fixed effect

Yes

No

Yes

N

326

326

326

statistics in parentheses; *

Note t p < 0.05, *** p < 0.01. The other control variables were not shown in the table. The same applies below Table 4.4 Results of administrative expense on RSE correction

p < 0.1, **

(1)

(2)

(3)

Fe

Panel SUR

PCSE

DV

Unedu

Unedu/Illi

Unedu

fiSEadmper

16.668*

7.162***

8.215

(1.909)

(4.934)

(1.562)

soSEadmper

−0.767

0.096

−1.098

(−0.560)

(0.254)

(−0.744)

fiBEadmper

−5.210

1.990

−10.691**

(−0.859)

(0.920)

(−2.194)

soBEadmper

4.490

2.993***

-0.464

(1.369)

(4.723)

(−0.202)

Year

No

Yes

Yes

Fixed effect

Yes

No

Yes

N

326

326

326

112

4 The Influencing Factors of the Imbalance of Rural Long Tail …

Table 4.5 Results of welfare expense on RSE correction

(1)

(2)

(3)

Fe

Panel SUR

PCSE

DV

Unedu

Unedu/Illi

Unedu

fiSEwelper

5.836***

3.984***

2.447

(3.279)

(6.162)

(1.023)

soSEwelper

2.255

1.255*

−0.263

(0.710)

(1.931)

(−0.120)

fiBEwelper

2.410

−3.947***

−11.251***

(0.694)

(−3.379)

(−3.507)

−0.395

−2.142***

−2.545

(−0.174)

(−3.127)

(−1.227)

Year

No

Yes

Yes

Fixed effect

Yes

No

Yes

N

326

326

326

soBEwelper

that W el GT has negative effects on the correction of the RSE mismatch. With the largest proportion of total financial RSE expenditures, the motivation and maintenance of governmental welfare expenses for RSE faculty members are blurred and elusive. The underlying reasons for this finding may be that the ‘long tail’ nature of RSE causes the information pertaining to RSE teaching faculty members and other resources to be distorted, dissolved and misallocated. Despite the development of RSE infrastructure investment in these years, RSE faculty members still face weak professional accomplishment and planning, as shown in previous studies. However, the opposite is (approximately) true for the public good characteristic of RSE. The coefficients of governmental welfare expenses for RBE (fiBEwelper) at least support the idea that the publicity of RSE still exists and drives welfare effects. This occurs even though the coefficients are not significantly negative in all models (even the positive coefficients are smaller in absolute value than that of fiSEwelper). The coefficients of other social sectors’ welfare expenses for RSE and RBE (soSEwelper and soBEwelper) are again not significant in most models. This finding may be related to the sources of social expenditure. As additional social expenditures on RSE are from private donations or corporate profits, they have no legal and moral duties in RSE provision. Based largely in charity and altruism, those welfare expenditures have private attributes and are targeted to specific regions, groups and individuals. Hence, welfare expenses from other social sectors have minor effects on the mismatch. Finally, the results of scholarship expenses on the correction are obtained, as shown below (Table 4.6). The results show that the coefficients of both governmental scholarship expenses for RSE and RBE (fiSEschper and fiBEschper) are not significant in most models. This result supported neither Hypothesis 4a nor Hypothesis 4b and obscured the

4.3 The Factors of Imbalance: Special Education Table 4.6 Results of scholarship expense on RSE correction

113 (1)

(2)

(3)

Fe

Panel SUR

PCSE

DV

Unedu

Unedu/Illi

Unedu

fiSEschper

5.063

−0.234

3.381

(0.641)

(−0.173)

(0.768)

soSEschper

−1.376

−1.039**

−2.074

(−0.671)

(−2.246)

(−1.249)

fiBEschper

0.769

7.371***

1.356

(0.137)

(3.863)

(0.321)

−2.290

−1.358**

0.196

(−1.616)

(−2.231)

(0.147)

Year

No

Yes

Yes

Fixed effect

Yes

No

Yes

N

326

326

326

soBEschper

double-edged sword. However, the absolute value of fiBEschper showed overwhelmingly smaller coefficients than those of fiSEschper, thus reminding us to see this issue from the dual ‘long tail’ nature of RSE. At least the nature of the public good in the form of basic education is rooted in sufficient rationality, specifically to cover more of the financial burdens for uneducated RSE children. However, the ‘long tail’ scattering of RSE distorts distribution and creates information asymmetry. Especially in poor rural areas, the high opportunity cost and unclear value of education has led to many families being unaware of the need to receive and support education. Many such families are unwilling to receive aid for their special needs children, even when the offer of aid comes from rich donors. This dampening mechanism could be given more support from the coefficients of other social sectors’ scholarship expenses for RSE and RBE (soSEschper and soBEschper). The ‘massive amateur producer’ (Anderson, 2007) represented by numerous and various NGOs could directly interact with citizens’ real demands in the form of grassroots and the ‘long tail aggregator’. The NGOs could collect and filter information through platform sharing and have access to exclusive and accurate donors for RSE students. Hence, the coefficients of soSEschper and soBEschper are significantly negative in most models. These results are consistent with those of Anderson (2007), indicating that the aggregation of the ‘long tail’ acts as the source of democratic distribution. This aggregation effect is the most significant in scholarship expenses, because this category of demand can potentially be the best at exerting the relevant information and cost advantages, as the aggregation effect essentially faces the exact RSE demanders. The sensitivity and mechanism analysis will be divided into four parts. First, due to the variety and complexity of different disability types, public expenditures may have differentiated influences on their RSE mismatch and correction. Hence, we separate RSE children into those with visual disability, hearing disability, speech disability,

114

4 The Influencing Factors of the Imbalance of Rural Long Tail …

physical disability, intellectual disability, psychiatric disability and multiple disabilities. Furthermore, the interaction effects of RSE with other categories are not only between special and basic education. Professional education, as an alternative path, also has mediation influences as the ‘head’ on the ‘long tail’ of RSE. Thus, the system estimation of RSE and professional education is set. Third, due to the change in the policy implementation of RSE in 2009, the policy is viewed here as a quasi-random experiment to perform difference-in-differences (DID). A synthetic control method (SCM) of public expenditures on RSE is also conducted, as a robustness check. Finally, the time dynamic effect of the correction mechanism is tested by changing independent variables into current periods and second-order time lags. (1)

Different expenditures on the subtypes of RSE correction

Due to the ‘long tail’ nature of RSE, its demand is fragmented and individualized, with various symptoms and disabilities. Hence, a need exists to analyze RSE’s correction mechanism in specific subtypes. Based on the official website of the China Disabled Persons’ Federation, RSE children are divided into categories characterized by six kinds of disabilities: visual, hearing, physical, intellectual, psychiatric and multiple disabilities. From the distribution of the proportions of different RSE subtypes, as shown below, the authors found that the proportions of RSE do have differentiated degrees of prevalence. Intellectual disabilities account for the largest proportion of the whole population of students in RSE. This finding proves that these disabilities are widespread. However, visual and psychiatric disabilities are relatively minor in all subtypes (Fig. 4.3). The MEML method is used to measure the correction of different expenditures on the following subtypes of RSE. The total number of uneducated RSE children for each subtype is chosen as the DV, for a robustness check. The results of capital expenditures on the various subtypes of RSE correction are shown below (Table 4.7). The authors found that, on the one hand, for infrastructure investment and construction, the ‘long tail’ nature of different subtypes of RSE on the correction is still robust (the negative significance of SEconper). In addition, the ‘head’ nature of RBE (BEconper) is also negative on the correction, although some of the RBE are not significant. Regarding different subtypes of RSE, the absolute values of the coefficients of psychiatric and multiple disabilities were found to be smaller than those of other subtypes. Compared to other disabilities, psychiatric children have intrinsic difficulties in their education. Some of these individuals face much higher education costs and are enrolled in full-time daycare. Hence, for these children, there is a lack of service availability and motivation, even with the promotion of capital investment. On the other hand, multiple disabilities can be seen to represent a comprehensive category that has mediation effects on other subtypes of disabilities. Accordingly, the coefficients for children with multiple disabilities are in the range of the coefficients of other children with special needs with regard to absolute value. In addition, the coefficients of visual and hearing are most significantly negative in all models (fiSEcapper and soBEcapper). Most visually- and hearing-impaired children with special needs have the capacity to study and can be educated, if they are able to

4.3 The Factors of Imbalance: Special Education

115

Fig. 4.3 Proportion distribution of different sub-types of RSE. Note Data source is the China Disabled Persons’ Federation official website. Drawn by STATA 15.0 Table 4.7 Results of capital expenditures on subtypes of RSE correction (1)

(2)

(3)

(4)

(5)

Visual

Hearing

Physical

Intellectual

Psychiatric

Multiple

−360.859*

−407.691**

−198.023

−212.250

−100.116

−89.171

(−1.918)

(−2.105)

(−0.855)

(−0.786)

(−1.202)

(−0.619)

soSEcapper −78.173

−51.721

51.952

197.948

−26.870

−55.894

(−0.728)

(−0.470)

(0.390)

(1.288)

(−0.578)

(−0.679)

fiSEcapper

SEconper fiBEcapper

−820.271*** −903.147*** −596.098*** −873.989*** −312.116*** −418.977*** (−4.371)

(−4.668)

(−2.585)

(−3.242)

(−3.735)

(−2.919)

353.128

−511.589

603.822

175.714

130.264

264.414

(0.708)

(−1.004)

(0.972)

(0.247)

(0.608)

(0.691)

786.528***

501.455*

767.286**

105.889

131.938

(1.940)

(3.683)

(1.913)

(2.563)

(1.195)

(0.819)

−2.1e + 03**

−1.4e + 03

−2.3e + 03*

−550.462

−821.156**

−1.2e + 03

(−2.150)

(−1.420)

(−1.919)

(−0.396)

(−1.962)

(−1.560)

326

326

326

326

326

326

soBEcapper 406.943* BEconper

N

(6)

Note The other control variables were not shown in the table. The same applies below

116

4 The Influencing Factors of the Imbalance of Rural Long Tail …

Table 4.8 Administrative expenditures on subtypes of RSE correction (1)

(2)

(3)

(4)

(5)

(6)

Visual

Hearing

Physical

Intellectual

Psychiatric

Multiple

1045.185**

981.101**

719.817

144.129

439.513**

448.898

(2.242)

(2.150)

(1.292)

(0.234)

(2.205)

(1.298)

soSEadmper

79.822

35.320

−76.520

−204.264

48.315

−1.012

(0.607)

(0.275)

(−0.493)

(−1.189)

(0.847)

(−0.010)

fiBEadmper

−1.4e + 03**

−3.1e + 03***

−2.4e + 03***

−3.9e + 03***

−763.666***

−1.3e + 03***

(−2.576)

(−5.735)

(−3.675)

(−5.437)

(−3.331)

(−3.256)

soBEadmper

315.218

717.855***

64.879

744.778**

−10.462

−94.645

(1.375)

(3.187)

(0.234)

(2.439)

(−0.109)

(−0.554)

326

326

326

326

326

326

fiSEadmper

N

receive suitable auxiliary teaching equipment that helps them overcome their physical inconveniences. In China, most of these children are being educated independently in special schools for individuals who are deaf, blind and mute. However, in rural areas, the auxiliary teaching equipment needs can hardly be fully satisfied. This is the underlying reason why capital expenses have larger marginal effects on visually- and hearing-impaired disabled RSE children. We discuss the results of administrative expenditures on subtypes of RSE correction below (Table 4.8). Consistent with the main models, the ‘long tail’ nature of different RSE was still found to cause administrative expenses to play separate roles in both RSE and RBE (the positive significance of fiSEadmper and the negative significance of fiBEadmper). A moderate range of administrative expenses is necessary, especially considering the attribute of RSE as a public service (such as RBE). However, the excessive burden of administrative expenses causes the scattering distribution of the ‘long tail’, making the demands difficult to satisfy and hampering the recognition of those demands. Regarding the different subtypes of RSE, the coefficients of the psychiatric and multiple subtypes are still smaller in absolute value. In addition, the insignificance of the coefficients of physical disabilities is noteworthy. Except for movement inconvenience, RSE children with physical disabilities could still compete equally with their normal RBE counterparts. In China, a considerable proportion of physically-disabled children with special needs (such as crippled children) are being educated in the pattern of ‘follow-up in class’ with other normal students. Under these circumstances, they may be the least-affected RSE group among all subtypes. The results of welfare expenditures on subtypes of RSE correction are discussed below (Table 4.9). In contrast to the main model, a positive significance of fiBEwelper was found in some models. This finding implies that, in line with the subdivision of RSE categories, the public good nature of the positive motivation effects for RSE faculty members will

4.3 The Factors of Imbalance: Special Education

117

Table 4.9 Results of welfare expenditures on subtypes of RSE correction (1)

(2)

(3)

(4)

(5)

Visual

Hearing

Physical

Intellectual

Psychiatric Multiple

fiSEwelper

927.478***

874.023***

631.579***

650.380***

305.490***

(5.528)

(5.187)

(3.099)

(2.802)

(4.105)

(3.444)

soSEwelper

47.475

−22.523

−68.436

−169.921

−24.646

−99.522

(0.250)

(−0.118)

(−0.296)

(−0.646)

(−0.295)

(−0.692)

fiBEwelper

185.212

1037.818***

640.588**

1490.802***

63.150

94.182

(0.841)

(4.689)

(2.377)

(4.873)

(0.660)

(0.565)

soBEwelper 63.663

48.273

112.543

−11.502

72.742

272.585**

(0.353)

(0.267)

(0.513)

(−0.046)

(0.918)

(1.998)

326

326

326

326

326

326

N

(6) 437.080***

fade and shift to its ‘long tail’. Especially when dealing with children with psychiatric and multiple special needs, the mentality and chronicity of their disabilities, along with the mediation effects, can easily cause RSE faculty members to deem their efforts to be fruitless and discouraging. Finally, we turn to the results of scholarship expenditures on the various subtypes of RSE correction, as shown below (Table 4.10). Most coefficients of scholarship expense were still found to be not significant. However, the coefficients of soBEschper become significantly negative. This finding supports the idea that other social sectors (such as NGOs) can customize and lock onto specific recipients with the largest marginal benefit of scholarships, due to these sectors’ information advantages. Although the whole proportion of soBEschper is not significant in the main models, with the extension of the ‘long tail’, heterogeneous Table 4.10 Scholarship expenditures on subtypes of RSE correction

fiSEschper

(1)

(2)

(3)

(4)

(5)

Visual

Hearing

Physical

Intellectual

Psychiatric Multiple

273.721

189.506

61.280

429.870

152.412

−186.051

(0.558)

(0.385)

(0.100)

(0.634)

(0.776)

(−0.502)

soSEschper −9.251 fiBEschper

−22.372

4.232

−230.510

52.291

61.631

(−0.055)

(−0.133)

(0.021)

(−1.007)

(0.768)

(0.491)

−313.008

−982.078**

−1.0e + 03**

−2.2e + 03***

−74.456

156.515

(−0.733)

(−2.292)

(−1.976)

(−3.895)

(−0.408)

soBEschper −283.257** −559.192*** −325.121** −517.422*** −88.747* N

(6)

(0.495) −216.916**

(−2.419)

(−4.756)

(−2.346)

(−3.322)

(−1.751)

(−2.507)

326

326

326

326

326

326

118

4 The Influencing Factors of the Imbalance of Rural Long Tail …

demands will be more apt to place other social actors in the role of the ‘long tail aggregator’. In summary, different subtypes of RSE exert divergent effects on the correction mechanism. With regard to visual and hearing disabilities, the best solution is to highlight the investment in auxiliary teaching facility members, in order to help promote the availability of RSE to children with such disabilities. With regard to physical disabilities, the existing pattern of ‘follow-up in class’ could help these children enjoy equal opportunities to access RBE, without any artificial entrance barriers. With respect to individuals with intellectual and psychiatric disabilities, due to the relatively higher costs of entrance to education, they are less affected by expenses from governments and other social sectors. Multiple disabilities could be seen as a comprehensive category and as a method of mediation for other subtypes of disabilities. Thus, their marginal effects are also moderate. In addition, due to the ‘long tail’ nature of specific subtypes of RSE, other social sectors could provide more assistance in locating, matching and sponsoring the accuracy of unsatisfied RSE demands and act as the correction mechanism in those subfields. (2)

Interaction and mediation effects of RSE with rural vocational education (RVE)

To date, the authors have only considered the interaction and mediation effects of RSE and RBE, which act similarly to the complementation between the ‘head’ and the ‘long tail’ in the power-law distribution. Due to the intrinsic physical and/or mental disadvantages of RSE children, it is sometimes more important for them to master vocational skills. That type of education could better prepare them to earn a livelihood than they could by acquiring comprehensive knowledge. Hence, it is necessary to consider the role of rural vocational education (RVE), which acts as a substitution for (or a competing option against) RSE, especially the good public nature of RSE. However, RBE is different, because RVE is also at the ‘long tail’ end of the power-law distribution of all education categories. Therefore, the interaction and mediation effects of RSE and RVE will be less complementary but more competitive. To prove this statement, the different proportions of different expenditures on RVE (HVCs) are set as the alternative ‘proxy variable’; this is done instead of testing RBE. The fixed-effects, FGLS, PCSE and MEML models are still set. In addition, all of the missing values have been changed to zero. Due to the hard convergence of panel SUR between RSE and RVE variables, the panel SUR model is ignored. The results of capital expenses on the correction are presented below (Table 4.11). The coefficients of SEconper were still found to be significantly negative, which supports the idea that the expense of infrastructure construction has the largest marginal effect on the correction of RSE. However, this time, the coefficients of RVE (fiVEcapper, soVEcapper, VEconper) were no longer significant. This finding is consistent with the interaction between RSE and RVE: both are on the ‘long tail’ end of the curve. Hence, this finding proves they have a more competitive than complementary relationship, which means that both show a homogeneous ‘long tail’ nature. Then, the administrative expenditures on RSE correction with RVE are examined, as shown below (Table 4.12).

4.3 The Factors of Imbalance: Special Education Table 4.11 Results of capital expenditure on RSE correction with RVE

119 (1)

(2)

Fe

PCSE

DV

Unedu

Unedu

fiSEcapper

−3.075

0.142

(−1.110)

(0.076)

soSEcapper

0.200

0.437

(0.162)

(0.484)

SEconper

−8.077***

−4.572**

(−3.563)

(−2.466)

−0.984

0.570

fiVEcapper

(−0.337)

(0.340)

soVEcapper

1.067

0.795

(1.046)

(0.857)

VEconper

−1.962

5.540**

(−0.508)

(2.128)

Year

No

Yes

Fixed effect

Yes

Yes

N

236

236

Note Due to data limitations, only the RVE data from 2007 to 2014 could be obtained. The same applies below Table 4.12 Administrative expenditure on RSE correction with RVE

(1)

(2)

Fe

PCSE

DV

Unedu

Unedu

fiSEadmper

16.371*

8.713*

(1.963)

(1.671)

soSEadmper

−1.630

−2.617*

(−1.062)

(−1.933)

fiVEadmper

−6.654

−6.576**

(−1.148)

(−2.245)

soVEadmper

1.110

0.151

(0.827)

(0.105)

Year

No

Yes

Fixed effect

Yes

Yes

_cons

11.504***

15.933***

(3.109)

(8.463)

N

236

236

120

4 The Influencing Factors of the Imbalance of Rural Long Tail …

Table 4.13 Results of welfare expense on RSE correction with RVE

(1)

(2)

Fe

PCSE

DV

Unedu

Unedu

fiSEwelper

6.159***

−1.188

(3.507)

(−0.494)

soSEwelper

2.570

−3.154

(1.012)

(−1.468)

fiVEwelper

2.565

−1.156

(1.100)

(−0.663)

−3.231*

−1.112

(−1.773)

(−0.664)

Year

No

Yes

Fixed effect

Yes

Yes

_cons

8.386*

17.409***

(2.013)

(7.159)

236

236

soVEwelper

N

Only the coefficients of fiVEadmper are found to be significantly negative in these models. This finding indicates that a moderate investment in administrative expenses will not harm the positive spillover effects of RSE; instead, the investment helps to correct the mismatch, as other education categories participate. In addition, the coefficients of fiSEadmper are still significantly positive, which is consistent with the main models above. Next, the results of welfare expenses on RSE correction with RVE are analyzed (Table 4.13). The coefficients of fiSEwelper were still found to be significantly positive. This finding implies that, even when considering substituting RVE for RBE, the motivation effects of welfare expenses on RSE faculty members were still distorted and misleading. As Koutrouba et al. (2006) argued, the considerable hesitation and prejudice of RSE families and societies is still widespread, even to the extent of affecting the development and inclusion of RVE. The natural feelings of inferiority and judgment related to vocational education in China causes RSE and RVE faculty members to prefer to stay in comprehensive colleges (Feng, 2012; Smit et al., 2014). A similar situation can also be seen in the results of scholarship expenses on RSE correction with RVE, as shown below (Table 4.14). This time, the coefficients of neither fiSEschper nor fiVEschper are significant. The weak efficacy of welfare and scholarship on either faculty members or students proves that the mismatch of RSE is not the direct result of insufficient investment. There are more complex, underlying reasons related to keeping all stakeholders financially and spiritually motivated. In all, the interaction of RSE with RVE is not as functional as that with RBE, due to the sharing ‘long tail’ nature of both RSE and RVE. However, the main

4.3 The Factors of Imbalance: Special Education Table 4.14 Results of scholarship expense on RSE correction with RVE

121 (1)

(2)

Fe

PCSE

DV

Unedu

Unedu

fiSEschper

8.238

2.224

(1.227)

(0.434)

soSEschper

−0.757

−1.465

(−0.303)

(−0.784)

fiVEschper

0.221

−0.702

(0.114)

(-0.423)

−2.369

−1.630

(−0.731)

(−0.757)

Year

No

Yes

Fixed effect

Yes

Yes

N

236

236

soVEschper

regression results still support the hypothesis that capital expenses (mainly infrastructure construction) have the largest marginal effects on the correction. In addition, administrative expenses have dual effects on the dual nature of RSE. Given RSE’s ‘long tail’ nature, the governance efficiency of RSE provision could be dampened. For the nature of public goods, a moderate proportion of administrative expenses is necessary for the normal operation of the correction mechanism. The other two categories (welfare and scholarship) of expenses have obscure and blurry effects on the correction, due to the distortion and misallocation of those expenses. Questions remain regarding whether the financing of and expenditure on school construction have differential impacts on various special needs children (Dhuey & Lipscomb, 2013). Due to the different special education needs, which are in turn due to physical and intellectual disabilities, and hearing and visual impairment, children’s satisfaction with rural special needs education is not high (Banks et al., 2015; Franck & Joshi, 2017). Therefore, RSE schools should be tailored for special needs children living in areas that are remote from educational facilities, public transportation, childcare facilities and consistent, highly-qualified healthcare service personnel (Meyers et al., 2015). Cullen (2003) found that the special education of children with non-physical and physical disabilities faces issues of divergent fiscal incentives and identification by local authorities. However, little consensus exists on the most suitable way to evaluate the existing construction of RSE schools in terms of supporting divergent needs and ensuring the customized distribution of RSE services. The authors also examine an important aspect of school construction for RSE, one that has been neglected in the policy debate. Specifically, what are the differential impacts of school construction and investment on different special needs children? On the one hand, the ability to train and educate visually, hearing- and limb-impaired children with special needs may be highly dependent on investment in RSE schools. On the other hand, challenges related to the long-term daily healthcare of mentally

122

4 The Influencing Factors of the Imbalance of Rural Long Tail …

handicapped and multiple-special needs children may dampen the effects of school construction on the children’s skills development and knowledge acquisition (Tran, 2014). While there is ample room for academic discussion regarding the advantages and disadvantages of financing school construction, such contradictions reflect the controversy associated with many related policies. In addition, most studies have used quantitative methods, such as ordinary least squares (OLS) and instrumental variable approaches, to evaluate the policy effects (Kim et al., 2018). These methods cannot eliminate endogeneity issues and the reciprocal causation between school construction and RSE dropout rates (Tompkins, 2006). The financial capital used to support the building of rural special education schools in 2009 is used in a quasi-natural experiment. The aim is to investigate the differential impact of educational infrastructure on the dropout rate of children with special needs and its mechanism. This book compares changes in the dropout rates of various special needs children across different regions as a means to measure the effects of school construction. The difference-in-differences (DID) method is used to estimate changes in the dropout rates of various special needs children across different regions, after the policy implementation in 2009. This strategy allows us to control for alternative regional factors that may affect local RSE dropout rates, such as total education budget (Brownell et al., 2005) or education provision by nonprofit organizations, which serves as a substitute (Tilak, 2002). This book reveals substantial decreases in RSE dropout rates in those regions that received investment in school construction, compared to other regions, after the policy implementation in 2009. Investment in school construction produced an average of 500–600 additional special needs children with physical impairments being served after the policy implementation, compared with an average of 180–280 mentally handicapped and multiple special needs children. One caveat with regard to these results is that the investment in RSE schools in China may not have been exogenous, even though the policy’s timing and the types of technologies and economic factors available for construction were exogenous. To address these potential problems, the data were subjected to a series of robustness checks. Endogeneity tests account for unobservable characteristics that may have encouraged policy implementation in areas with relatively poor RSE budgets and a lesser degree of civil participation. A low degree of correlation with the RSE budget and social groups suggests that the underlying policy implementation is less likely to be a bias source. During the development of special education in the past few decades, China has adopted the following two methods of educational arrangements for children with special needs: the first approach is to study in regular classes. This is the simplest method, but this method can easily cause children with special needs to become isolated, which has an impact on their psychology. The second method is to set up special education schools. This is a better method than having special needs children study in regular classes. As to the actual situation in China, special education schools, whether funded by the government or donated by society, have become the main

4.3 The Factors of Imbalance: Special Education

123

source of education for children with special needs, especially in large and mediumsized cities. This is happening because the number of children with special needs has reached a certain scale, and society is now paying more attention to children with special needs. However, due to the scattered distribution of children with special needs in rural areas, the government’s financial capacity and social attention in these areas is not as high as in the cities. This is especially the case in the central and western regions, where traffic is not convenient. Most children with special needs in these regions can only study in regular classes, or they may choose not to receive a compulsory education. As a result, the education of children with special needs presents obvious urban–rural and regional differences. In 2009, the Chinese government issued a document entitled “Circular on Views on Further Speeding Up the Development of Special Education”. In this document, local governments in the central and western regions of China made specific provisions to support the establishment of special education schools. The document held that a special education school should be built independently in counties with a population of more than 0.3 million, or in those counties with a relatively large number of children with special needs but no special education school. In addition, the document stated that one or more special education schools should be built, as a whole, within the scope of cities in counties with a population of less than 0.3 million. At the same time, the document also required that the enrollment rate of three types (visual, hearing and intellectual disabilities) of children with special needs should be increased yearby-year in the central and western rural areas that have already been “nine-year compulsory education” areas. Next, compulsory education for children with special needs should be taken as an important part of China’s universal nine-year compulsory education; the enrollment rate of the three types of children with special needs should also reach about 70% in the central and western rural areas without the nine-year compulsory education. Local governments in the central and western rural areas are now required to create conditions to ensure that school-age children with special needs receive compulsory education. Obviously, the Chinese government hopes to increase infrastructure investment in special education schools in the central and western regions, so as to increase the enrollment rate of children with special needs in those regions. However, is this government really achieving the expected goal? In other words, does investment in education infrastructure reduce the dropout rate of children with special needs? This study examines this policy background and policy objectives. Rural public services in China have undergone considerable adjustments since 2003. Chinese central governments have adopted the strategy of “public finance covering the rural” (Yang & Tian, 2009). Due to data constraints and other factors, this book uses panel data from 30 provinces (excluding Hong Kong, Macau, Taiwan and Tibet) in China, covering the 2003–2014 period. All data in the models are from the China Education Statistical Yearbook, Statistical Yearbook of China’s Educational Expenditure, Statistical Yearbook of the Cause of Disabled People in China, China Rural Statistical Yearbook, China Civil Affairs Statistical Yearbook, China Population and Employment Statistical Yearbook, China Statistical Yearbook, the official website of the China Disabled Persons’ Federation and the Report on the

124

4 The Influencing Factors of the Imbalance of Rural Long Tail …

Marketization Index of China’s Sub-provinces, 2016. All the missing values were changed to zero. Following Imbens and Wooldridge (2009) this policy implementation is taken as a quasi-natural experiment, and the DID method is used to check the implementation effects in this study’s main model. According to the document “Circular on Views on Further Speeding Up the Development of Special Education”, the local governments in central and western regions are required to invest in special education schools, in order to increase the enrollment rate of children with special needs. Here, the western and central regions are taken as the treatment group, and the eastern regions are taken as the control group to build the DID model. By controlling other factors, this method can be used to compare the differences between the treatment and control groups, before and after policy implementation, to estimate the policy implementation effect (average treatment effect (ATE)). The variable Treat is used to distinguish between the treatment and control groups: Treat = 1 for the treatment group (the western and central regions); Treat = 0 for the control group (eastern regions). The variable Period is used to reflect the policy implementation (Treat = 1 for during and after policy implementation; Treat = 0 for before policy implementation). To evaluate the policy, the intersection of Treat * Period is set (Treat * Period = 1 when Treat = 1 and Period = 1). In the main models, Unedu is used as the outcome variable, and Edu (the proportion of educated RSE students) is used as the robust variable, in order to measure the local RSE dropout rate. The year 2009 is set as the time node of policy implementation. Hence, the DID empirical model is as follows: U nedu = β0 + β1 T r eat + β2 Period + β3 T r eat ∗ Period + Contr ol + μ where β1 controls the difference between the treatment and control groups, β2 controls the common shock of time for the treatment and control groups, β3 is the coefficient that reflects the policy effects, Contr ol represents other control variables, and μ is the random error. After these treatments, the general factors that affect all local RSE dropout rates can be eliminated, such as geographic and demographic characteristics, the macro-environment, and economic growth. The key aim of the DID method is to determine whether there is a significant difference between groups in a region, before policy implementation. In a DID analysis, only the parallel trend can reveal the causal effects of a policy. Hence, the mean value of Unedu for the treatment and control groups is obtained first, in order to test the parallel trend, as shown below (Fig. 4.4). The above figure shows that the treatment and control groups had similar trends in Unedu before 2009. However, after 2009, the treatment group (western and central provinces) sharply decreased in Unedu numbers and again achieved a steady trend in parallel with the control group (eastern provinces). Furthermore, the interaction coefficients of the years and treatment groups were set, in order to test the parallel trend. Below, Fig. 4.2 shows the dynamic effects of the 2009 policy on Unedu (Fig. 4.5). The coefficients of the interaction fluctuated at around 5 before 2009. However, after 2009, the coefficients decreased sharply to 0 and were steady. This parallel trend indicates that the treatment and control groups are comparable. The effects of the

4.3 The Factors of Imbalance: Special Education

125

Fig. 4.4 Variation Trend of the RSE Dropout Rates. Note The data sources are the China Rural Statistical Yearbook and the official website of the China Disabled Persons’ Federation. The y-axis is the coefficient value of Unedu. The x-axis is the year. Drawn by STATA 15.0. The same applies below

2009 policy could also last for an extended period. Hence, these premises support the performance of the DID analysis. The explanatory variable is the RSE dropout rate (unedu). The calculation method is the number of rural children with special needs who are registered but not attending school, divided by the total number of rural children with special needs, aged 0–14. Considering that RSE in China is more in the stage of basic education (primary and junior high school), the age of special needs children is set at from 0–14 years old. The assumption is made that all children with special needs have the potential, ability and right to receive compulsory education. Therefore, those rural children with special needs who have not registered for school are considered to be drop out children, and their education needs have not been effectively met. In addition, considering the different educational requirements of different types of special needs children, based on the official website of the China Disabled Persons’ Federation divides special needs children into five categories: visually-impaired children with special needs, hearing-impaired children with special needs, physically disabled children with special needs, mentally handicapped children with special needs and children with comprehensive, multiple special needs. The total number of different types of special needs children is calculated for a robustness test. This study focuses on whether the construction of special education school buildings and other teaching infrastructure for rural special education will affect the local rural special education dropout rate. Through the time variable (Period) of policy

126

4 The Influencing Factors of the Imbalance of Rural Long Tail …

Fig. 4.5 Dynamic effects of the 2009 policy on RSE dropout rates. Notes The y-axis is the coefficient value of the interaction of the years and the treatment group. The x-axis is the year. The vertical bands represent + (−)1.96 times the standard error of each point estimate, and Current indicates 2009. Other years are in relation to 2009

and the experimental variable (Treat) of the processing sample, the interaction item (Treat* Period) is set as the core explanatory variable, which is used to measure the effect of the quasi-natural experiment. In the analysis process, policy implementation was set as Treat* Period_2009 in 2009. At the same time, in order to conduct the placebo test, the assumption was made that the policy would be implemented in 2008 and 2010. In addition, the analysis sets two variables, Treat* Period_2008 and Treat* Period_2010. For the control variables of the DID model, one must take into account, as much as possible, the relevant variables that affect the dropout rate of rural special education and the construction of rural special education school buildings. Referring to relevant research, this book finally selects rural per capita income, rural per capita cultural NGOs, rural special education computer room area, per capita special education budget expenditure and marketization index as the control variables that affect the rural special education dropout rate and rural special education school building construction. Among the above, rural per capita income reflects the local rural economic level; rural per capita cultural NGOs reflect the local rural social capital activity; rural special education computer room area reflects the local rural special education technology level; per capita special education financial budget expenditure

4.3 The Factors of Imbalance: Special Education

127

reflects the local government’s education input, and the marketization index reflects the local market-oriented activity. There are 357 valid samples; the average dropout rate is 7.107%, with a minimum value of 0.038% and a maximum value of 26.568%. From the perspective of different types of special needs children, the number of physically disabled children with special needs accounted for the largest proportion, followed by those with a hearing disability, visual disability, comprehensive disabilities and mental disability. The average value of Treat * Period_2009 is 0.495, which shows that the samples were almost evenly distributed before and after the implementation of the policy. Among other control variables, the degree of marketization is limited by data, and the 327 samples are slightly less than those of the other variables. In Models (1)-(3), Unedu was selected as the outcome variable, while in Models (4)-(6), the proportion of RSE children educated (Edu) was used as the sensitivity check. Furthermore, Models (1) and (4) were the baseline models without any covariates. Models (2) and (5) added income level, RSE budget, computer room, social groups, and marketization degree as covariates. Models (3) and (6) further added the year dummy variables. Models (7) and (8), 2008, and 2010 were chosen as false time nodes of policy implementation, in order to facilitate the placebo test’s performance. All the DID results are listed below (Table 4.15). The results show that the coefficients of Treat * Period_2009 were significant (especially for Models (1) and (2) at the 5% level). The interaction coefficients were negative (approximately −2 to −3), which means that investment helped decrease the RSE dropout rate. In Models (4)–(6), the interaction coefficients were significantly positive (approximately 2–3 at the 5% level), which means that policy implementation also helped increase the RSE enrollment ratio. These findings support this study’s hypothesis that the construction and renovation of RSE teaching infrastructures and facilities, based on governmental expenditures, have corrective effects on RSE’s imbalance and improvement. However, the nonsignificant coefficients of Treat* Period_2008 and Treat* Period_2010 in Models (7) and (8), respectively, support the authors’ comments about the parallel trend. In the 2009 policy implementation, investment in RSE included construction and the capital expenses for teaching infrastructure, equipment (such as computers), and building repairs. It is not difficult to understand how these investment types could enhance the quality of supply and have more substantial positive externalities than other expenses. This means their crowding-out effects will not hamper the efficiency level of supply. These results support the authors’ idea that greater investment in RSE infrastructure has large spillover effects in terms of correcting RSE’s insufficient supply. The underlying reason for this finding may be the signaling failure of individual RSE demands, which, with their heterogeneous and latent characteristics, makes trial and error with repetitive investment necessary to address dropout. Special-needs children with various symptoms and disabilities are geographically isolated and distributed in rural areas in a fragmented manner. It is therefore necessary to analyze the impact of RSE school construction on dropouts in specific RSE categories. Based on the China Disabled Persons’ Federation, RSE children were divided into five categories: those with visual disability, hearing disability, limb

−3.268**

(−3.05)

(−2.81)

*

(−3.62)

(1.32)

(−6.59)

−2.218**

Note t statistics are in parentheses; p < 0.1,

**

p < 0.05,

(34.67)

336

(4.16)

9.193***

_cons

370

12.09***

No

Year

N

No

No

Yes

−13.89**

1.272

***

370

336

(6.26)

15.75***

No

Yes

(1.96)

2.921*

(−1.72)

−2.243

(5)

Edu

336

(2.67)

12.39*

Yes

Yes

(2.33)

3.787**

(−1.38)

−4.461

(6)

Edu

336

(−0.17)

−1.076

Yes

Yes

(0.53)

0.515

(−4.05)

−14.97***

(7)

Unedu

336

(−0.13)

−0.828

Yes

Yes

(−0.66)

−0.748

(−3.05)

−12.77**

(8)

Unedu

p < 0.01. Other control variables are not shown in the table. The same applies below

336

(31.98)

8.990***

(−0.20)

No

−1.239

No

(1.79)

2.935*

(−1.55)

−2.441

(4)

Edu

Yes

Yes

(−0.13)

−0.126

(3)

(2)

(1)

−1.640***

Covariant

Treat*Period

Period_2010

Treat*Period

Period_2008

Treat*Period

Period_2009

Unedu

Unedu

Unedu

Table 4.15 Basic DID results

128 4 The Influencing Factors of the Imbalance of Rural Long Tail …

4.3 The Factors of Imbalance: Special Education

129

impairment, mental disability and multiple disabilities. The DID model above was still used, and the number of dropouts in the different categories of RSE children were taken as the robust outcome variables. The results of all RSE categories are shown below (Table 4.16). The coefficients of Treat * Period_2009 are still significantly negative, which means that the influence of infrastructure investment and construction on the RSE dropout rate is robust. Regarding the different categories of RSE, the authors found that the absolute values of visual, hearing, and physical disabilities were approximately 500–600 more than those of mental and multiple disabilities (approximately 180–280) on average. The nature of various RSE categories explains this difference in marginal effects. Most visually- and hearing-impaired children with special needs have the potential and capacity to be educated, if they can access suitable auxiliary teaching equipment that will help them overcome their physical limitations. These special needs children can be taught to attend to the phonological features of words through phonemic awareness activities, auditory discrimination activities, or phonic analysis instruction. The differences in abilities may be accounted for by the differences in school investment. The differences in abilities may also be accounted for by the differences Table 4.16 DID results for all RSE categories Visual Period_2009 Treat*Period Income Computer Group SEbud

Hearing

Physical

Mental

Multiple

(1)

(2)

(3)

(4)

(5)

255.2*

189.4

157.6

74.44

119.9

(2.12)

(1.67)

(1.37)

(1.54)

(1.64)

−541.7***

−622.7***

−511.6**

−183.5**

−287.1**

(−3.98)

(−4.55)

(−3.36)

(−3.11)

(−2.86)

−23,786.8

468.6

−3672.2

−3852.0

−21,377.9

(−0.67)

(0.01)

(−0.11)

(−-0.27)

(−0.95)

−1,645,534.5** −1,973,522.6*** −1,694,476.0** −739,581.8* −998,393.7* (−3.05)

(−3.79)

(−3.57)

(−2.55)

(−2.69)

−46,092.9

−47,707.6

−25,266.7

−14,237.6

−25,537.0

(−1.92)

(−1.99)

(−1.01)

(−1.33)

(−1.48)

18,935.4

14,580.4

9284.7

2562.3

9197.6

(0.80)

(0.62)

(0.34)

(0.25)

(0.49)

Marketization −25.93

−50.67

−41.96

−9.619

−8.784

(−0.83)

(−1.25)

(−0.97)

(−0.82)

(−0.35)

1282.8***

1538.0***

1755.2***

536.7***

965.5***

(6.42)

(5.68)

(6.33)

(5.40)

(5.88)

327

327

327

327

327

_cons N

Note income, group, computer, SEBud, and marketization are the rural per capita income, the rural per capita amount of social cultural organizations, the rural computer room area for special education, the budget expenditure of special education per RSE student, and the marketization degree, respectively

130

4 The Influencing Factors of the Imbalance of Rural Long Tail …

in school investment teaching (Schirmer & McGough, 2005). Despite the finding that special needs students can access phonological information and visual codes, the lack of instructional intervention is still widespread. In China, most of these children are being educated independently, in special schools that are exclusive to children with visual and hearing losses. However, in rural areas, providing all the necessary auxiliary teaching equipment is difficult, which is why infrastructure investment has larger marginal effects on these categories. Similarly, except for concerns related to movement inconvenience, limb-impaired children with special needs can still compete equally with their normal counterparts. Personal, social, environmental, policy and program-related barriers and facilitators all influence the amount of education that physically handicapped children with special needs receive (Shields et al., 2012). These students are also reported to have fewer choices of adequate, accessible, or convenient facilities and transportation. In China, a considerable proportion of physically handicapped children with special needs are being educated in follow-up classes with other normal students. In this situation, the special needs students may be more affected by investment in schools and other teaching infrastructures. Compared with the three categories above, mentally handicapped children with special needs have intrinsic difficulties in education. Specifically, they face much higher education costs and tend to require full-time daycare. Even with the promotion of school construction, there is a lack of service availability and motivation. In China, mental disability accounts for approximately 9% of all disabilities (Zheng et al., 2011). As China is undergoing a rapid socioeconomic transition, the provision of education to mentally handicapped children with special needs should be considered based on multiple social integration factors, such as the community, health care systems, and poverty reduction (Li et al., 2012). The term “multiple disabilities” represents a broad category that mediates the effects of other disability categories. Accordingly, the coefficients for children with multiple disabilities are in the range of the coefficients for other children with special needs with regard to absolute value. These children probably have a learning disability or a delay in intellectual development, as well as speech problems. Following examination, diagnosis and treatment planning, much of the required preventive, simple treatment and oral health promotion can be performed by trained auxiliaries (Desai et al., 2001). However, due to the chronicity of the disabilities of mentally handicapped and children with multiple special needs, as well as the mediation effects, it is easy for RSE teachers and school managers to deem their own efforts to be fruitless and discouraging. The most important difficulties these children experience in education are their learning facilities, the lack of empathy from their student peers and barriers in the physical environment (Tran, 2014). In summary, school investment and construction exert different impacts on different categories of RSE children. Regarding visually- and hearing-impaired children with special needs, the best solution is to highlight investment in auxiliary teaching infrastructure, in order to help promote RSE availability. Regarding physically handicapped children with special needs, the existing pattern of “follow-up in class” can help them enjoy equal opportunities to access basic education, without

4.3 The Factors of Imbalance: Special Education

131

being overburdened by artificial entry barriers. Mentally handicapped children with special needs are less affected by school construction, due to the higher entry costs of education. Children with multiple special needs can be viewed as a comprehensive category and as mediation for other disability categories. Thus, in terms of the absolute value, their marginal effects among the other categories are also moderate. The dominant role of government investment in teaching infrastructure may crowd out the civil participation of social groups (Tilak, 2002). These factors, if present, could confound the authors’ argument that school construction investment reduces RSE dropout rates. Hence, this study follows Tanaka (2015) and takes the RSE budget and social groups (SEbud and group) as outcome variables, in order to test the trend difference’s significance. The model results are as follows (Table 4.17): The coefficients of Treat*period_2009 were found to be not significant and virtually indistinguishable from zero. A low degree of correlation with RSE budget and social groups suggests that the underlying policy implementation is less likely to be a bias source. This finding could increase our confidence in the policy’s exogeneity, Table 4.17 Results of the endogeneity test SEbud

SEbud

Group

Group

(1)

(2)

(3)

(4)

Treat*Period

−0.000108

−0.0000299

0.0000676

−0.000243

(−1.45)

(−0.20)

(1.63)

(−1.88)

Income

0.142**

0.154

−0.0504

−0.0744

(3.43)

(1.92)

(−1.53)

(−1.22)

−0.503*

−0.511*

0.0533

0.0162

(−2.57)

(−2.08)

(0.28)

(0.09)

−0.0000310

−0.0000234

−0.0000774

0.0000392

(−0.68)

(−0.30)

(−1.59)

(0.72)

Computer Marketization Unedu Period_2009 Group

0.0000139

−0.0000128*

(1.95)

(−2.15)

−0.0000534

0.000243

(−0.11)

(0.83)

0.457

0.505

(1.09)

(1.01) 0.415***

SEbud

0.351**

(4.10)

(3.54)

Year

No

Yes

No

Yes

_cons

−0.0000512

−0.000301

0.000652*

0.000105

(−0.24)

(−0.71)

(2.64)

(0.35)

N

336

336

336

336

Note Models (1) and (3) are baseline models. Models (2) and (4) add more variables, including Unedu, Treat * Period_2009 and the year dummies

132

4 The Influencing Factors of the Imbalance of Rural Long Tail …

Table 4.18 QDID results Quantile Unedu (1)

Unedu

Unedu

Unedu

Unedu

(2)

(3)

Quantile Unedu (1)

(2)

(3)

−1.077

−1.558

−1.968*

−3.858***

−4.058***

(−1.538)

(−5.659)

10

−0.491 (0.684)

(0.747)

(−1.75)

(0.850)

20

−1.193

−2.418*** −2.786*** 70

−2.297*

−3.709*** −4.434***

(0.841)

(0.788)

(−1.72)

1.009

30

−2.528*** −3.081*** −3.675*** 80

−3.109*

−3.563*** −4.430***

(0.810)

(0.795)

(−5.217)

(−1.75)

1.309

(−3.753)

40

−2.838***

−3.562***

−3.704***

−6.627***

−5.404***

−5.882***

(0.706)

(0.598)

(−5.967)

(2.263)

1.781

(-3.162)

50

−3.000*** −3.503*** −3.882*** (1.133)

(0.724)

60

(−4.207)

90

(−5.619)

(−4.529)

Note Model (1) is the baseline model without any covariates. Model (2) adds the income, SEbud, computer, marketization and group covariates. Model (3) also adds the year dummies

because the DID value of these covariates did not change, strongly suggesting that this quasi-experimental design is unlikely to be biased by changes in unobservable variables. In some instances, policy-makers are more concerned about the heterogeneous effects of policies on different quantiles of the population distribution, namely, the quantile treatment effect. A QDID analysis, the DID method estimated by preset quantiles of the dependent variables (Callaway & Li, 2017), addresses heterogeneous treatment effects. To further decrease the estimation bias of individual heterogeneity and to analyze the effects of policy implementation under different covariant distributions, this research follows Lucas and Mbiti (2012) and uses the QDID method to study the treatment effects of different quantiles, rather than the average. The year 2009 was still set as the time node of policy implementation. The results are as follows (Table 4.18): The results show that the 2009 policy implementation had progressively increasing negative effects on the treatment group (western and central provinces). The higher the value of Unedu of the province (higher quantiles) dropout’s correction effects was, the larger the dropout’s correction effects were after the policy implementation. This finding is consistent with existing studies. The results show that, with increasing investment in teaching infrastructure and construction (the main policy measure of 2009), especially concerning governmental expenditures, RSE is affected by a correction mechanism. This is especially true in provinces with higher RSE dropout rates. Unfortunately, the DID analysis also has inherent shortcomings when evaluating policy implementation. First, the control group’s selection involves some subjectivity and arbitrariness, which may harm the selection’s persuasiveness. Second, policy implementation may be endogenous, which means that the pilot regions have

4.3 The Factors of Imbalance: Special Education

133

systematic differences from other regions (Abadie et al., 2010). If sufficient reasons to eliminate the policy endogeneity do not exist, then the DID analysis may cause biased errors. Abadie and Gardeazabal (2003) proposed the Synthetic Control Method (SCM) for policy evaluation. The SCM provides a data-based method for selecting control groups and evaluating policies, and this method has three advantages. First, SCM broadens the traditional DID method by taking a nonparametric perspective. Second, it determines the data’s weight when constructing control groups, which could reduce the subjective judgment. Third, because a DID analysis cannot resolve the endogeneity issue caused by various unobserved factors with time variation, the SCM method allows for these factors with time variation. The SCM method constructs a counterfactual state to present the similarity between the treatment and control groups before policy implementation. This counterfactual state is a weighted mean to which the control group contributes; to avoid excessive extrapolation, the sum of the weighting is 1 (Temple, 1999). This study follows Abadie et al. (2010), using the SCM to measure the influential roles of different expenditures in Unedu with policy implementation. No interaction effects were thought to exist between the treatment and control groups, and all covariates were predetermined and exogenous (which will be tested later). Traditionally, SCM is more concerned with the aggregate information of one region when evaluating policy implementation. The mean value of all treated groups (western and central provinces) was first set as one treatment unit. Then, all other control groups (eastern provinces) were used to match this treatment unit as the SCM’s synthetic control unit. Figure 4.6 shows the variation trend of Unedu between the treatment unit and the synthetic control unit. The results show that, before 2009, the two groups were nearly parallel. The mean value of Unedu for the treatment group was approximately 4 higher than that of the synthesized control group. However, after 2009, the groups exhibited a noticeable divergent trend. The treatment group’s Unedu value sharply decreases and intersects with that of the synthesized control group in 2011. After 2011, the treatment group’s dropout rate was even lower than that of the synthesized control group. This finding further supports the idea that the implementation of the 2009 infrastructure investment and construction policy created differences between the treatment unit and the synthetic control unit. The differential impacts of financial investment in RSE school construction in China on local dropout rates were examined by employing a quasi-experiment to incorporate the contextual factors associated with RSE infrastructure and motivation. The difficulties expressed by those belonging to different RSE categories include the limitations of learning facilities, which are clearly more prominent than policy implementation limitations. The regions receiving financial support were also found to have advantages in terms of correcting and mitigating the dropout rate. The explanatory model and theories developed in this book are convincing, since they were replicated in different RSE categories with different measurement methods and mediation effects. The authors found that children with different categories of

134

4 The Influencing Factors of the Imbalance of Rural Long Tail …

Fig. 4.6 Variation trend of the RSE dropout rate. Note The data sources are the China Education Statistical Yearbook, Statistical Yearbook of China’s Educational Expenditure, China Rural Statistical Yearbook, China Civil Affairs Statistical Yearbook, China Population and Employment Statistical Yearbook, China Statistical Yearbook, the official website of the China Disabled Persons’ Federation and the Report on the Marketization Index of China’s Sub-provinces (2003–2014); Unedu is used as the outcome variable, while income, SElbud, computer, group and marketization are used as the observable covariates for synthesis. The same applies below

various disabilities vary significantly with regard to their educational needs. In particular, mentally handicapped children and children with multiple special needs were less sensitive to infrastructure investments than were children with physical disabilities. The underlying mechanism may be that, with regard to being educated, mentally handicapped children with special needs have congenital disadvantages. They are more likely to (and may prefer to) receive healthcare than to receive an education. The marginal effects on children with multiple special needs include the mediation of other disability categories. These findings should be viewed as part of an evolving understanding of RSE conditions, based on Conlin and Jalilevand (2015). The findings, representing more than a decade’s worth of research, should also be part of a more general discussion of financing children with special needs (Cullen, 2003). Methods that can be used to address RSE dropout rates are revealed, and they partly originate from a lack of infrastructure investment, especially in developing countries. To overcome this situation and improve RSE, policymakers should consider software development and the construction of physical infrastructure, which is currently the best way to increase enrollment. School settings are the key factor in improving the social inclusion process of RSE children and in aiding those people who supporting these children in overcoming the difficulties they encounter in schools. The authors recommend promoting any activities that increase the educational and social awareness of the challenges faced by children with special needs.

4.3 The Factors of Imbalance: Special Education

135

Generally, due to the time-lag recognition and provision of RSE, different categories of expenditures from various suppliers have divergent influences on the intertemporal correction mechanism. Specifically, capital expenses (especially infrastructure construction) have the largest positive marginal effects on the dynamic correction of RSE mismatch. This finding is true for both the ‘long tail’ and the public nature of RSE, and when considering RSE’s interaction with other education categories. Second, governmental administrative expenses have significant negative impacts on the dynamic correction of RSE. However, the administrative expenses of other social sectors are not significant, due to their minor proportions. The proportion of administrative expenses can be seen as the indicator of governance efficiency in the provision. Due to the public nature of RSE, similar to other education categories, a moderate proportion of administrative expenses is necessary for the normal operation of RSE. However, with regard to RSE’s ‘long tail’ nature, an excessive investment of administrative expenses could crowd out other kinds of expenses and have side effects on the correction. Hence, related policies should consider such an investment more seriously and control that investment reasonably. Third, the influences of welfare expenses for faculty members from both governments and other social sectors were not significant for the dynamic correction of RSE in this study. The underlying reason may be the distortion and misallocation of these resources feeding into the weak motivation of RSE faculty members. Due to the institutional, cultural and social barriers they face, RSE faculty members are still insensitive to financial incentives, due to their blurry and discouraging career expectations. Fourth, the influence of scholarship expenses for students and their families, from both governments and other social sectors, was also not significant in the dynamic correction of RSE. In addition to the similarity in the motivation of faculty members with regard to welfare expenses, the cognitive deficiency of RSE families and the high entry cost of education also decrease the effects of financial investments provided directly to the families. However, the aggregation of the ‘long tail’ could act as a democratic distribution mechanism and best exert the information and cost advantage, as the long tail essentially faces the exact RSE demanders. In the sensitivity analysis, children with psychiatric disabilities were found to be less sensitive to financial investments than children with other physical disabilities (including vision- and hearing-related disabilities). The underlying mechanism is that children with psychiatric disabilities have congenital disadvantages to being educated and are more inclined to receive healthcare than education. In addition, the marginal effects on children with multiple special needs are the mediation of children with physical and psychiatric disabilities. In the test of mediation of RSE with RVE, the results were not as functional as those with RBE, due to the sharing ‘long tail’ nature of RSE and RVE. The ‘long tail’ nature could dampen the governance efficiency of RSE provision. This finding implies that most RSE children in China are accepting basic education rather than vocational education. Hence, the relationship between RSE and RVE is not complementary, but rather continuous on the ‘long tail’ curve.

136

4 The Influencing Factors of the Imbalance of Rural Long Tail …

The policy evaluation of infrastructure construction in the western and central regions of China in 2009 supports the significance of allocating capital expenses as being the chief task required to correct the supply insufficiency of RSE. The verification of the time dynamic effect supports the idea that this recognition and correction have time-lag properties. Hence, the investment of RSE expenditure must consider this utility lag and place greater emphasis on the improvement of existing RSE conditions. Committed separation of the design of the ‘long tail’ nature of RSE is a critical academic and empirical need. This is especially true in light of the fact that those public demands have the dual attributes of both heterogeneous privacy and universal externalities, similar to other education categories. This book creatively draws parallels from the roles and functions of RBE on illiteracy to the public good nature of RSE on its dropout rate as the ‘proxy variables’, thereby dividing this dual nature. Both of these results raise serious challenges to the study of Pulkkinen and Jahnukainen (2016), which found that local stakeholders targeted the resources of special education differently with respect to the trade-off between the provision of special education and basic education. This conflict needs to be resolved in the very near future. Actually, an increasing amount of research supports the relationship between different expenditures, financing and RSE provision efficiency. Various researchers have argued that the interaction of different stakeholders in RSE could influence the results of RSE provision (Baker & Ramsey, 2010; Mahitivanichcha & Parrish, 2005). However, despite these theoretical and empirical consequences, the division and customization of specific subtypes and expenses of RSE based on RSE’s unique characteristics appear to be relatively scarce. Few scholars that raise the idea argue that the combination of uniqueness and commonness of RSE and its corresponding dynamic correction interact with other education products. Inspired from these perspectives, as well as from frameworks that have been widely applied to other areas (such as business management), this study proposes a functional analysis of the dynamic correction and mediation effects of the ‘long tail’ and ‘head’ in public education as a whole. This research also contributes to the fiscal implementation of divergent expenditures, according to their various motivation effects on the correction. Given that existing studies imply that RSE demands are easy to ignore and mismatch, the results of this research offer several indirect suggestions regarding how some expenditures (such as capital expenses) help to dynamically alleviate the imbalance and the extent of marginal effects. In particular, the authors have proposed and found that the ‘long tail’ nature of RSE can directly influence the interaction and mediation of other education categories in the correction. These findings suggest that the balance of RSE not only benefits from financial expenditures by governments, but also may benefit from other social sectors (though not significantly in some models). Given that some categories of expense may be distorted and/or misallocated, the authors do not intend to advise relevant stakeholders to cancel these investments, the nature of which can be improved by a better understanding and designing of the motivation channel for RSE faculty members and students. Rather, efforts should be directed towards fostering the formal instant recognition and provision mechanism,

4.3 The Factors of Imbalance: Special Education

137

which can be achieved through cooperating with other NGOs acting as the ‘long tail aggregator’ and formalizing the network platform to share information, thus considering the dual nature of RSE together from the perspective of implementing system detection. Moreover, given the findings that different subtypes of RSE have divergent performances associated with different expenses, the detailed and tailored pattern of ‘cater to all disabilities’ may be necessary for exclusive or comprehensive RSE schools. More importantly, our studies suggest that a new perspective could be adopted with regard to the balance and equilibrium of RSE from system theory. In particular, rather than regarding the supply of RSE as originating solely statically, we may want to further investigate the dynamic cause of these interactions and determine whether they reflect the internal nature of RSE. Causes such as scattering distribution and demand heterogeneity could be worthy of further investigation. Given that governments are likely to engage in the infrastructure construction of RSE as a way to restore balance, this book examines whether different expenses are reflective of different stakeholders’ attempts to resolve the mismatch, as well as intertemporal tendencies to resort to broader investment in construction, by addressing these issues.

4.4 Factors of the Imbalance: Special Health The term “rural special health service” refers to the heterogeneous and minority medical public services provided to the rural special needs population. In turn, the term “rural special needs population” refers to the population with a higher probability of medical risk or risk loss, due to the special physical or economic conditions of individuals in rural areas (such as the disabled or residents with rare diseases). This part of the population is a group not only facing a high probability of “poverty caused by illness”, but also the direct impact on the imbalance of supply and demand of rural long tail public services. Many of these special health services have the nature of long tail need and long-term uninterrupted treatment (cyclical and chronic); the treatment cost and frequency are also relatively high. When enjoying the new rural cooperative medical scheme, this group is limited by the number of and average cost per visit. The compensation level is also limited, which can easily lead to insufficient supply. China has begun to explore new rural cooperative medical schemes for special diseases coverage, as well as a compensation policy. However, due to the lack of unified norms and guidance, the relevant policy system will have difficulty keeping up with the changing medical needs of rural residents. The imbalance of rural long tail special health public services is affected by China’s geographical conditions, natural resource endowment, economic development, government security, social participation, marketization degree and the individual heterogeneity of residents. In order to further empirically analyze the influencing factors of the imbalance of rural long tail special health public services, the data in this section come from the 2008–2018 CHARLS (China health and retirement longitudinal study). Also, individual rural residents are used as samples to construct

138

4 The Influencing Factors of the Imbalance of Rural Long Tail …

panel data (panel fixed effect). Among them, middle-aged and elderly rural residents with disabilities (including physical and mental disabilities) or more serious infectious diseases (such as rabies, schistosomiasis, AIDS, Japanese encephalitis, dysentery, measles, brucellosis, gonorrhea, syphilis, etc.) were selected as the main samples of special health service recipients. The question asked to determine if a person is a rural resident was: “What is the type of your household registration at present?” The individual residents who chose rural household registration were taken as the sample population of rural demand. The question asked to screen for special health residents was: “Do you have the following disability problems (a list was given)?” Rural residents with one or more of the physical disabilities of brain damage, mental retardation, blindness or semi-blindness, deafness or semi-deafness, mute or severe stuttering were selected. Another question was, “Have you ever been told that you have the following infectious diseases (a list was given)?” The rural residents who chose one or more of tuberculosis, hepatitis B, malaria, rabies, schistosomiasis, AIDS, Japanese encephalitis, dysentery, measles, brucellosis, gonorrhea and syphilis were also chosen as the sample of rural long tail special health public services. The explained variable was the degree of mismatch of rural long tail special health services, which was measured by different subjective and objective indicators (including difficulty in meeting the basic necessities of life, service satisfaction, reasons for not seeing a doctor, etc.). The degree of difficulty with regard to clothing, food, housing and transportation is based on whether the resident receives help with daily behaviors such as dressing, bathing, eating, getting out of bed, going to the toilet, housework, cooking, shopping, making phone calls, taking medicine and so on. Hence, the dummy variable need-help was constructed. Service satisfaction is the subjective judgment of the rural special population with regard to the level and quality of medical service supply. A dummy variable unsatis was constructed (satisfied with service = 0; dissatisfied with service = 1). Other explanatory variables include a series of individual variables of rural special needs, such as health level, marriage status, insurance participation, type of work (divided into three categories: employed, self-employed, and farming), income level, education level, etc. The specific calculation method and descriptive statistics of each variable are shown below (Tables 4.19 and 4.20). Considering that the explained variables are discrete binary variables, the panel binary selection model is used in each model. A panel probit model is used in Models (1) and (2). A panel logit model is used in Models (3) and (4), and a year dummy variable is added to control the time effect. At the same time, the self-employed occupation is selected as the benchmark group, in order to estimate the comparison coefficient of the other two types of occupation (employed and farming). As shown in the following table, Models (1) and (3) select needhelp as the explained variable, and Models (2) and (4) select unsatisfaction as the explained variable. The regression results of all models are as follows (Table 4.21): By observing the regression results, the coefficient of health in all models can be seen to be significantly negative (at least at 5% level), consistent with people’s expectations. Generally speaking, the higher the individual health level of rural residents

4.4 Factors of the Imbalance: Special Health

139

Table 4.19 Variable list of rural long tail special health service (I) Name

Calculation method

Source

Health

Healthy set 1; not healthy set 0

DA001 What do you think of your health?

Married

Married set 1; not married set 0

BE001 What is your current marital status?

Insurance

If enjoy the new rural cooperative EA001 Do you have the following medical insurance (cooperative medical insurance? medical insurance) or the medical insurance for urban and rural residents, set 1; if not set 0

Employed

If being employed set 1; if not set 0

FL001 Are you employed?

Farming

If farming work set 1; if not set 0

FL001 Was your last job in farming?

Self-employed If yes set 1; if no set 0

FL001 Were you self-employed in your last job?

Wage

If yes set 1; if not set 0

GA001 In the past year, have you received any salary?

Education

If above university education, set 1; if not set 0

BD001 What is the highest education you have obtained?

Table 4.20 Statistical description of variables in special health service (I) Variable

Obs

Mean

Std. dev

Min

Max

Needhelp

3,871

0.0529579

0.2239784

0

1

Unsatis

3,871

0.1198657

0.3248463

0

1

Health

3,871

0.1097908

0.3126692

0

1

Married

3,871

0.7889434

0.4081112

0

1

Insurance

3,871

0.9377422

0.2416544

0

1

Employed

3,871

0.0284164

0.1661808

0

1

Farming

3,871

0.2030483

0.4023202

0

1

Wage

3,871

0.1410488

0.3481169

0

1

Education

3,871

0.6078533

0.4882922

0

1

Data source 2008–2018 CHARLS

with special needs is, the smaller the demand for special medical care and the lower the degree of mismatch and imbalance will be. This is especially the case when the supply level remains unchanged. This relationship is not only reflected in the need for help from others, but also in the satisfaction with medical services. The coefficient of married is also significantly negative in most models (at least at the level of 10%). This finding indicates that a change in marriage status can significantly affect the subjective judgment and attitude of rural special needs individuals towards the supply and demand balance of long tail public services. Compared with unmarried rural residents, married rural residents not only can help each other through

140

4 The Influencing Factors of the Imbalance of Rural Long Tail …

Table 4.21 Analysis of impact factors in the supply–demand disequilibrium of rural long tail special health service (I) (1)

(2)

(3)

(4)

Need help

Unsatisfaction

Need help

Unsatisfaction

−1.888**

−0.839***

−4.035**

−1.259***

(−2.121)

(−3.154)

(−2.440)

(−3.005)

Married

−0.885**

−0.324*

−1.803**

−0.493*

(−2.073)

(−1.845)

(−2.190)

(−1.868)

Insurance

−0.638

−0.144

−1.409

−0.226

(−0.973)

(−0.477)

(−1.044)

(−0.496)

Employed

3.854*

0.279

9.750**

0.396

(1.792)

(0.674)

(2.556)

(0.638)

−0.675

−0.211

−1.226

−0.333

(−0.840)

(−1.114)

(−0.816)

(−1.154)

Wage

−0.314

0.134

−0.643

0.192

(−0.489)

(0.646)

(−0.572)

(0.616)

Education

−0.161

0.060

−0.587

0.088

(-0.247)

(0.379)

(-0.435)

(0.374)

Health

Farming

Year

Y

Y

Y

Y

_cons

0.532

−2.614***

0.436

−4.307***

(0.563)

(−6.811)

(0.221)

(−7.676)

529

3871

529

3871

N

marriage, but married couples also have incomparable advantages in terms of information and cost when identifying their own special medical needs (between spouses). These advantages will reduce married couples’ dependence on external help. On the one hand, the “tail” special medical needs respond to the problem of high cost. On the other hand, marriage has a positive impact on the well-being of residents’ current lives„ which in turn can cover the satisfaction of residents with special needs with long tail medical services. Therefore, compared with unmarried individuals, married individuals are more likely to have their medical demands satisfied. The coefficient of insurance is not significant. This shows that, for the special population in rural areas, due to the populations’ minority and heterogeneous long tail medical service demand, general medical insurance has difficulty covering these needs. On the contrary, insurance has a crowding out effect on other types of medical security and expenses. This leads to a negative impact on the satisfaction with medical security of rural residents, and increases the difficulty of satisfying their medical treatment needs. The evaluation of medical treatment satisfaction of the special medical population is affected by objective factors, such as service quality, attitude, price, and convenience, as well as subjective factors, such as treatment purpose and service expectations. Some problems, such as medical services and drug types, cause the

4.4 Factors of the Imbalance: Special Health

141

long tail patients’ overall satisfaction with medical services to be lower, especially in terms of the coverage, diagnosis and treatment levels. From the perspective of occupations of those with different special needs, compared with self-employed, the coefficient of employed is positive in all models; the coefficient of farming is negative in all models (though not significant). This finding shows that, compared with self-employment, individuals employed by others are less satisfied with the supply of rural long tail special medical services. The reason behind this may be that, compared with those who are employed by others, the self-employed and those with agricultural work tend to live and work in the traditional natural economy; that is, the family mode of “self-insurance” being used to supply medical needs. This kind of family supply self-insurance mode can disperse the special medical needs within a family by expanding the family’s size, so as to control the cost of rural special medical services within a desirable financing range. Elderly patients whose original occupation was farming are more likely to use medical services to meet their senior medical needs than those who did not work in farming. The difference is that those who are employed are more market-oriented in life, and they are more inclined to have medical services provided through the mode of commercial insurance. However, as mentioned above, the medical insurance model represented by the new rural cooperative medical scheme has difficulty giving full play to the targeted information advantages of the family self-supply model. Insurance can only compensate for more general and homogeneous medical needs, which is also the reason why the employed coefficient is significantly positive. Wage and education coefficients are not significant in all models, indicating that the imbalance of rural long tail special medical services is not directly affected by the income and education level of the demanders. The fragmentation and discretization of rural special medical care are widely distributed in individuals, with a certain degree of power-law distribution characteristics. Therefore, no significant linear relationship exists between the degree of imbalance and the income and education levels. Generally speaking, compared with rural special education, the supply level, quality and quantity of rural special health are more affected by the health level and marital status of the demanders. However, the demand subject’s occupation choice, insurance status, education and income level have limited influence. Rural special health is more inclined to be provided by the family through the traditional natural economy, thus resisting the natural risk of medical demand. Comparatively speaking, a more commercialized employment demand subject is more vulnerable to special health risks and has lower satisfaction in terms of their medical needs. However, the limitation of the above analysis lies in that the analysis only considers the impact of special rural needs’ individual factors on the imbalance. The analysis does not take into account the macro-regional factors that the individual needs live with and produce. This may also have indirect effects on the characteristics of micro individual heterogeneity. Even if only the micro influencing factors of individual demand are considered and only a few characteristic variables are selected, exhausting all the micro characteristic heterogeneity of individual demand is impossible.

142

4 The Influencing Factors of the Imbalance of Rural Long Tail …

Therefore, in order to further improve and supplement the existing empirical research, and considering the impact of regional factors on the imbalance for individuals with special needs in rural areas, this book also selects China family panel studies (CFPS) as the data source. The sample of village residents was established by asking the question, “What is the main reason you are not working?” Those who chose “incapacity due to disability disease” were chosen as the special needs of rural long tail medical service sample. Other regional data sources are from various databases. The explanatory variables used to measure the imbalance of rural long tail medical services are dissatisfaction, dissatisfaction2, unhealth and unhealthlevel. The question used to determine the level of unsatisfaction and unsatisfaction2 was, “What’s your overall satisfaction with the conditions of medical treatment?” In this study, unsatisfaction is a multivariate discrete variable (very satisfied set 1; satisfied set 2; so-so set 3; dissatisfied set 4; very dissatisfied set 5). In addition, unsatisfaction2 is a binary discrete variable (0 for very satisfied or satisfied, and 1 for so-so, unsatisfied or very unsatisfied). The question used to determine the levels of unhealth and unhealthlevel was, “What do you think of your health status?” Next, unhealth is a multivariate discrete variable (very healthy set 1; healthy set 2; relatively healthy set 3; so-so set 4; unhealthy set 5). Finally, unhealthlevel is a binary discrete variable (very healthy, healthy or relatively healthy set 0; so-so or unhealthy set 1). The sample sources and descriptive statistics of other explanatory variables are shown in the table below (Tables 4.22 and 4.23). Because the dependent variables are discrete at the individual level, and because the independent variables are at different levels (regional level and individual level), the common multivariate or binary discrete variable model will cause errors. This study uses the cross-level multiple ordered logistic model meologit and cross-level mixed panel-data binary logistic model xtmelogit. These models can better control the impact of regional differences on the characteristics of individual demand level. Therefore, unsatisfaction, unsatisfaction2, unhealth and unhealthlevel are selected as dependent variables to set Models (1)–(4), in which Models (1) and (3) are meologit models, and Models (2) and (4) are xtmelogit models. The random intercept variance is significant at the level of 5%.This finding indicates that significant differences exist in the imbalance of rural long tail medical demand among different provinces (or regions), proving the rationality of the cross-level model. All regression results are shown below (Table 4.24). From the analysis of individual micro variables, the income level variable in all models is significantly negative (0.2–0.3, significant at 1% level). This shows that, as the income and economic level of individuals with special needs increase, their dissatisfaction with the supply of rural special long tail medical services decreases; their subjective evaluation of their own health also increases significantly. A mutual causality exists between the health status and income of rural residents. The better the health status is, the higher the personal income appears to be. Similarly, the higher the personal income is, the better the health status appears to be (Krug, & Eberl, 2018). The impact of rural residents’ income levels on their own health is different in different regions (Ma et al., 2019). These studies are consistent with our results.

4.4 Factors of the Imbalance: Special Health

143

Table 4.22 List of all variables in rural long tail special health service (II) Variable

Method

Source

Income

The values from low to high, respectively, are 1–5

QN8011: What is your income level?

Insurance

If new rural cooperative medical scheme, public medical system or supplementary medical insurance, set 1; if not set 0

P605: What kind of insurance do you have?

Work

Yes set 1; no set 0

CFPS_everwork: Have you worked before?

Healthagency

Number of medical institutions in the province (or region)

China Health Statistical Yearbook

Marketization

Marketization index of the province (or region)

According to Fan et al. (2003)

NGOemployee

Total number of employees of NGOs in the province (or region)

China Civil Affairs Statistical Yearbook

Propgdp

Per capita GDP of the province (or region)

China Statistical Yearbook

Proexp

Fiscal expenditure of the province (or region)

Propopu

Total population of province (or region)

Table 4.23 Statistical description of all variables in rural long tail special health service (II) Variable

Obs

Mean

Std. dev

Min

Max

Dissatisfaction

1067

2.339269

7,547,129

1

5

Dissatisfaction2

1169

0.3892216

0.4877824

0

1

Unhealth

1078

4.441558

0.9547896

1

5

Unhealth level

1169

0.7510693

0.4325786

0

1

Income

1067

2

1.150903

1

5

Insurance

1078

0.9267161

0.2607228

0

1

Work

604

0.6721854

0.4698059

0

1

NGOemployee

695

204,756.5

145,838

627

632,390

Marketization

695

6.404029

1.862782

−0.3

10.92

Healthagency

695

35,329.83

25,652.72

1322

81,403

Propgdp

695

42,089.26

22,129.36

5222

115,053

Proexp

695

4257.088

2905.682

151.2

13,414.4

Propopu

695

5240.702

3013.923

277

10,999

144

4 The Influencing Factors of the Imbalance of Rural Long Tail …

Table 4.24 Analysis of impact factors in the supply–demand disequilibrium of rural long tail special health service (II) (1)

(2)

(3)

(4)

Dissatisfaction

Dissatisfaction2

Unhealth

Unhealthlevel

−0.229***

−0.229***

−0.379***

−0.377***

(−3.137)

(−2.822)

(−4.896)

(−4.026)

0.285

0.287

−0.095

0.382

(0.911)

(0.810)

(−0.260)

(0.953)

Work

−0.128

−0.181

−0.079

−0.052

(−0.741)

(−0.966)

(−0.399)

(−0.215)

Healthagency

−0.000**

−0.000**

−0.000**

−0.000***

(−2.488)

(−2.407)

(−2.268)

(−2.689)

−0.139

−0.268**

−0.404**

Income Insurance

Marketization

−0.162 (−1.633)

(−1.270)

(−2.037)

(−2.558)

NGOemployee

0.000

0.000

−0.000

−0.000

(0.603)

(0.778)

(−0.573)

(−0.235)

Propgdp

0.000

0.000

0.000

0.000

(1.425)

(1.246)

(0.412)

(1.215)

−0.000

−0.000

−0.000

−0.000

(−0.552)

(−0.513)

(−0.005)

(−0.318)

0.000**

0.000**

0.000

0.000*

(2.470)

(2.144)

(1.360)

(1.878)

Proexp Propopu

−0.327

_cons

3.730***

(−0.538) N

604

604

(3.697) 604

604

Especially with regard to the underdeveloped rural areas in the central and western regions, an improvement in income level has a greater marginal effect on the health status and satisfaction of rural residents with long tail medical needs. The coefficients of insurance and work are not significant, which is consistent with the results of CHARLS. Based on the long tail of rural special medical needs, the impact of the job choice and insured status of the demand subject on the imbalance is limited but not obvious. From the analysis of the independent variables at regional level, the coefficient of healthagency is significantly negative (1% level). However, the absolute value of the coefficient is relatively small, related to the small value range (the maximum is only 1–5). Some scholars have pointed out that, due to the influence of information disadvantage and weakening incentive, the government can easily find itself in a state of conflict with medical patients when setting up and operating public medical institutions, because the government cannot accurately identify and provide medical services (Chen et al., 2019). This is consistent with the conclusion of this study’s

4.4 Factors of the Imbalance: Special Health

145

analysis. Especially for the rural long tail medical services, the number of medical institutions has a certain distortion and failure in terms of the subjective evaluation and objective health status of the individuals in need. In addition, the coefficient of marketization is negative in all models (0.1–0.4, significant at 5% level). This finding shows that, the higher the degree of marketization is, the higher the satisfaction and subjective health evaluation of rural special medical needs will be. As marketization is directly proportional to the activity and contribution rate of local private economies, the development of private hospitals has a significant positive impact on the individual health level of rural residents (Li & He, 2019). This effect varies between regions, especially in the eastern regions in China. There, a higher degree of marketization is more conducive to playing a positive role in health performance. The coefficient of NGOemployee is not significant, and the absolute value is very small. This may be related to the fact that the development of NGOs in rural areas in China is still limited by economic, social and institutional factors. In practice, NGOs can promote the resolution of medical disputes and improve the quality of medical services through information communication and an interest expression mechanism (Lucas, 2008). However, on the one hand, some problems still exist, such as a lagging legislation process, imperfect laws and regulations, non-standard organizational behavior, and a serious lack of social support. On the other hand, it is very important for rural medical NGOs to survive and win development space by establishing a good public image (Cheng et al., 2010). In the supply of long tail medical services in rural areas, the number of NGOs in different regions is especially divergent. This leads to the coefficient of NGOs being not significant, and NGOs will therefore have difficulty in effectively playing the role as the third sector. Generally speaking, the empirical results confirm the previous theoretical analysis, which holds that the influencing factors of the imbalance of rural long tail medical public services include not only the heterogeneity of individual characteristics, but also the factors at the regional level. The latter include the public choice and irrationality of government medical investment, immature NGOs, and the lack of social support. Therefore, for this special type of rural long tail medical public service, we must further set up a mechanism design and analyze the interaction between different stakeholders, so as to achieve the optimal equilibrium of the supply state.

4.5 Factors of Imbalance: Elderly Care Due to the limitations of China’s rural economic level, as well as the concept of filial piety in traditional culture and social intergenerational support, most of the elderly residents in China’s rural society still adopt the traditional mode of homebased care. The demand for rural care institutions represented by nursing homes for the aged is more discrete and fragmented, in line with the requirements of long tail public services. These rural nursing homes for the aged are more limited in terms of their ability to meet the needs of rural residents’ daily lives and hospice

146

4 The Influencing Factors of the Imbalance of Rural Long Tail …

care; these homes generally have the characteristics of being administrative and semi-professional services (Milligan, 2012). The rural elderly residents’ demand for public elderly care services, like old people’s home in different areas (and especially those with special needs), is affected by their own health status, economic condition and family care situation. Moreover, for the rural elderly care service with a long tail nature, the elderly rural residents’ health status is not optimistic; many are badly in need of treatment. In addition, compared with urban communities, the infrastructure construction and software supporting services related to public elderly care in rural areas are relatively backward. Meanwhile, the traditional home-care mode is gradually being weakened by the impact of urbanization. Therefore, rural special needs elderly residents’ demand for a new pension mode has become a problem that cannot be ignored. Some scholars have pointed out that, after controlling the individual characteristic factors such as age, marriage, living conditions and income level, an increasing proportion of rural senior residents with special needs are not satisfied with the traditional home-care mode (Wong & Leung, 2012; Zhu, 2015). Although the longterm care demands brought about by chronic diseases encourage the elderly with special needs in rural areas to choose public-based elderly care, the relationship between the residents’ self-care ability and public elderly care demand intensity may not be monotonous linear and can be affected by other non-individual characteristics (Keimig, 2021; Wang, 2018). These studies highlight the fact that, when exploring the imbalance of rural long tail elderly care services (as represented by the new type of public elderly care), in addition to considering the individual characteristics of the needs, more factors at the regional level should be added. In this chapter, CHFS 2013 (China household financial survey) was selected as the sample source (considering the inconsistency of samples from before 2013, panel data were not selected). In the questionnaire, question [A2022] asked, “What is your current household registration type?” This study selected the rural household sample. Question [A2025D] asked, “Have you been unable to work or live normally in the past year due to any physical reasons?” The rural household samples who chose Yes were taken as the rural samples with special needs. Question [H3047] asked, “What form of retirement do you want in the future?” The rural sample who chose living in nursing homes was taken as the sample of public long tail elderly care demands. Due to physical disability or discomfort, these rural residents often find it difficult to have their elderly care needs fully met by family members (such as the support of their children). Based on their heterogeneity with long tail demand, these residents were selected as the samples for empirical research. The imbalance of public long tail elderly services is measured by the level of trust in the governmental provision of elderly care services. The variable Trust is obtained from question [H3042], which asked, “If you buy endowment insurance from the government, do you believe that the government will give you the money as promised in the future?” The answers were rated at from 1 to 5, or from least to most trustworthy, respectively. Residents’ trust in the government’s endowment insurance is closely related to the government’s construction of a sound social credit system (Zhang & Qiu, 2015). Based on the level of trust in the government elderly

4.5 Factors of Imbalance: Elderly Care

147

care service within the village, rural social network communication and capital can enhance rural residents’ expectations and then encourage them to participate in public elderly care services. Therefore, the residents’ degree of trust in the government’s public elderly care service with regard to special needs residents is used as a measure of the imbalance of rural long tail public elderly care service. The explanatory variables at the individual level of rural special elderly care needs include education, married, insurance (whether or not you have health insurance) and age. The regional level of rural special elderly care needs includes marketization (marketization degree of the province or region), NGOemployee (the employee number of NGOs), GDP (per capita GDP) and population (local population). The specific measurement methods, data sources and descriptive statistics of each variable are shown in Tables 4.25 and 4.26. Table 4.25 List of all variables in rural long tail special care for the aged (I) Variable

Measurement

Question

Education

Numbers 1–9 are, in order, never attended school, primary school, junior middle school, senior high school, technical secondary school/vocational high school, junior college/vocational high school, university undergraduate, master’s degree and doctor’s degree

[A2012] What is your education level?

Married

Married set 1; unmarried, separated, divorced, widowed and other set 0

[A2024] What is your marital status?

Insurance

If respondent has any endowment insurance, set 1; if none set 0

[F1001b] Which kind of retirement salary or social endowment insurance do you have?

Age

2013 minus birth year

[A2005] What is your birth year?

Source CHFS; variables at the regional level are explained and measured in the same way as above

Table 4.26 Statistical description of variables in rural long tail elderly care (I) Variable

Obs

Mean

Std. dev

Min

Max

Trust

258

4.496124

0.8380451

1

5

Education

451

2.390244

1.072191

1

7

Married

451

0.8758315

0.3301399

0

1

Insurance

451

0.7161863

0.4513482

0

1

Age

451

54.66962

12.85542

20

91

Marketization

490

6.20349

1.909272

-0.3

10.92

NGOemployee

490

182,598.7

136,337

627

632,390

GDP

490

40,268.7

23,154.7

5222

115,053

Population

490

4364.32

2657.564

277

10,999

Source CHFS

148

4 The Influencing Factors of the Imbalance of Rural Long Tail …

Table 4.27 Impact factors in the disequilibrium of rural long tail special care for the aged (I) (1)

(2)

(3)

(4)

Education

−0.542***

−0.286***

−0.542***

−0.281***

(−4.009)

(−3.647)

(−3.720)

(−3.363)

Married

1.197***

0.595**

1.256***

0.610**

(2.850)

(2.465)

(2.860)

(2.427)

0.349

0.268

0.337

0.255

(1.104)

(1.462)

(0.976)

(1.302)

Age

0.004

0.001

0.014

0.007

(0.377)

(0.222)

(1.154)

(1.030)

Marketization

0.311**

0.126

0.325*

0.136

(2.125)

(1.581)

(1.822)

(1.380)

−0.000

−0.000

−0.000

−0.000

(−1.284)

(−1.267)

(−1.330)

(−1.306)

GDP

−0.000

−0.000

−0.000

−0.000

(−1.339)

(−1.298)

(−1.249)

(−1.125)

Population

0.000

0.000

0.000

0.000

(0.421)

(0.946)

(0.376)

(0.678)

0.629

0.207

Insurance

NGO employee

var (cons[province]) N

258

258

(1.539)

(1.613)

258

258

Considering the multivariate ordered discrete characteristics of the dependent variable, the Ologit and Oprobit Models (1) and (2) are first set. Further considering the influence of different levels of independent variables (individual and regional levels) on the dependent variables, the cross-level multivariate ordered discrete models Meologit and Meoprobit (3) and (4) are considered. The results of all models are as follows (Table 4.27): Taking into account the factors at the individual level of special needs, the individual education level is significantly negative in all models (the coefficient size is 0.2–0.5, significant at the 1% level). This finding indicates that individuals with special needs who have a higher education level are more inclined not to trust the government to provide elderly care services, resulting in a more serious imbalance. Rural residents in China have a significant age divide in terms of when they choose to participate in endowment insurance. Education level is negatively correlated with farmers’ participation behavior; the longer the education period is, the lower the participation rate of farmers will be. That is to say, with an improved the education level, rural residents have less trust in the elderly care service provided by governments, especially the special needs with the property of long tail. These residents are more inclined to satisfy these heterogeneous needs through other channels (such as commercial insurance).

4.5 Factors of Imbalance: Elderly Care

149

Marital status (married) was significantly positive across all models (coefficient sizes ranged from 0.5 to 1.2 and significant at the 1% level). This finding indicates that marriage is beneficial to improving the degree of trust of rural individuals special elderly care needs with regard to the elderly care service provided by the government, so as to effectively correct the imbalance. Marriage can influence the special needs of the long tail old-age care services in two possible ways. First, the complementary effect of the married couples is equivalent to providing family assistance for individual elderly care, which in turn can reduce the demand intensity for public elderly care services. Second, marriage is highly related to children. The role of children as “raising children to provide security in old age” is still relatively obvious in rural areas of China. Studies have pointed out that, in current rural network families, the level of support provided by children to their elderly parents is relatively low, but the social capital brought in by children’s networks can significantly enhance the homebased elderly care support received by parents (Horsfall et al., 2012; Johansson et al., 2012). As the elderly care function of social capital brought about by marriages tends to weaken, the function and role of rural public elderly care service will become more and more important. From the perspective of explanatory variables at the regional level, only the marketization degree is significantly positive in Models (1) and (3) (the coefficient size is about 0.3). This finding indicates that the enhancement of marketization is conducive to the correction and alleviation of the imbalance of rural long tail special elderly care service. With the continuous strengthening of marketization, China’s rural old-age security system is also beginning to move toward socialization and marketization (Han, 2020). Some scholars have pointed out that, in the context of industrialization and marketization, China’s rural elderly care services are facing constant changes in terms of the space and boundary of elderly care (Feng et al., 2012). In particular, the trend of couples having “fewer children” and the acceleration of social mobility is causing rural residents with special long tail needs to break away from intergenerational needs and seek social support for elderly care. The trend shows the overflow and replacement of elderly care demands, from family to community and community to society (White, 2018). For rural residents with special needs, this mode of “joint endowment” between the government and family is closely related to the joint production mode of agricultural marketization. In addition, the married couple plays the mutual complement and transformation role. Unfortunately, other explanatory variables at the regional level were not significant; their coefficients were close to zero. This finding indicates that a certain disconnection exists between macro-economic indicators and micro-individual demand characteristics; this needs further verification and analysis. In this paper, the 2015 China general social survey (CGSS) was selected as the sample source for a robustness test. Question A18 asked, “What is your current household registration status?” Those who responded “agricultural household registration” were chosen as the sample of rural residents. Question D35 asked, “What was the main reason you left your last job?” Other questions included D6, “In your opinion, what is the main reason you have been treated unfairly?” and A54 “What’s

150

4 The Influencing Factors of the Imbalance of Rural Long Tail …

the reason you didn’t work last week?” Rural residents with (permanent) disabilities, health problems and an incapacity to work were selected as individuals with special needs in rural areas. Next, question B14 asked, “Do you think the government should or has the responsibility to provide the following benefits (a list was given, including elderly care)?” Individuals who chose absolutely responsible or may have responsibility were set as rural long tail samples with special elderly care needs. The dependent variables used to measure the imbalance of rural long tail elderly care services were satisfaction and subjective. Here, satisfaction measures the satisfaction of residents with rural long tail elderly care needs with the life security provided by the government. Question B15 asked, “Are you satisfied with the government’s performance in providing adequate living security for the elderly?” Satisfaction scores were assigned from a scale of 1 to 5 (very satisfied, satisfied, fair, dissatisfied, and very dissatisfied, respectively). Next, subjective measures the subjective score of the government’s provision of elderly care services. Question B16 asked, “What is your rating of the elderly care provided by the government?” The scores ranged from 0 to 100, according to the level of evaluation. Other measurement methods and descriptive statistics of explanatory variables at the micro-individual level and regional level are shown in Table 4.28 (Table 4.29). Ordinary single-layer Models (1) and (2) were first set as the reference control; multi-level nested Models (3) and (4) were set in consideration of the differential influence of explanatory variables at different levels. Among them, the dependent variables of Models (1) and (3) are satisfaction; the dependent variables of Models (2) and (4) are score. Considering that satisfaction is a multivariate ordered discrete variable, Model (1) adopts the ordered discrete model Ologit; Model (3) adopts the multi-layer ordered discrete model Meologit. Also, score is a continuous variable, so Model (2) is OLS, while Model (4) is linear regression mixed with a cross-layer mixing effect. Among them, the random intercept variance of Models (3) and (4) is significant at the 5% level, thereby proving the rationality of the cross-layer model. Other empirical regression results are as follows (Table 4.30): Table 4.28 List of all variables in rural long tail special care for the aged (II) Variable

Measurement

Question

Age

2015 minus birth year

A3. What’s your date of birth?

Married

Married set 1; unmarried, separated, divorced, widowed and other set 0

A69. What is your marital status now?

Work

Currently working set 1; currently not working set 0

A58. What is your work status?

Health

Relatively healthy or very healthy set 1; less healthy or very unhealthy set 0

A15. How do you feel about your current state of health?

Expense

Place province (area) government finance expenditure

China Statistical Yearbook

Source 2015 CGSS

4.5 Factors of Imbalance: Elderly Care

151

Table 4.29 Statistical description of variables in rural long tail elderly care (II) Variable

Obs

Mean

Std. dev

Min

Max

Satisfaction

711

2.476793

0.9430002

1

5

Score

711

68.90436

21.28656

0

100

Age

711

67.25176

12.79458

24

93

Work

711

0.161744

0.368475

0

1

Health

711

0.4824191

0.5000426

0

1

Married

711

0.6174402

0.4863543

0

1

NGO employee

776

195,643.4

144,917.3

627

632,390

Marketization

776

6.597848

1.698869

-0.3

10.92

Proexp

776

4450.753

2648.274

151.2

13,414.4

Propopu

776

5305.2

2773.818

277

10,999

Source 2015 CGSS Table 4.30 Impact factors of rural long tail special elderly care (II) (1)

(2)

(3)

(4)

Satisfaction

Score

Satisfaction

Score

−0.014***

0.297***

−0.014**

0.224***

(−2.583)

(4.733)

(−2.328)

(3.579)

Work

0.192

−5.942***

0.148

−5.245**

(0.984)

(−2.753)

(0.729)

(−2.560)

Health

−0.330**

−0.318

−0.484***

2.069

(−2.311)

(−0.203)

(−3.180)

(1.376)

Married

−0.135

1.497

−0.098

−0.667

(−0.916)

(0.925)

(−0.632)

(−0.426)

0.000

−0.000

0.000*

−0.000

(1.421)

(−1.467)

(1.725)

(−0.744)

Marketization

0.074

−0.218

−0.028

0.618

(1.233)

(−0.327)

(−0.301)

(0.577)

Expense

0.000

0.000

0.000

0.000

(0.189)

(0.628)

(0.050)

(0.691)

−0.000

0.002***

−0.000

0.001

(−1.463)

(3.267)

(−0.052)

(0.832)

Age

NGO employee

Population

0.366**

var cons[province])

(2.101) N Source 2015 CGSS

711

711

711

711

152

4 The Influencing Factors of the Imbalance of Rural Long Tail …

By observing the empirical results, one can find that the age of rural special needs individuals is significant in all models (at least 5% level). Models (1) and (3) were significantly negative (coefficient size 0.01), while Models (2) and (4) were significantly positive (coefficient size 0.3). This finding indicates that the satisfaction of rural long tail individuals with special elderly care needs with public elderly care services tends to increase in line with the increase in their age. These services then have higher subjective evaluation scores, thus reducing the subjective degree of the imbalance. Some scholars have pointed out that age has a weak negative impact on the demand for and willingness to avail of elderly care services (Zhang et al., 2020). In other words, the positive correlation between the age of individuals with special needs and the satisfaction with the long tail elderly care services is more on the demand side. This effectively alleviates the insufficient supply of elderly care service, by reducing the subjective intention of individuals with special needs. The reason behind this phenomenon is directly related to the depth and influence of traditional culture and the conservatism of the residents with special needs at different age levels. As the age of rural residents increases, and the stronger the degree of family dependence, the more elderly residents are inclined to (and prefer) the home-based elderly care lifestyle, even if the family conditions make it difficult to meet their individual heterogeneity preferences. In addition, whether individuals with special needs participate in work is positive in Models (1) and (3) (the coefficient is about 0.1, although not significant), and significantly negative in Models (2) and (4) (the coefficient is about 5). This indicates that job participation (including agricultural work and non-agricultural work) has a negative impact on the subjective evaluation of the imbalance of rural long tail elderly care public services. Compared with non-working rural residents, working rural residents are more likely to have the imbalance. The reason behind this finding may lie in the fact that rural residents, represented by farming, are often exposed to natural risks. For rural residents who have lost the ability to work, the income they required to maintain a normal life and production in the past will be especially affected by their own conditions. Some scholars have pointed out that, the longer farmers work in agriculture, the more they rely on agriculture as their income source, and the more urgent is their requirement for home-based care (Yeoh & Huang, 2009). Therefore, an increase in the level and degree of demand will further aggravate the imbalance of the long tail elderly care services. In addition, the explanatory variables at local level, including NGOempolyee, expense and population, still have small coefficients and tend to zero. On the other hand, marketization has a relatively large coefficient, meaning this variable has a great influence on individuals with micro-special needs (although the coefficient is not significant). The robustness test further illustrates the complexity and the multi-dimensional nature of rural long tail elderly care demand.

4.6 The Factors of Imbalance: Special Finance

153

4.6 The Factors of Imbalance: Special Finance Although rural financial products are mostly provided through market-oriented services, the regulation and access of the financial market represented by these products, as well as the corresponding infrastructure construction of rural financial products, are public and have a spillover effect. In particular, due to the impact of incomplete rural financial market and transaction costs, a rural financial market has the nature of a public service, meaning the intervention and participation of governments is required to reduce transaction costs, alleviate information asymmetry and promote the implementation of contracts (Owusu-Antwi & Antwi, 2010). Restricted by the level of rural economic development and traditional culture, more rural residents choose family savings to conduct financial management and to maintain value (Wang et al., 2014). Relatively speaking, the financial market represented by stocks, bonds, funds and other financial products has a relatively discrete and minority demand distribution for institutional norms and convenience of supporting facilities. This financial market belongs to the category of rural long tail public services. Some scholars have put forward that the coverage of financial outlets in rural areas of China is low, and the competition is insufficient. This makes meeting the service demands of farmers to the maximum extent difficult (Turvey & Kong, 2010). Therefore, how to better regulate and supply more reasonable and sound rural financial products, and how to provide high-quality supporting infrastructure construction is an issue of vital importance. In order to further analyze the influencing factors of the imbalance of rural long tail financial public services, the 2013 CHFS was chosen as the sample source, where question A2022 in the questionnaire asked, “What is your current household type?” Those residents who responded “agricultural household registration” were set as the sample of rural residents. Question D3101 asked, “Does your family hold stock accounts?” Question D4100a asked, “Does your family hold bonds?” Question D5102 asked, “Does your family hold funds?” Question D6102 asked, “Does your family own any financial derivatives?” The individuals who chose “No” for the last four questions were set as the sample of the imbalance of rural long tail financial services. The dependent variables used to measure the degree of imbalance are stock, bond, fund and derivative. The source question for stock was question D3102, which asked, “What is the reason why your family does not have stock accounts?” If the interviewee chose burdensome procedure as their answer, stock was set at 1; other answers were set at 0. The source question for bond was question D4100B, which asked, “What is the reason why your family did not buy bonds?” If the interviewee chose “burdensome procedure”, bond was set at 1; other answers were set at 0. The source question for fund was question D5102a, which asked, “What is the reason why your family did not buy funds?” If the interviewee chose “burdensome procedure” as their answer, fund was set at 1; other answers were set at 0. The source question for derivative was question D6102A, which asked, “What is the reason why your family does not buy financial derivatives?” If the interviewee chose “burdensome procedure” as

154

4 The Influencing Factors of the Imbalance of Rural Long Tail …

their answer, derivative was set at 1; other answers were set at 0. The measurement methods and descriptive statistics of other explanatory variables at individual and regional levels are shown in Tables 4.31 and 4.32. Firstly, this paper only considers the influencing factors at the level of individual demand. Since all the dependent variables of rural long tail financial public services are discrete binary variables, a logistic model was selected to construct stock, bond, fund and derivative, (1)–(4), respectively. The empirical results of all models are as follows (Table 4.33): By observing the empirical results, one can find that the education level of the long tail demand individuals is significantly positive in all models (the coefficient is 0.3, significant at the 1% level). This finding indicates that the educational level of individual demand is positively correlated with the imbalance of rural long tail financial services. In other words, the higher the educational level is, the greater the demand for rural long tail financial services will be, and the more inclined people with a higher education level are to believe that the supply of financial services cannot meet their demands. The reason behind this is that an improved education level can also improve the feasibility of individual wealth accumulation by increasing individual knowledge. Thus, on the one hand, these individuals could have the potential to demand more financial public services and thereby realize the preservation and appreciation of their wealth accumulation (Fungácová & Weill, 2015). On the other hand, rural demand individuals with higher education have a higher cognitive level and a comparative advantage in terms of obtaining information related to financial services (Brand & Xie, 2010). Such people can better identify the different types of long tail financial services that are suitable for their own conditions. Table. 4.31 List of all variables in rural long tail financial service (I) Variable

Measurement

Question

Education

Numbers 1–9 are, in order, never attended school, primary school, junior middle school, senior high school, technical secondary school/vocational high school, junior college/vocational high school, university undergraduate, master’s degree and doctor’s degree

[A2012] What is your education level?

Married

Married, set 1; unmarried, separated, divorced, or widowed, set 0

[A2024] What is your marital status?

Member

Number of household members

[A2000] Number of household members?

Age

2013 minus birth year

[A2005] Birth year

Work

Yes, set 1; no, set 0

[A3000] Are you currently employed, including in farming?

Health

Very good, good, generally good, set 1; very bad, not good, set 0

[A2025b] How is your health now?

Source 2013 CFPS

4.6 The Factors of Imbalance: Special Finance

155

Table 4.32 Statistical description of all variables in rural long tail financial service (I) Variable

Obs

Mean

Std. dev

Min

Max

Stock

28,687

0.0119218

0.108536

0

1

Bond

28,687

0.0104228

0.1015606

0

1

Fund

28,687

0.0092376

0.0956693

0

1

Derivative

28,687

0.0079827

0.0889902

0

1

Education

28,687

2.490222

1.125701

1

9

Age

28,684

50.61913

14.22514

17

113

Member

28,687

4.062886

1.762245

1

19

Health

28,687

0.3844947

0.4864841

0

1

Married

28,687

0.9059504

0.2919028

0

1

Work

28,686

1.268354

0.4431105

1

2

Marketization

28,713

7.03029

1.745628

-0.3

10.92

NGO employee

28,713

225,744.7

179,443.1

627

632,390

Propgdp

28,713

54,711.34

21,075.01

5222

115,053

Propopu

28,713

5413.029

2788.527

277

10,999

Source 2013 CFPS Table 4.33 Impact factors in the disequilibrium of rural long tail financial service (I) (1)

(2)

(3)

(4)

Stock

Bond

Fund

Derivative

Education

0.317***

0.370***

0.365***

0.348***

(7.141)

(7.934)

(7.361)

(6.585)

Age

−0.023***

−0.022***

−0.019***

−0.026***

(−4.656)

(−4.177)

(−3.524)

(−4.348)

0.282**

0.129

0.122

0.193

(2.260)

(0.995)

(0.891)

(1.297)

Member

−0.072**

0.003

−0.083**

−0.047

(−2.039)

(0.078)

(−2.062)

(−1.085)

Health

0.216*

0.104

0.080

0.208

(1.814)

(0.819)

(0.592)

(1.426)

_cons

−4.263***

−4.738***

−4.597***

−4.655***

(−11.712)

(−12.357)

(−11.212)

(−10.532)

N

28,684

28,684

28,684

28,684

Insurance

Source 2013 CFPS

156

4 The Influencing Factors of the Imbalance of Rural Long Tail …

In addition, age is significantly negative in all models (the coefficient size is 0.02, significant at the 1% level). This finding indicates that the age of those with individual demand is negatively correlated with the imbalance of rural long tail financial public services. The older the individual is, the smaller the demand for long tail financial services will be, and the lower the imbalance. The reason behind this is that, generally speaking, the older rural residents are, the more inclined they are to avoid financial risks (Glaser et al., 2014). They are more inclined to pursue the stability of wealth, which causes them to have less of a demand for financial planning and investment services (Sahi, 2013). Such differences in financial asset allocation preferences brought about by aging within families reflect the discrete and fragmented distribution of the long tail demand. In order to further analyze and verify the robustness of the model results, regionallevel explanatory variables (such as marketization, NGOemployee, GDP and population) were further added. Melogit (1)–(4), a cross-level logistic model, was built with stock, bond, fund and derivative as the dependent variables. The results of all models are as follows: Looking at Table 4.34, one can find that the coefficients of education and age at the individual level are still robust. Meanwhile, the coefficient of GDP at the regional level is significantly positive (though with a small coefficient) in most models. Consistent with relevant studies (Liu et al., 2021; Serrano-Cinca & Gutiérrez-Nieto, 2014), after considering explanatory variables at the regional level, the higher the level of regional economic development is, the higher the demand for rural long tail finance will be, and thus, the greater the degree of imbalance. The multiple levels of rural economic development between regions, as well as the internal and external imbalance, are what lead to the pluralism of the development of rural financial public services. However, the influence of this kind of macro-economic indicator on the micro-individual financial demand is indirect, resulting in a small coefficient. Similarly, one can infer that the influence mechanism of local marketization, NGO development and population level on the micro-individual long tail demand preference is also indirect, especially the interaction between different actors. This phenomenon requires further theoretical deduction.

4.7 Summary Based on different measurement methods and regulatory effects, this chapter regards the imbalance of different types of rural long tail public services as a part of the services’ historical evolution. The results play an important role in terms of creating a broader understanding of how to effectively supply rural long tail public services. A strong theoretical need exists to effectively define and distinguish rural long tail and head public services in an empirical design, especially considering the heterogeneity and fragmentation of long tail public needs. More and more researchers believe that in the supply of rural public services, a correlation does indeed exist between different types of suppliers. However, few

4.7 Summary

157

Table 4.34 Impact factors in the disequilibrium of rural long tail financial service (II) (1)

(2)

(3)

(4)

Stock

Bond

Fund

Derivative

Education

0.301***

0.344***

0.346***

0.331***

(6.577)

(7.145)

(6.764)

(6.049)

Age

−0.019***

−0.020***

−0.019***

−0.025***

(−3.964)

(−3.837)

(−3.360)

(−4.052)

Member

−0.025

0.043

−0.061

−0.013

(−0.677)

(1.136)

(−1.444)

(−0.281)

0.157

0.050

0.006

0.146

(1.297)

(0.385)

(0.043)

(0.984)

Married

−0.273

−0.171

0.047

0.102

(−1.606)

(−0.923)

(0.229)

(0.456)

Work

−0.034

0.113

0.049

0.031

(−0.253)

(0.816)

(0.330)

(0.192)

−0.045

0.021

0.114

0.081

(−0.426)

(0.214)

(1.112)

(0.718)

NGO employee

−0.000

−0.000

−0.000

−0.000*

(−0.455)

(−0.889)

(−1.290)

(−1.675)

GDP

0.000**

0.000**

0.000

0.000*

(1.995)

(2.142)

(1.366)

(1.880)

0.000

0.000

0.000

0.000

(1.095)

(0.932)

(1.518)

(1.512)

var (cons[province])

0.180*

0.091

0.066

0.084

(1.921)

(1.621)

(1.228)

(1.416)

N

28,683

28,683

28,683

28,683

Health

Marketization

Population

Source 2013 CFPS

studies have classified and refined rural public services according to their inherent characteristics; nor have they analyzed the impact of different types of expenditures on public services based on different providers. In fact, considering the dual nature of rural long tail public services, there is a possibility of realizing self-dynamic correction. Inspired by these problems, this chapter puts forward the functional theory of the dynamic correction and adjustment effect of the long tail and head of rural public services. In addition, the authors makes some theoretical and empirical contributions to the marginal effect of different types of expenditures, based on the differentiated incentive motivations of different stakeholders. However, like all studies, these empirical analyses still have many limitations that need to be addressed. First of all, as stated in the previous theory, the long tail nature of rural public services means that the inclusive supply and satisfaction of rural public services must not rely solely on the government; other social participants also have

158

4 The Influencing Factors of the Imbalance of Rural Long Tail …

an incentivized motivation to supply. Based on the “long tail aggregator” represented by NGOs, decentralized and minority rural public services can be satisfied through multiple customized services, and they can play a functional role. However, most of the empirical analyses in this chapter do not prove this path to any significant effect. Moreover, this chapter proposes that stakeholder behavior (as represented by the government) will not produce instantaneous utility. Rather, recognition and supply will dynamically lag, based on political pressure and the government’s aim to maximize social welfare. Therefore, the dynamic correction mechanism of rural long tail public services needs to be compared with the substitution effects among these complementary choices, rather than studying the mechanism and effects together. Second, this chapter only investigates the influencing factors of the imbalance of several representative rural long tail public services. However, other types of long tail services also have continuous diversified functions, either complementing or competing with the services listed in this chapter. From this perspective, this study is limited to the theorizing effect of different types of service imbalance in discrete situations. In the future, the function of continuity along the long tail curve should be examined from a life cycle perspective. Third, although this chapter supports the beneficial effects of different types of regional factors, the over-marketization of rural long tail public services is likely to exacerbate this imbalance in certain contexts. In addition, the crowding out effect among individual demands will also affect the imbalance of rural long tail public services. This chapter provides an examination of the distortion of this welfare effect. Future research should also further investigate the time effectiveness of this incentive failure, and design more effective incentive mechanisms based on different types of providers. Fourth, this chapter believes that the dynamic regulation mechanism of rural long tail public services may have a convergence trend. However, can the existence of a self-correction mechanism be proven simply by observing such a convergence trend? Further proof is needed, which could be provided by introducing more variables in the future. Moreover, although the time lag effect of different suppliers is emphasized, this chapter does not include the lag term of the dependent variable as the explanatory variable based on endogenous consideration. The potential side effects behind this mechanism are also not explored. This chapter measures the marginal effect of different individual demand characteristics and regional factors on alleviating the imbalance. In addition, more rapid, timely and efficient formal identification and the establishment of a supply mechanism is the direction of future research. By cooperating with other NGOs, the government can act as a “long tail aggregator”, forming a networked supply platform to share information and detect changes in real time. The next chapter analyzes the degree of imbalance of different types of rural long tail public services, from the perspective of empirical measurement.

References

159

References Abadie, A., Diamond, A., & Hainmueller, J. (2010). Synthetic control methods for comparative case studies: Estimating the effect of California’s tobacco control program. Journal of the American Statistical Association, 490(105), 493–505. Abadie, A., & Gardeazabal, J. (2003). The economic costs of conflict: A case study of the Basque Country. American Economic Review, 93(1), 113–132. Anderson, C. (2007). The long tail: How endless choice is creating unlimited demand. Random House. Baker, B. D., & Ramsey, M. J. (2010). What we don’t know can’t hurt us? Equity consequences of financing special education on the untested assumption of uniform needs. Journal of Education Finance, 245–275. Banks, J., Frawley, D., & McCoy, S. (2015). Achieving inclusion? Effective resourcing of students with special educational needs. International Journal of Inclusive Education, 19(9), 926–943. Blackwell, J. L. (2005). Estimation and testing of fixed-effect panel-data systems. The STATA journal, 5(2), 202–207. Brand, J. E., & Xie, Y. (2010). Who benefits most from college? Evidence for negative selection in heterogeneous economic returns to higher education. American Sociological Review, 75(2), 273–302. Brownell, M. T., Bishop, A. M., & Sindelar, P. T. (2005). NCLB and the demand for highly qualified teachers: Challenges and solutions for rural schools. Rural Special Education Quarterly, 24(1), 9–15. Buchanan, J. M. (1978). Cost and choice: an inquiry in economic theory. University of Chicago Press. Buchanan, J. M. (2014). Public finance in democratic process: Fiscal institutions and individual choice. UNC Press Books. Buchanan, J. M., & Musgrave, R. A. (1999). Public finance and public choice: two contrasting visions of the State. MIT press. Callaway, B., & Li, T. (2017). Quantile treatment effects in difference in differences models with panel data. Chen, Y., Ding, S., Xu, Z., Zheng, H., & Yang, S. (2019). Blockchain-based medical records secure storage and medical service framework. Journal of Medical Systems, 43(1), 1–9. Cheng, J. Y., Ngok, K., & Zhuang, W. (2010). The survival and development space for China’s labor NGOs: Informal politics and its uncertainty. Asian Survey, 50(6), 1082–1106. Collins, B. C., & Ludlow, B. L. (2018). Best practices for students with moderate and severe disabilities: a rural retrospective. Rural Special Education Quarterly, 37(2), 79–89. Conlin, M., & Jalilevand, M. (2015). Systemic inequities in special education financing. Journal of Education Finance, 41(1), 83–100. Cullen, J. B. (2003). The impact of fiscal incentives on student disability rates. Journal of Public Economics, 87(7–8), 1557–1589. Desai, M., Messer, L. B., & Calache, H. (2001). A study of the dental treatment needs of children with disabilities in Melbourne. Australia. Australian Dental Journal, 46(1), 41–50. Dhuey, E., & Lipscomb, S. (2013). Funding special education by total district enrollment: advantages, disadvantages, and policy considerations. Education Finance & Policy, 8(3), 316–331. DiMaggio, P. J., & Anheier, H. K. (1990). The sociology of nonprofit organizations and sectors. Annual Review of Sociology, 16(1), 137–159. Edmonds, B. C., & Spradlin, T. (2010). What does it take to become a high-performing special education planning district? A study of Indiana’s special education delivery service system. Remedial and Special Education, 31(5), 320–329. Fan, G., Wang, X., & Zhu, H. (2003). NERI index of marketization of China’s provinces. National Economic Research Institute.

160

4 The Influencing Factors of the Imbalance of Rural Long Tail …

Fan, G., Wang, X. L., Zhang, L. W., & Zhu, H. P. (2003). Report on the relative progress of marketization in various regions of China. Economic Studies, 3, 9–18. (in Chinese). Feng, Y. (2012). Teacher career motivation and professional development in special and inclusive education: Perspectives from Chinese teachers. International Journal of Inclusive Education, 16(3), 331–351. Feng, Z., Liu, C., Guan, X., & Mor, V. (2012). China’s rapidly aging population creates policy challenges in shaping a viable long-term care system. Health Affairs, 31(12), 2764–2773. Franck, B., & Joshi, D. K. (2017). Including students with disabilities in education for all: Lessons from Ethiopia. International Journal of Inclusive Education, 21(4), 347–360. Fungácová, Z., & Weill, L. (2015). Understanding financial inclusion in China. China Economic Review, 34, 196–206. Glaser, J., Kuwayama, D., Stone, D., Schanzer, A., Eldrup-Jorgensen, J., Powell, R., … & Nolan, B. (2014). Factors that determine the length of stay after carotid endarterectomy represent opportunities to avoid financial losses. Journal of Vascular Surgery, 60(4), 966–972. Han, K. (2020). Development of China’s elderly welfare in the transitional period. In Social welfare in transitional China (pp. 187–219). Palgrave Macmillan. Horsfall, D., Noonan, K., & Leonard, R. (2012). Bringing our dying home: How caring for someone at end of life builds social capital and develops compassionate communities. Health Sociology Review, 21(4), 373–382. Imbens, G. W., & Wooldridge, J. M. (2009). Recent developments in the econometrics of program evaluation. Journal of Economic Literature, 47(1), 5–86. Johansson, S., Leonard, R., & Noonan, K. (2012). Caring and the generation of social capital: Two models for a positive relationship. International Journal of Social Welfare, 21(1), 44–52. Jung, H. S., & Thorbecke, E. (2003). The impact of public education expenditure on human capital, growth, and poverty in Tanzania and Zambia: A general equilibrium approach. Journal of Policy Modeling, 25(8), 701–725. Keimig, R. K. (2021). Growing old in a New China: Transitions in elder care. Rutgers University Press. Kim, Y. S., Joo, H. J., & Lee, S. (2018). School factors related to high school dropout. KEDI Journal of Educational Policy, 15(1), 59–80. Koutrouba, K., Vamvakari, M., & Steliou, M. (2006). Factors correlated with teachers’ attitudes towards the inclusion of students with special educational needs in Cyprus. European Journal of Special Needs Education, 21(4), 381–394. Krug, G., & Eberl, A. (2018). What explains the negative effect of unemployment on health? An analysis accounting for reverse causality. Research in Social Stratification and Mobility, 55, 25–39. Li, N., Pang, L., Du, W., Chen, G., & Zheng, X. (2012). Association between poverty and psychiatric disability among Chinese population aged 15–64 years. Psychiatry Research, 200(2–3), 917–920. Li, S. K., & He, X. (2019). The impacts of marketization and subsidies on the treatment quality performance of the Chinese hospitals sector. China Economic Review, 54, 41–50. Liu, Y., Ji, D., Zhang, L., An, J., & Sun, W. (2021). Rural financial development impacts on agricultural technology innovation: Evidence from China. International Journal of Environmental Research and Public Health, 18(3), 1110. Lucas, A. M., & Mbiti, I. M. (2012). Access, sorting, and achievement: The short-run effects of free primary education in Kenya. American Economic Journal: Applied Economics, 4(4), 226–253. Lucas, H. (2008). Information and communications technology for future health systems in developing countries. Social Science & Medicine, 66(10), 2122–2132. Ma, Z., Xue, Y., & Hu, G. (2019). Nonparametric analysis of income distributions among different regions based on energy distance with applications to China Health and Nutrition Survey data. Communications for Statistical Applications and Methods, 26(1), 57–67. Mahitivanichcha, K., & Parrish, T. (2005). The implications of fiscal incentives on identification rates and placement in special education: Formulas for influencing best practice. Journal of Education Finance, 1–22.

References

161

Marchenko, Y. V. (2005). Estimating variance components in Stata. State Journal, 6(1), 1–21. Mason-Williams, L. (2015). Unequal opportunities: A profile of the distribution of special education teachers. Exceptional Children, 81, 247–262. Meyers, A. B., Tobin, R. M., Huber, B. J., Conway, D. E., & Shelvin, K. H. (2015). Interdisciplinary collaboration supporting social-emotional learning in rural school systems. Journal of Educational & Psychological Consultation, 25(2–3), 109–128. Milligan, C. (2012). There’s no place like home: Place and care in an ageing society. Owusu-Antwi, G., & Antwi, J. (2010). The analysis of the rural credit market in Ghana. International Business & Economics Research Journal (IBER), 9(8). Pulkkinen, J., & Jahnukainen, M. (2016). Finnish reform of the funding and provision of special education: The views of principals and municipal education administrators. Educational Review, 68(2), 171–188. Sahi, S. K. (2013). Demographic and socio-economic determinants of financial satisfaction. International Journal of Social Economics. Schirmer, B. R., & McGough, S. M. (2005). Teaching reading to children who are deaf: Do the conclusions of the National Reading Panel apply? Review of Educational Research, 75(1), 83–117. Serrano-Cinca, C., & Gutiérrez-Nieto, B. (2014). Microfinance, the long tail and mission drift. International Business Review, 23(1), 181–194. Shields, N., Synnot, A. J., & Barr, M. (2012). Perceived barriers and facilitators to physical activity for children with disability: A systematic review. British Journal of Sports Medicine, 46(14), 989–997. Sindelar, P. T., Pua, D. J., Fisher, T., Peyton, D. J., Brownell, M. T., & Mason-Williams, L. (2018). The demand for special education teachers in rural schools revisited: An update on progress. Rural Special Education Quarterly, 37(1), 12–20. Smit, K., de Brabander, C. J., & Martens, R. L. (2014). Student-centred and teacher-centred learning environment in pre-vocational secondary education: Psychological needs, and motivation. Scandinavian Journal of Educational Research, 58(6), 695–712. Tanaka, S. (2015). Environmental regulations on air pollution in China and their impact on infant mortality. Journal of Health Economics, 42, 90–103. Temple, J. (1999). The new growth evidence. Journal of Economic Literature, 1, 112–156. Thornton, P. M. (2013). The advance of the party: Transformation or takeover of urban grassroots society? The China Quarterly, 1–18. Tilak, J. B. (2002). Determinants of household expenditure on education in rural India (No. 88). National Council of Applied Economic Research. Tompkins, R. B. (2006). The challenges and opportunities facing rural America: Finding answers in our public schools. Book presented at the Research and Policy Workshop of the Economic Research Service of the U.S. Department of Agriculture. Washington, D.C. Tran, K. V. (2014). Exploring the experience of children with disabilities at school settings in Vietnam context. Springer plus, 3(1), 103. Turvey, C. G., & Kong, R. (2010). Informal lending amongst friends and relatives: Can microcredit compete in rural China? China Economic Review, 21(4), 544–556. Wang, L. (2018). Study on the modern senior care service and security system. In The development of security and whole care system for the aged in China (pp. 57–125). Springer. Wang, P., Liu, Q., & Qi, Y. (2014). Factors influencing sustainable consumption behaviors: A survey of the rural residents in China. Journal of Cleaner Production, 63, 152–165. White III, S. (2018). Aging gracefully: Spiritual care for aging adults. WestBow Press. Wong, Y. C., & Leung, J. (2012). Long-term care in China: Issues and prospects. Journal of Gerontological Social Work, 55(7), 570–586. Yang, Z., & Tian, X. (2009). The transition of state-peasants relationship: From the fiscal perspective in three decades of reform in China. China Agricultural Economic Review, 1(4), 382–394. Yeoh, B. S., & Huang, S. (2009). Foreign domestic workers and home-based care for elders in Singapore. Journal of Aging & Social Policy, 22(1), 69–88.

162

4 The Influencing Factors of the Imbalance of Rural Long Tail …

Yin, Z., Kang, C., Wang, L., Geng, D., & Xiong, Z. (2017). Public security expenditure, education investment, and social stability: An empirical analysis based on provincial panel data from China. Revista De Cercetare Si Interventie Sociala, 59, 239. Zhang, L., Zeng, Y., Wang, L., & Fang, Y. (2020). Urban-rural differences in long-term care service status and needs among home-based elderly people in China. International Journal of Environmental Research and Public Health, 17(5), 1701. Zhang, Y., & Qiu, Z. (2015). The Analysis of the risks faced by China’s social endowment insurance under the background of rapid aging population. Open Journal of Business and Management, 3(02), 185. Zheng, X., Chen, G., Song, X., Liu, J., Yan, L., Du, W., & Zhang, J. (2011). Twenty-year trends in the prevalence of disability in China. Bulletin of the World Health Organization, 89, 788–797. Zhu, H. (2015). Unmet needs in long-term care and their associated factors among the oldest old in China. BMC Geriatrics, 15(1), 1–11.

Chapter 5

The Measurement of the Imbalance of Rural Long Tail Public Services

Abstract As mentioned above, the main reason for the imbalance of rural long tail public services is determined by both supply and demand. Especially from the perspective of the nature of the rural long tail public service itself, its discrete and fragmented characteristics, to a large extent, restrict the satisfaction of different providers for the rural long tail public demand. Therefore, this chapter starts with the degree of decentralization and fragmentation of rural long tail public services in different fields, and empirically measures the degree of imbalance. By measuring the imbalance of basic public services from the perspective of substitution of supply or demand, the underlying assumption is that demand can be uniformly satisfied through large-scale supply, while ignoring the special properties of fragmentation and atomization of heterogeneous demand. Relying on the property of rural long tail public services, this chapter calculates the supply–demand imbalance index from the degree of discretization in different fields. This chapter use the geographical distance of different types of rural long tail public services to measure the concentration degree of supply and demand. Specifically, the average distance between different types of public services and the central location (such as the urban geographic center) is calculated by Baidu Map to measure the concentration degree of public services. This measurement method can analyze the development process of diachronic geographic location, spatial scale and demand (supply) expansion or contraction from the perspective of spatial accumulation elements. The following is a detailed measurement from different fields which include rural special education, rural special health, rural special elderly care, rural special finance. Generally speaking, this chapter calculates the imbalance index of rural long tail public services in different areas of different prefecture level cities from the empirical perspective. It finds that there are great differences in the imbalance of long tail public services in different rural areas. It also synthesizes the imbalance indexes of various fields and sums them up to calculate the weighted average value, and constructs the comprehensive imbalance index of rural long tail public services.

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 J. Luo, Rural Long Tail Public Service and the Correction Mechanism, https://doi.org/10.1007/978-981-16-4023-0_5

163

164

5 The Measurement of the Imbalance of Rural Long Tail Public …

5.1 Measurement Method There are few researches on the measurement of the imbalance of rural public services. Some scholars calculate the unbalance between supply and demand of rural social endowment insurance: there is a huge gap between the target replacement rate and the demand replacement rate of rural residents’ pension (Cai & Cheng, 2014). But more studies have measured the extent of the imbalance in terms of residents’ satisfaction, supply efficiency or cost-output ratios. By measuring the imbalance of basic public services from the perspective of substitution of supply or demand, the underlying assumption is that demand can be uniformly satisfied through large-scale supply, while ignoring the special properties of fragmentation and atomization of heterogeneous demand. Relying on the property of rural long tail public services, this chapter calculates the supply–demand imbalance index from the degree of discretization in different fields. The specific calculation formula is as follows: U Ert =

Nrt SCrt Prt DCrt

U Ert represents the rural long tail public service imbalance index in R region in T period. SCrt represents the degree of decentralization of rural long tail public service supply in R region in T period. DCrt represents the degree of decentralization of rural long tail public service demand in R region in T period. Nrt is the number of supply institutions. Prt represents the number of demand people. Nrt is the weight of the imbalance degree, and represents the number of long tail Prt supply institutions per capita for people with special needs. The larger U Ert is, the discretization degree of supply of rural long tail public services is significantly greater than that of demand. The distribution domain of demand is a subset of the distribution domain of supply, and the supply can meet the demand within a reasonable space and time range, and the less unbalanced degree of supply and demand is. The smaller U Ert is, the discretization degree of supply of rural long tail public services is significantly less than the discrete degree of demand. This means that the distribution domain of supply is a subset of the distribution domain of demand: part of the long tail demand is in a more discrete space outside the supply feasible set, and the imbalance is more serious. This chapter use the geographical distance of different types of rural long tail public services to measure the concentration degree of supply and demand. Specifically, the average distance between different types of public services and the central location (such as the urban geographic center) is calculated by Baidu Map to measure the concentration degree of public services in R region in t period. This measurement method can analyze the development process of diachronic geographic location, spatial scale and demand (supply) expansion or contraction from the perspective

5.1 Measurement Method

165

of spatial accumulation elements. The following is a detailed measurement from different fields.

5.2 Measurement of Imbalance Index of Rural Long Tail Public Services: Rural Special Education The discrete index of rural long tail special education demand comes from China Labour Force Dynamics Survey 2016 (CLDS 2016). The latest data covers more than 100 prefecture-level administrative units in 29 provinces and municipalities in China. The survey targets the entire labour force (family members aged 15–64) in the sample households. In terms of sampling method, a multi-level probability sampling method proportional to the size of the labour force is adopted. In the way of tracking survey, rotation sample tracking method is adopted, which can not only better adapt to the drastic changes in China, but also take into account the characteristics of cross-sectional survey and tracking survey. Firstly, the community types in the questionnaire were screened, and the sample of rural community was retained as the rural sample. Secondly, according to the district-divided questionnaire R 2.6, the number of disabled persons in the village committee is weighted as the number of special educational needs in the village. According to the questionnaire R 58.1. The distance between the village and the nearest county seat/district government? Take each prefecture-level administrative unit as the unit, and the straight-line average distance between the village and the nearest county seat/district government of each individual sample is obtained as the distance of special education needs (km). We multiply the distance of special educational needs by the number of special needs, and the discrete index of special education needs in each prefecture-level city is obtained (see Table 5.1). According to the analysis of the above table, cities with large distance between special education needs are mainly distributed in third-tier and fourth-tier cities in central and western China (such as Pu’er, Honghe, Xiangtan, Xuzhou, Laibin, Baise, Hengyang, Diqing, Yan’an, Zhongwei, etc.). In these cities, due to geographical factors such as topography and landforms, the distribution of different villages is relatively dispersed. Or the population distribution of different villages is relatively dispersed due to the more even population distribution (“urban sprawl” development). From the perspective of the sample number of disabled people, the prefecture-level cities with a large number (such as Hefei, Luzhou, Dalian, Huanggang, etc.) also ranked the top in the country in terms of their total population in 2016 (the 2017 China City Statistical Yearbook). From the perspective of the dispersion index distribution of special education needs, the cities with high dispersion degree are also dominated by these cities (the top 10 are Pu’er, Hefei, Honghe, Baise, Panzhihua, Xiangtan, Laibin, Zhongwei, Fuzhou, and Maoming). The dispersion index of rural long tail special education supply is calculated from Baidu Map. First of all, search the keywords of “special education” and

166 Table 5.1 The demand discrete index of rural long tail special education

5 The Measurement of the Imbalance of Rural Long Tail Public … City

Average distance of special educational needs (km)

Sample number of disabled persons (people)

Discrete index of special education needs

Tianjin

60

30

1800

Tangshan

20.21

91

1839.36

Baoding

3

8

24

Zhangjiakou

4.60

93

427.80

Datong

30

31

930

Yuncheng

5

62

310

Linfen

20

11

220

Ulanqab

20

112

2240

Xilingol

21

50

1050

Dalian

20

150

3000

Anshan

10

105

1050

Panjin

15

44

660

Jilin

31.55

45

1419.82

Harbin

38.35

50

1917.68

Nanjing

8.31

42

349.33

Xuzhou

50

55

2750

Changzhou

8.30

168

1394.40

Suqian

25

55

1375

Ningbo

1

62

62

Wenzhou

30.16

50

1508.17

Jiaxing

18.07

136

2458.84

Taizhou

25

54

1350

Hefei

25.53

585

14,937.77

Wuhu

17

50

850

Lv’an

27.05

35

946.88

Bozhou

36

40

1440

Xuancheng

39

33

1287

Fuzhou

50

72

3600

Quanzhou

27.60

45

1242.18

Ganzhou

36.57

9

329.13

Yichun

10.35

7

72.46

Weifang

25.32

4

101.29

Jining

19.39

14

271.52

Taian

9.40

8.5

79.98 (continued)

5.2 Measurement of Imbalance Index of Rural Long Tail Public … Table 5.1 (continued)

City

Average distance of special educational needs (km)

Sample number of disabled persons (people)

167 Discrete index of special education needs

Linyi

6.32

30

189.61

Zhengzhou

20

17

340

Kaifen

22.39

30

671.90

Luoyang

33.09

50

1654.76

Xuchang

10.99

14.5

159.42

Shangqiu

25

33

825

Xinyang

17.79

22.5

400.37

Xiaogan

22

9

198

Huanggang

16.39

145

2377.68

Xianning

26.01

50

1300.94

Xiangtan

73.22

68

4978.98

Hengyang

40.38

3

121.15

Foshan

20.60

108

2224.99

Jiangmen

13.00

47.5

617.64

Maoming

35

100

3500

Zhaoqin

27

50

1350

Huizhou

16.25

29.67

482.15

Heyuan

11.61

129

1498.67

Yangjiang

22.63

78

1765.83

Chaozhou

1.78

4

7.15

Jieyang

21.95

69.33

1522.55

Baise

43.33

59.67

7757.09

Laibin

50

95

4750

Chongzuo

21.33

80

1707.05

Chongqin

28.88

19

548.88

Zigong

32

24

768

Panzhihua

34.59

149

5155.17

Luzhou

5

260

1300

Nanchong

15.59

38

592.74

Meishan

12

124

1488

Ziyang

20.02

53

1061.57

Liangshan

27

6

162

Baoshan

30

30

900

Pu’er

90.52

176

15,932.63 (continued)

168 Table 5.1 (continued)

5 The Measurement of the Imbalance of Rural Long Tail Public … City

Average distance of special educational needs (km)

Sample number of disabled persons (people)

Discrete index of special education needs

Honghe

76

123

9348

Diqing

40

13

520

Xi’an

22.65

64

1450.06

Baoji

12.85

102.67

1319.74

Yan’an

40

12

480

Zhangye

21.56

34

733.07

Longnan

34.32

34.50

1184.27

Zhongwei

38.58

108.50

4186.53

Source CLDS2016. Two decimal points are retained, and other prefecture-level cities are ignored due to lack of samples

other rural special education institutions through Baidu map, and locate these rural special education institutions in different regions (prefecture-level administrative units) through administrative divisions (located in town, village or suburb). At the same time, manually calculate the distance between their registered service address and the geographical center of prefecture-level administrative units. The average geographical distance and distance variance are calculated as the value of supply concentration. The concentration degree of rural special education institutions in different prefecture level cities where the demand is located is shown in the table below (the basis for selecting prefecture level cities is to match the demand, ignoring the data of other prefecture level cities) (Table 5.2): The prefecture level cities with a relatively large average distance from rural special education supply are mainly distributed in the areas with large land occupation and large population in Southwest China (including Diqing, Chongqing, Zigong, Baoshan, Honghe, etc.). From the perspective of the supply of rural special education, the top cities are Linyi, Luoyang, Fuzhou, Ganzhou, Jining, Yuncheng, etc. It can be found that these prefecture level cities have more counties under their jurisdiction. Chinese government stipulates that at least special education schools in counties with a population of more than 300,000, which is also the reason for the large number of schools in these areas. However, there are still many prefecture level cities that lack the supply data of rural special education (such as Datong, Ulanchab, xilingol, Panjin, Wuhu, Xuancheng, etc.), and these cities choose to establish special education schools in urban areas. From the perspective of the dispersion index of special education supply, the variance is still high, and the highest-ranking city is 60 times higher than the second. The cities ranking the top are mainly distributed in the areas with large population and area (such as Linyi, Luoyang, Fuzhou, Ganzhou, Zigong, Honghe, Baise, Nanjing, Suqian, Nanchong, Jining, etc.).

5.2 Measurement of Imbalance Index of Rural Long Tail Public … Table 5.2 The supply dispersion index of rural long tail special education

City

Average distance of special education supply (km)

Number of special education supplies

169 Dispersion index of special education supply

Tianjin

19.1

1

19.1

Tangshan

31.67

3

95.01

Baoding

47.3

1

47.3

Zhangjiakou

42.73

3

128.19

Datong

0

0

0

Yuncheng

40.12

5

200.6

Linfen

32.9

3

98.7

Ulanqab

0

0

0

Xilingol

0

0

0

Dalian

0

0

0

Anshan

66.8

1

66.8

Panjin

0

0

0

Jilin

24.2

1

24.2

Harbin

85.7

2

171.4

Nanjing

83.97

3

251.91

Xuzhou

73.2

1

73.2

Changzhou

43.9

2

87.8

Suqian

58.5

4

234

Ningbo

0

0

0

Wenzhou

18.93

3

56.79

Jiaxing

24.95

2

49.9

Taizhou

44.7

2

89.4

Hefei

49.2

2

98.4

Wuhu

0

0

0

Lv’an

61

2

122

Bozhou

57.6

1

57.6

Xuancheng

0

0

0

Fuzhou

65.72

6

394.32

Quanzhou

54.7

2

109.4

Ganzhou

57.73

6

346.38

Yichun

61.2

2

122.4

Weifang

67

1

67

Jining

44.76

5

223.8

Taian

72.3

1

72.3

Linyi

62.44

8

499.52 (continued)

170 Table 5.2 (continued)

5 The Measurement of the Imbalance of Rural Long Tail Public … City

Average distance of special education supply (km)

Number of special education supplies

Dispersion index of special education supply

Zhengzhou

39.5

1

39.5

Kaifen

45.85

2

91.7

Luoyang

60.05

8

480.4

Xuchang

0

0

0

Shangqiu

51.23

4

204.92

Xinyang

42.7

1

42.7

Xiaogan

39.95

2

79.9

Huanggang

0

0

0

Xianning

37

1

37

Xiangtan

10.2

1

10.2

Hengyang

21.4

1

21.4

Foshan

30

1

30

Jiangmen

57

1

57

Maoming

23.2

1

23.2

Zhaoqin

91.4

1

91.4

Huizhou

52.2

2

104.4

Heyuan

54.95

2

109.9

Yangjiang

0

0

0

Chaozhou

0

0

0

Jieyang

58.97

3

176.91

Baise

64.23

4

256.92

Laibin

0

0

0

Chongzuo

0

0

0

Chongqin

118.9

1

118.9

Zigong

110.67

3

332.01

Panzhihua

0

0

0

Luzhou

16.3

1

16.3

Nanchong

55.95

4

223.8

Meishan

41.8

3

125.4

Ziyang

52

1

52

Liangshan

8.3

1

8.3

Baoshan

109.45

2

218.9

Pu’er

0

0

0

Honghe

104.43

3

313.29

Diqing

129.2

1

129.2 (continued)

5.2 Measurement of Imbalance Index of Rural Long Tail Public … Table 5.2 (continued)

City

Average distance of special education supply (km)

Xi’an Baoji

171

Number of special education supplies

Dispersion index of special education supply

65

3

195

34.7

4

138.8

Yan’an

0

0

0

Zhangye

0

0

0

Longnan

0

0

0

Zhongwei

0

0

0

Source Baidu map, with two decimal places. Other prefecture level cities are ignored due to the lack of samples

Through the supply and demand dispersion index of rural special education, the imbalance index (divided by both) of rural special education can be calculated as follows (sorted by the size of imbalance index). It finds that the imbalance index of supply and demand of special education is not as different among different cities as imagined, and the variance is small. Among them, Linyi, the highest ranking City, is more than 200 times higher than Hefei. And many prefecture level cities have index value of 0 (supply is 0). The average value of the imbalance index of all prefecture level cities is 0.22, which proves that the imbalance is relatively serious nationwide (the smaller the value is, the higher the imbalance degree). Generally speaking, cities with small imbalance mainly distributed in cities with large population and area (Linyi, Baoding, Yichun, Ganzhou, Tai’an, Jining, Nanjing, Weifang, etc.), among which Shandong province is in the leading position in the supply and demand matching of special education. Shandong Special Education Promotion Plan (2014–2016) determined that by 2016, the compulsory education enrolment rate of children with special needs reached 95%, and the rate of children with disabilities entering kindergartens and receiving rehabilitation education in three years before school reaches 90% (Hua, 2014). Shandong Province, as the province with the largest population in China, has a population of more than 300,000 in most counties. Therefore, it can basically achieve full coverage of county-level special education schools, which can effectively alleviate the dispersion of rural special education needs (Table 5.3).

5.3 Measurement of Imbalance Index of Rural Long Tail Public Services: Rural Special Health The dispersion index of rural long tail special health service demand also comes from CLDS 2016. Firstly, the community types in the questionnaire were selected, and

172

5 The Measurement of the Imbalance of Rural Long Tail Public …

Table 5.3 The supply–demand imbalance index of rural long tail special education

City

Imbalance Index City

Imbalance Index

Linyi

2.63

Changzhou 0.06

Baoding

1.97

Tangshan

0.05

Yichun

1.69

Liangshan

0.05

Ganzhou

1.05

Ziyang

0.05

Taian

0.90

Bozhou

0.04

Jining

0.82

Wenzhou

0.04

Nanjing

0.72

Honghe

0.03

Weifang

0.66

Baise

0.03

Yuncheng

0.65

Xianning

0.03

Linfen

0.45

Xuzhou

0.03

Zigong

0.43

Jiaxing

0.02

Xiaogan

0.40

Jilin

0.02

Nanchong

0.38

Foshan

0.01

Zhangjiakou 0.30

Luzhou

0.01

Luoyang

0.29

Tianjin

0.01

Diqing

0.25

Maoming

0.01

Shangqiu

0.25

Hefei

0.01

Baoshan

0.24

Xiangtan

0.00

Chongqing

0.22

Datong

0.00

Huizhou

0.22

Ulanqab

0.00

Hengyang

0.18

Xilingol

0.00

Suqian

0.17

Dalian

0.00

Kaifen

0.14

Panjin

0.00

Xi’an

0.13

Ningbo

0.00

Lvan

0.13

Wuhu

0.00

Jieyang

0.12

Xuancheng 0.00

Zhengzhou

0.12

Xuchang

Fuzhou

0.11

Huanggang 0.00

Xinyang

0.11

Yangjiang

0.00

0.00

Baoji

0.11

Chaozhou

0.00

Jiangmen

0.09

Laibin

0.00

Harbin

0.09

Chongzuo

0.00

Quanzhou

0.09

Panzhihua

0.00

Meishan

0.08

Pu’er

0.00

Heyuan

0.07

Yan’an

0.00

Zhaoqing

0.07

Zhangye

0.00

Taizhou

0.07

Longnan

0.00 (continued)

5.3 Measurement of Imbalance Index of Rural Long Tail Public … Table 5.3 (continued)

173

City

Imbalance Index City

Imbalance Index

Anshan

0.06

0.00

Zhongwei

Source Baidu map, with two decimal places. Other prefecture level cities are ignored due to the lack of samples

the samples of rural community were retained as rural samples. Then through the community questionnaire C2_7, The number of mental patients in villages was taken as the sample demand of rural special medical treatment. Thirdly, the linear average distance between the township of each individual sample and the county of the local city (R58.1 distance between the village and the nearest county/district government) was calculated as the distance of special medical needs (km). We multiply the distance of special medical needs by the number of needs, the dispersion index of special medical needs of each prefecture level city is obtained (see Table 5.4). From the perspective of special medical demand distance, the top cities (with large demand distance) are still mainly the third and fourth tier cities in the central and western regions (Yichang, Sipin, Pu’er, Xiangxi, Honghe, Xiangtan, Kashgar, Dingxi, etc.), and there are some minority areas. However, the cities with the highest number of mental patients in the sample are not all the population gathering areas, but scattered in Guangdong, northeast and northwest (Pingxiang, Zhongshan, Zhuhai, Qiqihar, Wuwei, Zhaoqing, Yangjiang, Foshan, etc.). Correspondingly, the cities with higher dispersion index of special medical needs are also dominated by these cities (Pingxiang, Pu’er, Qiqihar, Yichang, Xiangxi, Zhaoqing, Fuzhou, Yangjiang, etc.). The dispersion index of rural long tail special medical supply comes from Baidu map. First of all, through Baidu map search keywords such as “special medical care”, “mental health care”, and through administrative divisions (located in town, village or suburb), these rural special medical institutions in different regions (prefecture level administrative units) are located. At the same time, we manually calculate the distance between their registered service address and the geographical center of prefecture level administrative units. The average geographical distance and distance variance are calculated as the value of supply concentration. The results show that the concentration degree of rural special medical institutions in different prefecture level cities where the demand is located is shown in the table below (the basis of selecting prefecture level cities is to match the demand, ignoring the data of other prefecture level cities) (Table 5.5): From the ranking of the special medical supply distance in the above table, the cities with large geographical level are mainly the cities with large population and large land occupation (such as Chongqing, Harbin, Tongren, Baise, Baoshan, Zhangjiakou, Yan’an, Lincang and Qiqihar). However, a large proportion of prefecture level cities lack supply (distance is 0), and there is no corresponding special medical institutions in rural areas. The logic behind this is that the high-risk population with rare diseases and mental diseases is a special group that needs to be considered in the overall medical insurance system for rural major diseases in China. However, there are still many gaps in the scope of compensation for special diseases,

174

5 The Measurement of the Imbalance of Rural Long Tail Public …

Table 5.4 The demand dispersion index of rural long tail special health service

City

Distance of Number of Dispersion special mental patients medical medical needs demand index (km)

Beijing

2

8

16

Tianjin

31

1.5

46.5

Tangshan

20.21

3.67

74.1707

Handan

6

2

12

Xingtai

3.93

5.5

21.615

Zhangjiakou 4.6

5

23

Datong

30

2

60

Jincheng

30

2

60

Yuncheng

5

2

10

Linfen

20

3.5

70

Tongliao

17

2

34

Ulanqab

20

6

120

Xilingol

21

7

147

Shenyang

15

2

30

Dalian

20

3

60

Anshan

10

10

100

Fushun

20

4

80

Jinzhou

6

0

0

Liaoyang

20

1.5

30

Panjin

15

7

105

Jilin

31.55

7

220.85

Sipin

100

2

200

Songyuan

10

0

0

Harbin

38.35

3

115.05

Qiqihar

36.2

20

724

Nanjing

8.31

3

24.93

Xuzhou

50

7

350

Changzhou

8.3

8

66.4

Yancheng

78

4

312

Yangzhou

15.59

1.5

23.385

Suqian

25

8

200

Hangzhou

7.23

10

72.3

Ningbo

1

10

10

Wenzhou

30.16

7

211.12

Jiaxing

18.07

19.5

352.365 (continued)

5.3 Measurement of Imbalance Index of Rural Long Tail Public … Table 5.4 (continued)

175

City

Distance of Number of Dispersion special mental patients medical medical needs demand index (km)

Taizhou

25

4

100

Hefei

25.53

9.5

242.535

Wuhu

17

6

102

Fuyang

35

16

560

Lvan

27.05

4.67

126.3235

Bozhou

36

16

576

Xuancheng

39

3

117

Fuzhou

50

8

400

Sanming

3

3

9

Quanzhou

27.6

7.5

207

Nanpin

32

7

224

Longyan

30

1

30

Pinxiang

40

30

1200

Ganzhou

36.57

1.5

54.855

Yichun

10.35

3.33

34.4655

Jinan

14.06

1.5

21.09

Qingdao

30.18

5

150.9

Weifang

25.32

2.5

63.3

Jining

19.39

6

116.34

Tai’an

9.4

4.33

40.702

Linyi

6.32

4.5

28.44 153

Dezhou

9

17

Zhengzhou

20

6

120

Kaifen

22.39

1.5

33.585

Luoyang

33.09

4

132.36

Pindingshan

20

2

40

Xuchang

10.99

4.67

51.3233

Nanyang

25

5

125

Shangqiu

25

4

100

Xinyang

17.79

8.5

151.215

Zhoukou

18

2

36

Yichang

120

5

600

Jinmen

50

3

150

Xiaogan

22

1

22

Huanggang

16.39

9.67

158.4913 (continued)

176 Table 5.4 (continued)

5 The Measurement of the Imbalance of Rural Long Tail Public … City

Distance of Number of Dispersion special mental patients medical medical needs demand index (km)

Xianning

26.01

Xiangtan

73.22

2.5

183.05

Hengyang

40.38

1.67

67.4346

Loudi

40.67

0.5

20.335

Xiangxi

84.87

6.67

566.0829

Zhuhai

10

21

210

Foshan

20.6

17

350.2

Jiangmen

13

5.75

74.75

Maoming

35

5

175

Zhaoqing

27

17

459

Huizhou

16.25

7

113.75

Heyuan

11.61

12.5

145.125

Yangjiang

22.63

17

384.71

Zhongshan

14.72

23

338.56

Chaozhou

1.78

6.5

11.57

Jieyang

21.95

12.67

278.1065

Guigang

8.77

8.5

74.545

Baise

43.33

5.75

249.1475

Laibin

50

4

200

Chongzuo

21.33

4.33

92.3589

Chongqing

28.88

7.67

221.5096

8

208.08

Zigong

32

2

64

Panzhihua

34.59

11

380.49

Luzhou

5

7

35

Nanchong

15.59

5.5

85.745

Meishan

12

12

144

Ziyang

20.02

3

60.06

Liangshan

27

0

0

Zunyi

34.2

7.5

256.5

Tongren

27.73

6

166.38

Yuxi

33

4

132

Baoshan

30

3

90

Pu’er

90.52

12

1086.24

Lincang

11

4

44

Honghe

76

2

152 (continued)

5.3 Measurement of Imbalance Index of Rural Long Tail Public … Table 5.4 (continued)

177

City

Distance of Number of Dispersion special mental patients medical medical needs demand index (km)

Diqing

40

0

Xi’an

22.65

2

45.3

Baoji

12.85

8.67

111.4095

Weinan

1

7

7

Yan’an

40

1

40

Wuwei

17

18

306

Zhangye

21.56

0.5

10.78

Pinliang

22.87

4.5

102.915

Dingxi

52

1

52

Longnan

34.32

4.25

145.86

Zhongwei

38.58

6.33

244.2114

Kashgar

63.99

2

127.98

0

Source Baidu map, with two decimal places. Other prefecture level cities are ignored due to the lack of samples

compensation methods and supporting system construction (Karttunen & Rautiainen, 2013). These rural areas are limited by geographical conditions, scattered living and high cost in all aspects, so it is difficult to realize the supply of specialization and customization. From the number of special medical institutions, the cities with high ranking mainly include the cities in the central and western regions with large land occupation and large population (such as Chongqing, Nanyang, Jining, Xinyang, Fuyang, Zhoukou, Zunyi and Pingdingshan). Because the national mental health plan advocates that every county should establish psychiatric hospitals, more prefecture level cities (Taking Chongqing as an example) are in the leading position in the number of special medical institutions. However, the variance between different prefecture level cities is not large (only 2.28) from the national distribution. From the perspective of the dispersion index of special medical supply, the cities with higher ranking also maintain consistency (such as Chongqing, Xinyang, Handan, Lincang, Qiqihar, Nanyang, Huanggang, Sanming and Zhangjiakou). In particular, Chongqing has an index value of more than 900 times that of Liangshan which ranks behind (cities with a value of 0 are not considered). This is related to the jurisdiction (26 districts, 8 counties, 4 autonomous counties), land area (824,000 km2 ) and the special medical (mental disease) institutions in almost every county under Chongqing. From the view of cities with low ranking of discretion index of special medical supply, many cities have a value of 0 (Datong, Ulanchab, xilingol, Shenyang, Dalian, Anshan, Fushun, Jinzhou, Liaoyang, Jilin, Siping, Changzhou, Yangzhou, Hangzhou, etc.). These cities can be divided into two types, which reflect a certain degree of

178

5 The Measurement of the Imbalance of Rural Long Tail Public …

Table 5.5 The demand dispersion index of rural long tail special health service

City

Special medical Number of supply distance special (km) medical institutions

Dispersion index of special medical supply

Beijing

62.9

3

188.7

Tianjin

69.75

2

139.5

Tangshan

50.3

3

150.9

Handan

53.48

10

534.8

Xingtai

55.5

5

277.5

Zhangjiakou

123.6

3

370.8

Datong

0

0

0

Jincheng

31.35

2

62.7

Yuncheng

55.08

4

220.32

Linfen

50.4

4

201.6

Tongliao

85.5

1

85.5

Ulanqab

0

0

0

Xilingol

0

0

0

Shenyang

0

0

0

Dalian

0

0

0

Anshan

0

0

0

Fushun

0

0

0

Jinzhou

0

0

0

Liaoyang

0

0

0

Panjin

27.93

3

83.79

Jilin

0

0

0

Sipin

0

0

0

Songyuan

7

2

14

Harbin

196.1

1

196.1

Qiqihar

105.13

4

420.52

Nanjing

48.1

1

48.1

Xuzhou

61.4

3

184.2

Changzhou

0

0

0

Yancheng

93.67

3

281.01

Yangzhou

0

0

0

Suqian

46.73

3

140.19

Hangzhou

0

0

0

Ningbo

13.6

2

27.2

Wenzhou

32.4

1

32.4

Jiaxing

26.2

4

104.8 (continued)

5.3 Measurement of Imbalance Index of Rural Long Tail Public … Table 5.5 (continued)

179

City

Special medical Number of supply distance special (km) medical institutions

Dispersion index of special medical supply

Taizhou

60.2

2

120.4

Hefei

20.7

2

41.4

Wuhu

60.17

3

180.51

Fuyang

53.8

6

322.8

Lvan

73.8

2

147.6

Bozhou

0

0

0

Xuancheng

0

0

0

Fuzhou

60.3

2

120.6

Sanming

93.7

4

374.8

Quanzhou

57

2

114

Nanpin

0

0

0

Longyan

52.74

5

263.7

Pinxiang

0

0

0

Ganzhou

46.03

3

138.09

Yichun

0

0

0

Jinan

0

0

0

Qingdao

0

0

0

Weifang

65.7

1

65.7

Jining

42.54

7

297.78

Tai’an

80.7

2

161.4

Linyi

54.9

5

274.5

Dezhou

66.65

2

133.3

Zhengzhou

0

0

0

Kaifen

48.8

4

195.2

Luoyang

45.73

3

137.19

Pindingshan

25.23

6

151.38

Xuchang

38.6

1

38.6

Nanyang

58.93

7

412.51

Shangqiu

38.47

3

115.41

Xinyang

98.95

6

593.7

Zhoukou

46.62

6

279.72

Yichang

57.2

2

114.4

Jinmen

105.7

2

211.4

Xiaogan

0

0

0

Huanggang

78.92

5

394.6 (continued)

180 Table 5.5 (continued)

5 The Measurement of the Imbalance of Rural Long Tail Public … City

Special medical Number of supply distance special (km) medical institutions

Dispersion index of special medical supply

Xianning

78

156

Xiangtan

18.5

1

18.5

Hengyang

49.17

3

147.51

Loudi

49.63

3

148.89

Xiangxi

77.8

1

77.8

Zhuhai

0

0

0

Foshan

0

0

0

Jiangmen

0

0

0

Maoming

0

0

0

Zhaoqing

0

0

0

Huizhou

57

6

342

Heyuan

96.9

1

96.9

Yangjiang

45.1

1

45.1

Zhongshan

0

0

0

Chaozhou

49.6

1

49.6

Jieyang

49.2

2

98.4

Guigang

61.68

4

246.72

Baise

150.4

2

300.8

Laibin

0

0

0

Chongzuo

74.1

1

74.1

Chongqing

210.85

13

2741.05

Zigong

0

0

0

Panzhihua

0

0

0

Luzhou

45

4

180

Nanchong

52.25

2

104.5

Meishan

32.97

3

98.91

Ziyang

80.6

3

241.8

Liangshan

2.9

1

2.9

Zunyi

32.7

6

196.2

Tongren

158.4

2

316.8

2

Yuxi

0

0

0

Baoshan

126.05

2

252.1

Pu’er

0

0

0

Lincang

108.25

4

433

Honghe

69.45

4

277.8 (continued)

5.3 Measurement of Imbalance Index of Rural Long Tail Public … Table 5.5 (continued)

181

City

Special medical Number of supply distance special (km) medical institutions

Dispersion index of special medical supply

Diqing

0

0

0

Xi’an

84.5

1

84.5

Baoji

69.3

1

69.3

Weinan

66.68

4

266.72

Yan’an

112.8

1

112.8

Wuwei

0

0

0

Zhangye

0

0

0

Pinliang

0

0

0

Dingxi

0

0

0

Longnan

8.8

1

8.8

Zhongwei

0

0

0

Kashgar

0

0

0

Source Baidu map, with two decimal places. Other prefecture level cities are ignored due to the lack of samples

comprehensiveness. Although the northeast prefecture level cities represented by Ulanchab, Xilingol, Shenyang, Dalian, Anshan, Fushun, Jinzhou, Liaoyang, Jilin and Siping have occupied a large area, they have lost a lot of population in recent years (Zhao & Liu, 2018), which makes the supply of special medical services insufficient. Changzhou City, Yangzhou City, Hangzhou city and other economically developed cities in the Yangtze River Delta, due to the frequent relocation of counties to districts in recent years, the number of special medical institutions below the county-level administrative units has decreased sharply. Based on the rural long tail special medical supply and demand dispersion index, the imbalance index of rural long tail special medical care is constructed, and the following table is arranged in descending order. It can be found that the imbalance index of 36% prefecture level cities is 0, which is mainly related to the lack of corresponding rural special medical institutions in these areas. In addition, the prefecture level cities with high imbalance index (i.e. the degree of imbalance is relatively light) are mainly in Hebei, Shanxi, Gansu provinces and municipality directly under the Central Government (including Handan, Weinan, Yuncheng, Zhangjiakou, Xingtai, Chongqing, Beijing, etc.). These cities have a large population density and many counties under their jurisdiction, which provides population space for the supply of rural special medical institutions. In addition, the variance of the rural long tail special medical imbalance index is 7.17 among different prefecture level cities, which shows a small volatility compared with other rural long tail services. This may be related to the unified standards of special medical treatment, especially mental disease medical treatment at

182

5 The Measurement of the Imbalance of Rural Long Tail Public …

county level in various regions of China. However, the particularity of special medical treatment is also reflected in the personalized needs of rural residents, such as roll call operation, overtime operation, whole course nursing, special ward, expert clinic and so on (Zhou & Huang, 2015). These differences in demand levels have a significant impact on the imbalance of special medical care in different regions (Table 5.6).

5.4 Measurement of Imbalance Index of Rural Long Tail Public Services: Rural Special Elderly Care China’s rural long tail elderly care public services are mainly reflected in the particularity of elderly care needs (for special groups such as the disabled elderly), atomization of distribution (the contradiction between the scattered residence of the elderly and the centralized construction of nursing homes), heterogeneity and minority (higher income rural residents have higher level and quality of elderly care needs). The reason behind this is related to the complexity of China’s rural institutional elderly care. China 13th Five Year Plan puts forward the “home-based care for the aged, community-based care for the aged, institutional care for the elderly as a supplement, and the construction path of the elderly care service system combined with medical care". It has the characteristics of high cost and high risk (high risk refers to the prevention of all kinds of accidents when taking care of the elderly; high cost refers to the qualified and professional nursing institutions) make it necessary to consider the concentration of demand when selecting the supply site. This kind of private elderly care institution has the characteristics of social enterprise, and needs to charge a certain fee to compensate the cost. However, the public rural elderly-care institutions mainly focus on the “Five Guarantees” for the elderly (In rural areas of China, social insurance is implemented for the members who have no ability to work and have no guarantee of life, namely, food, clothing, medical care and burial). It is difficult to ensure that the supply of aged-care services is at a high quality level. Therefore, some scholars have proposed that under the social background that the elderly care mode of public and private institutions cannot fully meet the elderly care needs of the rural elderly in China, the new elderly care mode of “charity + poverty alleviation + industry” is more suitable for realizing the customized supply of special groups of elderly care (Zhang & Han, 2018a, 2018). This is more important for the rural long tail elderly care with the characteristics of dispersion and minority. Under the proposal of this new elderly care mode, the demand and supply dispersion degree of rural elderly care long tail public services in different regions of China are calculated through Baidu map. The long tail index of China’s demand for elderly care services is derived from CLDS2016. Firstly, the community types in the questionnaire were selected, and the samples of rural community were retained as rural samples. Secondly, according to the age item in the individual questionnaire, the individual sample older than 50 years old is taken as the rural elderly sample.

5.4 Measurement of Imbalance Index of Rural Long Tail Public … Table 5.6 The imbalance index of rural long tail special health service

183

City

The imbalance index

City

The imbalance index

Handan

44.57

Baoji

0.62

Sanming

41.64

Qiqihar

0.58

Weinan

38.10

Fuyang

0.58

Yuncheng

22.03

Quanzhou

0.55

Zhangjiakou

16.12

Xuzhou

0.53

Xingtai

12.84

Jieyang

0.35

Chongqing

12.37

Fuzhou

0.30

Beijing

11.79

Jiaxing

0.30

Lincang

9.84

Yichang

0.19

Linyi

9.65

Hefei

0.17

Longyan

8.79

Wenzhou

0.15

Zhoukou

7.77

Xiangxi

0.14

Loudi

7.32

Yangjiang

0.12

Kaifen

5.81

Xiangtan

0.10

Luzhou

5.14

Longnan

0.06

Chaozhou

4.29

Songyuan

0.00

Ziyang

4.03

Liangshan

0.00

Tai’an

3.97

Datong

0.00

Xinyang

3.93

Ulanqab

0.00

Pingdingshan

3.78

Xilingol

0.00

Guigang

3.31

Shenyang

0.00

Nanyang

3.30

Dalian

0.00

Huizhou

3.01

Anshan

0.00

Tianjin

3.00

Fushun

0.00

Linfen

2.88

Jinzhou

0.00

Yan’an

2.82

Liaoyang

0.00

Baoshan

2.80

Jilin

0.00

Ningbo

2.72

Sipin

0.00

Jining

2.56

Changzhou

0.00

Ganzhou

2.52

Yangzhou

0.00 0.00

Tongliao

2.51

Hangzhou

Huanggang

2.49

Bozhou

0.00

Hengyang

2.19

Xuancheng

0.00

Tangshan

2.03

Nanpin

0.00

Nanjing

1.93

Pingxiang

0.00

Tongren

1.90

Yichun

0.00 (continued)

184 Table 5.6 (continued)

5 The Measurement of the Imbalance of Rural Long Tail Public … City

The imbalance index

City

The imbalance index

Xi’an

1.87

Jinan

0.00

Honghe

1.83

Qingdao

0.00

Wuhu

1.77

Zhengzhou

0.00

Harbin

1.70

Xiaogan

0.00

Jinmen

1.41

Zhuhai

0.00

Nanchong

1.22

Foshan

0.00

Baise

1.21

Jiangmen

0.00

Taizhou

1.20

Maoming

0.00

Lvan

1.17

Zhaoqing

0.00

Shangqiu

1.15

Zhongshan

0.00

Jincheng

1.05

Laibin

0.00

Weifang

1.04

Zigong

0.00

Luoyang

1.04

Panzhihua

0.00

Yancheng

0.90

Yuxi

0.00

Dezhou

0.87

Pu’er

0.00

Chongzuo

0.80

Diqing

0.00

Panjin

0.80

Wuwei

0.00

Zunyi

0.76

Zhangye

0.00

Xuchang

0.75

Pinliang

0.00

Xianning

0.75

Dingxi

0.00

Suqian

0.70

Zhongwei

0.00

Meishan

0.69

Kashgar

0.00

Heyuan

0.67

Source CLDS and Baidu Map

Again, according to the individual questionnaire I 1.20: Have you participated in the following endowment insurance (including the unit retirement system)? All the individual choosing No were selected. Then according to the individual questionnaire I 9.4.1: What do you think of your current health status? All the individual choosing “very unhealthy”, “relatively unhealthy” or “general” were selected as the sample of rural institutional elderly care needs. Taking each administrative unit at the prefecture level as the unit, the linear average distance between the township/village where each individual sample is located and the county of the city where it is located (in village questionnaire R 58.1: the distance between the village and the nearest county/district government) is obtained as the demand distance (km) of special elderly care. At the same time, the sample number of special elderly care needs is taken as the weight, and the two are multiplied to get the dispersion index of special elderly care needs of each prefecture level city (see the table below) (Table 5.7).

5.4 Measurement of Imbalance Index of Rural Long Tail Public … Table 5.7 The demand discretion index of rural long tail special elderly care

185

City

Distance of Number of special elderly special elderly care demand care needs (km)

Discrete index of special elderly care needs

Tianjin

60

7

420

Tangshan

14.86

15

223

Baoding

3

2

6

7

32.2

Zhangjiakou 4.60 Datong

30

11

330

Yuncheng

5

4

20

Linfen

20

4

80

Ulanchab

20

19

380

Xilingol

21

12

252

Dalian

20

16

320

Anshan

10

9

90

Panjin

15

1

15

Jilin

5

7

35

Songyuan

10

19

190

Harbin

65

23

1495

Nanjing

12

7

84

Xuzhou

50

14

700

Changzhou

8.30

3

24.9

Suqian

25

22

550

Hangzhou

5

6

30

Ningbo

1

3

3

Wenzhou

4.5

7

31.5

Jiaxing

13

8

104

Taizhou

25

11

275

Hefei

34

10

340

Wuhu

17

2

34

Lvan

10

1

10

Bozhou

36

10

360

Xuancheng

39

1

39

Fuzhou

50

9

450

Quanzhou

25

6

150

Ganzhou

40

17

680

Yichun

15

12

180

Shangrao

10

14

140

Weifang

30

6

180 (continued)

186 Table 5.7 (continued)

5 The Measurement of the Imbalance of Rural Long Tail Public … City

Distance of Number of special elderly special elderly care demand care needs (km)

Discrete index of special elderly care needs

Jining

23.75

12

285

Tai’an

4.63

8

37

Linyi

4

12

48

Zhengzhou

20

20

400

Kaifen

10

11

110

Luoyang

25

16

400

Xuchang

6.93

15

104

Shangqiu

25

1

25

Xinyang

11.52

13

149.7

Zhoukou

18

1

18

Xiaogan

22

2

44

Huanggang

10

5

50

Xianning

26

13

338

Xiangtan

64.29

7

450

Hengyang

70

10

700

Foshan

12

3

36

Jiangmen

15.67

27

423

Maoming

35

36

1260

Zhaoqing

27

17

459

Huizhou

15.93

55

876

Heyuan

20

21

420

Yangjiang

25.61

82

2100

Zhongshan

10

5

50

Chaozhou

1.2

17

20.4

Jieyang

24.04

57

1370

Laibin

50

33

1650

Chongzuo

30

15

450

Chongqing

40

13

520

Zigong

32

48

1536

Panzhihua

33.85

13

440

Luzhou

5

20

100

Nanchong

15

8

120

Meishan

12

24

288

Ziyang

22.47

17

382

Liangshan

27

21

567 (continued)

5.4 Measurement of Imbalance Index of Rural Long Tail Public … Table 5.7 (continued)

187

City

Distance of Number of special elderly special elderly care demand care needs (km)

Discrete index of special elderly care needs

Baoshan

30

30

Pu’er

80

17

1360

Honghe

76

11

836

Diqing

40

2

80

Xi’an

27.14

7

190

Baoji

13.63

8

109

Yan’an

40

2

80

Tianshui

16.5

20

330

Zhangye

40

6

240

Pinliang

15

6

90

Longnan

33.89

9

305

Zhongwei

30.88

16

494

1

Source CLDS and Baidu Map

It can be seen from the above table that the distribution characteristics of special elderly care demand distance in different prefecture level cities are similar to that of special education, mainly in the third and fourth tier prefecture level cities in central and Western China (Pu’er, Honghe, Hengyang, Xiangtan, Laibin, Ganzhou, Diqing, Yan’an, Zhangye, etc.). From the perspective of the number of special pension needs, many cities with high demand are located in Guangdong, Guangxi, Sichuan and other provinces and autonomous regions (Yangjiang, Jieyang, Huizhou, Zigong, Maoming, Laibin, Jiangmen, Meishan, Heyuan, etc.). From the perspective of the dispersion degree of special elderly care needs, the cities with higher dispersion degree also come from these prefecture level cities. The dispersion index of rural long tail elderly care service supply comes from Baidu map. First of all, search the keywords of “rural nursing homes”, “rural homes for the aged”, “rural welfare institutions” and other rural nursing institutions through Baidu map. Locate these rural nursing institutions in different regions (prefecture level administrative units) through administrative divisions (villages and towns or suburbs), and manually calculate the distance between the geographical center of the administrative unit at the prefecture level where the registered service is located. The average geographical distance and distance variance are calculated as the value of supply concentration. The concentration degree of rural elderly care institutions in different prefecture level cities where the demand is located is shown in the table below (the basis of selecting prefecture level cities is to match the demand, ignoring the data of other prefecture level cities) (Table 5.8): From the perspective of the average distance of rural long tail elderly care supply, the administrative units of prefecture level cities with higher average distance are

188

5 The Measurement of the Imbalance of Rural Long Tail Public …

Table 5.8 The supply dispersion index of rural long tail special care for the aged

City

Average distance of elderly care supply (km)

Number of rural institutions’ elderly care supply

Supply dispersion index

Tianjin

18.1

4

72.4

Tangshan

13.8

3

41.4

Baoding

11.02

5

55.1

Zhangjiakou

12.75

2

25.5

Datong

28.86

10

288.6

Yuncheng

14.53

4

58.12

Linfen

30.63

3

91.89

Ulanchab

60

4

240

Xilingol

222.73

3

668.19

Dalian

93.11

7

651.77

Anshan

15.43

4

61.72

Panjin

23.34

5

116.7

Jilin

13

2

26

Songyuan

10.72

5

53.6

Harbin

17.33

3

51.99

Nanjing

6.8

6

40.8

Xuzhou

17.49

7

122.43

Changzhou

25.29

8

202.32

Suqian

17.83

9

160.47

Hangzhou

41.24

9

371.16

Ningbo

54.62

13

710.06

Wenzhou

41.93

9

377.37

Jiaxing

25.85

13

336.05

Taizhou

49.76

11

547.36

Hefei

24.09

8

192.72

Wuhu

20.36

3

61.08

Lvan

27.61

11

303.71

Bozhou

6.75

2

13.5

Xuancheng

35.08

5

175.4

Fuzhou

18.84

15

282.6

Quanzhou

41.48

12

497.76

Ganzhou

8.95

2

17.9

Yichun

41.9

5

209.5

Shangrao

32.12

4

128.48

Weifang

33.6

6

201.6 (continued)

5.4 Measurement of Imbalance Index of Rural Long Tail Public … Table 5.8 (continued)

189

City

Average distance of elderly care supply (km)

Number of rural institutions’ elderly care supply

Supply dispersion index

Jining

51.82

18

932.76

Tai’an

43.68

4

174.72

Linyi

36.77

6

220.62

Zhengzhou

29.53

6

177.18

Kaifen

30.35

4

121.4

Luoyang

21.96

9

197.64

Xuchang

33.5

1

33.5

Shangqiu

36.53

6

219.18

Xinyang

15.13

3

45.39

Zhoukou

47.31

10

473.1

Xiaogan

56

1

56

Huanggang

0

0

0

Xianning

40.38

4

161.52

Xiangtan

17.71

9

159.39

Hengyang

23.78

6

142.68

Foshan

18.37

7

128.59

Jiangmen

24.86

5

124.3

Maoming

9.03

3

27.09

Zhaoqing

13.9

3

41.7

Huizhou

59.83

16

957.28 598.48

Heyuan

74.81

8

Yangjiang

30.7

5

153.5

Zhongshan

17.12

10

171.2

Chaozhou

25.9

5

129.5

Jieyang

22.3

6

133.8

Laibin

24.67

3

74.01

Chongzuo

36.03

3

108.09

Chongqing

42.05

9

378.45

Zigong

12.88

4

51.52

Panzhihua

17.65

2

35.3

Luzhou

16

4

64

Nanchong

58.92

14

824.88

Meishan

44.21

6

265.26

Ziyang

40.6

10

406

Liangshan

0

0

0 (continued)

190 Table 5.8 (continued)

5 The Measurement of the Imbalance of Rural Long Tail Public … City

Average distance of elderly care supply (km)

Number of rural institutions’ elderly care supply

Supply dispersion index

Baoshan

111.88

4

447.52

Pu’er

49.93

3

149.79

Honghe

105.83

4

423.32

Diqing

0

0

0

Xi’an

36.93

6

221.58

Baoji

44.73

4

178.92

Yan’an

53.9

4

215.6

Tianshui

59.44

7

416.08

Zhangye

63.28

5

316.4

Pinliang

70.6

2

141.2

Longnan

6.4

1

6.4

Zhongwei

0

0

0

Source Baidu Map

mainly distributed in northwest and southwest autonomous regions and multi-ethnic areas (such as Xilingol, Baoshan, Honghe, Pinliang, Zhangye, Ulanqab, etc.). These areas cover a large area and the population density is not high, which makes the distribution of different elderly care institutions more dispersed. From the perspective of the number of rural institutional endowment supply, the prefecture level cities with more supply institutions are mainly distributed in the eastern regions with higher population density and more developed economy (such as Jining, Huizhou, Fuzhou, Ningbo, Jiaxing, Quanzhou, Taizhou, etc.). This is consistent with previous studies that the development of rural elderly care supply in different regions (especially the eastern and central and western regions) is unbalanced (Song et al., 2020; Wang, 2016). From the perspective of the dispersion index of institutional elderly care supply, it reflects the comprehensive regional distribution. Prefecture level cities with large index include Huizhou, Jining, Ningbo, Dalian, Heyuan, Taizhou, Quanzhou and other eastern cities, as well as Xilingol, Nanchong and other western cities. However, the degree of supply dispersion behind these cities reflects a different logic: some are due to the distribution of institutions; some are due to the scattered distribution and long distance between institutions (multi-ethnic areas in the west). In addition, from the perspective of variance of supply dispersion index, the highest ranked city index value is nearly 150 times of the lowest, reflecting the larger variance. There are also a few cities without corresponding supply of rural elderly-care institutions (Liangshan, Diqing, Zhongwei, etc.), which makes the dispersion index 0.

5.4 Measurement of Imbalance Index of Rural Long Tail Public …

191

Furthermore, according to the calculated dispersion index of institutional elderly care, the above formula is used to divide them to get the imbalance index of institutional elderly care. At the same time, the index is ranked from high to low as follows (Table 5.9): It can be found the variance of imbalance index of rural long tail elderly care in different prefecture level cities is large (Ningbo is more than 10,000 times that of Longnan), which shows great regional differences. According to the calculation method of imbalance index, the larger the value is, the smaller the imbalance degree is. As one of the special economic zones of coastal reform and opening-up in China, Ningbo not only has a developed private economy and a high degree of marketization, but also has higher income of rural residents and higher demand for the elderly in rural institutions (Xu & Wang, 2013). In addition, Ningbo has a high degree of population aging, and the proportion of the elderly in rural areas is much higher than that in urban areas (Zuo & Zhou, 2009). Since 2006, Ningbo has been promoting the rural community elderly care mode, which can not only better meet the local rural “empty nester”, but also better alleviate the contradiction between the traditional social customs and the modern transformation of elderly care. This mode mainly relies on the collective economic strength to meet the basic living needs of the elderly in the community. Taking the village as an important environment to play the role, a number of home-based elderly caring service institutions are established in rural areas to provide life care, spiritual comfort, health care, economic assistance and other services for the elderly in rural areas with service needs. It is convenient for the elderly to enjoy local and nearby public long tail services (Sun & Cao, 2010). On the contrary, the degree of imbalance of rural elderly care institutions in more prefecture level cities is higher, and the dispersion degree of demand is higher than supply. This is related to the complexity of rural elderly care in China. On the one hand, public rural elderly care institutions (nursing homes, welfare homes, glory homes, etc.) are more targeted at the rural “Five Guarantees” farmers, such as the old, weak, and disabled farmers who have neither labour ability nor economic source. The infrastructure and living conditions of these elderly care institutions are poor, and they can only meet the basic needs of these rural residents at a lower level. It is difficult to attract the rural elderly with higher income conditions and demand for quality of life. In addition, due to the influence of Chinese traditional culture such as “raising children to guard against old age”, many rural residents think that it is “unfilial” for their parents to be fostered in local elderly care institutions, and their motivation to actively send their parents to local elderly care institutions is insufficient. Therefore, although many of these institutions in China have a wide coverage, they are facing the phenomenon of “vacancy” and oversupply. On the other hand, due to the fact that their children are working outside and other reasons, many rural elderly people with better economic conditions, more openminded and higher level of elderly care needs are difficult to really enjoy higher level of services due to the discrete living environment. Although more and more private elderly care institutions begin to settle in rural areas, they also choose to build in the counties or suburbs with high population density and concentrated residence on the

192

5 The Measurement of the Imbalance of Rural Long Tail Public …

Table 5.9 The imbalance index of rural long tail special elderly care

City

Imbalance index City

Imbalance index

Ningbo

236.687

0.875

Datong

Lvan

30.371

Zhangjiakou 0.792

Zhoukou

26.283

Jilin

0.743

Baoshan

14.917

Chongqing

0.728

Hangzhou

12.372

Anshan

0.686

Wenzhou

11.980

Luzhou

0.640

Baoding

9.183

Ulanqab

0.632

Shangqiu

8.767

Fuzhou

0.628

Changzhou

8.125

Hefei

0.567

Panjin

7.780

Honghe

0.506

Nanchong

6.874

Luoyang

0.494

Chaozhou

6.348

Nanjing

0.486

Taian

4.722

Xianning

0.478

Linyi

4.596

Zhengzhou

0.443

Xuancheng

4.497

Xiangtan

0.354

Foshan

3.572

Xuchang

0.322

Zhongshan

3.424

Xinyang

0.303

Quanzhou

3.318

Jiangmen

0.294

Jining

3.273

Suqian

0.292

Jiaxing

3.231

Songyuan

0.282

Yuncheng

2.906

Chongzuo

0.240

Yan’an

2.695

Hengyang

0.204

Xilingol

2.652

Tangshan

0.186

Dalian

2.037

Xuzhou

0.175

Taizhou

1.990

Tianjin

0.172

Wuhu

1.796

Pu’er

0.110

Baoji

1.641

Jieyang

0.098

Pinliang

1.569

Zhaoqing

0.091

Heyuan

1.425

Panzhihua

0.080

Zhangye

1.318

Yangjiang

0.073

Xiaogan

1.273

Laibin

0.045

Tianshui

1.261

Bozhou

0.038

Xi’an

1.166

Harbin

0.035

Yichun

1.164

Zigong

0.034

Linfen

1.149

Ganzhou

0.026

Weifang

1.120

Maoming

0.022

Kaifen

1.104

Longnan

0.021 (continued)

5.4 Measurement of Imbalance Index of Rural Long Tail Public … Table 5.9 (continued)

City

Imbalance index City

193 Imbalance index

Huizhou

1.093

Huanggang

0.000

Ziyang

1.063

Liangshan

0.000

Meishan

0.921

Diqing

0.000

Shangrao

0.918

Zhongwei

0.000

Source CLDS and Baidu Map. The supply index of Huanggang, Liangshan, Diqing and Zhongwei were 0

consideration of economic benefits. For the rural elderly living in other remote rural areas, the transportation cost of their residence and relocation is too high, resulting in a high “threshold” of supply to meet demand. There are still many residents with rural elderly care needs facing supply shortage. It is the coexistence of these two sides that causes the complexity of the imbalance of the long tail elderly care in rural China.

5.5 Measurement of Imbalance Index of Rural Long Tail Public Services: Rural Special Finance The dispersion index of rural long tail special finance demand comes from CLDS 2016. Firstly, the community types in the questionnaire were selected, and the samples of rural community were retained as rural samples. Secondly, according to F4.7 in the family questionnaire, does your family hold financial products? Such as stocks, funds, bonds. The sample choosing Yes are selected as the individuals with rural special financial needs, and calculate the number of samples as the number of needs. Taking each administrative prefecture level unit as a unit, the linear average distance between the township where each individual sample is located and the county seat of the prefecture level city is obtained (the distance between the village and the nearest county/district government in village questionnaire R 58.1) as the demand distance of rural finance (km). At the same time, the sample number of rural financial demand is taken as the weight, and the two are multiplied to get the dispersion index of rural special finance demand of each prefecture level city (Table 5.10). From the total distribution of the samples in different prefecture level cities, the samples with long tail financial needs only cover more than 40 prefecture level cities, which is closely related to the minority and high-level nature of this demand. From the perspective of demand distance, the third and fourth tier cities in the central and western regions are still ranked higher (such as Xiangtan, Baise, Loudi, Diqing, Ganzhou, Maoming, Panzhihua, Longnan, etc.). Different from that, in terms of the number of financial needs, the higher ranked cities are mainly eastern cities (such as Foshan, Hangzhou, Jining, Jiangmen, Nanjing and Wenzhou), and distributed in economically developed cities. This is consistent with the positive correlation between financial demand and economic development: financial demand may be

194

5 The Measurement of the Imbalance of Rural Long Tail Public …

Table 5.10 The demand dispersion index of rural long tail finance service

City

Demand distance of finance (km)

Number of financial needs

Demand dispersion index

Foshan

20.60

19

391.43

Baise

43.34

16

693.37

Hangzhou

7.24

13

94.08

Jining

19.39

12

232.73

Jiangmen

13.00

12

156.03

Nanjing

8.32

10

83.17

Wenzhou

30.16

9

271.47

2.00

8

16.00

36.57

8

292.56

9.41

8

75.28

Longnan

34.33

7

240.29

Fuzhou

50.00

5

250.00

Loudi

40.67

5

203.36

Jieyang

21.96

5

109.80

Jiaxing

Beijing Ganzhou Tai’an

18.08

4

72.32

Sanming

3.00

4

12.00

Xiangtan

73.22

4

292.88

Yangjiang

22.64

4

90.56

Sipin

100.00

2

200.00

Lv’an

27.05

2

54.11

Quanzhou

27.60

2

55.21

Xuchang

10.99

2

21.99

Maoming

35.00

2

70.00

Zhongshan

14.73

2

29.45

Chaozhou

1.79

2

3.58

Diqing

40.00

2

80.00

Tianshui

27.72

2

55.43

Wuwei

17.00

2

34.00

4.60

1

4.60

Zhangjiakou Ningbo

1.00

1

1.00

Kaifen

22.40

1

22.40

Pindingshan

20.00

1

20.00

Xinyang

17.79

1

17.79

Zhaoqing

27.00

1

27.00

Chongzuo

21.34

1

21.34

Panzhihua

34.60

1

34.60 (continued)

5.5 Measurement of Imbalance Index of Rural Long Tail Public … Table 5.10 (continued)

City

Ziyang

Demand distance of finance (km) 20.03

Number of financial needs 1

195 Demand dispersion index 20.03

Source CLDS; Other cities are ignored due to the lack of samples

caused by internal economic growth (Batuo et al., 2018). Therefore, the distribution of financial demand dispersion index among different prefecture level cities also reflects the comprehensiveness of the two (demand distance and demand number). The measurement of dispersion index of rural long tail financial service supply comes from Baidu map. First of all, Baidu map search “rural finance” and other keywords. At the same time, through the administrative divisions (located in villages or suburbs), these rural financial institutions in different regions (prefecture level administrative units) are located. The distance between their registered service address and the geographical center of prefecture level administrative units is manually calculated. The average geographical distance and distance variance are calculated as the value of supply concentration. The results show the concentration degree of rural financial institutions in different prefecture level cities where the demand is located (prefecture level cities are selected based on matching the demand, ignoring the data of other prefecture level cities) (Table 5.11). In terms of the average distance of financial supply, the higher ranking administrative units include Diqing, Baise, Ziyang, Tianshui and other prefecture level cities with large area in the central and western regions, and Hangzhou, Jining, Nanjing, Beijing, Ningbo, Fuzhou and other prefecture level cities with relatively developed economy in the East. The logic behind this is that the construction of public infrastructure such as rural financial institutions is closely related to the local economic development level, as well as the local geographical environment and social capital (Bongomin et al., 2018; Crowe, 2006). Some scholars have shown that the supply of rural finance can be divided into two types: the main supply objects of rural formal finance are relatively richer rural residents with higher social capital and higher financial demand; while relatively poor rural residents mainly obtain financial support from informal financial channels (Abraham, 2018; Sun et al., 2020). These two ways work together, which makes the rural long tail financial demand reflect the polarization difference. In terms of the supply number of rural finance, the top prefecture level cities have the characteristics of higher rural population density (such as Jining, Baise, Hangzhou, Ziyang, Zhongshan, Foshan, Pingdingshan, etc.). These cities are comprehensive in the dispersion index of financial supply (such as Hangzhou where Alibaba’s headquarter is located). This dispersion index reflects the multi-subjectivity (public institutions, market enterprises, NGOs) and informality of supply, related to the increase of local comprehensive demand for financial management and credit. From the imbalance index of rural long tail financial services, as shown in Table 5.12, Ningbo still ranks the highest. Consistent with the above analysis, Ningbo still

196

5 The Measurement of the Imbalance of Rural Long Tail Public …

Table 5.11 The supply dispersion index of rural long tail finance service

City

Average supply distance of finance (km)

Supply number of rural finance

Supply dispersion index

Foshan

19.53

14

273.42

Baise

102.94

18

1852.92

Hangzhou

118.11

16

1889.76

Jining

64.36

24

1544.64

Jiangmen

0

0

0

Nanjing

50.57

3

151.71

Wenzhou

27.23

3

81.69

Beijing

42.05

9

378.45

Ganzhou

11.52

6

69.12

Tai’an

0

0

0

Longnan

3.34

7

23.38

Fuzhou

31.8

5

159

Loudi

0

0

0

Jieyang

9

4

36

Jiaxing

9.23

3

27.69

Sanming

0

0

0

Xiangtan

7.67

3

23.01

Yangjiang

4.15

2

8.3

Sipin

0

0

0

Lv’an

25.11

6

150.66

Quanzhou

20.43

7

143.01

Xuchang

0

0

0

Maoming

0

0

0

Zhongshan

22.5

14

315

Chaozhou

9.1

2

18.2

Diqing

145.87

7

1021.09

Tianshui

28.62

5

143.1

Wuwei

3.07

4

12.28

Zhangjiakou

14.2

2

28.4

Ningbo

36.8

5

184

Kaifen

8.1

2

16.2

Pindingshan

22.82

10

228.2

Xinyang

10.23

3

30.69

Zhaoqing

5.3

1

5.3

Chongzuo

6.35

2

12.7

Panzhihua

23.76

5

118.8 (continued)

5.5 Measurement of Imbalance Index of Rural Long Tail Public … Table 5.11 (continued)

197

City

Average supply distance of finance (km)

Supply number of rural finance

Supply dispersion index

Ziyang

29.37

15

440.55

Source CLDS; Other cities are ignored due to the lack of samples

Table 5.12 The imbalance index of rural long tail finance service

City

Imbalance index City

Imbalance index

Ningbo

184.00

0.64

Beijing

23.65

Chongzuo 0.60

Ziyang

21.99

Jiaxing

0.38

Hangzhou

20.09

Wuwei

0.36

Diqing

12.76

Jieyang

0.33

Pindingshan

11.41

Wenzhou

0.30

Zhongshan

0.24

Fuzhou

10.70

Ganzhou

Jining

6.64

Zhaoqing

0.20

Zhangjiakou

6.17

Longnan

0.10

Chaozhou

5.08

Yangjiang 0.09

Panzhihua

3.43

Xiangtan

0.08

Lv’an

2.78

Jiangmen

0.00

Baise

2.67

Tai’an

0.00

Quanzhou

2.59

Loudi

0.00

Tianshui

2.58

Sanming

0.00

Nanjing

1.82

Sipin

0.00

Xinyang

1.73

Xuchang

0.00

Kaifen

0.72

Maoming

0.00

Foshan

0.70

Source CLDS; Other cities are ignored due to the lack of samples

performs the best in the matching of rural long tail financial management (with the highest index and the lowest imbalance). The reason behind this is that Ningbo, as a typical export-oriented city, adapt to develop the rural credit mode of land use right mortgage (rural land management right mortgage financing) (Sun, 2019). It could promote the improvement of rural financial system and credit market to increase farmers’ income (Yu, & Dai, 2014). In addition, the continuous innovation of Ningbo’s financial supply system (such as the establishment of the first rural cooperative bank in 2003, Ningbo Yinzhou Rural Cooperative Bank) provides policy and institutional soil for the effective supply of rural finance (Zhong, 2019). From the variance of the imbalance index of rural long tail financial services, the cities with the highest ranking are 2300 times higher than those with lower ranking,

198

5 The Measurement of the Imbalance of Rural Long Tail Public …

showing higher variance and regional differences. Those cities with vast land or large demand density but less supply, represented by Wuwei, Jieyang, Ganzhou, Zhaoqing, Longnan, show a higher level of imbalance.

5.6 Summary Generally speaking, this chapter calculates the imbalance index of rural long tail public services in different areas of different prefecture level cities from the empirical perspective. It finds that there are great differences in the imbalance of long tail public services in different rural areas. It also synthesizes the imbalance indexes of various fields and sums them up to calculate the weighted average value, and constructs the comprehensive imbalance index of rural long tail public services as follows (descending order) (Table 5.13): Generally speaking, the comprehensive imbalance index of rural long tail public service is different in different regions (variance is 9.78). The regions with small imbalance (large comprehensive index) are mainly distributed in prefecture level cities with large population, land occupation, high income and more counties in the Eastern and Central areas (such as Ningbo, Handan, Sanming, Beijing, Lu’an, Hangzhou, Ziyang, Yuncheng, Zhangjiakou, etc.). Especially Ningbo, as a sample of the supply of long tail public services in China, can meet the local rural long tail public demand through the participation of market forces, NGOs and governments. The cities with high imbalance (low comprehensive index, except for 0) are mainly in the third and fourth tier cities in the West (such as Wuwei, Songyuan, Longnan, Pu’er, Bozhou, Liangshan, etc.). These cities have large rural area, but due to the influence of geographical environment, the rural residents in these areas lived too scattered and difficult to meet their long tail needs. In order to further analyze the difference of the comprehensive imbalance index of rural long tail public service, this chapter takes the province (municipality directly under the central government and autonomous region) as the unit. The weighted average value of the comprehensive index of prefecture level cities within the provincial jurisdiction is used as the comprehensive index of the province. The map is drawn as follows (Fig. 5.1): This chapter finds that, from the perspective of the inter-provincial distribution of the imbalance index, the regions around the Yangtze River Delta and BeijingTianjin-Hebei are the main ones with low degree of imbalance, while the central and western provinces have high degree of imbalance. Generally speaking, the imbalance degree of rural long tail public services in different regions of China shows the trend of Eastern < Central < Western. This distribution pattern is related to the influencing factors of long tail demand mentioned in the last Chapter, which have a significant negative impact on the satisfaction and acquisition of rural residents in China. Therefore, it is urgent to design relevant correction mechanism to make up for this imbalance. The next chapter will introduce the mechanism design theory,

5.6 Summary

199

Table 5.13 The comprehensive imbalance index for rural long tail public service City

Special education

Special health

Elderly care

Finance

Comprehensive

Ningbo

0

2.72

236.68

184

105.85

0

10.41

Handan

44.57

Sanming

41.64

Weinan

38.10

Beijing Lv’an

11.14 9.52

11.79 0.13

Zhoukou Hangzhou

1.17

30.37

7.77

26.28

23.65

8.86

2.78

8.61 8.51

0.00

12.37

20.09

8.11

Ziyang

0.05

4.03

1.06

21.99

6.78

Yuncheng

0.65

22.03

2.90

Zhangjiakou

0.30

16.12

0.79

Baoshan

0.24

2.80

14.91

4.49

Linyi

2.63

9.65

4.59

4.21

3.78

1.56

11.41

4.29

6.34

5.08

3.92

0.00

3.42

10.70

3.53

Pindingshan Chaozhou

0

Zhongshan

6.39 6.17

5.84

4.19

Chongqing

0.22

12.37

0.72

Jining

0.82

2.56

3.27

6.64

3.32

Diqing

0.25

0.00

0

12.76

3.25

Wenzhou

0.04

0.15

11.98

0.30

3.11

Baoding

1.97

Shangqiu

0.25

Xintai

12.84

Lincang Tai’an

3.33

1.15

3.21 9.18

2.78

8.76

2.54

9.84 0.90

Longyan

3.97

2.46 4.72

0

8.79

2.39 2.19

Panjin

0

0.80

7.78

2.14

Nanchong

0.38

1.22

6.87

2.11

Changzhou

0.06

0.00

8.12

Kaifen

0.14

5.81

1.10

Loudi

7.32

2.04 0.72

1.94

0

1.83

Quanzhou

0.09

0.55

3.31

2.59

1.63

Xinyang

0.11

3.93

0.30

1.73

1.51

Lvzhou

0.01

5.14

0.64

Yan’an

0

2.82

2.69

Nanjing

0.72

1.93

0.48

1.44 1.37 1.82

1.23 (continued)

200

5 The Measurement of the Imbalance of Rural Long Tail Public …

Table 5.13 (continued) City

Special education

Special health

Elderly care

Xuancheng

0

0.00

4.49

Finance

Comprehensive 1.12

Linfen

0.45

2.88

1.14

1.12

Huizhou

0.22

3.01

1.09

1.08

Foshan

0.01

0.00

3.57

0.70

1.07

Jiaxing

0.02

0.30

3.23

0.38

0.98

Baise

0.03

1.21

2.67

0.97

1.26

2.58

0.96

0.24

0.95

3.43

0.87

Tianshui Ganzhou

1.05

2.52

0.02

Wuhu

0

1.77

1.79

Panzhihua

0

0.00

0.08

Guigang

3.31

Nanyang

3.30

0.89 0.82 0.82

Taizhou

0.07

1.20

1.99

0.81

Tianjin

0.01

3.00

0.17

0.79

Xi’an

0.13

1.87

1.16

0.79

Yichun

1.69

0.00

1.16

0.71

Weifang

0.66

1.04

1.12

0.70

Xilingol

0

0.00

2.65

0.66

Hengyang

0.18

2.19

0.20

0.64

Huanggang

0

2.49

0

0.62

Baoji

0.11

0.62

1.64

0.59

Honghe

0.03

1.83

0.50

0.59

Tangshan

0.05

2.03

0.18

0.56

Heyuan

0.07

0.67

1.42

0.54

Dalian

0

0.00

2.03

0.50

Tongliao

2.51

Tongren

0.62

1.90

0.47

Harbin

0.09

1.70

0.03

0.45

Luoyang

0.29

1.04

0.49

0.45

Meishan

0.08

0.69

0.92

Fuzhou

0.11

0.30

0.62

0.64

0.42

Chongzuo

0

0.60

0.41

0.80

0.24

Xiangxi

0.14

1.27

Jinmen

1.41

0.42

0.35 0.35

Zhangye

0

0.00

1.31

0.33

Xianning

0.03

0.75

0.47

0.31 (continued)

5.6 Summary

201

Table 5.13 (continued) City

Special education

Special health

Elderly care

Suqian

0.17

0.70

0.29

Xuchang

0

0.75

0.32

Jincheng

Datong

0

0.12

0.35

0.09

0.23 0.33

0.90 0

0.00

Dezhou

0.87

Zunyi

0.76

Jilin

0.02

Hefei

0.01

Anshan

0.06

Xuzhou

0.03

Ulanqab

0

0.00

0.26 0.26

0.91

Yancheng

Comprehensive 0.29

1.05

Shangrao Jieyang

Finance

0.22 0.22

0.87

0.21 0.21 0.19

0.74

0.19

0.17

0.56

0.18

0.00

0.68

0.18

0.53

0.17

0.18

0.00

0.63

0.15

Qiqihar

0.58

0.14

Fuyang

0.58

0.14

Zhengzhou

0.12

0.00

0.44

Xiangtan

0

0.10

0.35

0.14

Zigong

0.43

0.00

0.03

Xiaogan

0.4

0.00

Jiangmen

0.09

0.00

0.29

0

0.09

Zhaoqing

0.07

0.00

0.09

0.2

0.09

0.08

0.13 0.11 0.10

Wuwei

0.00

Songyuan

0.00

0.28

0.12

0.07

0.09

0.07

0.1

0.04

Yangjiang

0

Yichang

0.36

0.09 0.07

0.19

0.04

Longnan

0

0.06

0.02

Pu’er

0

0

0.11

0.02

Bozhou

0.04

0

0.03

0.02

Liangshan

0.05

0

0

0.01

Laibin

0

0

0.04

0.01

Maoming

0.01

0

0.02

0

0.008

Dingxi

0

0

Fushun

0

0

Jinan

0

0

Jinzhou

0

0 (continued)

202

5 The Measurement of the Imbalance of Rural Long Tail Public …

Table 5.13 (continued) City

Special education

Special health

Elderly care

Finance

Comprehensive

Kashgar

0

0

Liaoyang

0

0

Nanpin

0

0

Pinliang

0

0

Pinxiang

0

0

Qingdao

0

0

Shenyang

0

0

Sipin

0

Yangzhou

0

0

Yuxi

0

0

Zhongwei Zhuhai

0

0 0

0

0

0

0 0

Source CLDS and Baidu map. The weight of the imbalance index in each field is 0.25. The comprehensive imbalance index is obtained by summation. The blank space indicates no index, and the decimal point is reserved for two places

Fig. 5.1 The provincial distribution of the comprehensive imbalance index of rural long tail public service in China. Note The data of each province comes from the weighted average value of the comprehensive index of cities in the province; The default value of missing data is 0. Source The Chinese Map is from Resources and Environment Science and Data Center, Chinese Academy of Sciences (http://www.resdc.cn/)

5.6 Summary

203

based on the game between different actors, to design the correction mechanism of the imbalance of rural long tail public services.

References Abraham, T. W. (2018). Estimating the effects of financial access on poor farmers in rural northern Nigeria. Financial Innovation, 4(1), 1–20. Batuo, M., Mlambo, K., & Asongu, S. (2018). Linkages between financial development, financial instability, financial liberalisation and economic growth in Africa. Research in International Business and Finance, 45, 168–179. Bongomin, G. O. C., Munene, J. C., Ntayi, J. M., & Malinga, C. A. (2018). Exploring the mediating role of social capital in the relationship between financial intermediation and financial inclusion in rural Uganda. International Journal of Social Economics. Cai, Y., & Cheng, Y. (2014). Pension reform in China: Challenges and opportunities. Journal of Economic Surveys, 28(4), 636–651. Crowe, J. A. (2006). Community economic development strategies in rural Washington: Toward a synthesis of natural and social capital. Rural Sociology, 71(4), 573–596. Hua, L. (2014). The average funding for special education in Shandong will be raised to 6000 yuan next year. Modern Special Education, 9, 62–62. (in Chinese). Karttunen, J. P., & Rautiainen, R. H. (2013). Distribution and characteristics of occupational injuries and diseases among farmers: A retrospective analysis of workers’ compensation claims. American Journal of Industrial Medicine, 56(8), 856–869. Song, S., Wang, D., Zhu, W., & Wang, C. (2020). Study on the spatial configuration of nursing homes for the elderly people in Shanghai: Based on their choice preference. Technological Forecasting and Social Change, 152, 119859. Sun, H., Li, X., & Li, W. (2020). The nexus between credit channels and farm household vulnerability to poverty: Evidence from rural China. Sustainability, 12(7), 3019. Sun, L. (2019). Management behaviour study on agricultural industrial organization embedding, farmer heterogeneity and rural land management right mortgage financing-based on a survey analysis of 561 famers in Anhui Province. International Journal of Research in Humanities and Social Studies, 6(4), 41–48. Sun, Y. Y., & Cao, F. Q. (2010). The cultural choice of the mode of providing for the aged in the developed rural area: Some thoughts on providing for the aged in rural communities of Ningbo. Ningbo Economy, 9, 25–28. Wang, L. M. (2016). Construction strategy of unbalanced diversified rural old age security mode. Review of Economic Research, 71, 31–32. Xu, Z. L., & Wang, B. (2013). Performance evaluation of urban and rural social endowment insurance system based on residents’ satisfaction. Rural Economy, 5, 70–74. (in Chinese). Yu, L., & Dai, L. (2014). Study on practical experiences of rural land management right mortgage loan in China. In International Conference on Logistics Engineering, Management and Computer Science (LEMCS 2014) (pp. 141–144). Atlantis Press. Zhang M. S. & Han J. F. (2018a). Building a new rural pension model of "charity + poverty alleviation + industry". Academic Journal of Zhongzhou, 258(6):68–73. Zhang, X., & Han, L. (2018b). Which factors affect farmers’ willingness for rural community remediation? A tale of three rural villages in China. Land Use Policy, 74, 195–203. Zhao, F., & Liu, Y. J. (2018). Why does the population of the three northeast provinces lose?—Based on the factor time varying coefficient model. Population Journal, 40(4), 81–90. (in Chinese). Zhong, M. Z. (2019). New rural cooperative finance in China: The problem and its solution. APLPJ, 21, 1.

204

5 The Measurement of the Imbalance of Rural Long Tail Public …

Zhou, W. X., & Huang, X. M. (2015). Importance and measures of protecting patients’ privacy in special outpatient service. Journal of Qilu Nursing, 1, 54–55. (in Chinese). Zuo, J. Y., & Zhou, Z. H. (2009). Practice and thinking of home care service for the aged in rural areas of Ningbo. Ningbo Communication, 10, 50–51. (in Chinese).

Chapter 6

Correction Mechanism of the Imbalance of Rural Long Tail Public Services

Abstract Referring to the definition of mechanism design (Börgers and Krahmer in An introduction to the theory of mechanism design. Oxford University Press, USA, 2015), the design of the correction mechanism for the imbalance of rural long tail public services refers to choose the appropriate mechanism under the given economic and social environment, on the premise of not destroying the incentive motivation of all stakeholders (including individuals, governments and NGOs), so as to achieve the Pareto effect of social welfare and individual rational allocation results. In detail, this chapter includes basic mechanism design (basic setting, commitment mechanism, government regulation), incentive compatibility and information efficiency mechanism, interval design of imbalance correction mechanism, GrovesClark correction mechanism, Nash equilibrium correction mechanism, dynamic adjustment mechanism and accurate matching mechanism.

6.1 Basic Mechanism Design 6.1.1 Basic Setting As the moral provider of rural long tail public service, the local government wants to “outsource” the supply of this service to local NGOs based on its own information and cost constraints. For the local government, the benefits (mainly social benefits and economic benefits based on rational economic man) brought by this long tail public service supply quantity q is B(q). The cost of NGOs to provide this long tail public service is not observable for the local government. But for all NGOs, the supply cost is common knowledge, including external fixed cost F and internal variable cost V, which belongs to the set V = {E f f, I ne f f }. Where E f f is the efficiency of NGOs’ supply, and the corresponding probability is P. I ne f f is the inefficiency (or low efficiency) of NGOs’ supply, and the corresponding probability is 1 − P. Therefore, the cost function of NGOs to provide rural long tail public services is {C(q, E f f ) = q ∗ E f f + F i f P © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 J. Luo, Rural Long Tail Public Service and the Correction Mechanism, https://doi.org/10.1007/978-981-16-4023-0_6

(6.1)

205

206

6 Correction Mechanism of the Imbalance of Rural Long …

{C(q, I ne f f ) = q ∗ I ne f f + F i f 1 − p

(6.2)

Set E f f = I ne f f − E f f as the uncertainty of marginal cost for NGOs. This information structure is exogenous to all stakeholders. NGOs can get the cost of government purchase t by providing quantity q of long tail public services. Let ∃ be a feasible allocation set, then ∃ = {(q, t)q ∈ R, t ∈ R} Based on the long tail attribute of rural long tail public service, the demand information is asymmetric between the government and NGOs. The effective purchase level of public service supply depends on the equality between the marginal purchase value of the government and the marginal supply cost of NGOs. Therefore, the first order conditions of cost functions (6.1) and (6.2) are   C  qE f f = E f f

(6.3)

  C  q I ne f f = I ne f f

(6.4)

Social value VE f f and VI ne f f brought by complete information supply level q E f f and q I ne f f should be non-negative, i.e.VE f f > VI ne f f . It can be seen from (6.3) and (6.4) that in the rural long tail public service cost function, fixed cost F does not have any impact, so it can be simplified to 0. Because the marginal value of government purchase is decreasing, there is a monotony of supply quantity between the efficiency of NGOs with optimal supply, namely q E f f > q I ne f f . In this kind of efficient government outsourcing purchase of rural long tail public services, assuming that the opportunity cost OC of NGOs to undertake supply responsibility is given, the constraint condition for NGOs to choose to participate in supply is as follows t E f f − E f f ∗ q E f f ≥ OC

(6.5)

t I ne f f − I ne f f ∗ q I ne f f ≥ OC

(6.6)

In order to achieve the first-order optimization of supply, the government, as the buyer of long tail public services, can put forward the following conditions to NGOs: the government pay the expenses t E f f or t I ne f f in the corresponding supply quantity level q E f f or q I ne f f to make t E f f = E f f ∗ q E f f or t I ne f f = I ne f f ∗ q I ne f f . Therefore, no matter what type of NGO is (efficient or inefficient), it can achieve zero profit return (crucial for the non-profitability of NGOs). So the optimal supply under the condition of complete information is (t E∗ f f , q E∗ f f ) or (t I∗ne f f , q I∗ne f f ). Effective government purchase will not result in cost loss. The utility of the corresponding function of the first order is satisfied

6.1 Basic Mechanism Design

207

U E∗ f f = t E∗ f f − E f f ∗ q E∗ f f = 0 U I∗ne f f = t I∗ne f f − I ne f f ∗ q I∗ne f f = 0 However, based on the attribute of rural long tail public service, the cost of NGOs in the supply belongs to private information and is not easy to identify. When the supply and demand of rural long tail public services are out of balance, as the buyer (client) of public services, the choice of government supply to NGOs is (t E∗ f f , q E∗ f f ) or (t I∗ne f f , q I∗ne f f ), in order to achieve the incentive compatibility of imbalance correction mechanism. Especially when there is information asymmetry between the government and NGOs (a large number of NGOs have information advantages), the supply cost will rise due to participation restriction. In the optimal choice (t E∗ f f , q E∗ f f ) or (t I∗ne f f , q I∗ne f f ), NGOs with private information tend to choose efficient contract conditions (t E∗ f f , q E∗ f f ). When the government requires all types of NGOs to provide long tail public services, the information rent they must give up is UE f f = tE f f − E f f ∗ qE f f > 0 U I ne f f = t I ne f f − I ne f f ∗ q I ne f f > 0 Therefore, the optimal solution for the government to purchase rural long tail public services is to maximize social welfare max

(t E∗ f f ,q E∗ f f )(t I∗ne f f ,q I∗ne f f )

        p B q E f f − t E f f + (1 − P) B q I ne f f − t I ne f f

Because it is difficult for the government to identify the information rent brought by the accurate rural long tail demand, this chapter can use the information rent to express the purchase cost of the government’s objective function, i.e. (U E f f , q E f f ) or (U I ne f f , q I ne f f ). This can make this chapter focus on the allocation of unbalanced demand information, thus changing the government objective function to     p B qE f f − E f f ∗ qE f f (U E f f ,q E f f )(U I ne f f ,q I ne f f )       + (1 − P) B q I ne f f − I ne f f ∗ q I ne f f − pU E f f + (1 − p)U I ne f f max

The first two terms are expected supply and correction efficiency, while the third term is efficiency distortion (information rent) caused by information asymmetry. This optimal function must satisfy the constraint conditions U I ne f f = E f f q E f f UE f f = 0 By substituting the above two equations into the government objective function, we can get

208

6 Correction Mechanism of the Imbalance of Rural Long …

max

(q E f f ,q I ne f f )

        p B q E f f − E f f ∗ q E f f + (1 − P) B q I ne f f − I ne f f ∗ q I ne f f

− pE f f ∗ q E f f The existence of private information makes the optimal function of the government as the principal change because of the existence of information rent. Low efficiency NGOs cannot get information rent, while high efficiency NGOs can get information rent by “imitating” low efficiency Ineff because they have private information. So the first order condition becomes   B  q ES Bf f = E f f or q ES Bf f = q E∗ f f

(6.7)

  B   1 − P B  q ISne f f − I ne f f = pE f f

(6.8)

where SB and * represent the second-order optimal and the first-order optimal respectively, and (6.8) reflects the trade-off between correction efficiency and distortion (information rent) under asymmetric information. This chapter can briefly summarize that the design of this correction mechanism of government outsourcing NGO purchase service needs to meet the following points. For the type of efficient NGO, the first-order optimal of supply and output will not produce distortion, that is, q ES Bf f = q E∗ f f . But for the low efficiency NGO type, its efficiency distortion is satisfied  B  V  q ISne f f = I ne f f +

p E f f 1− p

Only efficient types of NGOs can get information rent, that is to say B U ES Bf f = E f f ∗ q ISne ff

The second-order optimal payment (service purchase) fees are B t ES Bf f = E f f ∗ q E∗ f f + E f f ∗ q ISne ff B SB t ISne f f = I ne f f ∗ q I ne f f

Faced with the efficiency loss caused by information asymmetry, the government tends to make corresponding purchase plans for different types of NGOs. The Revelation Principle of mechanism design can ensure that based on the cardinal nature of different types of NGOs, the government, as the principal, sets up a general mechanism to meet different types of needs and incentives. This direct display mechanism is the mapping g(·) from the NGO type set ϕ to the behavior set ω. It can be expressed as g(E f f ) = [q(E f f ), t(E f f )]

6.1 Basic Mechanism Design

209

When NGOs (agents of long tail service imbalance correction) reveal their own types (Eff or Ineff ), the government (clients of the long tail service imbalance correction) promises to achieve the supply and correction of q(E f f ) level by purchasing transfer payment fee t(E f f ). When this direct display mechanism can effectively motivate different types of NGOs to reveal their real types, such as satisfy t(E f f ) − E f f ∗ q(E f f ) ≥ t(I ne f f ) − E f f ∗ q(I ne f f ) t(I ne f f ) − I ne f f ∗ q(I ne f f ) ≥ t(E f f ) − I ne f f ∗ q(I ne f f ) Assume that the information space set available to all NGOs is M. The standard of the correction mechanism for the imbalance of rural long tail public services is defined as the mapping g(E f f ) = [q(E f f ), t(E f f )] from the information set M to the behavior set ω. It enables the government to effectively encourage NGOs to accurately identify and supply long tail public services by purchasing services and paying fees, so as to achieve dynamic correction and balance of supply and demand. The optimal information m ∗ (E f f ) selected by different types of NGOs satisfy     t m ∗ (E f f ) − E f f q m ∗ (E f f ) ≥ t(m) − E f f q(m), m ∈ M This mechanism includes allocation rules w(E f f ) = [q(m ∗ (E f f )] , t(m ∗ (E f f )] to match efficiency type set ϕ and behavior set ω. Therefore, the Revelation Principle of this correction mechanism is that the allocation rule w(E f f ) with mechanism [M, g(·)] can be executed by direct display mechanism. Next, we consider a semi-linear utility function, in which NGOs act as agents to purchase services, and the objective function is set as U = t − C(q, E f f ) Incentive-compatible distribution supply meets the participation constraints     U E f f = t E f f − C q E f f , E f f ≥ t I ne f f − C q I ne f f , E f f     U I ne f f = t I ne f f − C q I ne f f , I ne f f ≥ t E f f − C q E f f , I ne f f U E f f ≥ 0 U I ne f f ≥ 0 According to the above constraints, we can deduce U E f f ≥ U I ne f f + δ(q)     δ(q) = C q I ne f f , I ne f f − C q I ne f f , E f f These incentives and participation constraints can ensure the second-order optimization, and make the information rent obtained by the efficient type of NGOs to be

210

6 Correction Mechanism of the Imbalance of Rural Long …

U E f f = δ(q) Therefore, based on the general preference of Spencer-Mills Attribute, the optimal mechanism of the government as the principal satisfies the following points. (1)

For the types of NGOs with efficient supply, there is no distortion effect in the first-order optimal correction q ES Bf f = q E∗ f f     B  q E∗ f f = Cq q E∗ f f , E f f For inefficient NGOs, there is a diminishing output distortion effect in the B ∗ correction mechanism, which makes them satisfy q ISne f f < q I ne f f . At the same time, the first and second order optimal conditions satisfy     B  q I∗ne f f = Cq q I∗ne f f , I ne f f  B   SB  B  q ISne f f = C q q I ne f f , I ne f f +

(2)

p  S B  δ q I ne f f 1− p

Only efficient NGOs can get positive information rent  B  U ES Bf f = δ q ISne ff

(3)

The second-order optimal transfer payments of government purchase services are    B  t ES Bf f = C q E∗ f f , E f f + δ q ISne ff  SB  SB t I ne f f = C q I ne f f , I ne f f

Based on the diversity and fragmentation of rural long tail public services, the degree of imbalance is often multi-dimensional in terms of demand varieties and types, which makes the phenomenon of multi-dimensional supply of NGOs in undertaking the task of purchasing services. The same NGO corresponds to the type bundle set of public services Q = (q1 + q2 + · · · + qn ). The cost function of NGO as agent is strictly convex to the origin of supply quantity q, while the welfare utility function of government as principal is strictly concave to the origin of supply quantity q. In this incentive mechanism of multi-dimensional supply and correction, the government has the incentive to pay attention to and supervise the multi-dimensional supply activities of NGOs at the same time. The information rent obtained by efficient NGOs can be written as U E f f = δ(q), δ(q) = C(q, I ne f f ) − C(q, E f f ). The corresponding first-order optimal supply-output vector is q ES Bf f = q E∗ f f   ∗   B q E f f = C q E∗ f f , E f f

6.1 Basic Mechanism Design

211

For the type of low efficiency NGO, the supply level of the second-order optimal q ES Bf f satisfies the condition  B   SB  B q ISne f f = C q I ne f f , I ne f f As for incentive compatibility, this chapter can consider the restriction condition     U E f f ≥ U I ne f f + δ q I ne f f and U I ne f f ≥ U E f f − δ q E f f as       δ q E f f = C q E f f , I ne f f − C q E f f − E f f     ≥ C q I ne f f , I ne f f − C q I ne f f − E f f   = δ q I ne f f When different types of NGOs participate in the correction of rural long tail public services, sometimes the NGOs themselves do not know whether they are in high efficiency or not in the supply. But they complete the purchase or actively supply based on charity nature and social drive. Sometimes the government purchases services to NGOs before the imbalance occurs, so the optimal correction mechanism design also depends on the risk preference of NGOs in the supply. If the NGO is risk neutral in the supply, its prior participation constraints can be written as PU E f f + (1 − P)U I ne f f ≥ 0 As the government is the principal, its objective function is decreasing for the information rent of NGOs. It has incentive to reduce the information rent in the correction of rural long tail public services. From the perspective of the government, the first-order optimal supply output is monotonous for incentive compatible execution conditions. NGOs obtain corresponding economic rewards by showing their own supply and correction efficiency. Therefore, this chapter can conclude that when the NGO is risk neutral, the pre-designed correction mechanism can implement the first-order optimal supply results. When the optimal mechanism is designed as     ∗ t E f f , q E∗ f f , t I∗ne f f , q I∗ne f f   where t E∗ f f = B q E∗ f f − T ∗ , t I∗ne f f = B q I∗ne f f − T ∗ , T ∗ is the one-time transfer payment for the purchase of services by the government. The incentive compatibility of this mechanism satisfies   t E∗ f f − E f f q E∗ f f = B q E∗ f f − E f f q E∗ f f − T ∗   > B q I∗ne f f − E f f q E∗ f f − T ∗ = t I∗ne f f − E f f q I∗ne f f This incentive compatibility mechanism has strict differences for different types of NGOs, in which the one-time transfer payment T ∗ can meet the participation

212

6 Correction Mechanism of the Imbalance of Rural Long …

constraints of NGOs, that is         T ∗ = p B q E∗ f f − E f f q E∗ f f + (1 − p) B q I∗ne f f − I ne f f q I∗ne f f This first-order optimal incentive compatibility enables different types of NGOs to realize all the welfare values of the long tail service bundle and the balance between the value and cost of supply as the maximizer of efficiency, similar to the residual value requester in the principal-agent theory.

6.1.2 Commitment Mechanism In order to solve the problem of incentive compatibility, this chapter assumes that the government, as the principal of purchasing services, has strong commitment ability. It can not only display information through the aggregation of “long-term” effect and the distribution of information rent, but also reduce the cost of the display mechanism through the distribution efficiency. This kind of correction mechanism design is in a dynamic balance, which is constantly updated according to the change of demand. It can be seen as the adjustment and redesign of their own behavior based on Pareto improvement among different stakeholders. Based different types of NGOs in high effi characteristics, when   on their own SB SB SB B ciency t E f f , q E f f and low efficiency t I ne f f , q ISne f f shows their preference, the government can reveal the dynamic adjustment mechanism based on information, so as to overcome the failure of the supply of inefficient NGOs. The increase of social welfare comes from the adjustment of its supply level from the second-order optimal q ES Bf f to first order optimal q E∗ f f . By sharing the increase of social welfare, the government must at least achieve the utility level before the adjustment. For the NGOs with low supply efficiency before, B SB the purchase cost of transfer payment they received from t ISne f f = I ne f f q I ne f f goes ∗ ∗ up to t I ne f f = I ne f f q I ne f f , which can maintain the participation constraint to zero. However, the promotion of the transfer payment fee may worsen the ex-ante incentive compatibility constraints for the efficient NGOs. For efficient NGOs, they have incentives to get more transfer payments by hiding their efficiency types, thus destroying the principle of credible commitment display in imbalance correction. Therefore, this dynamic imbalance correction mechanism faces the trade-off between the weakening of ex-ante incentives and the improvement of ex-post efficiency. In addition, this mechanism design has self-renewal and negation on the basis of dynamic adjustment. When NGOs reveal their real efficiency types by choosing the optimal supply scheme, the government has incentives to “crowd out” information rent based on the complete type of information obtained. On the other hand, NGOs tend to give up the mechanism design that the level of ex-post correction utility is negative. Therefore, the threat of correcting the difference between ex-ante and expost makes the participation of NGOs disturbed by the design of signal mechanism.

6.1 Basic Mechanism Design

213

It is assumed that both NGOs and the government can observe the ex-post signal S based on efficiency type Eff in the rural long tail public service correction mechanism. This mechanism design should be able to provide useful information of different correction status according to the preference disclosure and observable signals of long tail demand. Suppose that this signal has only two values S1 and S2 . The conditional probabilities of different values are   μ1 = Pr S E f f = S1 ≥ 1/2   μ2 = Pr S I ne f f = S2 ≤ 1/2 Suppose that the ex-post information rents are u 11 = t(E f f, S1 ) − E f f q(E f f, S1 ) u 12 = t(E f f, S2 ) − E f f q(E f f, S2 ) u 21 = t(I ne f f, S1 ) − I ne f f q(I ne f f, S1 ) u 22 = t(I ne f f, S2 ) − I ne f f q(I ne f f, S2 ) By identifying the long tail needs, NGOs obtain their own efficiency type signals, and implement the correction mechanism before the signals are sent out. Based on the realization of the signal, the incentive constraint can be expressed as μ1 u 11 + (1 − μ1 )u 12 ≥ μ1 (u 21 + E f f q21 ) + (1 − μ1 )(u 22 + E f f q22 ) μ2 u 22 + (1 − μ2 )u 21 ≥ μ2 (u 12 − E f f q12 ) + (1 − μ2 )(u 11 − E f f q11 ) where qi j , i, j = 1, 2 are the supply quantity of NGOs based on different efficiency types under different signaling mechanisms. The participation restriction can be expressed as μ1 u 11 + (1 − μ1 )u 12 ≥ 0 μ2 u 22 + (1 − μ2 )u 21 ≥ 0 Now, assume the signal S that is not verifiable can be ex-ante identified for the government. Before the incentive mechanism can be provided, the government can use Bayesian Law to calculate the ex-post beliefs of efficient NGOs under different signaling mechanisms, namely   P1 = Pr S E f f = S1 =  P2 = Pr S E f f

pμ1 pμ1 + (1 − p)(1 − μ2 )  p(1 − μ1 ) = S2 = p(1 − μ1 ) + (1 − p)μ2

The optimal signal mechanism design include under the condition of signal S1 or B S2 , the descending distortion effect of the supply level q ISne f f of inefficient NGOs

214

6 Correction Mechanism of the Imbalance of Rural Long …

 B  B  q ISne f f (S1 ) = I ne f f +

P1 E f f 1 − P1  B  P2 B  q ISne E f f f f (S2 ) = I ne f f + 1 − P2 In this chapter, we can regard μ as the information contained in the signal mechanism itself. Signal S1 means that NGOs are efficient, which will lead to a lower B supply level of q ISne f f and information rent. Especially, when μ is large enough, the government can change and adjust the mechanism so that it has incentive effect only for efficient NGOs. On the contrary, signal S2 means that NGOs are inefficient, and the government tends not to reduce the information rent excessively, but to maintain the incentive level of the previous information signal.

6.1.3 Government Regulation As the regulator of the correction mechanism of rural long tail public services, the government tends to weight averagely the social surplus B(q)−t and regulatory cost U = t − E f f ∗ q brought by the participation of NGOs in the supply. The objective function of the government in the regulatory mechanism can be written as V = B(q) − E f f ∗ q − (1 − α)U α is the weight and less than 1. Under the condition of incentive compatibility and participation constraints, maximizing the expected social welfare makes the supply level of efficient NGOs q ES Bf f = q E∗ f f , and the supply level of inefficient NGOs satisfy  B  B  q ISne f f = I ne f f +

p (1 − α)E f f 1− p

It can be seen from the above formula that when α increases, the government (as the regulator) will weaken its attention to the distribution of information rent. So a higher level of α can reduce the distortion effect of the imbalance correction. When the efficiency type of NGO is continuous rather than discrete, that is, the type set ∅ = [I ne f f, E f f ], and the cumulative distribution function is F(E f f ), the mechanism design of direct display {q(θ ), t(θ )} is still credible and satisfies t(θ1 ) − θ1 q(θ1 ) ≥ t(θ2 ) − θ1 q(θ2 ) (θ1 , θ2 ) ∈ ∅ (θ1 − θ2 )(q(θ2 ) − q(θ1 )) ≥ 0

6.1 Basic Mechanism Design

215

This means that the supply level of incentive compatibility is non increasing. Based on the first-order optimal condition of efficiency set ∅, the infinity of incentive compatibility can reduce the difference of efficiency space and converge to monotonicity. This chapter still defines the utility of information rent U (E f f ) = t(E f f ) − E f f ∗ q(E f f ), and the optimization problem of the government as the principal in the supervision mechanism with continuous NGO efficiency is I

ne f f

[B(q(E f f )) − E f f ∗ q(E f f ) − U (E f f )] f (E f f )d(E f f )

max

U (·),q(·) Ef f

Consistent with the discrete efficiency space of NGOs, continuous incentive compatibility implies that most inefficient NGOs can implement the participation constraints. For most efficient NGOs, this kind of continuous regulatory mechanism design does not cause distortion effect, but can obtain positive information rent

θ2 U

SB

(E f f ) =

q S B (t)dt θ1

In general, this chapter set the basic correction mechanism design to determines the balance of rural long tail public service. Based on the information advantage of NGOs, the government (as the trustor) and NGOs (as agent) exist trade-off and conflict between information rent and supply efficiency in the interaction and purchasing service. Based on the single incentive compatibility and participation constraint mechanism, this conflict can be mitigated through dynamic adjustment, information mechanism and government regulation. Especially in the adverse selection of multi-dimensional supply objectives of NGOs, on account of efficiency type dependence, limited liability and random participation (liquidity), this chapter can use the principle of optimal mechanism display to effectively encourage NGOs to participate in the supply services most suitable for their own characteristics through nonlinear transfer payment costs.

6.2 Incentive Compatibility and Information Efficiency Mechanism The last section emphasizes that the government, as the moral bearer of public service, can transfer the NGOs through the way of purchasing services based on their own information disadvantage and financial constraints. However, because of the information asymmetry between government and NGOs, there is a trade-off between the distribution of information rent and the efficiency of correction between the principal and the agent. Although the diversified NGOs have the non-profitability and charity

216

6 Correction Mechanism of the Imbalance of Rural Long …

nature, they also selectively “produce” public services based on their own characteristics in the supply and have differences in efficiency. This section introduces a large number of NGOs as agents. Information asymmetry exists not only between government and NGOs, but also among various NGOs. This complex and multidimensional interaction makes the social mechanism decision-making have different interests. As the social welfare planner, the government sometimes lacks incentives to distinguish the real long tail demand and free-riding behavior due to public choice. For NGOs, it is possible to improve organizational welfare by falsely reporting their own efficiency. Therefore, the effective participation and corrective behavior of NGOs include not only incentive compatibility, but also limitation from resource dependence. Therefore, based on the long tail attribute of rural public service, the key of mechanism design is to achieve the expected social welfare goal (dynamic balance of supply and demand) and meet the incentive compatibility of individual NGOs. Conflicting goals and scattered demand information are the basic elements of incentive theory. This incentive compatibility mechanism includes five elements: external environment; social goals (dynamic balance of supply and demand); economic mechanism rules; solutions to individual “rational economic man” behavior of NGOs; implementation of social choice goals (incentive compatibility between individual interests and social goals). The details are as follows: ei = (Z i , wi , ≥i , Yi ): The economic characteristics of individual i of NGOs include behavior collection, initial endowment, demand preference and production capacity. (e1 , e2 , . . . , ei ) ∈ E, E is a set of all economic characteristics. U = U1 × · · · × Un is the set of all feasible utility functions. Based on the specific environment of rural long tail demand, all feasible external environment sets include initial endowment and production level. As a social planner, the government does not know the utility function of every NGO, and the individual NGOs do not know their respective utility functions. Z = Z 1 × . . . × Z n is all behavior spaces, where A ∈ Z is all feasible sets. F : E → A is the corresponding social choice function. Due to the long tail attribute of rural public services, it is difficult for the government to accurately identify the private characteristics of different demand individuals. Therefore, it needs to design an appropriate incentive mechanism to coordinate the optimal social goals and the interests of all stakeholders (including demand subjects and NGOs), so that different stakeholders have incentive choice behavior (including supply level and demand level) to achieve social optimum. In order to do so, social planners (mechanism designers) should formulate corresponding rules based on all available private information, including information space and behavior outcome function. Mi is information space of NGO i. M = M1 × · · · × Mn is information space for effective interaction and communication. h : M → Z is the function of information set transformed into behavior result. π = (M, h) is the corresponding mechanism

6.2 Incentive Compatibility and Information Efficiency Mechanism

217

design. This mechanism design is similar to the game between different stakeholders, and the results of all information spaces correspond to the utility functions. However, once the individual characteristics of different NGOs are determined, the implementation theory of mechanism design requires that the mechanism can realize the incentive compatibility for individual characteristics and social expected goals on the self-adaptation and adjustment behavior of different NGOs. Therefore, the supply level of different NGOs in the correction of rural long tail public services depends on their timely feedback of external resources and economic environment based on their own information and effective demand identification. Suppose b(e, π) is the supply equilibrium strategy of NGOs based on selfadjustment and adaptation, and the corresponding equilibrium result is the game rule function of this equilibrium strategy. Given mechanism π and equilibrium strategy b(e, π), this chapter studies the cross relationship between F(e) and b(e, π) based on the implementation mechanism of social choice rule F. This chapter defines the correction mechanism of the imbalance of rural long tail public services as (M, h). It satisfies that for all e ∈ E, the corresponding social choice in the correction strategy b(e, π) to achieve the balance is optimal and can be fully implemented. Where b(e, π) = ∅, there is a correction equilibrium solution. h[b(e, π)] = F(e), the individual interests of NGOs are consistent with social goals. This correction mechanism (M, h) can implement rule F in b(e, π), so it is the incentive compatible social choice function corresponding to b(e, π). For the rural long tail public service, because of its dual attributes of public goods and private goods, its mechanism design involves two aspects. On the one hand, it has the attribute of continuous public goods, and its supply level depends on the capital collected by the rural community as a whole (including user fees, taxes and donations). Let y be the quantity of the imbalance of this long tail service, and c(y) be the cost of correcting the imbalance, then the feasible set of this mechanism is   A = y, z 1 (y), z 2 (y), . . . z n(y) ∈ z i (y) = c(y) i∈N

where z i (y) is the share of supply and correction of long tail public service y undertaken by individual i, while the total social welfare to achieve correction is r(y). Therefore, the total surplus of social correction of rural long tail service is V(y) = r(y) −



ci (y)

i∈N

On the other hand, from the perspective of the divisible private nature of rural long tail public service, the number of consumers of each specific type of rural long tail public service may (with divisibility and exclusiveness). Let the  be very limited

N result space be Z = y ∈ {0, 1}: i=1 yi = 1 , where yi = 1 means that the net surplus of the consumer group serving for this particular rural long tail is

218

6 Correction Mechanism of the Imbalance of Rural Long …

vi (y) = vi yi Based on the dual attributes of rural long tail public service, the optimal correction level depends on the real value function vi (·) of NGOs. Let H: V → Z be the decision rule. If the following conditions are met i∈N

vi [h(vi )] ≥

   vi h vi i∈N

Then we call the mechanism (M, h) efficient. When for each NGO, the obtained information space is the collection of its external environment, that is, M = E, then the rural long tail public service has the corrective display mechanism. Among them, the most convincing correction mechanism is that in equilibrium, all NGOs, as agents of purchasing services, have the incentive to show their real efficiency. For this correction mechanism (E, h), if it satisfies e ∈ b(e, π ) h(e) ∈ F(e) Then the correction mechanism can effectively and truly implement the social choice rule F in b(e, π). Although the information space faced by the correction mechanism is uncertain, the Revelation Principle shows that as long as the information space of the revelation mechanism contains the individual characteristics and external environment information of all NGOs, the real revelation of their own efficiency characteristics is the dominant strategy for all NGOs. This display mechanism (E, h) can truly implement the social choice function rule F in the dominant strategy, and has the characteristics of strong incentive compatibility of individual NGOs. Therefore, the Revelation Principle establishes a mapping relationship between the correction mechanism (M, h) and the dominant equilibrium of real information features. In addition, the correction mechanism of the imbalance of rural long tail public services also needs to meet the information efficiency, that is, to minimize the information cost and transaction cost needed for the correction. The “market mechanism” of rural public service club goods (this kind of market is a hypothetical market, which can partly reflect the demand relationship through the service cost) can better realize the minimization of information cost and transaction cost. In this correction mechanism, there is a process of timely interaction, adjustment and processing of massive discrete demand information, and the balance must realize the dynamic stability of information exchange and identification. Based on the long tail attribute of rural public service, information processing and exchange have the characteristics of decentralization, which challenges the information efficiency of correction mechanism. It is assumed that the long tail of rural public demand individuals (rural residents) are j; its demand space is X j ; the utility function of consumption is R j . The

total consumption of rural long tail public services as a whole is N = Nj X j .The

6.2 Incentive Compatibility and Information Efficiency Mechanism

219

production possibility set of NGO i as the main body of imbalance correction is Yi ,and its initial resource endowment is Wi .Therefore,the economic features of the whole correction mechanism is E i j = X j , R j , Yi , Wi . In the transmission of demand information within the mechanism, effective mechanism design can accurately convey the demand information to every NGO and mechanism planner (government) without information distortion. In this process, the loss of information is reflected in the increase of transaction costs. Let the information space of NGO i be m i . The information set of the whole mechanism is Mi = in m i . In the process of dynamic adjustment and correction, different NGOs exchange demand information with other organizations based on their own demand information, and the information efficiency response function is   m i (t + 1) = ϑi m(t), E i j When information interaction and adjustment in the mechanism achieve dynamic stability, m ∗ becomes the steady equilibrium point of information response function   m i = ϑi m ∗ , E i j The result of resource allocation (long tail public service) is Z = h(m ∗ ), where h(·): M → Z is the result function of resource allocation. Therefore, the information efficiency of correction mechanism includes (M, Z, ϑ). The dynamic mechanism of information transmission and resource allocation is determined by the information supply and demand mechanism. Resource allocation rules can fully realize the mapping relationship between information and resources, so as to connect social planners, NGOs and demand stakeholders in the economic environment. In the correction mechanism, NGO i adjusts its correction behavior at t + 1 time dynamically only through its own information at time t, and has nothing to do with other NGOs and demand stakeholders. This chapter calls this information feedback mechanism as information dispersion correction mechanism: m i (t + 1) = ϑi (m(t), E i ) Although this information dispersion correction mechanism can guarantee individual information privacy, it may cause multi-dimensional and unbalanced tasks in a specific social optimal planning goal, resulting in too large information space and low efficiency. Therefore, only when the information space of the correction mechanism is the smallest among all the mechanisms that lead to Pareto Efficiency (the information cost is the smallest), the information mechanism is efficient. And the performance of this correction mechanism should satisfies Z = h(m) Under the principle of complete market commodity exchange, the “market mechanism” of rural public service club goods can effectively correct the imbalance. Define

220

6 Correction Mechanism of the Imbalance of Rural Long …

the resource allocation and information efficiency mechanism (M, Z, ϑ) under the economic environment E, which satisfies the following conditions: (1) (2) (3)

Dispersion of demand information Pareto resources allocation Meet the needs of individual rationality

Under the market-oriented conditions (who pays who enjoys), the rural long tail imbalance correction mechanism can achieve information efficiency (Tian, 2006). Therefore, the market-oriented competition mechanism, with private property rights of NGOs as the premise and diversified competition as the means, meets the requirements of individual rationality, effective allocation and incentive compatibility at the same time. It is similar to the utilization of information efficiency and cost by Lindahl mechanism (Hurwicz, 1979). Generally speaking, the correction mechanism of rural long tail public service includes two dimensions, one is the market information displayed by demand, the other is the resource allocation of public service. Based on the multi-dimensional and discrete nature of rural long tail public demand, it is difficult to achieve the integration of heterogeneous consumption utility function only through the government’s directive correction mechanism. Only when the information cost and transaction cost are small enough, the mechanism interval design based on market rule competition and division of labor is reasonable and effective.

6.3 Interval Design of Imbalance Correction Mechanism According to the transaction cost theory of new institutional economics, the boundary of an organization depends on the comparison between the transaction cost saved by the organization replacing the market and the internal governance cost caused by the existence of the organization (Coase, 1937). When the cost of internal transaction is equal to the cost of market transaction, the organization achieves the best scale. Williamson (1979) proposed three dimensions (or signs) to distinguish transactions, namely asset specificity, uncertainty and transaction frequency. In this chapter, we think that the marginal cost and long-term benefit of rural public service are also affected by the type of marginal compensation. When NGOs choose the optimal scale boundary, they will be affected by the agglomeration degree of local rural long tail public demand, and consider the balance of marginal compensation cost and marginal social benefit as for the asset specificity. In this correction mechanism, the higher the agglomeration degree of long tail public demand is, the greater the positive spillover it is satisfied with, and the greater the marginal social benefit is. The marginal benefit curve MB will tilt slightly to the upper right and convex to the origin (M B  (D A ) > 0, M B  (D A ) < 0, slope K M B is smaller, as shown in Fig. 6.1). According to the new institutional economics, the degree of distribution of demand is related to the transaction cost and governance cost of supply. The higher the degree of demand agglomeration is, the lower the

6.3 Interval Design of Imbalance Correction Mechanism

221

Fig. 6.1 The interval of correction mechanism on aggregation degree of long tail public demand

marginal cost of the organization replacing the market to meet the demand (to meet the economies of scale) (Coase, 1937). However, the decreasing trend tends to be stable with the saturation of agglomeration degree MC  (D A ) < 0, MC  (D A ) < 0. When the government supplies public services, it will give priority to the “head” public demand (Douglas, 1987). In the marginal cost of government correction MC g = M B, the intersection B realizes cost compensation and profit-loss balance. It meets the internal interval Dg of the mechanism with corresponding equilibrium marginal cost MC g∗ . For NGOs, their inherent attributes make them have comparative advantages in meeting the long tail public demand with less agglomeration degree. On the one hand, the non-distributive constraints (Steinberg, 2003) make them pay more attention to the social benefits rather than economic benefits of demand satisfaction. On the other hand, their mobile and information costs (Salamon, 1987) make their marginal cost of public service supply MCs smaller, so its cost compensation point (intersection A with MB) is smaller than that of B in the degree of public demand agglomeration. Institutional economics believes that the advantages of NGOs are reflected in the improvement of asset specificity (specialization for a certain type of demand), the reduction of uncertainty (local “embeddedness” and social capital) and the increase of “transaction frequency” (customized supply and path dependence) (Williamson, 1979). From the perspective of cost compensation, in this correction mechanism, NGOs are suitable to correct the supply and demand imbalance of interval Ds between marginal cost compensation point A and B. And in the mechanism interval Di to the left of A, as the degree of demand agglomeration is too small, even NGOs cannot

222

6 Correction Mechanism of the Imbalance of Rural Long …

make ends meet through internal marketization, which has become a neglected long tail public demand imbalance. In this kind of correction mechanism, it is assumed that NGOs in different regions carry out “customized correction” according to the specific long tail public needs of their regions. In the long tail public demand interval Ds of NGOs, the corresponding marginal compensation cost of demand with different degree of agglomeration is also different. First of all, this is reflected in the fragmentation of geographical environment in the region. As shown in Fig. 6.2, compared with flat areas such as plains and hills, the geographical environment in mountainous and plateau areas is more fragmented. The relatively high transaction cost (including transportation convenience) of these areas makes the long tail public demand more dispersed and fragmented (Xie et al., 2019). It is shown in Fig. 6.2 the movement from point P to point M along the curve MCs . And the corresponding marginal compensation cost in mountainous and plateau areas MC M is higher than in plain and hilly area MC P . Institutional economics holds that by expanding the scale of individual production and supply, organizations can internalize external costs, reduce transaction costs and governance costs, and thus reduce marginal production costs. NGOs in mountainous and plateau areas with larger marginal compensation cost MC M need to rely on larger individual scale to reach or even surpass MC M . Therefore, the scale of individual organization in mountainous and plateau areas S M should be larger than the scale of individual organization in plain and hilly areas S P .

Fig. 6.2 The impacts of aggregation degree of long tail public demand on the correction marginal cost of NGOs

6.3 Interval Design of Imbalance Correction Mechanism

223

In addition, the agglomeration degree of public demand in urban and rural areas is also different. Compared with cities with concentrated residence and more convenient transportation, rural areas are more vulnerable to the impact of long tail public demand agglomeration due to more fragmented transportation and communication, higher compensation point of marginal cost correction (Gao et al., 2019). In addition to the agglomeration degree of demand, the asset specificity of different types of NGOs will also have an impact on their internal governance costs and the optimal individual size of organizational development. Institutional economics believes that with the improvement of the asset specificity of organizations, the internal organization can realize the internalization of external costs by coordinating transaction costs, so as to reduce the internal governance costs of production (Williamson, 1998). For NGOs with higher degree of asset specificity, they have higher opportunity cost after entering the corrective “market mechanism”. They are more motivated to improve the demand agglomeration by summarizing the demand, so as to make them along the internal governance cost curve I C S moves down right to reduce the cost of internal governance, namely I C S (K ) < 0, where K is the degree of asset specificity (as shown in Fig. 6.3, K 1 to K 2 ; suppose K 2 > K 1 ). NGOs with different asset specificity can adapt to the changes of internal governance costs by adjusting the scale of individual organizations. Organizations with higher asset specificity tend to have smaller individual scale due to smaller internal governance cost and marginal cost compensation point. To sum up, the differences in geographical environment, urban–rural distinction, organization type, transaction cost and other aspects in different regions have an

Fig. 6.3 The impacts of asset specificity on the internal governance cost of correction mechanism (K 2 > K 1 )

224

6 Correction Mechanism of the Imbalance of Rural Long …

impact on the long tail public demand agglomeration and organizational asset specificity, thus putting forward different requirements on the organizational scale required for the marginal cost compensation of supply. Based on this, the optimal correction scale function of NGO constructed in this chapter is: S = ϕ[MC(D A , K ), I C(K )] For the same subject of correction: S  (MC) > 0, S  (I C) > 0, MC  (D A ) < 0, MC  (K ) < 0, I C  (K ) < 0 When the scale of individual correction is optimal,         MC ∗ D ∗A , K ∗ = M B ∗ D ∗A , K ∗ , I C ∗ K ∗ = M B ∗ K ∗ The agglomeration degree and asset specificity of long tail public demand affect the marginal compensation cost and internal governance cost of demand satisfaction through different mechanisms, so as to affect the optimal scale boundary of correction mechanism. Based on the relevant theories of new institutional economics, this chapter holds that when NGOs correct the imbalance between supply and demand of rural long tail public services, there is also a logical relationship between the optimal individual size and supply cost. Only when the degree of demand (including quantity and quality) reaches a certain level, and the marginal benefit equals to the marginal cost of supply, can NGOs have the motivation to take the form of organizations instead of the “market” (or government) to correct, so as to realize the compensation of marginal cost. Therefore, the optimal scale of this correction mechanism depends on the internal governance cost, the marginal compensation cost of external costs, as well as the opportunity cost of entering the “market”.

6.4 Groves-Clark Correction Mechanism Based on the public goods attribute of rural long tail public service, its decentralized supply mode in the correction mechanism is easy to cause accurate identification distortion and free-riding problems. Especially for its long tail attribute, private information of demand is easy to lead insufficient supply. The Groves Clark correction mechanism of rural long tail public service can be used to solve the supply imbalance problem of discrete public goods (Bergemann & Välimäki, 2002, 2019). Assuming that there are n NGOs as agents in the whole correction mechanism, ci is the cost of providing and correcting long tail public services for NGOs. ri is the return of NGO i in supply, including certain economic compensation, social reputation, altruism, etc. For NGO i, the net surplus obtained by its participation in the supply and correction of specific rural long tail public services is vi = ri − ci . From the perspective of

6.4 Groves-Clark Correction Mechanism

maximizing the welfare surplus of the whole society, when

225

N

i=1 vi

=

N

ri −ci ≥ 0,

i=1

the correction mechanism of rural long tail public service is the most effective. However, for NGO i, the obtained net surplus ri is personal information and not easily recognized. The Groves Clark correction mechanism of rural long tail public service is used to solve the problem of encouraging different NGOs to show their real net surplus. Assuming that based on the private attribute of rural long tail public service, the net utility function of NGO i is semi-linear, then its utility level can be expressed as u i (ti , y) = ti + vi y where ti is the transfer payment obtained by undertaking the purchase of services, and y is the supply output. In this Groves Clark mechanism, NGOs are required to report their net surplus to the government (or public information platform). We can define this mechanism π = (M1 , M2 , . . . Mn , t1 , t2 , . . . tn , y) (1)

(2)

bi ∈ Mi . When each NGO undertakes the purchase of rural long tail public services, it reports its own net surplus value vi (not necessarily its true net surplus vi ), and provide a “bid price” bi to the government. Each NGO i receives a transfer payment ⎧ n ⎨ b when b ≥ 0 j i ti = j =i i=1 ⎩ 0

Therefore, in the Groves Clark correction mechanism, the utility function of NGO i is ⎧ n ⎨ v + b when b ≥ 0 i j i ui = j =i i=1 ⎩ 0 It can be proved that this mechanism can encourage each NGO to report its own real net surplus as the dominant strategy. However, it is difficult to truly reveal the net surplus through the tax level because Chinese government encourages NGOs to implement tax exemption policy. As mentioned above, in terms of mathematical definition, the demand level of rural long tail public service may not be discrete individuals, but continuous and integral on the long tail curve. Therefore, based on the mathematical integrability of its public attributes, this section can further rewrite the budget limit of NGO i as

226

6 Correction Mechanism of the Imbalance of Rural Long …

ci (y) = di + ti where di is the amount of social donations (including various human, material and financial resources) obtained for NGOs. As u i (ti , y) = ti + vi y, this chapter can draw the conclusion that for the total social cost of the correction mechanism C(y) =

n

di +

i=1

n

ti

i=1

Therefore, the equilibrium of Pareto effective allocation of this correction mechanism is max wi u i (ti , y) n

s.t. C(y) =

di +

i=1

n

ti

i=1

where wi is the weight of NGO i in the Groves Clark correction mechanism. For the semilinear utility function of NGO, the corresponding weight wi of all NGOs must be equal to ρ, and ρ is the Lagrange multiplier corresponding to Pareto efficient allocation. The maximum utility function can be transformed into   n (ti + vi (y)) Max i=1

The Lindahl-Samuelson equilibrium condition is n ∂u i (y) i=1

∂ yk

=

∂c(y) ∂ yk

In the Groves-Clark mechanism, in order to effectively realize the correction mechanism based on the imbalance of rural long tail public services, the government can claim that the mechanism designed by the government is to maximize the level of public services max y

n

bi (y)

i=1

Therefore, the Groves-Clark correction mechanism is (V, H), where V = V1 × · · · × Vn contains the information space set bi (y) brought by all utility function. h = (t1 (b) × · · · × tn (b), y(b)) are all behavior result functions, which are determined by the following conditions:

6.4 Groves-Clark Correction Mechanism

(1) (2)

227

After identifying the local specific rural long tail public demand, each NGO i reports its value function bi (y) (not necessarily equal to real vi (y)). The imbalance level of rural long tail public service y(b) satisfy max y

(3)

n

bi (y)

i=1

In the process of NGO i participating in the supply and correction of the imbalance, the transfer payment fee paid by the government is ti =

j =i

b j (y)

The net surplus of NGOs participating in supply is vi (y) + j =i b j (y). In this mechanism design, every NGO i has the motivation to truly report its own value

funcb j (y). tion bi (y) = vi (y), which maximizes the net surplus of NGOs vi (y) + j =i

Through reporting bi (y) = vi (y), NGOs are convinced that social welfare can be consistent with individual interests under the Lindahl-Samuelson condition. Therefore, report bi (y) = vi (y) is a dominant equilibrium strategy.

6.5 Nash Equilibrium Correction Mechanism The Hurwitz impossibility theorem (Hurwicz, 1979) holds that Pareto efficiency and mechanism design of individual rational allocation cannot realize the theory of dominant strategy implementation. Therefore, this section uses Nash equilibrium to describe the mechanism design based on the self-adaptive and dynamic adjustment of individual interest behavior of NGOs, and realizes Pareto optimal distribution and incentive compatibility. This Nash equilibrium correction mechanism (M, h) satisfies     h i m ∗ ≥ h i m i , m ∗−i where m ∗ is the Nash equilibrium solution of NGO I; m ∗−i denotes the Nash equilibrium solution of NGO except I; ≥ denotes weak superiority. Therefore, this kind of correction mechanism (M, h) is Nash-implementable under the social choice rule F, that is, N E(e, π ) = ∅, h(N E(e, π ) ∈ F(e). NE means Nash equilibrium. For the rural long tail public service revelation mechanism (E, h), when the real long tail demand information revelation set e∗ is Nash equilibrium, it meets   h ei∗ , e−i ≥ h(ei , e−i )

228

6 Correction Mechanism of the Imbalance of Rural Long …

However, it is important to note that when this correction mechanism is Nash equilibrium, all social selection rules must be fully implemented. The social selection rules under the Nash incentive compatibility mechanism must satisfy Maskin’s monotonality (Maskin, 1999). Based on the long tail attribute of rural public services, the recognition and implementation mechanism of different demand individuals is more complex. Especially for discrete rather than continuous long tail demand, the small difference in the supply and correction level of NGOs may lead to the deviation of optimal distribution results. This information space containing preference with differentiation and heterogeneity is multidimensional. That is, the optimal Nash equilibrium of rural long tail public service correction mechanism includes both the equilibrium solution of private goods dimension and public service dimension. Based on the dual attributes of rural long tail public service, this section can divide it into two different kinds of goods for analysis and solution. Set the private attribute of rural long tail public service Pr, the public goods attribute Pu, and the utility function v. Mi is the supply level that NGO i is willing to undertake voluntarily, while G i (M) is the voluntary supply level Mi that NGOs supply based on self-adaption and self-adjustment. Therefore, the equilibrium solution of this mechanism is N

Pr +

i=1

N i=1

G i (M) =

N

Pr + Pu

i=1

In this case, the net residual function of NGO is vi (M) = u i (xi (m), y(m)). When u i is the concave to the origin, the first-order optimization can achieve the Nash equilibrium of the correction mechanism N

∂u i ∂ Pu ∂u i i=1 ∂ Pr

Namely

=1 N

M RS Pu Pr = M RT S PuV

i=1

Therefore, the Lindahl-Samuelson equilibrium condition means that the Nash equilibrium distribution of the correction mechanism can achieve Pareto efficiency. If we further consider the rational distribution principle of individual NGOs, we n m i is the public can set the supply function y = f (v) = v. And Pu(m) = i=1 attribute of rural long tail public service; pi (m) is the differentiated evaluation of the same rural long tail public service based on the individual heterogeneity of divergent needs. Ii (m) = pi (m)Pu(m) is the supply cost of NGO i, and Pr (m) is the private consumption of rural long tail public services. Based on the above assumption, the budget limit of NGO i is Pr(m) + pi (m)Pu(m) = ti + di

6.5 Nash Equilibrium Correction Mechanism

229

In this case, the equilibrium solution of rural long tail public service correction mechanism can achieve Pareto efficiency on the basis of individual rational distribution of NGOs. The net surplus condition of NGO i is vi (m) = u i (Pr, Pu) Based on its dual attributes, the first-order optimal condition of rural long tail public service is ∂vi ∂u i ∂u i pi (m) + =0 =− ∂m i ∂ Pr ∂ Pu ∂u i ∂ Pu ∂u i ∂ Pr

= pi (m)

When u i is concave to the origin, the design of this mechanism can realize both Lindahl equilibrium and Nash equilibrium. Finally, this section considers that information asymmetry occurs not only between the government (as principal) and NGOs (as agent), but also between different individuals within NGOs. Based on the self-adaptation, interaction and adjustment among individuals of different NGOs, this section can set the Bayesian Nash equilibrium of rural long tail public service correction mechanism: Although individuals of different NGOs do not know each other’s demand preferences and economic features, they know the distribution probability of each feature. The incentive compatibility mechanism of Bayesian Nash equilibrium sets the utility function of each NGO as u i = (x, θi ), where θi is the type set of different NGOs. All NGOs know the different types of set θi is prior distribution. Based on their own type sets θi , every NGO can get that the conditional distribution probability of other NGOs is D(θ−i |θi ) =  θn θ1

D(θi , θ−i ) D(θi , θ−i )dθ−i

For mechanism (M, h), select m i as the function of θi , namely gi : θi → m i . Suppose Si is the set of all the strategies of individual i, then the expected utility function of supply and correction of NGO i in θi is

θn Wi (gi , θi ) =

 u i h(gi (θ ), θi ]D(θ−i |θi )dθ−i

θ1

When gi , θi satisfy   Wi (gi , θi ) ≥ Wi gi , g−i , θi

230

6 Correction Mechanism of the Imbalance of Rural Long …

Then the correction mechanism of rural long tail public service is Bayes Nash equilibrium mechanism, which depends on the social and economic environment (e, π) of long tail demand. Bayes-Nash incentive compatibility mechanism involves the relationship between F(e) and (e, π). When B(e) ∈ F(e), this Bayes-Nash equilibrium implements the social choice rule F. Therefore, based on the discreteness of rural long tail public service demand, the Bayesian-Nash equilibrium of this correction mechanism is implementable only when the social choice rule F satisfies Bayesian monotonicity and is compatible with Bayesian Nash incentive (Tian, 1999, 2004).

6.6 Dynamic Adjustment Mechanism As mentioned above, there is a dynamic adjustment and feedback based on the information, resource exchange and interaction of different stakeholders in the correction mechanism of the imbalance of rural long tail public services, rather than a static mechanism. Suppose that in this dynamic adjustment mechanism, different NGOs have divergent initial endowment Wi , feasible technical constraints Yi , information space m i and commitment ability. In the process of dynamic multi period game, NGOs can show their real information space in period t, but they may not have the ability of future correction commitment, which makes the information rent of period t + 1 invalid. Based on this dynamic volatility, NGO i has incentive to not really reveal or fully reveal its information space in period t, thus increasing its game ability of information rent in period t + 1. In addition, there is a trade-off between information rent and information efficiency. When the efficiency type of all NGOs θi are fully revealed, social planners have incentives to eliminate or overcome these information distortions, which may change the incentive compatibility and participation constraints of other NGOs. Referring to the model of Athey and Segal (2013), this section assumes that there are two adjacent periods t and t + 1, where σ is the discount factor. It is assumed that NGO i has private information θi = {θ1 , θ2 }, which needs accurate identification based on its own information advantages. The probabilities of these two kinds of information are p and 1 − p respectively. The long tail public service supply of NGO i in t period is xit , the social value (including social reputation, altruism and philanthropy) obtained through supply and correction is vit (xit ). The production cost is cit (xit ), and the sum of transfer payment, social donation and usage fee is pit (xit ). Therefore, from the total optimization of the whole time dimension t and t + 1, the intertemporal dynamic utility function of NGOs is U(xit , xit+1 ) =

2 t=1

σ t−1 (θi vit (xit ) + pit (xit ) − cit (xit ))

6.6 Dynamic Adjustment Mechanism

231

In this dynamic mechanism, the expected utility obtained by the mechanism planner (governments) through the purchase services is π( pit (xit ), pit+1 (xit+1 )) =

2

σ t−1 pt ( pθ1 vit (xit ) + (1 − p)θ2 vit (xit ))

t=1

Based on the information advantage of NGOs in providing rural long tail public services, they have partial commitment ability in the mechanism design of purchase services. Therefore, this mechanism can be designed as: in the period t, government as the client design and purchase services ( pit (xit ), pit+1 (xit+1 )), and NGOs choose to purchase services xit . In the period t + 1, the government adjusts the purchase services scheme based on thedynamic  feedback of demand information, and adds a   xit+1 in the period t + 1. The initial purchase service new mechanism design pit+1 mechanism pit+1 (xit+1 ) coexist with the  new one. In period t + 1, NGO i can choose   xit+1 and pit+1 (xit+1 ). between two mechanisms pit+1 It can be proved that there is always a refined Bayesian equilibrium mechanism in the design of this dynamic intertemporal purchase mechanism, which makes the dynamic equilibrium mechanism of intertemporal negotiation consistent with the static supply mechanism (Ausubel et al., 2002; Liu et al., 2018). In this refined Bayesian equilibrium mechanism, NGOs in the category of private information θ1 and θ2 meet the incentive compatibility and participation constraints, so they can effectively implement the mechanism. In addition, due to the partial commitment capacity of NGOs, the commitment of mechanism planners in period t + 1 is not credible. Therefore, this dynamic adjustment mechanism has different degrees of information advantages for NGOs. Mechanism designers can design sequential screening mechanism to realize the real revelation of private information of NGOs on the basis of mechanism learning (Courty & Li, 2000). On the one hand, the earlier the private information is screened, the lower the information rent of the mechanism. On the other hand, because the information is constantly updated, the earlier the private information is screened, which may lead to the deviation of resource allocation. Therefore, there is a trade-off between information rent and resource allocation efficiency based on time fluctuation. This dynamic adjustment mechanism can also achieve the budget balance between the government as the principal and the NGO as the agent. Due to the difference in the behavior choices of stakeholders in different periods, the incentive restraint of the mechanism will change. Assume that the cost of NGO i participating in the supply correction imbalance in period t is Cit (θit , xit ) = θit xit When all the private information is fully revealed, the optimal supply level of NGO i in period t satisfy max xit − θit xit xit

232

6 Correction Mechanism of the Imbalance of Rural Long …

The expected optimal level of NGO’s self-supply from period t to t + 1 satisfy max xit+1 − θit xit xit+1

In order to realize that the incentive compatibility under this dynamic adjustment mechanism is still effective, we refer to Athey and Segal (2013) to establish an effective dynamic mechanism of budget balance, and coordinate the trade-off between incentive compatibility and budget balance through transfer payment. Specifically, let incentive compatible transfer payment fees in t and t + 1 Tt (θt ) and Tt+1 (θt+1 ). It can be proved that the optimal incentive compatible transfer payment in the effective dynamic mechanism of budget balance is Tt+1 (θt+1 , θt ) = −

 1 (θt+1 )2 − (Eθt+1 )2 θt

where Eθt+1 is the efficiency type that in period t NGOs expect themselves to be in period t + 1, and its first-order optimal condition satisfies Eθt+1 = θt+1 . So it is consistent with the optimal incentive compatibility. On the other hand, the transfer payment fees between  NGO i  for budget balance  i  i  ij j j ij j j i i and j in t and t + 1 are Tt = Tt θt − Tt θt , Tt+1 = Tt+1 θt+1 − Tt+1 θt+1 ij

ij

respectively. Although NGOs i and j interact with each other Tt and Tt+1 . But when both sides show their own type θ, Eθt and Eθt+1 is independent and unaffected. Therefore, ij

ij

Eθt+1 Tt , Tt+1 = 0 The total transfer payment of correction mechanism as a whole is ij

ij

Tt (θt , θt=1 ) = Tt (θt ) − Tt+1 (θt , θt+1 ) ij

Tt+1 (θt , θt=1 ) = −Tt (θt , Eθt+1 ) This transfer payment can satisfy both incentive compatibility and participation constraints. The transfer payment of individual i to j of NGOs is equal to the change of the present value of expected utility in dynamic corrective behavior adjustment for other NGOs. Therefore, this correction mechanism based on sequential screening and dynamic information adjustment can effectively achieve Bayesian equilibrium optimization.

6.7 Accurate Matching Mechanism

233

6.7 Accurate Matching Mechanism Due to the lack of market mechanism (market failure) in rural long tail public service, it is difficult to show the real preference of demand and accurately match through price mechanism. For rural long tail public service, its demanders and suppliers (mainly NGOs) meet the bilateral multi-to-multi matching mechanism (each has a large number, with its own demand or efficiency preference). Suppose that the long tail public service demand in economic environment is D = (D1 , D2 , . . . Dn ), and the supplier is S = (S1 , S2 , . . . Sm ), The preference order of the demander and the supplier is O D and O S . Preference limit (sort number type) is Q D and Q S . Define the precise matching mechanism of rural long tail public service demand as M: D ∪ S → 2D∪S This bilateral multi-to-multi matching satisfies the following requirements M(D) ∈ S, M(D) ≤ Q D M(S) ∈ D, M(S) ≤ Q S M(D) ∈ S only when M(S) ∈ D For this kind of precise matching mechanism, when it satisfies M(a) = Ra (M(a), >a ), ∀a ∈ D ∪ S Ra is the rational choice of stakeholder a (demander and supplier). If >a represents the strict preference of a, we call this demand identification matching mechanism satisfying individual rationality. When the matching mechanism satisfies individual rationality, and there is no incentive mechanism to prevent pairing (D, S), we call the matching pairwise stable, denoted as S(O D , O S ). When this matching mechanism meets the responsive preference (Roth, 1985), this stable matching can achieve multi-to-one core setting. The same NGO accurately identifies and matches multiple consumers with the same type of long tail demand preference. However, when the multi-to-multi precision matching mechanism satisfies the  individual rationality, and there is no block set A, M  , where A ∈ D ∪ S, M  (a)/M(a) ∈ A, ∀a ∈ A M  (a) >a M(a)   M  (a) = Ra M  (a), >a The precision matching mechanism has the property of set stability, and all preference order O D and O S can keep stable matching. For the incentive compatibility of accurate multi-to-multi matching mechanism, the NGO set could meet the type limit saturation, and each stakeholder (demander and supplier) prefer to meet the

234

6 Correction Mechanism of the Imbalance of Rural Long …

maximum-minimum standard. The optimal rural long tail public service precise matching can achieve the weak Pareto efficiency, and it is the dominant strategy to show the truly efficiency type preference (Jiao & Tian, 2015). In the supply process of rural long tail public services, NGOs can realize the compensation of supply cost by charging a certain usage fee (not for profit), and better identify and reveal the real demand preference of rural residents. The NGOs match with the demanders in correcting the imbalance, and the obtained service usage fee is Fe S . The utility obtained by charging a certain usage fee is U DS (Fe S ). U DS (Fe S ) meets the following condition U DS (Fe S ∪ D) − U DS (Fe S ) ≤ FeUDS where FeUDS represents the maximum usage fee acceptable to the long tail consumer. This assumption means that for each NGO i, the marginal use cost of any long tail demand matched by its supply does not exceed the maximum consumption fee acceptable to the consumer. By charging a certain usage fee, stable core matching allocation can be realized (Hatfield & Milgrom, 2005). In addition, the stable matching of this mechanism design needs to meet the non-substitutability or complementarity of individual needs. Especially when the long tail needs meet the agglomeration (the larger the demand opportunity set, the greater the willingness of NGOs to accurately match), this matching mechanism can meet the incentive compatibility and participation constraints. In order to achieve this matching mechanism, we can use deferred-acceptance algorithm (Floréen et al., 2010; Liu et al., 2017) to calculate and achieve stable matching. When the long tail demand subject proposes a demand satisfaction scheme, the matching obtained by the deferred-acceptance algorithm can satisfy the incentive compatibility in the preference disclosure of the supply subject, and it is also optimal for the demand subject. In this accurate matching mechanism, deferred-acceptance algorithm is, for the sake of simplification, assuming that the demander is m and l; the supplier is i and j; and their respective preference order satisfy Si >m S j , Si >l S j Dm >i Dl >i ∅ (Dm , Dl ) > j Dm > j Dl > j ∅ Therefore, the possible accurate matching set is   X = (Dm , Si ), (Dm , S j ), (Dl , Si ), (Dl , S j ) Assume that the needs are matched by the NGOs to identify accurately. The algorithm starts from X D (0) = X, X S (0) = ∅.

6.7 Accurate Matching Mechanism

235

 In the first stage, NGO i chooses matching (Dm , Si );NGO j selection matching Dm , S j . Then the unmatched set in the first stage are Dm , S j and (Dl , Si ). The  matching set received by all NGOs in the first stage is X S1 = (Dm , Si ), Dl , S j . In the second stage, NGOs i still choose matching (Dm , Si ), and NGO j continues stage to choose match (Dm , S j ). The unmatched set   is (Dl ,Si ). The  in the second  matching set received by all NGOs is X S2 = (Dm , Si ), Dl , S j , Dm , S j . In the third stage, the long tail demanders choose to accept the matching set issued by all NGOs in the second stage, then      X D2 = (Dm , Si ), Dl , S j , Dm , S j where (X D2 , X S2 ) constitute the iterative fixed points of the matching set. Similarly, it can be proved that when the long tail demanders actively match based on the supply level of NGOs, the symmetry of the matching set makes the iterative fixed point satisfy the following conditions (1)

(2)

The deferred-acceptance algorithm based on maximum initial matching (X D , X S ) = (X, ∅) converges monotonically to the maximum iterative fixed point X DMax , X SMax . Its matching X DMax ∩ X SMax is the optimal stable matching for the long tail demand subject. The deferred-acceptance algorithm based on minimum initial matching (X D , X S ) = (∅, X) converges monotonically to the minimum iterative fixed point X DMin , X SMin . Its matching X DMin ∩ X SMin is the optimal stable matching for the long tail demand subject.

In general, this deferred-acceptance algorithm can be better applied to the accurate matching mechanism of rural long tail public services, so as to better realize the dynamic adjustment and correction of the imbalance between supply and demand. This is the key for the demand dispersion and indivisibility of rural long tail public services (services cannot be separated from service providers, such as long tail special education and health).

References Athey, S., & Segal, I. (2013). An efficient dynamic mechanism. Econometrica, 81(6), 2463–2485. Ausubel, L. M., Cramton, P., & Deneckere, R. J. (2002). Bargaining with incomplete information. In Handbook of game theory with economic applications (Vol. 3, pp. 1897–1945). Bergemann, D., & Välimäki, J. (2002). Information acquisition and efficient mechanism design. Econometrica, 70(3), 1007–1033. Bergemann, D., & Välimäki, J. (2019). Dynamic mechanism design: An introduction. Journal of Economic Literature, 57(2), 235–274. Börgers, T., & Krahmer, D. (2015). An introduction to the theory of mechanism design. Oxford University Press. Coase, R. H. (1937). The nature of the firm. Economica, 4(16), 386–405. Courty, P., & Li, H. (2000). Sequential screening. Review of Economic Studies, 67(4), 697–717.

236

6 Correction Mechanism of the Imbalance of Rural Long …

Douglas, J. (1987). Political theories of nonprofit organization. In The nonprofit sector: A research handbook (Vol. 43, p. 45). Floréen, P., Kaski, P., Polishchuk, V., & Suomela, J. (2010). Almost stable matchings by truncating the Gale-Shapley algorithm. Algorithmica, 58(1), 102–118. Gao, T., Erokhin, V., & Arskiy, A. (2019). Dynamic optimization of fuel and logistics costs as a tool in pursuing economic sustainability of a farm. Sustainability, 11(19), 5463. Hatfield, J. W., & Milgrom, P. R. (2005). Matching with contracts. American Economic Review, 95(4), 913–935. Hurwicz, L. (1979). Outcome functions yielding Walrasian and Lindahl allocations at Nash equilibrium points. The Review of Economic Studies, 46(2), 217–225. Jiao, Z., & Tian, G. (2015). The stability of many-to-many matching with max–min preferences. Economics Letters, 129, 52–56. Liu, A., Song, H., & Blake, A. (2018). Modelling productivity shocks and economic growth using the Bayesian dynamic stochastic general equilibrium approach. International Journal of Contemporary Hospitality Management. Liu, Y., Zhang, L., Tao, F., & Wang, L. (2017). Resource service sharing in cloud manufacturing based on the Gale-Shapley algorithm: Advantages and challenge. International Journal of Computer Integrated Manufacturing, 30(4–5), 420–432. Maskin, E. (1999). Nash equilibrium and welfare optimality. The Review of Economic Studies, 66(1), 23–38. Roth, A. E. (1985). The college admissions problem is not equivalent to the marriage problem. Journal of Economic Theory, 36(2), 277–288. Salamon, L. M. (1987). Of market failure, voluntary failure, and third-party government: Toward a theory of government-nonprofit relations in the modern welfare state. Journal of Voluntary Action Research, 16(1–2), 29–49. Steinberg, R. (2003). Economic theories of nonprofit organizations. In The study of the nonprofit enterprise. Springer. Tian, G. (1999). Bayesian implementation in exchange economies with state dependent preferences and feasible sets. Social Choice and Welfare, 16(1), 99–119. Tian, G. A. (2004). Unique informationally efficient allocation mechanism in economies with consumption externalities. International Economic Review, 45(1), 79–111. Tian, G. (2006). The unique informational efficiency of the competitive mechanism in economies with production. Social Choice & Welfare, 26(1), 155–182. Williamson, O. E. (1979). Transaction-cost economics: The governance of contractual relations. The Journal of Law and Economics, 22(2), 233–261. Williamson, O. E. (1998). The institutions of governance. The American Economic Review, 88(2), 75–79. Xie, X., Zhang, A., Wen, L., & Bin, P. (2019). How horizontal integration affects transaction costs of rural collective construction land market? An empirical analysis in Nanhai District, Guangdong Province, China. Land Use Policy, 82, 138–146.

Chapter 7

The Application of the Correction Mechanism: Internet + NGO

Abstract This chapter is the application of the correction mechanism: Internet + NGO. The supporting NGOs (umbrella organizations and hub NGOs) integrates the information, resources, social capital and other essential elements of the supply and demand sides. Although it does not participate in the supply of specific rural long tail public services, it is committed to providing intelligence, human resources, information and other support for other NGOs. It establishes the communication platform and security mechanism between NGOs and the government, and play the role of network platform for operation mechanism. First, this chapter analyzes the network externality of imbalance correction mechanism. In the view of “Internet + NGO”, the corrective mechanism of imbalance in rural long tail public services has network externalities. The network externality of this mechanism is reflected in that the more dispersed the long tail demand is, the larger the extension scope of the network is. The consumption of the long tail demand is affected by the complementary consumer goods of geography, industry and kinship in the network. Second, the “Internet+” economy has become full of information and information asymmetry will also be weakened. At the same time, this network economy has externality. As mentioned in the previous section, network externality refers to the network effect that market equilibrium fails to fully reflect the cost and benefit. This externality of the network comes from the interaction between its systematic and internal information flow, and the monopoly of network infrastructure. The network platform organization, represented by supporting NGOs (umbrella organizations and hub NGOs), integrates the information, resources, social capital and other essential elements of the supply and demand sides. Although it does not participate in the supply of specific rural long tail public services, it is committed to providing intelligence, human resources, information and other services and support for other NGOs. It could establish a communication platform and security mechanism among different NGOs and between NGOs and the government, playing the role of balance network platform operation mechanism for rural long tail public service. The practical mechanism of “Internet + NGO” for rural long tail public service has dual layers. The first layer is the networking information and resource exchange platform. The second layer is the networking demand and supply matching platform. Finally, this chapter further elaborates the practical operation of the network platform model of the rural long tail public service supply and demand equilibrium through the case study of the Internet platform of © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 J. Luo, Rural Long Tail Public Service and the Correction Mechanism, https://doi.org/10.1007/978-981-16-4023-0_7

237

238

7 The Application of the Correction Mechanism: Internet + NGO

JD.com public welfare foundation. The JD.com public welfare foundation is not only an individual of the foundation (NGO), but also a public service Internet platform based on its big data Internet of things technology and the supply chain of JD.com Mall. With the integration of social resources, the JD.com Public Welfare Foundation has carried out social innovation in the fields of rural poverty alleviation, disaster relief, education, environmental protection and social innovation. Through its public fund and material collection platform, the information exchange and supply-demand matching mechanism between rural long tail demand and fragmented rural public service supply can be effectively realized.

7.1 Network Externality of Imbalance Correction Mechanism In the view of “Internet + NGO”, the corrective mechanism of imbalance in rural long tail public services has network externalities. The network externality of this mechanism is reflected in that the more dispersed the long tail demand is, the larger the extension scope of the network is. The consumption of the long tail demand is affected by the complementary consumer goods of geography, industry and kinship in the network. It is assumed that NGO i can accurately identify the long tail demand of consumers (rural residents) with continuous preference characteristics in the network platform, and effectively supply the long tail public services with network externalities in rural areas. The utility function of NGO i in the network platform is V (q) = N V (Q) + Fe(q) + D + T (q) − cq    where q is the demand quantity of long tail accurately identified; Q = j f θj q is the information scale of network platform; θ V (q) is the private value of long tail public service consumption; N V (Q) is the network value of long tail public service consumption. Fe is the usage fee (similar to club goods) for consuming a specific type of long tail public service. D is the social donation (including human, material and financial resources) obtained by NGOs. T is the transfer payment for governments’ purchasing service, and C is the cost of providing a specific long tail public service. It is assumed that the long tail demand preference of rural residents   is θ j . The proportion of long tail demand individuals with preference θ j is f θ j . θ = θ j − θ j−1 is the difference between the two adjacent long tail demand preferences. Suppose the function distribution function   distribution    andcumulative  of θ j are f θ j = Pr θ = θ j and F θ j = Pr θ ≤ θ j . The utility function of long tail consumers θ j is U = θ V (q) + N V (Q) − Fe

7.1 Network Externality of Imbalance Correction Mechanism

239

where q is the accurately identified long tail demand quantity, that is, the consumption quantity of long tail demand. Suppose that the network externalities faced by different types of NGOs are equal in the network information platform. NV(q) has no relations with personal consumption q and type θ. This chapter consider that this internet platform is neutral (N V  (Q) = 0). And based on the multidimensional “Internet + NGO” mode, the marginal utility of consumers with single long tail demand is not affected by other consumers. In this network platform of incentive compatibility correction mechanism, in order to meet the utility function of long tail demand consumers, the supply of NGOs is constrained by participation V (q) ≤ U The design of corrective mechanism in network platform can satisfy both incentive compatibility and participation constraint θ V (q) + N V (Q) − Fe ≥ 0 N V (Q) + Fe(q) + D + T (q) − cq ≥ 0 When network externalities and asymmetric information exist at the same time, this long tail imbalance can still achieve the suboptimal state of long tail consumption non-distortion effect, so as to achieve the balance. At the same time, the information rent obtained by NGOs in the network platform can be reduced to zero. Although there is welfare loss for NGOs (as agents), the social welfare of the whole network platform can be significantly improved (to overcome information distortion). When the network externality and complete information exist simultaneously, the social planners only need to consider the participation constraints of NGOs, then the best supply in the correction mechanism can meet the requirements N V  (Q) + Fe (q) + T  (q) = c On the other hand, in this network platform, the government (social planners) has incentive motivation to gather all kinds of scattered long tail demand consumption to increase the scale effect of network value NV(Q). However, based on network neutrality, the improvement of this consumption level can still achieve the secondary optimal result of correction equilibrium.

240

7 The Application of the Correction Mechanism: Internet + NGO

7.2 Internet + NGO 7.2.1 Characteristics of Network Economy In essence, the Internet economy under the “Internet+” perspective is the Internet production and life style driven by information technology and knowledge (Leamer & Storper, 2014). It has the following characteristics: the supply and demand are changing instantaneously; the scale effect can be achieved by geometric series; the winner takes all; production, supply and circulation can be completely transparent; traditional hierarchy is weakening; transaction cost and information cost have decreased significantly or even reduced to zero; the agglomeration of demand becomes more convenient, leading to the ubiquitous competition and cooperation; the coexistence of virtualization and service personalization (Kogut, 2003). The “Internet+” economy has become full of information and information asymmetry will also be weakened. At the same time, this network economy has externality (Wang et al., 2005). As mentioned in the previous section, network externality refers to the network effect that market equilibrium fails to fully reflect the cost and benefit. This externality of the network comes from the interaction between its systematic and internal information flow, and the monopoly of network infrastructure (Mueller, 2017). Whether the network externality is positive or negative, it will lead to the deviation of equilibrium efficiency. The optimal result is difficult to achieve and the suboptimal result becomes a rational choice. This “Internet+” emerging economy mode integrates information flow, logistics and capital flow, and the distribution of information plays a core role. Based on the positive feedback effect of network economy, path dependence and transfer cost make the setting-up of initial mechanism very important. In this design based on the “Internet+” mechanism, we must consider the minimum demand and supply scale supporting the effective operation of the network, that is, the critical capacity of the network. Different from the traditional economic model, there are still marginal returns and increasing utility in this network economy, which has become the influencing factors that drive more stakeholders to join the network demand and supply mechanism. The increasing income based on the demand side scale economy is more important in the network economy. Combining with the scale economy of the supplier, it can produce two-way influence and superposition effect (Mei et al., 2018). Within the network economy, due to the development of network technology, information processing and transmission can break through the time and space constraints, greatly reducing the cost of information processing and transaction. Network economy can reduce the degree of information asymmetry between supply and demand, which is conducive to the overall allocation efficiency of social resources. Through the direct interaction between demand and supply and the dissolution of intermediary organizations, network economy is embodied in the customization, personalized service and production of “zero marginal cost” of unlimited replication and supply (Morrar et al., 2017). The development of network economy also

7.2 Internet + NGO

241

follows Matthew effect and Metcalfe’s law (Alabi, 2017; Eckhardt et al., 2019). The scarcity of economic resources has been greatly alleviated in the network economy, and the sharing of resources and information can create greater social value. In the era of network economy, discrete demand has the characteristics of spontaneous agglomeration (Li et al., 2019). This gathering of demand information and resources is not limited by time and space, and becomes a democratic and equal demand “public pond” (Cui, 2020). At the same time, with the great abundance of information, the demand has the characteristics of reverse transmission, that is, the overlapping of the transmission from high-level demand to low-level demand. In the consumption process of the network economy, consumers’ evaluation of the goods they need depends more on the interactive and experiential evaluation mechanism, such as public comments, shared post bar, search engine and other public sharing evaluation platforms. This mechanism acts as the “collector of evaluation information”. And consumers’ purchase behavior is more inclined to meet their own personalized needs by customized services. The supply and production subject in the era of network economy has the characteristics of self-adaptive and cooperative division of labor. With the development of flat, virtualized and decision-making decentralized supply, network organization has become the product of adapting to the characteristics of network economy. Network organization is a consortium (alliance) formed by multiple independent individuals based on the same goals. It realizes the organizational goals through interactive cooperation and timely feedback on the basis of division of tasks of the members of the organization. This kind of network organization is composed of many nodes to transmit and feedback information, and each node is composed of heterogeneous NGOs, which play the role of information intermediary, processing, communication and feedback in the network. Generally speaking, the network alliance has relative stability and dynamic adjustment, which can realize cost sharing, information sharing, lean production, benefit integration and utility optimization (Kale, 2017; Rezaei, & Behnamian, 2020). The market structure in the network economy is more reflected in the mode of monopoly competition. On the platform of monopoly standardization and homogenization, the products and services provided by different competitive organizations are highly different. This monopolistic competition mode based on the network sharing economy can effectively promote a more equitable distribution of resources and the universality of the information society (Pazaitis et al., 2017). In addition, the cost of trial-and-error correction in the market structure of network economy has decreased significantly, which makes the game between the demander and the supplier gradually change from “multi-to-multi” to “one to many” or even “one to one”. With the functional differentiation and price personalization of resource allocation in the market structure, the network economic market has gradually changed from the optimal equilibrium to the suboptimal equilibrium as the goal (Buldú et al., 2019).

242

7 The Application of the Correction Mechanism: Internet + NGO

7.2.2 Practical Mechanism The network platform organization, represented by supporting NGOs (umbrella organizations and hub NGOs), integrates the information, resources, social capital and other essential elements of the supply and demand sides (Feldner & Fyke, 2016; Scott & Jabbar, 2014). Although it does not participate in the supply of specific rural long tail public services, it is committed to providing intelligence, human resources, information and other services and support for other NGOs. It could establish a communication platform and security mechanism among different NGOs and between NGOs and the government (Janssen & Estevez, 2013), playing the role of balance network platform operation mechanism for rural long tail public service. Supporting NGO is an organization with mobile advantages and integration ability. It can integrate all kinds of knowledge and resources in cross-border and cross domain, and coordinate the specialized division of labor among different NGOs (Montes et al., 2005). It can also maintain the comparative advantages of different NGOs in providing specific rural long tail public services, and form a multidimensional and comprehensive supply system. This kind of “long tail aggregator” formed by the integration of single NGO’s resource dependence and comparative advantage not only accurately identify the demand according to the supporting NGO’s own information and mobile advantage, but also aggregate the discrete demand and allocate the supply tasks according to the individual endowment differences of NGOs. Therefore, it can greatly reduce the coordination and information cost of rural long tail public service, and improve the supply efficiency and form scale effect. From the perspective of social division of labor, supporting NGOs and other NGOs are the result of the gradual separation of functions and professional division of labor based on the continuous development and evolution of social public demand. In the network supply platform, it is similar to the division of political official and clerical official in western civil servant system (Kubátová & Kubát, 2020). Supporting NGOs play the role of political official, chief information officer and chief knowledge officer (CIO or CKO). They have comparative advantages in information acquisition, social capital and government communication. However, due to their different endowments, other NGOs have the advantage of professional technology. They can reduce the influence and interference of unproductive Affairs (such as information search, information identification, and public relations) through supporting NGOs. These NGOs concentrate on the role of clerical official, chief technology officer and chief operation officer (CTO or COO) of long tail demand on the basis of professional division of labor. The logic behind the evolution of specialization is that the government “outsource” or transfer the rural long tail public services to NGOs. Other NGOs further transfer the non-productive functions such as information acquisition and supply coordination to the “second division of labor” of supporting NGOs (Ma, 2017). Since the benefits of this specialization are higher than the transaction and coordination costs of individual actions of different NGOs, it is a Pareto improvement from the perspective of social welfare.

7.2 Internet + NGO

243

This network alliance mechanism based on the formation of supporting NGOs has greater advantages for the uncertainty and complementarity of rural NGOs. In the “self-cooperation” supporting NGOs (and their network platforms), all kinds of resources form a joint agglomeration through the “exchange” behavior based on equality, and finally achieve complementary advantages (Yang, 2015). At the same time, with the support of NGO platform, new resources are generated automatically under the exchange of existing resources. With the continuous change of public demand of rural long tail, the new resources formed by NGOs can not only make up for the loss caused by the adjustment of variable costs through sharing, but also realize the dynamic transfer of long tail equilibrium through the re-integration and allocation of supporting NGOs. This kind of network platform operation mechanism, represented by supporting NGOs (and their alliances), is a flexible network platform based on professional division of labor. With the purpose of maximizing the total social surplus, it could meet the scattering demand of the long tail in the rural areas, which has the characteristics of consensus, integration and credibility. Specifically, the network operation mechanism with supporting NGOs as intermediary organizations includes two levels of planning: networking communication and resource exchange platform; networking demand and supply matching platform (Fig. 7.1). (1)

The first layer: networking information and resource exchange platform

In the first level of the bilevel planning, the supporting NGOs play the role of specialized coordination cost and information resource exchange platform. The knowledge, social capital and professional advantages of their personnel in non-productive affairs make supporting NGOs accurately identify and observe regional long tail needs. They also have the distinctive ability to collect and analyze information: these supporting NGOs classify and summarize different types of information and resources, and then shares and feeds back to other NGOs in real time with the help of the Internet platform. Supporting NGOs are born with information advantages. In the process of initial establishment, subsequent development and communication with other NGOs and governments, they tend to take the real needs of local rural residents as the driving force, and strive to achieve accurate identification of various needs. The grassroots and “demand-oriented” development mode of supporting NGOs enables them to accurately identify their needs through daily communication with local rural residents. Through the obtained first-hand demand information, supporting NGOs reasonably arrange different response needs and assign tasks to other NGOs. Due to the endogeneity and embeddedness of supporting NGOs, organization sponsor and response subject are also part of the main body of long tail demand. They have incomparable advantages in identifying their own needs. They can achieve the identity and accurate matching of supply and demand without free riding, adverse selection and moral hazard. After receiving the supply tasks arranged by the supporting NGOs, different NGOs rely on their own technical advantages to provide directional supply, and give local rural residents full rights of independent choice and “voting with their feet” (Canales, 2020; Cordero-Guzmán et al., 2008). Rural residents tend to show their real demand

7 The Application of the Correction Mechanism: Internet + NGO

244

Public Sector

Resource Acquisition

Resource Allocation

First Layer Networking communication & resource exchange platform

Signal Mechanism

Disclosure Mechanism

Intermediary Purchase Service

Supporting NGO

Correction

Long Tail Service

Agglomeration

Networking demand & supply matching platform Customized Supply

Dynamic Supply

Second Layer

Other NGOs Fig. 7.1 The practical mechanism of “Internet + NGO” for rural long tail public service

preferences in the initial selection process, and choose the supply subject that best matches their own preferences to meet the demand. After the completion of “primary distribution and identification”, supporting NGOs further correct the subject, degree, quantity and quality of this primary supply response distribution by collecting feedback information and different public services being selected. In this process, the results of various public services being selected produce a “signal” mechanism to reveal the real demand preference of long tail. The supporting NGOs play the role of intermediary and bridge, connecting the real needs of the aggregate information and network platform as the resource endowment of the system. Generally speaking, the accurate identification and information and resource exchange mechanism of supporting NGOs for rural long tail public services is a process of information trial-and-error correction. Through the continuous cycle of selection and correction, the dynamic balance and accurate identification of long tail

7.2 Internet + NGO

245

demand of rural residents in different regions can be realized. By the information network platform, supporting NGOs can share real-time information about the real preferences of different regions, different groups and different residents, so as to reduce the fixed and variable costs of individual NGOs responding to the needs. The government can also get the most real demand information of rural residents through communication with supporting NGOs. As a long tail aggregator, supporting NGOs form a “public pond” of real demand information to reduce the search cost. The key of this information and resource exchange mechanism is the mobility and accessibility of information. (2)

The second layer: networking demand and supply matching platform

Based on the free exchange of information and resources in the first layer, the effective supply and demand balance mechanism depends on the reasonable operation of the second layer networking demand and supply matching platform. Because of their own endowments and information rent, different NGOs can target the customized long tail demand. This is the best at and the most efficient in supply, so as to become the “representative” of this specific demand. These NGOs could realize the “natural monopoly” within the customized supply on the basis of complete competition among different NGOs. As an important node of network demand and supply matching platform, supporting NGOs reasonably arrange “point-to-point” customized supply allocation tasks of NGOs, and select different supply “representatives” through information integration and optimal allocation of resources. At the same time, through the real-time coordination of supply, it can reduce the resource mismatch and cost fragmentation, and realize the internal scale economy of customized supply. The demand-oriented flexible supply feedback and matching mechanism, represented by supportive NGOs, focuses on capturing the details of demand changes and individual needs, and adopts various information technologies to track and identify residents’ choices and preferences in detail. It gives timely feedback to other NGOs, so as to achieve the goal of “lean supply” with shorter supply cycle and lower inventory level (Moyano-Fuentes et al., 2019; Tortorella et al., 2017). “Lean supply” means that different NGOs form close cooperation and division of labor through the information network shared by supporting NGOs, thus reducing the marginal cost of supply. In this kind of network platform, other NGOs can learn from each other, resist complex and changeable external risks together, and reduce the risk of information rent. Supporting NGOs should provide small and mediumsized NGOs with information and experience sharing. Supporting NGOs should also reduce friction and promote exchanges and cooperation among NGOs (Alshaerb et al., 2017; Zhu, 2016). Generally speaking, the networking matching platform refers to abandoning and overcoming the phenomenon of repeated supply caused by the unclear division of labor among homogeneous NGOs through intensive unified action and specialized division of labor. It could reduce the friction and ineffective competition of NGOs in obtaining resources and supply services. It could also realize the “natural monopoly” effect and the “representative” supply mechanism of the long tail demand for the

246

7 The Application of the Correction Mechanism: Internet + NGO

same kind of rural long tail demand. This supply mechanism reduces transaction cost through internal vertical integration, which makes the external fixed input of each NGO supply become internal variable input, and realizes the amplification of public resources. This networking matching platform is like an umbrella or a bunch of grapes, connecting all NGOs directly involved in the supply of long tail public services through the network, forming a division of labor system (Kenney, & Zysman, 2016; Xu, 2010).

7.3 Case Study: JD.com Public Welfare Foundation Platform Then, this chapter further elaborates the practical operation of the network platform model of the rural long tail public service supply and demand equilibrium through the case study of the Internet platform of JD.com public welfare foundation. The JD.com public welfare foundation is not only an individual of the foundation (NGO), but also a public service (social relief) Internet platform (with the function of supporting NGO) based on its big data Internet of things technology and the supply chain of JD.com Mall. With the integration of social resources, the JD.com Public Welfare Foundation has carried out social innovation in the fields of rural poverty alleviation, disaster relief, education, environmental protection and social innovation. Through its public fund and material collection platform, the “material-love connection (Wu-ai Xianglian)” information exchange and supply-demand matching mechanism between rural long tail demand and fragmented rural public service supply (love relief) can be effectively realized. This new public welfare supply mode of “One click donation and direct delivery of materials” can give full play to the advantages of JD.com self-supporting ecommerce. It could also provide comprehensive support and transparent user experience for rural public services from the aspects of service supply, logistics distribution, technical operation and customer service. The public welfare materials collection platform of JD is based on the mobile Internet development. As a fragmented long tail demand provider, the public selects the materials required for public services through the Internet client of JD.com, and can complete the supply purchase and donation by one click. Based on the supporting NGO platform, the demand information and resources are gathered, directly located and distributed to the demand area through the internal logistics system of JD.com. The logistics personnel play the node in the supply chain to realize the terminal service of supply. Meanwhile, fragmented long tail public service providers (buyers, such as Internet users) can query the logistics status of the purchased services in real time through mobile phone clients, so as to monitor the whole process of the whole public service supply chain in real time. Based on the special needs of rural long tail public service in terms of timeliness and regionality, JD.com public welfare platform helps cooperative public charity organizations to do well in project management, monitoring and other

7.3 Case Study: JD.com Public Welfare Foundation Platform

247

process work, so as to ensure timely delivery of demand goods and services and feedback of information (Cao, 2017; Zheng et al., 2020). The Internet platform of JD.com public welfare foundation can better meet the wide participation, openness and efficiency of long tail demand and supply. This balanced supply mode of Internet platform can effectively overcome the complex process of cooperation among various departments in the whole supply and donation chain. It organically combines information collection, material collection, warehouse transportation, terminal distribution and service, public supervision and other links (Yang et al., 2016). It not only improves the aggregation effect of long tail demand, but also improves the specialization of long tail supply professionalism and credibility, conducive to the correction of the imbalance in areas with too discrete demand. Specifically, this network platform mode includes self-service donation, logistics direct delivery, data monitoring, volunteer service and other links. By giving play to its technical advantages in the field of intelligent logistics, we can reasonably adjust its public goods inventory according to the supply data of fragmented suppliers’ love purchase and donation, and create supply chain collaboration and flexible production (Singh et al., 2018). At the same time, based on the user traffic data, it subdivides the long tail supply preference, sets differentiated and customized consumption demand strategies to accurately information push, and improves service efficiency. In addition, through the analysis of the long tail demand area location and order users, it adjusts the product layout, accurate positioning, and form the regional differentiation supply strategy (Shen et al., 2020). From the perspective of rural long tail public service, the stakeholders included in the JD.com public welfare Internet platform include long tail demanders, fragmented providers (consumer netizens), internal employees, government regulatory agencies, value chain partners, community environment, etc. (Millar, 2015). Each stakeholder has different incentive preferences and information mechanism in the network platform. Among them, the long tail demanders mainly focus on the accuracy, reliability and security of the products and services required. Their information mechanism is continuous offline communication and media information feedback. Fragmented consumers mainly focus on their information privacy, purchase experience and efficiency, and their information system is continuous online supervision and feedback. Internal employees mainly focus on their employment level, and their information mechanism is network training activities and appeal mechanism. Government regulatory agencies mainly focus on green supply, social benefits and information loss, and their information mechanism is irregular online supervision and offline investigation. Value chain partners mainly focus on social responsibility, industry development and supply chain management, and their information mechanism is regular online assessment. Community environment mainly focuses on the inclusiveness and externality of supply, and their information mechanism is daily media transmission and offline volunteer activities. Based on the differentiated incentive preference and information mechanism of different stakeholders in the Internet platform, JD.com public welfare realizes the scene connection, data connection and value exchange of rural long tail public service

248

7 The Application of the Correction Mechanism: Internet + NGO

balance through the supply mode of “unbounded retail” (Gümü¸s et al., 2016). It reduces social transaction and logistics costs through B2B, cross-border and crowdsourcing supply chains. At the same time, the online and offline Internet platform will be organically combined with the real economy to promote the supply side reform of rural long tail public services. Relying on the technology and information empowerment of JD.com cloud, the nationwide coverage of demand data will be realized. Especially for the rural long tail financial and poverty alleviation needs, it solved the financial needs of rural residents through JD.com agricultural loan, data agricultural loan, “warm JD.com” (nuan-dong) public welfare crowdfunding platform. This could effectively incubate the rural youth e-commerce training center, realize the satisfaction of rural entrepreneurship under the positioning of targeted poverty alleviation (Luo & Niu, 2019). From the perspective of community development of rural long tail public service, JD.com public Internet platform pursues green supply chain system in the supply of public services, and reduces resource consumption and environmental pollution in storage, transportation, transmission terminals and other aspects. Through the establishment of storage photovoltaic station and new energy transportation system, it could reduce the waste of traditional resources in the supply. At the same time, through the “green-flow plan”, the third-party professional recycling agencies are introduced to promote the green packaging of public services. Through information sharing on the Internet platform, the sustainable production and consumption of public services that meet the environmental and social friendly standards are promoted, and the standard system for the sustainable development of the industry is established. In general, the Internet platform of JD.com public welfare foundation can realize the integration processing and feedback satisfaction based on the fragmentation of rural long tail demand. It is committed to the sustainable development of social responsibility and the coordination and interaction with various stakeholders. By giving full play to its big data platform, supply chain information and resource advantages, it can continuously improve the user experience of demand. It could also provide sustainable supply for value chain partners in order to establish a self-built logistics and supply system covering the rural areas in China. In addition, the information mechanism on its Internet platform can reduce the distortion and “rent-seeking” in information transmission. This could realize the integration analysis of supply and demand information for upstream (government regulatory departments) and downstream (demand residents), which play the intermediary role of supporting NGOs. Finally, the platform, logistics, distribution, financial services and other businesses in the long tail public service supply chain are continuously opening and developing. Fragmented suppliers (Internet users) can optimize the purchase and donation experience through Internet terminals, reduce the cost of information search and screening, and maximize the effectiveness of network supply through this demand information open platform. In the future, JD.com public welfare network platform will further focus on keeping watch on the hot spots of social demand, help public welfare activities, and constantly explore innovative solutions for rural long tail public services (Wang, 2020).

References

249

References Alabi, K. (2017). Digital blockchain networks appear to be following Metcalfe’s Law. Electronic Commerce Research and Applications, 24, 23–29. Alshaerb, I. M. A., Al-Hila, A. A., Al Shobaki, M. J., & Abu Naser, S. S. (2017). Governance of public universities and their role in promoting partnership with non-governmental institutions. International Journal of Engineering and Information Systems (IJEAIS), 1(9), 214–238. Buldú, J. M., Pablo-Martí, F., & Aguirre, J. (2019). Taming out-of-equilibrium dynamics on interconnected networks. Nature Communications, 10(1), 1–9. Canales, K. L. S. (2020). Voting with their feet: Do people choose residential destinations based on naturally occurring advantages or man-made advantages of locations? (Doctoral dissertation). The University of North Carolina at Charlotte. Cao, P. H. (2017). Material-love linked, innovative material donation of JD.com public welfare material collection platform, and Zhang Zetian as the platform love ambassador. Society and Public Welfare. (in Chinese). Cordero-Guzmán, H., Martin, N., Quiroz-Becerra, V., & Theodore, N. (2008). Voting with their feet: Nonprofit organizations and immigrant mobilization. American Behavioral Scientist, 52(4), 598–617. Cui, J. (2020). Collaborative governance of local governments in China. Routledge. Eckhardt, G. M., Houston, M. B., Jiang, B., Lamberton, C., Rindfleisch, A., & Zervas, G. (2019). Marketing in the sharing economy. Journal of Marketing, 83(5), 5–27. Feldner, S. B., & Fyke, J. P. (2016). Rhetorically constructing an identity at multiple levels: A case study of social entrepreneurship umbrella organizations. International Journal of Strategic Communication, 10(2), 101–114. Gümü¸s, M., Kaminsky, P., & Mathur, S. (2016). The impact of product substitution and retail capacity on the timing and depth of price promotions: Theory and evidence. International Journal of Production Research, 54(7), 2108–2135. Janssen, M., & Estevez, E. (2013). Lean government and platform-based governance—Doing more with less. Government Information Quarterly, 30, S1–S8. Kale, V. (2017). Agile network businesses: Collaboration, coordination, and competitive advantage. CRC Press. Kenney, M., & Zysman, J. (2016). The rise of the platform economy. Issues in Science and Technology, 32(3), 61. Kogut, B. M. (Ed.). (2003). The global internet economy. MIT Press. Kubátová, H., & Kubát, M. (2020). The priest and the state: Clerical fascism in Slovakia and theory. Nations and Nationalism. Leamer, E. E., & Storper, M. (2014). The economic geography of the internet age. In Location of international business activities (pp. 63–93). Palgrave Macmillan. Li, J., Webster, D., Cai, J., & Muller, L. (2019). Innovation clusters revisited: On dimensions of agglomeration, institution, and built-environment. Sustainability, 11(12), 3338. Luo, X., & Niu, C. (2019). E-Commerce participation and household income growth in Taobao Villages. The World Bank. Ma, H. B. (2017). Supporting NGOs: A platform for the secondary division of social demand and supply. Social Work, 2, 72–81. (in Chinese). Mei, Y., Ye, J., & Zeng, Z. (2018). Two-way scheduling optimization of the supply chain in one-ofa-kind production based on dynamic production capacity restrictions. Journal of Manufacturing Systems, 47, 168–178. Millar, M. (2015). Global supply chain ecosystems: Strategies for competitive advantage in a complex, connected world. Kogan Page Publishers. Montes, J. L. I., Vela, F. L. G., & Megías, M. G. (2005). Supporting NGO modelling in cooperative work using patterns. In International Conference on Computer Supported Cooperative Work in Design (pp. 112–121). Springer.

250

7 The Application of the Correction Mechanism: Internet + NGO

Morrar, R., Arman, H., & Mousa, S. (2017). The fourth industrial revolution (Industry 4.0): A social innovation perspective. Technology Innovation Management Review, 7(11), 12–20. Moyano-Fuentes, J., Bruque-Cámara, S., & Maqueira-Marín, J. M. (2019). Development and validation of a lean supply chain management measurement instrument. Production Planning & Control, 30(1), 20–32. Mueller, M. (2017). Will the internet fragment? Sovereignty, globalization and cyberspace. Wiley. Pazaitis, A., De Filippi, P., & Kostakis, V. (2017). Blockchain and value systems in the sharing economy: The illustrative case of backfeed. Technological Forecasting and Social Change, 125, 105–115. Rezaei, S., & Behnamian, J. (2020). A survey on competitive supply networks focusing on partnership structures and virtual alliance: New trends. Journal of Cleaner Production, 125031. Scott, J., & Jabbar, H. (2014). The hub and the spokes: Foundations, intermediary organizations, incentivist reforms, and the politics of research evidence. Educational Policy, 28(2), 233–257. Shen, M., Tang, C. S., Wu, D., Yuan, R., & Zhou, W. (2020). JD.com: Transaction-level data for the 2020 MSOM data driven research challenge. Manufacturing & Service Operations Management. Singh, H., Garg, R., & Sachdeva, A. (2018). Supply chain collaboration: A state-of-the-art literature review. Uncertain Supply Chain Management, 6(2), 149–180. Tortorella, G. L., Miorando, R., & Marodin, G. (2017). Lean supply chain management: Empirical research on practices, contexts and performance. International Journal of Production Economics, 193, 98–112. Wang, C. C., Hsu, Y., & Fang, W. (2005). Acceptance of technology with network externalities: An empirical study of internet instant messaging services. Journal of Information Technology Theory and Application (JITTA), 6(4), 4. Wang, X. (2020). Blockchain chicken farm: And other stories of tech in China’s countryside. FSG Originals. Xu, Y. S. (2010). Structural innovation of NGOs: The growth of supporting institutions. Chinese NGOs, 8, 28–31. (in Chinese). Yang, K. (2015). Discussion on the key factors of the success of self-cooperation among NGOs: A case study of Shaanxi NGO disaster relief alliance after the May-12 Wenchuan earthquake. Chinese Public Administration, 2015(8), 66–70. (in Chinese). Yang, Q., Zhao, X., Yeung, H. Y. J., & Liu, Y. (2016). Improving logistics outsourcing performance through transactional and relational mechanisms under transaction uncertainties: Evidence from China. International Journal of Production Economics, 175, 12–23. Zheng, K., Zhang, Z., & Song, B. (2020). E-commerce logistics distribution mode in big-data context: A case analysis of JD.com. Industrial Marketing Management, 86, 154–162. Zhu, J. B. (2016). The role of supportive NGOs in social governance. Journal of Fujian Provincial Party School of CPC, 2, 44–50. (in Chinese).

Conclusion: Efficiency-Fairness Complementation

Through the methods of monopoly and competition, the combination of Internet + NGO is regarded as a fair and efficient allocation of resources and information (with the government as the platform owner and NGOs as the information owner). This kind of “public pond” is based on the infinity of network capacity and has no congestion effect while maintaining network externality. This supply-demand equilibrium model shares the fixed costs and variable costs of the differentiated long tail demand of the whole rural area. The model effectively encourages a large number of NGOs with various types to participate in the supply and correction of long tail public services by creating information rent, in order to realize the dynamic equilibrium of individual demand and the macro-economy. This kind of network supply and demand equilibrium system has the complementary mechanism of both efficiency and fairness. From the point of efficiency, the ownership and the right to use are alternate in the long tail public service “Internet + NGO” mechanism. The platform owners obtain the right to use the demand information by transferring the right to use the platform. The information owners also acquire the right to use the platform by sharing the right to use information. This efficiency, which is based on fairness, can give full play to the specialized economies of scale, and meet the incentive compatibility and participation constraints in the form of information rent. Through information sharing, a network platform can create social value, improve the efficiency of diversified supply and obtain information productivity on the basis of “value-added” personalized demand. The biggest beneficiaries of this “grassroots supply” and “grassroots demand” sharing mechanism are the supply and demand subjects of rural long tail public services, namely the rural residents. This kind of point-to-point, many-to-many network supply structure can achieve higher efficiency than the market mechanism that gathers price signals. The marginal utility of discrete supply can also be increased by sharing the fixed cost. An accurate identification information mechanism and two-way matching mechanism can effectively solve the

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 J. Luo, Rural Long Tail Public Service and the Correction Mechanism, https://doi.org/10.1007/978-981-16-4023-0

251

252

Conclusion: Efficiency-Fairness Complementation

problem of information asymmetry and realize the democratization of rural long tail public service. From the perspective of fairness, the supply and satisfaction of rural long tail public services are fragmented and discrete. From the perspective of human freedom and all-round development, fairness based on human self-identity (including selfrealization) is truly inclusive and free. “Fairness” is a narrow view of development, which measures the utility of public service only by scale effect and recipient size. According to Sen (2004), in the supply and demand equilibrium of rural long tail public services, the development of freedom is manifested as the inclusion of “qualitative” differences in the two-dimensional system of the quantity-price of public services. This development not only includes the personalization and networking caused by information and resource allocation, but also includes the experiential and social capital caused by interest interaction. Supply freedom, from the perspective of network sharing, is a fair view based on efficiency. When fairness is in the high information rent range (where the choice of diversification is located) (extending infinitely along the long tail demand curve), the improvement of the overall net surplus of the society that is brought about by fairness cannot be achieved and satisfied by traditional head public services (Jiang, 2018). In this inclusive distribution of social and economic surplus (primary distribution and secondary distribution), the participation of multiple NGOs is itself a kind of social net surplus. The improvement of the welfare level is organically combined with the interaction of information, resources and interests, so as to provide grassroots groups with a monopoly platform and a fair right to participate in “value-added” services. Generally speaking, fairness and efficiency, in the balance of rural long tail public services, are not opposite; each can promote the other. This is a kind of peopleoriented, real demand-oriented supply and development view. Through the creation of information conditions (to overcome the exclusiveness of knowledge and technology), fairness and efficiency can reduce the threshold of public participation (rural residents and NGOs). Using the efficiency principle of “who suits, who customizes” to carry out the primary supply distribution, and the supervision and error correction of network platform, the balance of efficiency and fairness can help achieve secondary supply distribution. This kind of network distribution mode creates a welfare mode of “monopoly platform + complete competition participation” on the basis of efficiency-fairness complementation. This is in line with Sen’s (2004) value, which holds that “choice itself (i.e. freedom) is better than choice result (development)”. In this sense, a rural long tail public service supply and demand balance mechanism enables everyone to benefit from the activities of creating economic net surplus through everyone’s sharing and participation. Therefore, the real meaning behind the balance of rural long tail public services is not simply based on the simple copy and paste of head homogeneous demand, but on the innovation oriented to the minority public demand. This is a diversified, personalized and socialized form of self-realization that is based on the balanced development of use rights and ownership rights.

Appendix

See Tables 1, 2. See Figs. 1, 2 and 3. See Table 3.

Table 1 The list of all variables in rural special education Sign

Name

Calculation method

Unedu

Proportion of uneducated rural special children

Number of registered rural children with special needs without education ÷ total rural population aged 0–14

Illi

Illiteracy rate

Total rural illiterate population ÷ total rural population over 6 years old

Visual

Number of children with visual disability without education

Registered children with visual disability without education * proportion of rural population aged 0–14

Hearing

Number of children with hearing impairment without education

Registered children with hearing disability without education * proportion of rural population aged 0–14

Intellectual

Number of children with intellectual disability without education

Registered children with intellectual disabilities without education * 0–14 years old rural population proportion (continued)

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 J. Luo, Rural Long Tail Public Service and the Correction Mechanism, https://doi.org/10.1007/978-981-16-4023-0

253

254

Appendix

Table 1 (continued) Sign

Name

Calculation method

Physical

Number of children with physical disability without education

Registered children with physical disability without education * proportion of rural population aged 0–14

Psychiatric

Number of children with psychiatric disability without education

Registered children with psychiatric disability without education * proportion of rural population aged 0–14

Multi

Number of children with comprehensive disability without education

Registered children with comprehensive disability without education * proportion of rural population aged 0–14

FiSEcapper

Governmental capital expense on special education (excluding infrastructure construction)

Financial capital expenditure of special education (equipment purchase, school repair, etc.) ÷ total financial expenditure of special education

SoSEcapper

Social capital expense of social special Social capital expenditure of special education education (equipment purchase, school building repair, etc.) ÷ (total expenditure for special education − total financial expenditure for special education)

FiBEcapper

Governmental capital expense on rural Rural basic education (including basic education (excluding middle school and primary school, the infrastructure construction) same below) financial capital expenditure (equipment purchase, school repair, etc.) ÷ total financial expenditure of rural basic education

SoBEcapper

Social capital expense on rural basic education

Social capital expenditure on rural basic education (equipment purchase, school repair, etc.) ÷ (total expenditure on rural basic education − total financial expenditure on rural basic education)

FiVEcapper

Government capital expense of rural vocational high school (excluding infrastructure construction)

Financial capital expenditure (equipment purchase, school repair, etc.) of rural vocational high school ÷ total financial expenditure of rural vocational high school

SoVEcapper

Social capital expense of rural vocational high school

Social capital expenditure of rural vocational high school (equipment purchase, school repair, etc.) ÷ (total expenditure of rural vocational high school − total financial expenditure of rural vocational high school) (continued)

Appendix

255

Table 1 (continued) Sign

Name

SEconper

Expense on infrastructure construction Expenditure on infrastructure of special education (mainly from the construction of special government) education ÷ total expenditure on special education

Calculation method

BEconper

Expense on infrastructure construction Expenditure on infrastructure of rural basic education (mainly from construction of rural basic the government) education ÷ total expenditure on rural basic education

VEconper

Expense on infrastructure construction Infrastructure construction expenditure of rural vocational high schools of rural vocational high school ÷ total (mainly from the government) expenditure of rural vocational high school

FiSEadmper

Governmental administrative expense Financial administrative expenditure of special education of the government for special education (official expenses, service fees, etc.) ÷ total financial expenditure on special education

SoSEadmper

Social administrative expense on social special education

Social administrative expenditure on special education (official expenses, service fees, etc.) ÷ (total expenditure on special education − total financial expenditure on special education)

FiBEadmper

Governmental administrative expense on rural basic education

Financial administrative expenditure of rural basic education finance (official expenses, service fees, etc.) ÷ total financial expenditure of rural basic education

SoBEadmper

Social administrative expense on rural basic education

Social administrative expenditure on rural basic education (official expenses, service fees, etc.) ÷ (total expenditure on rural basic education − total financial expenditure on rural basic education)

FiVEadmper

Governmental administrative expense on rural vocational high schools

Financial administrative expenditure of rural vocational high school (official expenses, service fees, etc.) ÷ total financial expenditure of rural vocational high school

SoBEadmper

Social administrative expense on rural vocational high schools

Social administrative expenditure of rural vocational high school (official expenses, service fees, etc.) ÷ (total expenditure of rural vocational high school − total financial expenditure of rural vocational high school) (continued)

256

Appendix

Table 1 (continued) Sign

Name

Calculation method

FiSEwelper

Governmental welfare expense on special education

Financial welfare expenditure for special education (basic salary, supplementary salary, welfare work and allowance for teaching staff, etc.) ÷ total financial expenditure for special education

SoSEwelper

Social welfare expense on special education

Social welfare expenditure for special education (basic salary, supplementary salary, welfare work and allowance for teaching staff, etc.) ÷ (total expenditure for special education − total financial expenditure for special education)

FiBEwelper

Governmental welfare expense on rural basic education

Financial welfare expenditure for rural basic education (basic salary, supplementary salary, welfare work and allowance for teaching staff, etc.) ÷ total financial expenditure for basic education

SoBEwelper

Social welfare expense on rural basic education

Social welfare expenditure for rural basic education (basic salary, supplementary salary, welfare work and allowance for teaching staff, etc.) ÷ (total expenditure for basic education − total financial expenditure for basic education)

FiVEwelper

Governmental welfare expense of rural Financial welfare expenditure of rural vocational high school vocational high school (basic salary, supplementary salary, welfare work and allowance for teaching staff, etc.) ÷ total financial expenditure of rural vocational high school

SoVEwelper

Social welfare expense of rural vocational high school

FiSEschper

Governmental scholarship expense for Financial scholarship expenditure of special education special education (scholarship and grant to students, etc.) ÷ total financial expenditure for special education

Social welfare expenditure of rural vocational high school (basic salary, supplementary salary, welfare work and allowance for teaching staff, etc.) ÷ (total expenditure of rural vocational high school − total financial expenditure of rural vocational high school)

(continued)

Appendix

257

Table 1 (continued) Sign

Name

Calculation method

SoSEschper

Social scholarship expense for special education

Social scholarship expenditure of special education (scholarship and grant to students, etc.) ÷ (total expenditure for special education − total financial expenditure for special education)

FiBEschper

Governmental scholarship expense on rural basic education

Financial scholarship expenditure of rural basic education (scholarship and grant to students, etc.) ÷ total financial expenditure for rural basic education

SoBEschper

Social scholarship expense on rural basic education

Social scholarship expenditure of rural basic education (scholarship and grant to students, etc.) ÷ (total expenditure for rural basic education − total financial expenditure for rural basic education)

FiVEschper

Governmental scholarship expense of rural vocational high school

Financial scholarship expenditure of rural vocational education (scholarship and grant to students, etc.) ÷ total financial expenditure for rural vocational education

SoVEschper

Social scholarship expense of rural vocational high school

Social scholarship expenditure of rural vocational education (scholarship and grant to students, etc.) ÷ (total expenditure for rural vocational education − total financial expenditure for rural vocational education)

Income

Rural per capita income

Rural per capita income ÷ GDP inflation elimination index (10 million yuan per person)

Group

Number of cultural NGOs per capita in Number of cultural NGOs in rural rural areas areas ÷ rural population (10,000 per person)

Computer

Area of computer room in rural special Direct apply from raw data (100 km2 ) education

SEbud

Per capita budget expense of rural special education

Total budget expenditure for special education ÷ (total number of rural students * 1000)

Marketization Marketization index

Fan et al. (2003)

BEbud

Per capita budget expense of rural basic education

Total budget expenditure of basic education ÷ total number of students in rural primary and secondary schools

Rurapop

Rural population

Direct apply from raw data (a million)

258

Appendix

Table 2 The statistic description of all variables for rural special education Variable

Obs

Mean

Std. Dev.

Min

Max

unedu

357

7.10702

5.215005

0.0385852

26.56842

illi

357

1.188133

2.166472

0.0116183

18.76768

visual

357

512.9236

691.1267

0

4558.614

hearing

357

572.4153

753.1095

0

4204.157

intellectual

357

998.4957

1052.541

0.3038585

6148.554

physical

357

927.891

982.8844

0.5064309

5504.857

psychiatric

357

228.9013

275.1201

0

1601.605

multi

357

512.7749

533.5723

0.1012862

2744.542

fiSEadmper

357

0.1318472

0.0563777

0.0102482

0.4101175

soSEadmper

357

0.2024637

0.1383536

0

0.9098361 0.7673993

fiSEwelper

357

0.5090991

0.1243886

0.0567159

soSEwelper

357

0.1665162

0.1592334

0

0.7850163

fiSEschper

357

0.1377826

0.0676744

0.0159894

0.3689919

soSEschper

357

0.2246669

0.1928869

0

0.8429715

fiSEcapper

357

0.152025

0.1136325

0.0085221

0.8508184

soSEcapper

357

0.4061389

0.2151026

0

1

SEconper

357

0.0568255

0.0984161

0

0.6013585

fiBEadmper

357

0.1132947

0.0592606

0.006082

0.2917688

soBEadmper

357

0.1931151

0.1179527

0.017348

0.5848703

fiBEwelper

357

0.6531947

0.1520218

0.3820043

0.956934

soBEwelper

357

0.213845

0.164165

0.0119222

0.8009072

fiBEschper

357

0.1063493

0.0843946

0.0000249

0.3314167

soBEschper

357

0.2510808

0.2471642

0.000028

0.8578329

fiBEcapper

357

0.1005923

0.0520529

0.0170487

0.291999

soBEcapper

357

0.2869179

0.1420838

0.0468524

0.8607732 0.2307642

BEconper

357

0.0245548

0.0257385

0

fiVEadmper

248

0.1195169

0.0855835

0

0.6554511

soVEadmper

248

0.2705165

0.1880046

0

1

fiVEwelper

248

0.4187416

0.2001796

0

0.9911215

soVEwelper

248

0.184076

0.1779395

0

0.7986076

fiVEschper

248

0.1974757

0.1327643

0

1

soVEschper

248

0.106154

0.0987335

0

0.618451

fiVEcapper

248

0.1211864

0.1228106

0

0.7375578

soVEcapper

248

0.291549

0.2335068

0

0.9933211

VEconper

248

0.0288521

0.0529082

0

0.3612942

income

357

0.004188

0.0023187

0.0013815

0.0154251 (continued)

Appendix

259

Table 2 (continued) Variable

Obs

Mean

Std. Dev.

Min

Max

computer

357

0.0002789

0.0002288

3.00e−06

0.0011907

group

357

0.0000964

0.0004083

1.71e−06

0.0055688

SEbud

357

0.000215

0.000719

8.70e−07

0.0090366

marketizat~n

327

6.234924

1.732427

2.53

10.92

BEbud

357

1.816999

3.501129

0.0240503

29.37518

rurapop

357

2319.241

1587.114

194

7037

Fig. 1 The distribution of uneducated rural special children 2003–2014. Note It was drawn from STATA15.0, and the same below

260

Fig. 2 The density distribution of different categories of expense

Fig. 3 The density distribution of different categories of rural special education

Appendix

Appendix

261

Table 3 The sensitivity check of time variation effect

DV

(1)

(2)

(3)

(4)

(5)

Fe

Panel SUR

FGLS

PCSE

MEML

Unedu

Unedu/Illi

Unedu

Unedu

Unedu

−2.848**

−3.601**

−0.503

−2.464

(1) Current capital expenditure C.fiSEcapper

−2.645 (−1.115)

(−2.227)

(−2.472)

(−0.310)

(−1.525)

C.soSEcapper

0.162

0.725

0.320

0.592

0.165

(0.109)

(1.093)

(0.456)

(0.705)

(0.183)

0.909

0.424

−0.463

0.478

0.547

(0.404)

(0.272)

(−0.281)

(0.262)

(0.328)

C.fiBEcapper

−5.629

0.945

8.826***

4.958

−0.289

(−0.850)

(0.511)

(2.991)

(1.184)

(−0.064)

C.soBEcapper

2.660

−1.811***

0.375

2.260

1.936

(0.915)

(−3.042)

(0.372)

(1.418)

(1.037)

C.BEconper

−38.008

−8.919***

−11.112

−29.386*** −34.766***

(−1.634)

(−3.023)

(−1.342)

(−3.024)

(−4.068)

Year

No

Yes

No

Yes

No

Fixed effect

Yes

No

Yes

Yes

Yes

cons

13.719***

13.355***

−1.389

15.829***

(5.585)

(15.884)

(−0.460)

(9.797)

327

327

327

327

2.717

−0.098

2.267

3.949

C.SEconper

N

327

(2) Current administrative expenditure C.fiSEadmper

3.261 (0.325)

(0.847)

(−0.033)

(0.459)

(0.951)

C.soSEadmper

−0.056

1.200

0.960

−1.405

−0.172

(−0.039)

(1.140)

(0.911)

(−0.867)

(−0.120)

1.236

12.396***

−3.434

−5.121

0.553

(0.172)

(5.763)

(−1.240)

(−1.059)

(0.115)

C.soBEadmper

3.084

0.589

−1.438

−2.356

1.678

(1.108)

(0.833)

(−1.000)

(−1.001)

(0.787)

Year

No

Yes

No

Yes

No

Fixed effect

Yes

No

Yes

Yes

Yes

cons

11.002***

14.374***

19.366***

13.631***

N

327

C.fiBEadmper

(4.148)

(16.868)

(10.165)

(8.331)

327

327

327

327

3.707***

4.904***

1.558

3.751**

(3) Current welfare expenditure C.fiSEwelper

3.994**

(continued)

262

Appendix

Table 3 (continued) (1)

(2)

(3)

(4)

(5)

Fe

Panel SUR

FGLS

PCSE

MEML

DV

Unedu

Unedu/Illi

Unedu

Unedu

Unedu

(2.182)

(2.792)

(3.987)

(0.654)

(2.421)

C.soSEwelper

1.672

−1.399

0.626

−0.448

2.036

(0.571)

(−0.980)

(0.481)

(−0.190)

(1.075)

C.fiBEwelper

1.632

0.459

−2.477*

−12.122*** −0.174

(0.599)

(0.401)

(−1.869)

(−3.825)

(−0.088)

0.762

−0.107

0.037

−1.566

0.248

(0.331)

(−0.182)

(0.027)

(−0.724)

(0.140)

Year

No

Yes

No

Yes

No

Fixed effect

Yes

No

Yes

Yes

Yes

cons

9.359**

12.965***

28.501***

12.840*** ara>

(2.525)

(14.042)

(8.057)

(7.360)

327

327

327

327

−1.208

−1.553

4.535

−0.441

C.soBEwelper

N

327

(4) Current scholarship expenditure C.fiSEschper

−0.024 (−0.003)

(−0.421)

(−0.619)

(0.991)

(−0.100)

C.soSEschper

0.260

−1.008

−0.865

−1.028

0.130

(0.132)

(−1.069)

(−0.939)

(−0.757)

(0.094)

−1.337

−4.637***

0.960

−1.646

−0.047

(−0.318)

(−2.864)

(0.418)

(−0.400)

(−0.014)

C.soBEschper

−0.935

0.907*

−0.129

−0.430

−0.677

(−0.698)

(1.891)

(−0.170)

(−0.308)

(−0.669)

Year

No

Yes

No

Yes

No

Fixed effect

Yes

No

Yes

Yes

Yes

cons

12.730***

14.100***

−4.570

14.779***

(15.704)

(−1.338)

(8.942)

N

327

327

327

327

C.fiBEschper

(3.829) 327

(5) Capital expenditure lagging second phase L2.fiSEcapper L2.soSEcapper L2.SEconper

−2.240

−3.189***

−3.135*

0.299

−2.217

(−1.078)

(−2.939)

(−1.956)

(0.169)

(−1.289)

0.977

1.254**

2.183***

0.863

1.199

(1.146)

(2.114)

(3.027)

(0.952)

(1.302)

−7.592***

−4.190***

−6.551***

−3.401*

−7.424*** (continued)

Appendix

263

Table 3 (continued) (1)

(2)

(3)

(4)

(5)

Fe

Panel SUR

FGLS

PCSE

MEML

DV

Unedu

Unedu/Illi

Unedu

Unedu

Unedu

(−5.803)

(−3.314)

(−4.142)

(−1.882)

(−4.751)

L2.fiBEcapper

2.695

7.739**

12.440***

12.077***

7.368

(0.504)

(2.365)

(3.910)

(2.625)ara> (1.605)

L2.soBEcapper

4.087

5.598***

−0.125

3.828**

3.710*

(1.610)

(4.661)

(−0.113)

(2.297)

(1.932)

L2.BEconper

−45.641***

−19.219***

−12.261

−35.715***

−41.095***

(−3.182)

(−3.773)

(−1.561)

(−3.407)

(−4.932)

Year

No

Yes

No

Yes

No

Fixed effect

Yes

No

Yes

Yes

Yes

cons

13.701***

13.067***

−6.852**

15.008***

(6.158)

(16.840)

(−2.227)

(10.407)

296

296

296

8.079**

8.473

20.653***

N

296

296

(6) Administrative expenditure lagging second phase L2.fiSEadmper

19.883**

11.038***

(2.639)

(3.683)

(2.460)

(1.521)

(4.640)

L2.soSEadmper

0.609

−0.207

−0.090

−0.988

0.463

(0.370)

(−0.238)

(−0.074)

(−0.477)

(0.339)

L2.fiBEadmper

−18.732**

−8.391**

−12.346***

−11.972**

−15.377***

(−2.651)

(−2.039)

(−4.130)

(−2.381)

(−2.781)

L2.soBEadmper

5.527

3.683***

0.978

1.206

4.963**

(1.625)

(3.011)

(0.708)

(0.505)

(2.349)

Year

No

Yes

No

Yes

No

Fixed effect

Yes

No

Yes

Yes

Yes

cons

11.902***

14.589***

15.151***

13.301***

N

296

(4.033) 296

(16.947)

(7.845)

(7.593)

296

296

296

(7) Welfare expenditure lagging second phase L2.fiSEwelper L2.soSEwelper L2.fiBEwelper

8.461***

5.482***

5.955***

5.068**

7.241***

(3.721)

(4.558)

(4.826)

(2.105)

(4.595)

2.063

0.692

−1.503

−1.136

1.725

(0.500)

(0.621)

(−1.168)

(−0.531)

(0.999)

3.277

−2.600

−1.708

−12.052*** 0.702 (continued)

264

Appendix

Table 3 (continued) (1)

(2)

(3)

(4)

(5)

Fe

Panel SUR

FGLS

PCSE

MEML

DV

Unedu

Unedu/Illi

Unedu

Unedu

Unedu

(1.039)

(−1.208)

(−1.212)

(−3.600)

(0.322)

L2.soBEwelper

0.406

−1.174

1.708

−1.760

−0.015

(0.161)

(−0.987)

(1.334)

(−0.884)

(−0.009)

Year

No

Yes

No

Yes

No

Fixed effect

Yes

No

Yes

Yes

Yes

cons

2.535

11.368***

24.434***

8.805***

(11.756)

(6.359)

(4.599)

N

296

296

296

296

(0.706) 296

(8) Scholarship expenditure lagging second phase 10.443*

1.755

−1.262

2.107

6.090

(1.748)

(0.732)

(−0.489)

(0.466)

(1.383)

−2.786

−1.785**

−0.572

−4.444**

−2.284

(−1.437)

(−2.101)

(−0.603)

(−2.533)

(−1.503)

L2.fiBEschper

2.258

11.384***

5.571**

4.071

3.770

(0.433)

(3.178)

(2.236)

(0.985)

(1.016)

L2.soBEschper

−3.503*

−1.458

−2.107***

−0.387

−3.104***

(−1.877)

(−1.344)

(−2.766)

(−0.301)

(−3.032)

L2.fiSEschper L2.soSEschper

Year

No

Yes

No

Yes

No

Fixed effect

Yes

No

Yes

Yes

Yes

cons

13.105***

13.974***

−6.939**

15.375***

(15.480)

(−2.051)

(9.082)

N

296

296

296

296

(5.353) 296

Note “C.*” represents the current value of the variable. “L2.*” refers to the lag phase II of all variables

Appendix

265

References Fan, G., Wang, X., & Zhu, H. (2003). NERI index of marketization of China’s provinces. National Economic Research Institute, Beijing. Jiang, Q. P. (2018). Sharing economy: Political economy of monopoly competition. Tsinghua University Press. (in Chinese). Sen, A. (2004). Rationality and freedom. Harvard University Press.