India Migration Report 2020: Kerala Model of Migration Surveys [1 ed.] 9780367625566, 9781003109747

Citation preview

India Migration Report 2020

India Migration Report 2020 examines how migration surveys operate to collect, analyse and bring to life socio-economic issues in social science research. With a focus on the strategies and the importance of information collected by Kerala Migration Surveys since 1998, the volume: • • • •

Explores the effect of male migration on women left behind; attitudes of male migrants within households; the role of transnational migration and its effect on attitudes towards women; Investigates consumption of remittances and their utilization; asset accumulation and changing economic statuses of households; financial inclusion of migrants and migration strategies during times of crises like the Kerala floods of 2018; Highlights the twenty-year experience of the Kerala Migration Surveys, how its model has been adapted in various states and led to the proposed large-scale India Migration Survey; and Explores issues of migration politics and governance, as well as the return migration strategies of other countries to provide a roadmap for India.

The volume will be of interest to scholars and researchers of development studies, economics, demography, sociology and social anthropology, and migration and diaspora studies. S. Irudaya Rajan is Professor at the Centre for Development Studies, Thiruvananthapuram, Kerala, India. With more than three decades of research experience in Kerala, he has coordinated eight major migration surveys (1998, 2003, 2007, 2008, 2011, 2014, 2016 and 2018) in Kerala (with Professor K.C. Zachariah), led the migration surveys in Goa (2008), Punjab (2011) and Tamil Nadu (2015), and provided technical support to the Gujarat Migration Survey (2010). He has published extensively in national and international journals on demographic, social, economic, political and psychological implications of international migration. Professor Rajan is currently engaged in several projects on international migration with the New York University, UAE Exchange Centre, India Centre for Migration of the Ministry of External Affairs and World Bank. He worked closely with the erstwhile Ministry of Overseas Indian Affairs, government of India; Department of Non-Resident Keralite Affairs (NORKA), government of Kerala and Kerala State Planning Board. He is currently co-chairing the working group on NORKA for the thirteenth five-year plan (2017– 2022) of Kerala State Planning Board, government of Kerala, and is initiating the Kerala Migration Survey 2021, funded by the Department of NORKA, government of Kerala. He is Editor of two Routledge series, India Migration Report (annual) since 2010 and South Asia Migration Report (biennial), and is the founder and Editor-inChief of the journal Migration and Development. Currently, he is the Chair of the KNOMAD (The Global Knowledge Partnership on Migration and Development, World Bank) working group on internal migration and urbanization. He is one of the expert committee members to advise the government of Kerala on COVID-19.

India Migration Report Editor: S. Irudaya Rajan

Centre for Development Studies, Thiruvananthapuram, Kerala, India

This annual series strives to bring together international networks of migration scholars and policy-makers to document and discuss research on various facets of migration. It encourages interdisciplinary commentaries on diverse aspects of the migration experience and continues to focus on the economic, social, cultural, ethical, security, and policy ramifications of international movements of people. India Migration Report 2010 Governance and Labour Migration India Migration Report 2011 Migration, Identity and Conflict India Migration Report 2012 Global Financial Crisis, Migration and Remittances India Migration Report 2013 Social Costs of Migration India Migration Report 2014 Diaspora and Development India Migration Report 2015 Gender and Migration India Migration Report 2016 Gulf Migration India Migration Report 2017 Forced Migration India Migration Report 2018 Migrants in Europe India Migration Report 2019 Diaspora in Europe India Migration Report 2020 Kerala Model of Migration Surveys India Migration Survey 2021 (forthcoming) Migration and Health

India Migration Report 2020 Kerala Model of Migration Surveys Edited by S. Irudaya Rajan

First published 2021 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 52 Vanderbilt Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2021 selection and editorial matter, S. Irudaya Rajan; individual chapters, the contributors The right of S. Irudaya Rajan to be identified as the author of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book has been requested ISBN: 978-0-367-62556-6 (hbk) ISBN: 978-1-003-10974-7 (ebk) Typeset in Sabon by Apex CoVantage, LLC

Contents

List of figuresviii List of tablesx List of contributorsxv Prefacexviii Acknowledgementsxxv   1 Large-scale migration surveys: replication of the Kerala model of migration surveys to India Migration Survey 2024

1

S. IRUDAYA RAJAN, K. C. ZACHARIAH AND ASHWIN KUMAR

  2 Keeping up with Kerala’s Joneses

23

BALASUBRAMANYAM PATTATH

  3 Remittances, health expenditure and demand for healthcare services

55

C. VALATHEESWARAN AND M. IMRAN KHAN

  4 Remittances and asset accumulation among the left-behind households

72

POOJA BATRA

  5 Impact of male migrants and their return, on women left behind: the case of Kerala, India

97

JURANY RAMIREZ GALLEGO, S. IRUDAYA RAJAN AND ARJUN S. BEDI

  6 Socio-economic dynamics of Gulf migration: a panel data analysis SHIBINU S.

120

vi  Contents   7 Why do migrants transfer money? Motivations of remittance from emigrants

136

ARYA RACHEL THOMAS AND ARCHANA K. ROY

  8 Reintegration and future plans of return migrants

147

AFSAL K. S. AND RESHMI R. S.

  9 Emigration and its effect on the labour force participation of women in the left-behind household

162

ROSHAN R. MENON AND R. B. BHAGAT

10 Transnational migration and gender attitudes: an exploratory analysis

177

S. IRUDAYA RAJAN, GINU ZACHARIA OOMMEN, SYED USMAN JAVAID, GEORGE JOSEPH, JENNIFER SOLOTAROFF AND LUIS ALBERTO ANDRES

11 Labour migration from Kerala: decision to migrate and choice of destination

204

ANU ABRAHAM

12 Social remittances of Keralans in neoliberal circulation

228

KELLEE S. TSAI

13 The alcohol paradox: revisiting the Kerala model of development

253

AJIT DAYANANDAN, RAJESH MANY AND GEORGE JOSEPH

14 Financial cost of international migration from South India: what does data spanning three decades tell us?

268

GANESH SESHAN

15 Exploitative or ethical? Understanding the labour recruitment practices in the 21st century from Kerala

280

C. S. AKHIL AND S. IRUDAYA RAJAN

16 Remittances and overseas migrants’ economic shocks: evidence from Kerala’s recipient households after the 2008 global crisis EUGENIA CANESSA

293

Contents vii 17 Cultural production of narratives as counter-archives in Kerala Gulf emigration studies

319

PRIYA MENON

18 Migration and asset accumulation in South India: comparing gains to internal and international migration from Kerala

326

GANESH SESHAN

19 Tracing the changing economic status of Kerala households: the KMS experience

359

UDAYA S. MISHRA AND S. IRUDAYA RAJAN

20 The Janus-faced dilemma of Migration as Adaptation: the impact of rapid onset floods on labour migration in Kerala, India, 2018

372

JESSICA LI

21 Climate change, migration and development: the aftermath of the 2018 floods in Kerala

385

S. IRUDAYA RAJAN, STEVE TAYLOR, ANEETA SHAJAN AND LISA HARDISTY

22 Impact of migration on financial inclusion and local financial systems in Kerala

407

K. JAFAR

23 International migration and global diplomacy

426

ALWYN DIDAR SINGH

24 Governance of emigration today

437

T. K. MANOJ KUMAR AND M. C. LUTHER

25 Empowerment of female return migrant initiatives in Indonesia: lessons for India

459

MUNA YASTUTI MADRAH, WAHYU ARIF RAHARJO, UMMU SYAFIQAH AZLAN, FIRMAN BUDIANTO, AROKKIARAJ HELLER AND ANTONIUS MARIA INDRIANTO

Index469

Figures

4.1 6.1 6.2 7.1 7.2 8.1 9.1 13.1 14.1 14.2 14.3 14.4 14.5 14.6 14.7 15.1 15.2 15.3 15.4 15.5 15.6 15.7 18.1

Utilization of foreign remittances by LBH based on frequency of receiving Change in the usage of cooking fuel, 2016 Change in the type of house, 2011 and 2016 Hypothetical model of change in motivation to remit Purpose for sending remittance to the household Countries last visited among return migrants in Malabar and Kerala Trends in labour force participation among women in emigrant households and non-emigrant households of Kerala Revenue from alcohol in Kerala’s revenue, 1996–2013 Real migration cost, 1980–2013 Real migration cost by destination countries in the GCC Real migration cost – selected components Real migration cost by gender Real migration cost by education attainment Real migration cost by number of other HH members working abroad Real migration cost by rate of migration at the sub-district level Cost of migration from 1970 to 2015 Average cost of documentation in million INR from 1970 to 2015 Average real cost of transportation from 1970 to 2015 Cost of recruitment from 1970 to 2015 Breakdown of total cost of migration in 1970 Breakdown of total cost of migration in 2015 Comparison between the average cost of network migration and recruitment agent-driven migration under various heads in the migration cost Overall assets index against landholdings in 1998

86 126 127 140 141 151 168 254 271 272 273 274 275 275 276 285 286 287 287 288 288 289 340

Figures ix 18.2

Change in overall assets index by migration status (1998–2003)340 18.3 Cumulative density function of change in overall assets index by migration status (1998–2003) 341 18.4 Education distribution of Kerala’s migrants in 2014 350 18.5 Income distribution by migration status in 2014 351 19.1 Index shift across quintiles 366 19.2a Index shift across quintiles – districts 367 19.2b Index shift across quintiles – districts 367 19.3 Index shift across quintiles – migrant and non-migrant households368 21.1 Breakdown of the casualties caused due to floods 388 21.2 Distribution of the damage caused across sectors 391 21.3 Conceptual framework to study the impact of floods on migration393 21.4 Age group distribution of illness 401 24.1 Emigration from the top nine labour sending states444 24.2 Migration trends from Indian states 445 24.3 Workers migrating to important destinations abroad 446

Tables



1.1 1.2 1.3 1.4

Various rounds of the Kerala Migration Survey, 1998–2008 6 Number of emigrants from Kerala 1998–2018 11 Number of return migrants into Kerala 1998–2018 12 Total remittance estimates and remittances per household in Kerala 1998–2018 12 1.5 KMS old and new panels 1998–2018 14 1.6 India Migration Survey 2024: sample frame 18 2.1 Breakup of households based on migration status 38 2.2 Change of household type by migration status between 2011 and 2016 38 2.3 Out-migration of household member 41 2.4 Out-migration of household member 43 2.5 Out-migration of household member 44 2.6 Change from zero migrants in 2011 to at least one migrant in 2016 45 2.7 Change from zero migrants in 2011 to at least one migrant in 2016 46 2A.1 Correlations between the relative deprivation measures and consumption expenditure 52 2A.2 Correlation between Yitzhaki indices and consumption expenditure52 2A.3 List of variables and their definitions 53 2A.4 Results of interviews 54 3.1 Descriptive statistics 61 3.2 The effect of remittances receipt on per-capita health expenditure64 3.3 Impact of remittances on household per-capita health expenditure66 4.1 The top remittance-receiving countries in the world 73 4.2 Descriptive statistics of international migrant households 78 4.3 Descriptive statistics of sampled households 79 4.4 Mode-wise remittance transfers to left-behind households 80 4.4(a) Mode-wise remittance transfers to left-behind households in rural areas 80

Tables xi 4.4(b) Mode-wise remittance transfers to left-behind households in urban areas 80 4.5 Purpose of international remittances by left-behind households81 4.5(a) Purpose of international remittances by left-behind households based on the location 82 4.5(b) Purpose of international remittances by left-behind households based on the geographical division 82 4.5(c) Purpose of international remittances by left-behind households based on the sex of the household head 83 4.5(d) Purpose of international remittances by left-behind households based on the sex and education of the household head 83 4.6 SUR estimates of expenditure on instruments of savings and investment for international migrant households 89 4.7 PSM estimates of expenditure on instruments of savings and investment for international migrant households 91 5A.1 Descriptive statistics 112 5A.2 Differences in means of the dependent variables by status of migration 114 5A.3 Differences in means of household characteristics by status of migration 115 5A.4 Logits for decision-making 116 5A.5 Logits for freedom of movement 117 5A.6 Logits for justification of domestic violence 118 5A.7 Description of variables 118 6.1 Household type of current migrant households (2016) in 2011 124 6.2 Change in family size, 2011 and 2016 124 6.3 Change in household headship, 2011 and 2016 125 6.4 Proportion of emigrants in 2011 and 2016 by destination 125 6.5 Households having possessions in 2016 127 6.6 Change in remittances 2011 and 2016 128 6.7 Logistic regression analysis to find out the impact of migration129 7.1 Description of variables used in the econometric model 139 7.2 Effect of transfer motives on purposes for which remittance is sent 142 7.3 Mean remittance received by the household income gap quintiles143 7.4 Motivational differences: Spline Tobit regression model showing the impact of motivation changes on total remittances sent 144 7.5 Marginal effects in Tobit model 145 8.1 Trends and distribution of return migration by districts in the Malabar region 149

xii  Tables

8.2 8.3 8.4 8.5

8.6 9.1 9.2 9.3 9.4 9.5 10.1 10.2 10.3 10.4 10A.1

11.1 11.2 11.3 11.4

11.5 13.1 13.2 13A.1 13A.2 3A.3 1 14.1 16.1 16.2 16.3 16.4

Profile of return migrants in Malabar 150 Reasons for return 152 Reasons for return migration by country last visited 152 Disease suffered and nature of work of return migrants in Malabar and Kerala 153 Economic activity before migration and at destination 155 Female labour force participation rate for Kerala (%) 163 Profile of women by emigration and remittancereceiving status in Kerala, 2018 167 District-wise level of labour force participation of women in the left-behind households of Kerala, 2018 169 Economic activity undertaken by women in emigrant and non-emigrant households of Kerala, 2018 170 Effect of emigration and remittances on the labour force participation of women in Kerala 172 Gender attitudes by religion, KMS 2013 184 Gender attitudes by religion, KMS 2013 188 Gender attitudes by wealth, KMS 2013 190 Gender attitudes by education, KMS 2013 191 A break-up of gender attitudes across different migrant groups (non-migrants vs OECD, Gulf, and domestic migrants)201 Description of explanatory variables 211 Descriptive statistics 213 Probit estimates of the decision to migrate 215 Multinomial logistic estimates of destination choice – internal or international 218 Destination choice of international migrants – Gulf vs non-Gulf 222 Demographic pattern of alcohol consumption behaviour in Kerala, 2012–13 259 Determinants of alcohol consumption in Kerala 261 Descriptive statistics of variables in the sample, 2012–13265 Correlation matrix (Pearson Correlation Coefficient) of major variables in the sample 266 Details of variables 267 Determinant of real migration costs 277 Initial characteristics of panel sample 300 Destinations of international migrants (September 2008) 302 Money brought home from labour migrants in response to unemployment and exchange rate shocks, fixed effect for 2008–2009 310 Dummy for receiving remittances from labour migrants in response to unemployment and exchange rate shocks, fixed effects for 2008–2009 311

Tables xiii 16.5 16.6 16.7 18.1 18.2 18.3 18.4 18.5 18.6 18.7 8A.1 1 18A.2 18A.3 19.1 19.2a 19.2b 19.3a 19.3b 19.4a 19.4b 19.5a 19.5b 21.1 21.2

Number of return migrants in response to unemployment and exchange rate shocks, fixed effect for 2008–2009 Remittances in response to unemployment and exchange rate shocks, fixed effect for 2008–2009, additional controls Remittances from labour migrants in response to unemployment and exchange rate shocks, fixed effect for 2008–2009 Summary statistics of panel households Principal components and summary statistics for constructed asset indices Prediction of international and domestic migration (first-stage regression) Impact of migration on household wealth Identification tests Impact of migration on household wealth – components of housing quality index Impact of migration on household wealth – components of durable asset index Impact on household wealth by duration of migration Impact of duration of migration on household wealth – components of housing quality index Impact of duration of migration on household wealth – components of durable asset index Transition in MPCE distribution of Kerala households during the period 1999–2012: NSSO rounds Household asset distribution in Kerala Migration Survey 1998 by districts Household asset distribution in Kerala Migration Survey 2018 by districts Household asset distribution in Kerala Migration Survey 1998 by type of households Household asset distribution in Kerala Migration Survey 2018 by household type Household asset distribution in Kerala Migration Survey 1998 by sex of head of household Household asset distribution in Kerala Migration Survey 2018 by sex of head of household Household asset distribution in Kerala Migration Survey 1998 by religion Household asset distribution in Kerala Migration Survey 2018 by religion Breakdown of sampling units Number of households and individuals affected by the floods

312 313 314 332 338 342 344 345 347 348 356 357 358 362 364 365 368 368 369 369 370 370 395 396

xiv  Tables 21.3 21.4

21.5 21.6 21.7 21.8 21.9 21.10



21.11 21.12 21.13 22.1 22.2 22.3

22.4 22.5 22.6

22.7 22.8 22.9 24.1 24.2 24.3

How did the floods affect jobs among affected individuals? Top ten activities that suffered during the floods and reasons for disruption (working-age population) Floods’ effect on land holdings/holders Losses incurred to businesses due to the floods Debts incurred Effect on health due to the floods Types of diseases which were suffered as a result of floods Experienced long-term psychological distress related to the floods Types of mental distress Reasons for migration Views about fisherfolk Distribution of migrant households across regions  Distribution of household remittance across regions  Land ownership pattern across migrant and nonmigrant households  Migration status and access to banking services  Migration status and household savings Migration status and engagement in Kurikkalyanam practice  Migration status and source of primary borrowing  Migration status and current liability of households  Migration status and most reliable source Emigration from the top nine labour sending states Migration trends from Indian states Workers Emigrating Abroad in 2017

396 397 398 399 399 400 400 402 402 403 403 412 412 414 415 416 417 418 419 420 443 445 446

Contributors

Anu Abraham is an assistant professor at the Narsee Monjee Institute for Management Studies (NMIMS), Mumbai, India. K. Afsal is a PhD student at the International Institute of Population Sciences, Mumbai, India. C. S. Akhil is a PhD student at the Centre for Development Studies, Thiruvananthapuram, India. Luis Alberto Andres is a lead economist at the World Bank. Ummu Syafiqah Azlan is a master candidate at the Department of Geography at National University of Singapore, Singapore. Pooja Batra is a PhD scholar in the economics department of Indian Institute of Management, Indore, India. Arjun S. Bedi is a professor of development economics at the International School of Social Studies (ISS) of Erasmus University Rotterdam. R. B. Bhagat is Professor in the Department of Migration and Urban Studies, International Institute of Population Sciences, Mumbai, India. Firman Budianto is a researcher at the Research Centre for Area Studies, Indonesian Institute of Science, Indonesia. Eugenia Canessa is a PhD student in development and local systems at the University of Trento and the University of Florence, Italy. Ajit Dayanandan is a professor of finance at the University of Alaska, Anchorage, United States of America. Jurany Ramirez Gallegos is a researcher at the International School of Social Studies (ISS) of Erasmus University Rotterdam. Lisa Hardisty is Principal Lecturer at Department of Social Sciences, Northumbria University, United Kingdom. Arokkiaraj Heller is a post-doctoral fellow at the Leibniz Science Campus, Eastern Europe-Global Area, Leipzig, Germany.

xvi  Contributors Antonius Maria Indrianto is a doctorate candidate at Institute for Population and Social Research at Mahidol University, Thailand. K. Jafar is an assistant professor at the Madras Institute of Development Studies, Chennai, India. Syed Usman Javaid is Operations Officer at the World Bank Pakistan Country Management Unit. George Joseph is a senior economist, Water Global Practice, World Bank. M. Imran Khan is an assistant professor at the Narsee Monjee Institute for Management Studies (NMIMS), Mumbai, India. Ashwin Kumar is a research scholar at the School of Interdisciplinary and Trans-Disciplinary Studies, IGNOU, New Delhi. He is also associated as a researcher with the Centre for Development Studies, Thiruvananthapuram, India. T. K. Manoj Kumar is an Indian civil servant, and has previously served as Joint Secretary in the Diaspora and Emigration Policy Divisions of the Ministry of Overseas Indian Affairs, Government of India. He has also held the position of Protector General of Emigrants. Jessica Li is currently a consultant with Future of Technology, Capgemini Invent. She had previously completed a master of arts in environmental policy from Sciences Po, France. M. C. Luther is a former Indian civil servant and previously served as the Joint Secretary to the government of India in the Overseas Employment Division of the Ministry of External Affairs as the Protector General of Emigrants (PGE). Muna Yastuti Madrah is a lecturer at the faculty of Islamic studies, Universitas Islam Sultan Agung, Indonesia. Rajesh Many is an assistant professor at the School of Gandhian Thought and Development Studies at the MG University, Kottayam, Kerala, India. Priya Menon is an associate professor in the English department at Troy University, Alabama, United States of America. Roshan R. Menon is M Phil scholar at International Institute of Population Sciences, Mumbai, India. Udaya S. Mishra is a professor at the Centre for Development Studies, Thiruvananthapuram. Ginu Zachariah Oommen is a member of the Kerala Public Service Commission, India and Director of the International Institute of Migration and Development (IIMAD).

Contributors xvii Balasubramanyam Pattath is currently a research associate at the Indian School of Business, Hyderabad, India. Wahyu Arif Raharjo is a lecturer at the Department of International Relations, Universitas Islam Indonesia, Indonesia. S. Irudaya Rajan is Professor at the Centre for Development Studies, Thiruvananthapuram, Kerala, India. Archana K. Roy is an associate professor in the Department of Migration and Urban Studies, International Institute of Population Sciences, Mumbai, India. Reshmi R. S. is an assistant professor in the Department of Migration and Urban Studies, International Institute of Population Sciences, Mumbai, India. Shibinu S. is Head of Department, Economics at P.S.M.O. College, Tirurangadi, Malappuram, Kerala, India. Ganesh Seshan is a senior economist in the Migration and Remittance Unit (DECMR) under the Development Economics Vice Presidency at the World Bank. Aneeta Shajan is a researcher at the Centre for Development Studies, Thiruvananthapuram, India. Alwyn Didar Singh is a former civil servant and served as secretary to the government of India in the Ministry of Overseas Indian Affairs (2009– 11), India. He also served as the secretary general of Federation of Indian Chamber of Commerce and Industry (FICCI). Jennifer Solotaroff is a senior social development specialist at the World Bank. Steve Taylor is a professor in the Department of Social Sciences at Nothumbria University, United Kingdom. Arya Rachel Thomas is a PhD student at the International Institute of Population Sciences, Mumba, India. Kellee S. Tsai is Chair Professor and Dean of Humanities and Social Science at the Hong Kong University of Science and Technology, India. C. Valatheeswaran is a postdoctoral researcher, Social Sciences Institute, National University of Ireland, Maynooth, Ireland. K. C. Zachariah is Honorary Professor at the Centre for Development Studies, Thiruvananthapuram, Kerala, India, and President of the International Institute of Migration and Development (IIMAD).

Preface

It is my immense pleasure to introduce the eleventh edition of the annual India Migration Report, which focuses on the importance of information collected by Kerala Migration Surveys since 1998. Looking at studies done through data collected from several rounds of migration surveys held in Kerala on a number of socio-economic issues as a consequence of migration, this volume acts as a unique repository of information, highlighting the importance of collecting large-scale data on this extremely important and dynamic phenomenon. The studies presented in this edition delve into varied topics such as conspicuous consumption of remittances and their utilization, the effect of male migration on left-behind women and the attitudes of male migrants within the household to a number of other socio-economic dimensions such as financial inclusion of migrants and return migration strategies, as well as migration strategies during times of crises like the Kerala floods of 2018. Through the various chapters, it also highlights the twenty-year experience of the Kerala Migration Surveys, which have been conducted by the Centre for Development Studies, Thiruvananthapuram and has taken place every five years since 1998 with eight completed rounds  – the only migration survey in the world to do so. Additionally, it also looks at issues of migration politics and governance, as well as return migration strategies of other countries to provide a roadmap for India. However, before I provide details about the chapters in the India Migration Report 2020, I would like to revisit the previous editions of the India Migration Report and the themes that they touched upon. The first edition of the India Migration Report, released in 2010, focused on the theme of governance and labour migration. Looking at the various governance structures and practices that govern international migration and labour recruitment, the volume shed light on the various issues plaguing workers, both skilled and unskilled, as they migrated for work in the Gulf Cooperation Council (GCC) countries. The volume touched upon the exploitative practices of recruitment agents and also looked at these issues through a gendered lens. The India Migration Report 2011 focused on the theme of migration, identity and conflict, and particularly on internal migration. The volume

Preface xix examines the changing nature of identities in a changing world brought about by migration and the violence and conflict that is a direct consequence of it. Set in the background of the Census 2011 and first talk about the implementation of the National Population Register of citizens, the importance of understanding cross-border and internal migration became a necessity. It looked directly at issues of conflict arising due to exploitative recruitment practices, poverty and gendered norms among vulnerable migrant workers, who are the backbone of the Indian economy. The India Migration Report 2012 was set in the aftermath of the Global Financial Crisis, which originated in the United States in 2008, but had gradually spread all over the world. The volume dealt with the socio-economic impact of this significant event on migration, remittance flows and development strategies, and the increasing cases of return migration and rehabilitation of these return migration in the wake of the crisis. The India Migration Report 2013 dealt with the important, yet often overlooked, aspect of social costs of migration – particularly the psychological and human costs on the people left behind in the migration process – mainly women, children and the elderly. The chapters dealt with topics such as children’s negotiation with parental migration, effects of migration on the elderly left behind, as well as women left behind and the larger demographic implications of migration. The India Migration Report 2014 looked at the role of the large Indian diaspora in various countries and their contributions to the homeland in the form of remittances and diaspora philanthropy. The volume contained chapters on the role of the diaspora in the development of the homeland, as well as the larger political and social contexts in which diaspora networking happens, and India’s policy towards it. The India Migration Report 2015 focused on the various aspects of gender and migration. Highlighting the fact that the increasing feminization of migration flows from India have largely been relegated to the background, the volume focused on the patriarchal norms through which policy is normally formed. The volume looked at the dynamics of female migration within India and ways in which policy mechanisms often overlook its complexities. The India Migration Report 2016 dealt specifically on the theme of Indian migration to the Gulf. The volume brought out the various policy and labour recruitment practices in both Indian and Gulf Cooperation Council (GCC) countries, with a view of the future of this very important migration corridor. The India Migration Report 2017 was based around the important theme of forced migration and looked at the various socio-economic consequences of it in India today. The volume contained chapters which critically examined issues relating to migration due to political conflict, natural disasters and development-induced forced migration, thereby bringing about a holistic view on the subject.

xx  Preface The India Migration Report 2018 focused specifically on the issues of Indian migrants in Europe, which has always been a preferred destination for high-skilled professionals and students after the United States. The various chapters looked at a number of issues relating to migration policies and politics in Europe and their effect on Indian migration, as well as the socioeconomic issues of already-settled Indian migrants in the region. The tenth edition of the India Migration Report in 2019 focused on the diaspora in Europe. The volume dealt with the various ways in which the Indian diaspora in Europe contributed to both the host countries and India as well. The changing dynamics of economic and political engagements, such as the fallout of Brexit on migration into Europe, were looked at in the chapters contained in the volume. The India Migration Report 2020 is, thus, the eleventh volume in the series and is outlined into twenty-five chapters, each looking at various aspects of migration using data collected through various migration surveys. The Introductory chapter, by S. Irudaya Rajan, K. C. Zachariah and Ashwin Kumar takes a look at a broad view of migration surveys around the world and in India. Detailing the type of information and the importance of such surveys, it then focuses on the vast experience gathered from the eight rounds of the Kerala Migration Survey over twenty years and its model being adapted in various states and leading to the proposed large-scale India Migration Survey. The second chapter, by Balasubramanyam Pattath, looks at the conspicuous consumption among migrant households in Kerala and how that determines household status among the community. The study relies on the panel data available through the Kerala Migration Surveys of 2011 and 2016, and finds that this conspicuous consumption also increases the tendency of these households to migrate. C. Valatheeswaran and M. Imran Khan use data from the 2010 Kerala Migration Survey to determine the effect of remittances on healthcare utilisation in Kerala. The study finds that, on the whole, remittance leads to greater spending in healthcare, it has a greater impact on lower-income households and those households belonging to the general category, but is not significant for households in the SC/ST category, who fail to reap the benefits of international migration. Using data from the Kerala Migration Survey of 2011, Pooja Batra examines whether increased remittances have led to an accumulation of assets within migrant households in Kerala. The study questions whether this leads to smooth long-run consumption of households in the absence of traditional savings and investments through income-generating sources. This is important, as this could lead to dependency on remittances and to “stunted development”. Jurany Ramirez Gallego, S. Irudaya Rajan and Arjun Bedi then look into the important aspect of left-behind women in the context of male migration to the Gulf. Using data from the 2016 Kerala Migration Survey, which had

Preface xxi a separate section on the effect of migration on gender attitudes, the study examines three aspects of women’s decision-making freedom, mobility and justification of domestic violence. In the sixth chapter, S. Shibinu examines data from the Kerala Migration Survey over the period of 2011–2016 to analyse the various socio-economic characteristics of migrant households in Kerala. The study creates a breakdown of characteristics, such as educational qualifications, region, household dynamics, religion and consumption needs of households, in order to determine factors of migration to the Gulf. Arya Rachel Thomas and Archana K. Roy examine the reasons for remittance-sending among individuals to households back home. Using data from the 2016 and 2018 Kerala Migration Surveys, and based on the analysis of income quintiles, the study tries to empirically analyse whether remittances are sent because of a selfish “exchange” reason, a selfless “altruistic” reason or a mix of the two – a “Veblen goods driven altruism” reason. The study finds that reasons vary significantly over different income quintiles. Afsal K. S. and Reshmi R. S., largely through an examination of the Kerala Migration Survey of 2018, focus on the socio-economic issues that return migrants suffer upon their return back into Kerala society. Based on both primary and secondary data, the study examines the trends, patterns and reasons of return migration, as well as the health issues faced by these migrants. Roshan R. Menon and R. B. Bhagat study the effect of emigration on the participation of women left behind labour force. While it is assumed that in a more empowered society, even more empowered in light of increased decision-making in the absence of males, the study finds the opposite and that the labour-force participation of women in migrant households is generally low due to a plethora of various factors, such as educational background, the status of women and religion. In the tenth chapter, S. Irudaya Rajan, Ginu Zachariah Oommen, George Joseph, Syed Usman Javaid, Jennifer Solotaroff and Luis Alberto Andres look at the role of transnational migration and its effect on attitudes towards women. Using data from the 2016 Kerala Migration Survey, the study covers a range of subjects, including attitudes to employment, gender-based violence, decision-making and general attitudes. The study finds that contrary to the myth of Kerala as a progressive society and the claim that migrants have generally more liberal attitudes than non-migrants, gender attitudes are generally more conservative than the non-migrant population, especially among migrants from the GCC countries. Using data from the Kerala Migration Survey of 2018, Anu Abraham analyses the determinants of international migration and the choice of destination. The study finds that individuals with a higher propensity to migrate are male, married, educated, unemployed and have a friend or relative already working outside. Muslims and individuals belonging to households with an existing emigrant and relatively wealthier households have a higher

xxii  Preface probability to emigrate, while emigrant networks reduce the probability of migrating internally. Kellee Tsai then looks at the phenomenon of migration from Kerala to the Gulf through the lens of the circulation of human and social capital in a neoliberal framework. She finds, that when caught between migration aspirations from Kerala to temporary migration in the Gulf, Non-Resident Keralites (NRKs) are often feted upon their arrival back in Kerala, and their movement is a source of prestige, which always acts as an attraction for migration. On the other hand, the study also finds neglect in the state infrastructure for the rehabilitation and re-integration of NRKs, which leads to the rise of NRK politics in order to put NRK issues in the limelight. Ajit Dayanandan, M. Rajesh and George Joseph use data from the 2014 Kerala Migration Survey to analyse a very important social scourge in Kerala society – the scourge of alcoholism. Based on a quantitative analysis of the determinants of alcoholism among Kerala society based on socio-economic characteristics, the study finds that gender, educational background, caste, marital status, religion and poverty play a key role in alcoholic tendencies. The high rate of alcoholism in Kerala thus proves to be dilemma for the much-touted Kerala Model of Development. Ganesh Seshan uses data from the Kerala Migration Surveys of 1998, 2007, 2008 and 2013 to examine the real costs of migration using a novel disaggregated time series method. Cost in real terms are found to be declining, driven by falling airfares and visa fees, and are lower for female migrants, for those with higher education and when other members of the household are working abroad. C. S. Akhil and S. Irudaya Rajan look at the twenty years of Kerala Migration Surveys in order to examine the trends and patterns in the cost of migrating to various destinations over the years. The study finds that through a spate of policy initiatives and regulations over the years at both the central and state level, the cost of migration has declined substantially, though with room for more reduction through appropriate policy responses. Using the Kerala Migration Re-survey of 2009, Eugenia Canessa studies the impact on remittances of macro-economic shocks at the destination. Using the context of the 2008 Financial Crisis and its effect on the Gulf, the chapter examines the role of unemployment at the destination, and the fluctuating exchange rates and the effect they have on remittance sending emigrants. The study finds that remittance flows increase in times of shocks, which can be seen as a shock-absorbing mechanism. Next, Priya Menon uses a literary analysis in order to highlight the human and affective experiences of Indian migrants in the Gulf. Mainly focusing on Benyamin’s famous novel Aadujeevitham, which chronicles the everyday life of labourer in the Gulf, the chapter shows how these “counter-archives”, which showcase the everyday feelings and thoughts of migrant workers, act as a valuable source of information and stand in complement to studies based on the socio-economic and legal contexts of these migrants.

Preface xxiii Ganesh Seshan looks into the role of migration in asset accumulation among households in Kerala. Using panel data from the Kerala Migration Survey over the period 1998–2003, the chapter compares the gains in asset accumulation from both international and internal migration in Kerala. The study finds migrant households, as a whole, experienced higher asset gains than non-migrant families over this five-year period. S. Irudaya Rajan and Udaya S. Mishra examine the changing economic status of Kerala households over a two-decade period from 1998–2018, using the novel approach of examining distributional shift based on a median marker using National Sample Survey Office (NSSO) and Kerala Migration Survey (KMS) data. This comparison confirms that the NSSO and KMS data conform to each other in showing the improvement brought on by migration in terms of region, migrant status, religion and headship of the household. In Chapter 20, Jessica Li uses the backdrop of the Kerala floods of 2018 to explore the impact of floods on migration in a climate-vulnerable geography dependent on multiple forms of labour migration to sustain its economy. The chapter looks at main migratory trends, the impact of floods on lives and livelihoods in Kerala, and analyses the role of migration as an adaptation strategy. S. Irudaya Rajan, Steve Taylor, Aneeta Shajan and Lisa Hardisty present the preliminary findings of the Kerala Flood and Migration Survey of 2019. The survey, which was held in the wake of the Kerala Floods of 2018, uses data from the Kerala Migration Survey of 2018 as a baseline and reports on the findings based on the human, physical and financial damage done by the floods, and the role of migration to ameliorate those losses. The study, therefore, provides a unique insight through quantitative data into the role of migration as a coping strategy during times of natural disasters. Jafar K. tries to understand the role of remittances in reshaping local financial arrangements. Using primary data through structured questionnaires, the study maps a range of formal and informal financial arrangements in place which reflect the dualistic nature of India’s financial system. It finds that migration status and access to remittances positively affect the process of financial inclusion and dependency of formal financial services. Alwyn Didar Singh looks at the geopolitics of migration governance, particularly within a multilateral framework. Looking at large binding agreements such as the New York Declaration on Large Movements of Refugees and Migrants, which led to the 2018 Global Compact on Migration, the chapter contends that these large agreements signed by many countries do not actually translate into freer movement for people across borders and is something that needs to be dealt with seriously within global diplomacy. T.K. Manoj Kumar and M.C. Luther, using their vast experience in the Overseas Employment Division of the government of India, look at the history of emigration clearance in India. Starting with its evolution during colonial times, this chapter looks at the way present institutions established

xxiv  Preface under the Emigration Act of 1983 are structured. The chapter also looks at specific challenges related to emigration of women and the latest measures employed by the government of India to protect workers abroad. The last chapter, by Muna Yastuti Madrah, Wahyu Arif Raharjo, Ummu Syafiqah Azlan, Firman Budianto, Arokkiaraj Heller and Antonius Maria Indrianto, presents the findings of a primary-level study which examined two policies aimed at rehabilitation and re-integration of return migrants in Indonesia. The chapter critically examines the policies – DEMISGRATIF and DESBUMI – and how they have been implemented by governance at each level in a particular village, providing examples for the same in India, as well as the return migration which is expected to increase in the coming years. The India Migration Report 2021 will focus on migration and health, while the India Migration Report 2022 will examine Indians in Canada. S. Irudaya Rajan

Acknowledgements

For the past ten years and eleven editions, the India Migration Report has been an annual series which has received overwhelming support and global recognition among a vast array of readers which include migration policymakers both in the government and private realms, development practitioners, activists and researchers working in the field. Their constant presence has turned this series into one of the most important works of reference in the field, and I  take this opportunity to thank them for all their support. I would also like to thank all the contributors who have worked hard to make this important and unique collection of works based on data collected through migration surveys possible. The series was conceived in 2008 and began in 2010 with the support and guidance of the erstwhile Ministry of Overseas Indian Affairs, which established the Research Unit on International Migration (RUIM) from 2006 to 2016 at the Centre for Development Studies (CDS) where I acted as Chair Professor. I express my deep gratitude to all the secretaries at the MOIA, especially S. Krishna Kumar, K. Mohandas and Dr. A. Didar Singh, without whose help, this series would not have begun and become what it is today. At CDS, I would like to thank the chairman, K.M. Chandrasekhar; director, Sunil Mani; registrar, Suresh Kumar; librarian, V. Sriram; and senior finance officer, S. Suresh, along with my esteemed colleagues, students, administrative and library staff for all their support and encouragement during my academic endeavours. I would also like to especially thank the previous directors, K.N. Nair, Pulapre Balakrishnan and Amit Shovon Ray, as well as chairpersons N.R. Madhava Menon and Bimal Jalan, who over the past ten years have provided great support to make this series a success. I  would also take this opportunity to thank my research team members, S. Sunitha, K.S. Sreeja, Ashwin Kumar, Aneeta Shajan, Sayed Migdad and Anand P. Cherian, whose hard work and enthusiastic support was integral in putting this report together. I am eternally grateful for the emotional support, patience and understanding I  have received from my wife, Hema, and our three children,

xxvi  Acknowledgements Rahul, Rohit and Mary Catherine, without which none of this would have been possible. Last, but not least, I would like to put on record my appreciation for the hard work of the editorial and sales team at Routledge for bringing out this report on time. S. Irudaya Rajan

1 Large-scale migration surveys Replication of the Kerala model of migration surveys to India Migration Survey 2024 S. Irudaya Rajan, K. C. Zachariah and Ashwin Kumar 1 Introduction Migration has become one of the most important and widely discussed topics in the world at the present time. There has been debate about the effects of migration, both in the context of emigration and immigration, and in the academic world, as well as country-level policymaking. Thus, it stands to good reason that communities and countries must keep track of how this phenomenon actually takes place. In order to ascertain the quantum of migration, which in turn gives a glimpse into the various facets that it contains, one has to understand the phenomenon at a closer level – by collecting basic data on exactly how many people are emigrating out of or immigrating into the country. The only way to collect this type of data is through the facilitation of large-scale surveys on migration. Despite calls to include migration modules on existing household survey databases or compiling and centralising existing labour force surveys (Santo Tomas et al., 2009), these efforts have largely fallen short of collecting the sort of broad-based data that encapsulates the phenomenon of migration and its socio-economic effects. This is especially the case when it comes to India. On the other hand, India has also provided examples of successful and long-standing migration surveys. This chapter takes a look at the various surveys that have provided migration trends and patterns, as well as the various facets of it. The chapter begins with an overview of the migration statistics and collection of migration databases around the world. Later, it particularly focuses on the migration surveys conducted in India, and highlights the eight rounds of the Kerala Migration Survey. The twenty-year experience of the various rounds of the Kerala Migration Study (KMS) is touched upon, highlighting its evolution with regard to the type of data it has collected over the years. Its adoption for migration surveys in a few states – namely Goa, Gujarat, Punjab and Tamil Nadu – is further discussed. The chapter ends with the conceptualisation of the India Migration Survey based on the experience of the various Kerala Migration Survey model.

2  S. Irudaya Rajan et al. This chapter, thus, looks to provide a wide look at the various ways data on migration is collected through the various migration surveys, and the importance of these surveys in ascertaining the way forward when it comes to a better method of migration data collection.

2  Migration surveys around the world Migration surveys have been conducted around the world with a wide variety of success. Apart from either destination or origin country-based surveys, there have also been multi-sited surveys, conducted at an inter-country level with the cooperation of various country governments as well as multilateral agencies, which shows the importance of collecting migration data all around the world (Beauchemin, 2014). The following is a look at a few of the large-scale migration surveys that have been conducted in various countries. 2.1 The NIDI-Eurostat Push-Pulls International Migration Project (1997–98) Organized by the  Netherlands Interdisciplinary Demographic Institute  (NIDI) and  Eurostat, the Push-Pulls International Migration Project collected data to study the determinants of international migration from countries of origin to the European Union. The countries of origin were Egypt, Ghana, Morocco, Senegal and Turkey, while the countries of destination were Spain and Italy. Specific immigrant groups were also targeted and non-migrants were also chosen in the sending countries in order to understand the determinants of migration. Therefore, in the sending countries, a total of 1550–1950 households were sampled, while around 500–670 households per immigrant group were selected in the receiving countries (Schoorl et al., 2000). It was the first multi-country project to involve the use of specialized household migration surveys to collect data in both sending and receiving countries. The NIDI-Eurostat survey provided the foundation for a number of inter-country surveys and forms the basis for a lot of the migration baseline statistics available with Eurostat. 2.2 The Latin America Migration Project (LAMP) (1998 and ongoing) The Latin American Migration Project (LAMP) came about as an extension of an earlier migration survey held in Mexico in 1982 called the Mexican Migration Project (MMP) to understand international migration from Mexico to the United States. This led to the extension of the project to the rest of the Central American states, who each have a considerable amount of emigration into the United States, resulting in the beginning of LAMP operations. LAMP involves surveying households in communities in Latin

Kerala model of migration surveys 3 American countries of origin (two to fourteen communities), as well as sometimes conducting in-depth data interviews with migrants in the United States or other destination countries from those same origin households and communities (Massey, 2004; Donato et al., 2010). It takes samples from rural areas with less than 2,500 inhabitants; pueblos or towns, having 2,500 to 10,000 inhabitants; mid-sized cities having 10,000 to 100,000 inhabitants; and, finally, a  metropolitan setting. In the pueblos and small rural areas, investigators conduct a complete census of dwellings and undertake random selection from the resulting list. In all cases, the neighbourhood must have at least 1,200 enumerated dwellings, from which a random sample of 200 is taken to enhance its representativeness.1 Additionally, the survey has been modified to quantify the migration from Paraguay to Argentina. Countries who have participated as of 2018 include Colombia, Costa Rica, Dominican Republic, Ecuador, El Salvador, Haiti, Nicaragua, Paraguay and Peru. The LAMP is currently based at Princeton University and the University of Guadalajara. 2.3 The Migration from Africa to Europe (MAFE) project (2005) The Migration from Africa to Europe (MAFE) project involved implementing surveys in households in three African countries of origin (Senegal, Ghana and Democratic Republic of Congo), and six western European countries of destination (France, UK, Italy, Spain, Netherlands and Belgium). This was a collaborative project among various organizations in the countries involved, from research institutes to civil-society groups. Migration histories were collected – providing more details than in the other projects mentioned – allowing a fuller appreciation of the patterns, causes and consequences of African migration. The MAFE project, therefore, generated a multi-sited, comparative and longitudinal survey (Beauchemin, 2012, 2014). The sample included 1,500 representative samples from each African nation and about 150 migrants per origin from each destination country in Europe. In total, there are 300 Congolese, 300 Ghanaians and 450 Senegalese in the MAFE European sample (MAFE, 2010). 2.4 The World Bank Africa migration project (2008–11) This survey was developed by the World Bank to collect data to study emigration of persons from households in six Sub-Saharan African countries (Senegal, Burkina Faso, Nigeria, South Africa, Kenya and Uganda) and the resulting remittances they received (Plaza and Ratha, 2012). As each country was also interested in internal migration, data were collected on both, with even immigration data collected in the case of South Africa. The questionnaire collected more detailed economic data than the other multicountry projects reviewed here, including household incomes/expenditures as well as remittances. The targeted sample was about 10,000 households

4  S. Irudaya Rajan et al. spread across the five countries, excluding South Africa, where 2012 samples were collected among immigrants (Plaza et al., 2011). 2.5 The Mediterranean Household International Migration Survey (MED-HIMS) project (2013 and ongoing) This MED-HIMS was a product of the European Commission’s MEDSTAT Programme in 2008 and aimed “to provide analysts, policy-makers and other stakeholders with detailed information on the determinants and consequences of migration to help them better target migration policy and development assistance”. Pilot surveys were implemented and analysed for the Palestinian Territory and Morocco. The first full MED-HIMS surveys were carried out in 2013 in Egypt with a total sample of 15,000 households comprised of 6,000 out-migrant households, 6,000 return migrant households and 3,000 non-migrant households; and in 2014 in Jordan with a sample of 7,480 households (Bilsborrow, 2013). It also involves in-depth analysis of the various findings from Egypt and Jordan, and it is anticipated that the next national country surveys will be carried out in Morocco, Tunisia, Algeria and Lebanon in the near future. Therefore, while these examples show that large-scale migration data collection has been attempted, both in the singular contexts of destinations and origins as well as in multi-sited contexts, migration data collection in a vast, socio-economically complex country such as India presents very different challenges.

3  Migration surveys in India: an overview There has been a rich history of data collection on socio-economic indicators in India. However, data on migration has only been tangentially collected in these datasets. Some of the main ones that collected data at a large-scale level and their attempts to collect data on migration in the country are: 3.1 Census The Census of India provides the most extensive database on migration, which is published in its D-Series and captured in the Census of 2011. The Census categorises people as migrants based on two criteria: place of birth and place of last residence. When a person is enumerated in the census at a place different from their place of birth, they are considered a migrant.2 It provides a broad-based picture into the stock and flows of migrants through various socio-economic dimensions at the time of enumeration, which is its main drawback, as it can only provide a picture of the migration scenario once every decade, and tracking migration patterns requires a more dynamic database. Another major limitation is the fact that it only covers data on internal migration and on immigration into India. It does not cover

Kerala model of migration surveys 5 emigration out of India in any depth, which is an important part of the migration story of India. 3.2  National Sample Survey (NSS) The National Sample Survey Organization conducts the largest household sample survey in the country, collecting socio-economic data across a vast cross-section in India. When it came to data specifically on migration, the 49th round of the NSS in 1993 included a section on migration and collected some characteristics of households with migrants. Similarly, the NSS’s 64th-round survey in 2007–08 covered employment and unemployment in India with a special reference to the state of migration in the country. NSS 49th- and 64th-round data were extensively used to calculate the state-wise migration scenario. The major limitation of the NSS rounds is their small sample size, which is not sufficient to estimate a reliable stock of migration and the remittance-sending behaviour of migrants. 3.3  Indian Human Development Survey (IHDS) The IHDS, which was conducted by the University of Maryland and the National Council of Applied Economic Research (NCAER), created a largescale multi-topic panel survey in 2004–05 consisting of 41,554 households across 1503 villages and 971 urban neighbourhoods. Most of the households were re-interviewed in 2011–12, which provided data for the second round of the IHDS. While the scope of the study was to cover topics like economic status, education, employment, marriage, gender relations, fertility, health and social capital, the IHDS data was used to estimate and bring out concerns interrelated with migration, albeit at a limited level. 3.4  Reserve Bank of India (RBI) RBI has been conducting surveys among authorised dealers (ADs) who act as intermediaries for remittances received by residents in India since 2006. Hence, through their surveys among authorised money transfer operators/ dealers operating in India, RBI estimates India’s inward remittances. As can be seen, while some databases have touched upon migration-related issues, they have done so rather tangentially and have provided only certain surface-level glimpses of this very complex phenomenon. Surveys covering the various dimensions and issues of migration in the Indian context were absent till the introduction of the Kerala Migration Survey in 1998.

4  The Kerala Migration Survey Initiated in 1998, the Kerala Migration Survey has been collecting data on the various facets on migration from the southern state of Kerala, a

6  S. Irudaya Rajan et al. historically important source of migrant labour, especially in the post-independence era. Kerala has seen a massive exodus of labour migrants in search of greener pastures, especially to the oil-rich Gulf states whose economy was rapidly expanding after the oil boom. After a couple of generations of this migration, there was still very little understanding of the consequences of this migration and the effect it had on the economy and society at large. Today, Keralites are found all over the world, but especially in the Gulf countries, where they have a sizeable and influential presence. The state is the largest recipient of remittances in the country, with it attracting 19 percent of the total remittances coming into the country.3 Hence, the first wave of the Kerala Migration Survey was conducted by the Centre for Development Studies, and was initially envisaged as a onetime study conducted from March–November 1998. Samples were collected from 10,000 households across 200 localities selected at random from the 14 districts. The KMS was thus the first large-scale migration survey held in post-independence India. The initial survey found that there was a total of 1.4 million Keralites abroad. This initial survey turned into a wave of studies, collected periodically over a period five years, 2003, 2008, 2013 and most recently in 2018; and three special surveys conducted in 2007, 2011 and 2016 at the behest of the Non-Resident Keralites Affairs Department of the Government of Kerala (NORKA). While initially focusing mainly on the issues of emigrants, the subsequent migration surveys dwelt on the issues of return migrants, out-of-state migrants and returning out-of-state migrants. As a result, the sample size was increased from 10,000 to 15,000 households in order to capture issues of all these categories. An abiding feature of the Kerala Migration Survey is that it has always adapted in order to capture the pressing issues of Kerala migration at the time. The KMS has thus been vital in providing necessary information and future trends that been valuable for the appropriate government interventions. A  look at the evolution in samples and methodologies shows the evolution of the understanding of the various issues that came to the fore. Table 1.1 Various rounds of the Kerala Migration Survey, 1998–2008 S. No

Survey

Year

No. of sampled households

1. 2. 3. 4. 5. 6. 7. 8.

Kerala Migration Survey Kerala Migration Survey Kerala Migration Survey Kerala Migration Survey Kerala Migration Survey Kerala Migration Survey Kerala Migration Survey Kerala Migration Survey

1998 2003 2007 2008 2011 2013 2016 2018

10,000 10,000 10,000 15,000 15,000 15,000 13,915 15,000

Kerala model of migration surveys 7 4.1  Kerala Migration Survey 1998 The first in the series of Kerala Migration Surveys was conducted between March and November of 1998 through the financial support of the Indian Council of Social Science Research (ICSSR) via its Indo-Dutch Programme on Alternatives in Development (IDPAD). The survey was conducted over a sample of 10,000 households spread over the 14 districts and 200 localities in Kerala, which were selected at random based on the 1991 Census. Five different schedules were used to collect the data on migrants and the effect of their migration on their households. This survey produced the first credible estimates of the magnitude of migration from Kerala. As mentioned before, this survey estimated that there were a total of 1.36 million migrants residing abroad, as well as an estimated 740,000 return migrants coming back into Kerala (Zachariah et al., 1999, 2000, 2001, 2003). This trend, however, was assumed to be a passing one, and it was thought that at the turn of the twenty-first century, the number of return migrants into the state would outnumber the number of migrants going out. The schedule was the first of its kind for Kerala, and indeed for India – being the first survey dedicated to the migration of a specified population in the Indian context. The Kerala Migration Study, as it was known then, was therefore among the first to define the concepts of emigrants (EMI), return emigrants (REM), out-migrants (OMI) and return out-migrants (ROM), thereby not only measuring the population of Keralites outside the country, but also those populations within India but out of the state (Zachariah and Rajan, 2015b). Further questions were also asked on the impact of migration on the women left behind, especially their autonomy on finances and decision-making within the household in the absence of their husbands. This led to the first study on the effects of migration on “Gulf wives”, something that would be a concern even in later migration surveys. These important categorisations became the foundation of subsequent Kerala Migration Surveys and surveys in other states using the KMS model. There were also special modules in the questionnaire related to the issues of migration and its effects on the elderly population left behind by the migrating generation. These findings were a significant step forward in the establishment of the importance of migration surveys not only in Kerala, but also India. 4.2  Kerala Migration Survey 2003 After conducting interim surveys on the issue of return migrants in India – the Return Migrants Survey, 2001 (Zachariah et al., 2006) – and to examine the employment, wages and working conditions in the Gulf on the request of NORKA (Zachariah et al., 2002), the second Kerala Migration Survey was initiated in 2003. This was part of a larger project called the “Economic Consequences of Emigration in South Asia: Case Studies of Kerala and Sri Lanka”. This project was funded by the South Asian Network of Economic

8  S. Irudaya Rajan et al. Institutes (SANEI), with additional grants received from the Human Development Report Project of the government of Kerala. A similar sample size of 10,000 households – the same as the 1998 survey – was selected at random through the multi-stage random sampling method. While carrying on with the collection of data from the first Kerala Migration Survey based on migration and remittances, the 2003 survey also, for the first time, collected information on the household expenses, the education of children and the costs of utilisation on healthcare expenses (Zachariah and Rajan, 2004, 2005). Another novel approach was the creation of panel data on migration details from 5,000 households and over 125 taluks, with 40 households each, where data from the first survey were collected again to present a trend in migration and its impact on Kerala society (Zachariah and Rajan, 2015b). This has led to the eventual creation of a twenty-year panel dataset, which is the first of its kind in the world and will be discussed in detail later in the chapter. The Kerala Migration Survey of 2003 was, thus, the first to establish, with the help of quantifiable and reliable data, that the emigration of Malayalis to the Gulf was a persistent and sustained corridor of migration. 4.3  Kerala Migration Survey 2007 The Kerala Migration Survey 2007, known as the Migration Monitoring Survey, was conducted through funding from NORKA. A similar sample of 10,000 households, as in the surveys before, was interviewed. In addition to the regular modules, the 2007 Kerala Migration Survey looked into the health status and educational qualifications of the household members as well as their outstanding loans, debts, health expenses and so on (Zachariah and Rajan, 2007). 4.4  Kerala Migration Survey 2008 With the success of the two Kerala Migration Surveys and one subsequent survey in 2007, the Kerala Migration Survey of 2008 was launched with a larger sample size of 15,000 households, with a minimum of 1,000 households over each district (Zachariah and Rajan, 2009, 2010a, 2010b, 2012). The KMS 2008 was held in the wake of two important international events – the surge in oil prices and the global financial crisis of 2008 – which saw a large amount of remittance flowing into India for both precautionary as well as investment purposes (ibid.), which was noted for the first time by the survey. As mentioned earlier, the Kerala Migration Survey of 2008 contained a special module on the effects of migration on the elderly. Questions were asked on living arrangements and family dynamics, economic and financial security, health status and food and nutrition, thereby providing a rich insight into this important issue, especially for an increasingly aging Kerala

Kerala model of migration surveys 9 society. The KMS 2008 also collected information on the effects of migration on left-behind children, making it the largest collection of information on this important, and often under-researched area (Rajan and Nair, 2013). The information thus gave a glimpse into the health, food and nutrition habits which children inculcate, as well as educational attainments of children left behind in the absence of their parents, which will have a major bearing on their later lives. More importantly, information was also collected on the psychological impacts of children who grow up in the absence of their parents, who would have otherwise been their major source of security and confidence. Thus, this survey laid the foundations of the study of transnational families, which will be an important aspect of study in the future. 4.5  Kerala Migration Survey 2011 The next wave of the Kerala Migration Survey was conducted in the same sample size and methodology of the 2008 survey, and was similarly funded by the Department of Non-Resident Keralite Affairs of the government of India and the Ministry of Overseas Indian Affairs of the government of India. The KMS 2011 was more important, as it coincided with the Census of 2011, and thus provided a comparative view of the two datasets, which were found to be similar in the case of Kerala (Rajan and Zachariah, 2017). Therefore, like the 2008  KMS, the KMS 2011 consisted of a sample of 15,000 households spread over the 14 districts and 300 localities (Zachariah and Rajan, 2011a; 2011b). The survey covered a wide variety of subjects and was divided into three types of schedules. As before, the survey covered basic household data and contained information on migration and remittances, as well as identifying return migrants and estimating emigrants and out-migrants. Interestingly, the Migration Survey of 2011 was the first survey to capture information on student migration from the state. It also recaptured information on the subjects of emigration and return migration. Also, a new section looked at attitudes towards emigration. An additional schedule consisted of the evaluation of the Rashtriya Swasthya Bima Yojana (RSBY) and its impact on the health of migrant households. In addition, data on household consumption, savings and investment behaviour was collected among the entire 15,000 households, which was earlier just restricted to 3,000 households. 4.6  Kerala Migration Survey 2013 The Kerala Migration Survey of 2013 was the sixth in the series of Kerala Migration Surveys and mostly followed the same methodology of the previous surveys. Data collection for KMS 2014 was carried out through field surveys between December 2013 and May 2014. The sample consisting of 14,575 households included 63,230 persons, i.e., 29,882 males and 33,341

10  S. Irudaya Rajan et al. females, with the female proportion in the sample being 52.7 percent (Zachariah and Rajan, 2015a, 2016). The survey also included a panel of 5,000 households from the previous surveys, thereby creating a longitudinal database on migration and remittances. The questionnaire consisted of six schedules and twenty-one blocks covering a wide variety of topics. While covering the topics of household and individual migration characteristics as well as remittance and overall consumption and saving behaviour, the schedule was unique in the sense that it included a section on the women left behind. Issues such as remittances and decision-making, as well as the prospects and problems that the women face due to the migration of the men in the house, were brought up. Similarly, it contained questions on the reproductive health of women, something that was not covered in previous surveys and which provided another facet of migration issues. This edition of the Kerala Migration Survey could thus enable one to find attitudes towards women, which is dealt with later in this edition in a separate chapter. Another issue that was covered was the prevalence of alcoholism among Kerala households, which is an important aspect to study given its widespread prevalence in Kerala society. 4.7  Kerala Migration Survey 2016 The seventh round of the Kerala Migration Survey, KMS 2016, was conducted by the Centre for Development Studies and had a similar sample size of 15,000 households over 14 districts in the state. What was unique about the 2016 survey was that it was the first of the Kerala Migration Surveys to be conducted as an entire panel of a previous survey, which was the KMS 2011 (Rajan and Zachariah, 2017; Zachariah and Rajan, 2018). While there were six panels created up to this period (1998–2003–2008– 2013, 2003–2008–2013, 1998–2008–2013, 2008–2013, 2003–2013 and 1998–2013), all of these datasets were only sub-sections of each panel with the same households. The KMS 2016, therefore, gathered information from 13,195 of the 15,000 households surveyed in 2011 (ibid.). 4.8  Kerala Migration Survey 2018 The latest wave of the Kerala Migration Survey was conducted in 2018 with a similar sample of 15,000 households. The localities, however, were increased to 500 from 300, with a more evenly intra-district distribution among the rural and urban strata (Rajan and Zachariah, 2019, 2020). The new topics covered included questions to understand the networks utilised by migrants, followed by the expectations that they have about the process of migration. There were also sections on the emigration of women, as well as a block on the pre-departure orientation of migrants. Like the previous migration surveys, this was also linked to them through a panel of

Kerala model of migration surveys 11 households creating a twenty-year panel database on migration and remittances, which was the first of its kind in the world. The KMS of 2018 also incidentally happened to serve as the baseline for impact assessments of the effects that flooding has on migration, as the catastrophic Kerala floods occurred only a few months after the conclusion of the fieldwork for the survey in the August of 2018. Thus, the KMS 2018 has proven to be a very valuable database in the assessment of natural disasters on lives and livelihoods, and migration as a strategy to ameliorate losses suffered to households and individuals, whether it be through voluntary efforts or through increased flows of remittances (Rajan et al., 2018).

5 An overview of twenty years of the Kerala Migration Survey The Kerala Migration Survey, through its twenty-year existence, has thus collected a wealth of information. Migration trends and patterns can be easily accessed through this period, and it provides a window into the various issues that large-scale migration has brought to the state. Twenty years of the KMS experience has shown an increase in emigration, although with a declining rate of growth – a trend which has been consistent throughout the various inter-survey periods. The 2018 survey found an estimate of about 2.12 million Keralites living out of the state, up from the 1.4 million first seen in 1998. The trend in emigration patterns, however, paints a slightly more complicated picture, as we see this has not always been a steady rate of growth. The inter-survey trend shows an increasing number of emigrants in every period, albeit at a decreasing pace, showing the declining incidence of emigration out of the state. However, the 2013–2018 period saw a reduction in the stock of emigrants out of Kerala by 11.6 percent since 2013. This shows that the phenomenon of return migration has already begun to take place in the context of Kerala, which is borne out by the fact that there was a population of almost 1.3 million return emigrants already in the state. As mentioned before, the vast majority of Kerala migrants migrate to the Gulf countries, from a base of about 93.4 percent of the total migrants based Table 1.2 Number of emigrants from Kerala 1998–2018 Year

No. of emigrants

Percentage increase/decrease

1998 2003 2007 2008 2011 2013 2018

1,361,919 1,838,478 1,848,000 2,193,412 2,280,543 2,400,375 2,121,887

— 25.9 0.5 18.7 3.8 5.0 −11.6

12  S. Irudaya Rajan et al. in the Gulf to about 85.4 percent in 2013, and back up to 89.2 percent in the 2018 survey. This period also saw a small distribution of migrants going to other destinations as well. The survey also found a change in destination preferences, with Saudi Arabia being the most preferred destination in 1998. However, by 2018, that position came to be dominated by the UAE, with it hosting the largest population of Keralites since 2003. The UAE currently hosts about 1.89 million Keralites (Rajan and Zachariah, 2019). The Kerala Migration Survey has also been instrumental in providing reliable estimates of remittances coming into the state. Thus, as we can see, the twenty-year profile of migration out of Kerala shows that emigration numbers have been progressively declining. This might be due to the combination of a number of reasons: an already ageing Malayali population, decline in wages in the Gulf states leading to a loss of allure in migrating there, the rising wages in the domestic economy, the drop in oil prices which has led to a slowdown in the Gulf economies, implementation of increasingly nationalist labour market policies in the Gulf states, as well as a large number of increasingly higher skilled Malayali migrants moving to other destinations in Europe and North America. On the other hand, remittances into the state have been progressively increasing. Coupled with the lessening emigration rates, this indicates an emigrant population Table 1.3 Number of return migrants into Kerala 1998–2018 Year

No. of return emigrants

Percentage increase/decrease

1998 2003 2008 2011 2013 2018

739,245 893,942 1,157,127 1,150,347 1,252,471 1,294,796

  17.3 22.7 −5.9 7.6 3.3

Source: Compiled through the various rounds of the Kerala Migration Survey.

Table 1.4 Total remittance estimates and remittances per household in Kerala 1998–2018 Year

Remittances

Remittances per household (Rs)

(In Rs. Cr.) 1998 2003 2007 2008 2011 2013 2018

13,652 18,465 24,525 43,288 49,695 71,142 85,092

21,469 24,444 32,467 57,215 63,315 86,843 96,185

Kerala model of migration surveys 13 moving up the social ladder in their places of destination, earning higher incomes and thus sending a larger quantum of remittances than in the past (Rajan and Zachariah, 2020). On the other hand, when it comes to out-migration, migrants from Kerala had been mainly moving to neighbouring states like Tamil Nadu, Karnataka and Maharashtra, although they have moved to different places since then. The 2018 survey found significant populations of Malayalis in every state and union territory in the country. More importantly, Kerala has over the years become a hub for migrants from other Indian states. It is estimated that there are about 3 million migrants from states like Bihar, Odisha, West Bengal, Assam and Uttar Pradesh, mainly due to the fact that Kerala offers the highest wages in the informal sector in the country (Rajan and Zachariah, 2019, 2020; Rajan and Sumeetha, 2018). This is a phenomenon that will need further monitoring with the further prevalence of this trend. Apart from creating a data bank on migration stock, KMS also touched on issues associated with migration. The survey looked at the various demographics of migrating populations like the educational and religious compositions of emigrants and, most importantly, the gender composition of migrants and also the various asset holdings of families over a period of time in order to glean the changes that were brought about by migration and remittances. The Kerala Migration Surveys have collected information on the social costs of migration by capturing the issues of women, children and the elderly left behind by husbands, parents and children, which is a very important consequence of migration. It has been repeatedly found that while the left behind fare better than their non-migrating counterparts in terms of material benefits, the additional strains of responsibilities and even feelings of loneliness and depression, and the resultant health issues are major issues to contend with (Rajan, 2015). There is also a greater need for sharing the available data to enhance the undertaking of migration from India. Keeping this in mind, data collected from the various migration surveys are available for open access. This step can be considered one of the significant achievements in the era of non-open access of databases to third parties. The electronic databases are readily available on the CDS webpage4 for anyone interested in utilising it for their research and policy work.

6  Creation of panel data Another feature of the Kerala Migration Survey has been the creation of a twenty-year panel dataset. The longevity and constant periodicity of the Kerala Migration Survey has enabled us to build a panel dataset over various time periods. The various panels aim to display characteristics of a total of 6,282 households over ten panels. This will be the first of its kind in the world, and presents a strong case for the Kerala Migration Survey as a

14  S. Irudaya Rajan et al. Table 1.5  KMS old and new panels 1998–2018 Old panel

No. of households

Panel no.

1998–2003–2008–2013–2018 1998–2008–2013–2018 1998–2013–2018 2003–2008–2013–2018 2003–2013–2018 2008–2013–2018 Total

405 176 297 1,010 634 1,279 3,801

New panel

No. of households

Panel no.

1998–2018 2003–2018 2008–2018 2013–2018 Total households

283 449 503 1,246 2,481

7 8 9 10

1 2 3 4 5 6

Source: Panels generated through the various rounds of the Kerala Migration Survey.

model to be followed. Table 1.5 presents a breakdown of the various panels that the Kerala Migration Survey data has enabled us to create.

7  Implementation of the KMS model in other states After nearly a decade following the first KMS model of migration surveys, states like Goa (2008), Punjab (2010), Gujarat (2011) and Tamil Nadu (2015 and a proposed one in 2020) have also successfully conducted statelevel surveys to estimate the quantum and effects of inter-state, intra-state and international migration on their respective states. 7.1 Goa migration survey (2008) As mentioned earlier, the success of the Kerala Migration Survey saw it initially being extended to the state of Goa. Given its history as a colony of Portugal, Goa has had a long, yet not very well understood, history of emigration before the study. Thus, the Goa Migration Survey was sanctioned in conjunction with the Research Unit on International Migration at the Centre for Development Studies in Thiruvananthapuram, as per the request of the Department of Non-Resident Indian Affairs of the government of Goa. The Goa Migration Survey was thus conducted in a sample of 718 households and found Goan diaspora in at least 43 countries, with the majority of them (56 percent) residing in the Gulf countries. The GMS also found significant Goan diaspora in Europe, South East Asia and North America. Interestingly, it was found that a very significant number of Goan emigrants,

Kerala model of migration surveys 15 comprising about 7 percent of the total emigrant population, were working aboard ships. Another important finding was that a large number among the Goan emigrant population was educated, with about 58  percent of the population having a minimum of a secondary-level qualification, as compared to the 28  percent among the general population. This has a bearing on the receipt of remittances into Goa as well, with Goa receiving about Rs. 700 crores as remittances, which is equivalent to 6  percent of the state’s total GDP. A special survey was also conducted on the left-behind wives of Goan emigrants, as well as the elderly left behind, which gave an insight into an often-overlooked consequence of migration. It was found that while the families of emigrants were more well off than the general population in terms of income and quality of lives, they were socially isolated, suffered from loneliness and were burdened with the additional responsibilities of the household (Rajan and Zachariah, 2011). With a large emigrant population, Goa gave ample insights into the nature and consequences of large-scale migration out of a state. 7.2 Punjab International Migration Survey (2010) Punjab is another state which has had a rich and varied history of international migration. Data, as in the case with most states in India, was virtually absent apart from fragmented state government statistics. The Punjab Internationals Migration Survey (PIMS) in 2010 was the first attempt to collect reliable, large-scale data on the determinants and the socio-effects of international migration from the state. The survey was carried out by the Centre for Research in Rural and Industrial Development (CRRID), Chandigarh with the support of the Centre for Development Studies, Thiruvananthapuram and funded by the Institut National d’Études Démographiques (INED), Paris. The sample population was fixed at 10,000 households (6,524 rural and 3,476 urban), divided over 17 districts among the three traditionally classified regions of Doaba, Majhaand and Malwa (Nanda et  al., 2020). As with the KMS model of surveys, a multi-stage stratified random sampling was used to identify the households. The primary objective of the study was to collect large-scale information on the nature and determinants of international migration from Punjab, investigate the types and uses of remittances in the state, and discover the attitudes and behaviour of the people with respect to international migration. The study also had an overview of the various state and civil society interventions and policies which affected migrants at the destinations and the people left behind in the process. The survey, interestingly, found that 11 percent of households in Punjab reported to have at least one member of the household as an emigrant, with rural households having twice the number of households with emigrants than urban ones. Most of the emigrants went to other Asian states, followed

16  S. Irudaya Rajan et al. by Europe and North America. When it came to remittances, about onetenth of the total sampled households reported to having received remittances, with the Doaba region reporting the highest incidence of remittance receipts. On the other hand, when it came to return migration, interestingly, it was seen that most of the return migrants belonged to younger age categories. Most return migrants cited reasons of tough working conditions, expiry of job contracts and loss of jobs as reasons for return (ibid.). 7.3 Gujarat Migration Survey (2012) The next wave of migration surveys was conducted in the state of Gujarat. This has also been a state which has seen historical migration to various parts of the world, most famously in East Africa and North America. It was for an understanding of this phenomenon that the Gujarat Migration Survey was conducted in 2011. It was sponsored by the Non-Resident Gujarati Foundation, the government of Gujarat and the Ministry of Overseas Indian Affairs, government of India. 10,000 households selected through stratified random sampling and 200 primary sampling units (120 rural and 80 urban) covering the entire state of Gujarat were included in this survey (Bhagat et al., 2017). The study found an emigration rate of 8 per 1,000 individuals in the state, with 27 households out of 1,000 having at least one member out of the state. 7.4 Tamil Nadu Migration Survey (2015) The Tamil Nadu Migration Survey (TMS) was initiated in 2015 in order to get an estimate of the quantum and understanding of the various challenges associated with the state. The state has seen a history of migration tracing back to the migration of indentured labourers in the nineteenth century to plantations in British colonies all over the world. In the post-independence era as well, Tamil Nadu has been a very significant contributor to the emigrant population from India, ranking second only to Kerala in this regard for a number of years. Thus, emigration from the state is an important component of Tamil Nadu society, and it was important to understand the dynamics of it in the same way as in the case of Kerala. The Tamil Nadu Migration Survey went about trying to understand this phenomenon. It was coordinated by the Centre for Development Studies in conjunction with the Loyola Institute of Social Science Research and Training (LISSTAR) and the Centre for Diaspora Studies, MS  University, Tirunelveli. It was commissioned by the Non-Resident Tamils (NRT) Welfare Board under the Commissioner of Rehabilitation, Tamil Nadu, through funding from the Tamil Nadu planning commission (Rajan et al., 2017a, 2017b). The TMS was, thus, conducted over a sample of 20,000 households spread over 32 districts in the state. The TMS estimated a total population 2.2 million Tamil emigrants living all over the world. The top destination

Kerala model of migration surveys 17 overseas for Tamil emigrants was found to be Singapore, with 410,000 Tamil migrants (ibid.). Total remittances into the state was estimated to be Rs. 61,843 crores, which was equivalent to 14  percent of the state GDP, with an average per capita remittance of Rs. 8500 (Rajan et  al., 2017a, 2017b, 2018). The next round of the Tamil Nadu Migration Survey is under discussion with the 2020 survey set to be launched in the near future.

8  Moving towards an India migration survey The experience of twenty years of the Kerala Migration Survey has shown the extent to which migration surveys have the potential to collect a wide range of data on various socio-economic characteristics of a population. Given the importance of migration with regard to its relationship to societies, it also paints a picture of the general well-being of a population. Migration affects the economic and social well-being of individuals, households, entire communities and eventually of nations themselves. The KMS, over the years, has collected information on a number of socio-economic indicators – from the more general, such as trends in demographic change in the society, change in education and literacy levels, and employment/unemployment rates in the society and how that affects society; to more nuanced pictures, such as the issues faced by women, children and the elderly left behind in the migration process, as well as changing family dynamics and even aspirations of a society as a whole. These are interconnected issues and the phenomenon of migration provides a kaleidoscopic lens through which to look, ponder and analyse them. Therefore, with the experience of the KMS and the implementation of the KMS model in different states in hindsight, migration surveys attain great importance, especially in developing economies and societies, where one can notice a high prevalence of migration. It is no surprise that even at an international level, issues related to migration have taken centre stage. The Sustainable Development Goals, announced in 2015, highlighted the importance of safe, orderly and legal migration, which was later emphasised with the signing of the Global Compact on Migration in 2018. India, being signatories to both documents, needs to take into account the prevalence and scale of migration that its citizens are engaged in for various reasons. However, one of the major issues leading to the incomplete understanding of migration trends and patterns in India is the lack of data at various macro and micro levels. It is with keeping this in mind that the recent Draft Emigration Bill of 2019 proposes that migration data be stored in a shared database open to all stakeholders and also to collect data on return emigrants to monitor their resettlement in their various states.5 Capturing reliable data is vital not only for understanding the society as a whole, but also for the making of reasonable, humane and safe policies for migration. However, data capturing larger socio-economic conditions brought about by migration within

18  S. Irudaya Rajan et al. Table 1.6 India Migration Survey 2024: sample frame States/union territories

Population (2011)

Number Number of of districts households (2011) (2011)

Uttar Pradesh 19,98,12,341 71 Maharashtra 11,23,74,333 35 Bihar 10,40,99,452 38 West Bengal 9,12,76,115 19 Madhya 7,26,26,809 50 Pradesh Tamil Nadu 7,21,47,030 32 Rajasthan 6,85,48,437 33 Karnataka 6,10,95,297 30 Gujarat 6,04,39,692 26 Andhra 4,93,86,799 13 Pradesh Odisha 4,19,74,218 30 Telangana 3,51,93,978 10 Kerala 3,34,06,061 14 Jharkhand 3,29,88,134 24 Assam 3,12,05,576 27 Punjab 2,77,43,338 20 Chhattisgarh 2,55,45,198 18 Haryana 2,53,51,462 21 Delhi 1,67,87,941 9 Jammu & 1,25,41,302 22 Kashmir Uttarakhand 1,00,86,292 13 Himachal 68,64,602 12 Pradesh Tripura 36,73,917 4 Meghalaya 29,66,889 7 Manipur 28,55,794 9 Nagaland 19,78,502 11 Goa 14,58,545 2 Arunachal 13,83,727 16 Pradesh Puducherry 12,47,953 4 Mizoram 10,97,206 8 Chandigarh 10,55,450 1 Sikkim 6,10,577 4 Andaman and 3,80,581 3 Nicobar Islands Dadra and 3,43,709 1 Nagar Haveli Daman and 2,43,247 2 Diu Lakshadweep 64,473 1 INDIA 1,21,08,54,977 640

Sample % Sample households household

3,34,48,035 2,44,21,519 1,89,13,565 2,03,80,315 1,50,93,256

71,000 35,000 38,000 19,000 50,000

10.71 5.28 5.73 2.87 7.54

1,85,24,982 1,27,11,146 1,33,57,027 1,22,48,428 1,27,18,976

32,000 33,000 30,000 26,000 13,000

4.83 4.98 4.52 3.92 1.96

96,37,820 83,03,612 78,53,754 62,54,781 64,06,471 55,13,071 56,50,724 48,57,524 34,35,999 21,19,718

30,000 10,000 14,000 24,000 27,000 20,000 18,000 21,000 9,000 22,000

4.52 1.51 2.11 3.62 4.07 3.02 2.71 3.17 1.36 3.32

20,56,975 14,83,280

13,000 12,000

1.96 1.81

8,55,556 5,48,059 5,57,859 3,96,002 3,43,611 2,70,577

5,000 7,000 9,000 11,000 5,000 16,000

0.75 1.06 1.36 1.66 0.75 2.41

3,02,450 2,22,853 2,41,173 1,29,006 94,551

5,000 8,000 5,000 5,000 5,000

0.75 1.21 0.75 0.75 0.75

76,458

5,000

0.75

60,956

5,000

0.75

11,574 5,000 24,95,01,663 6,63,000

0.75 100.00

Kerala model of migration surveys 19 communities in India are either too fragmented or too small to gain larger pictures, or are absent altogether. To collate and gain an overall understanding of migration within and out of India is the need of the hour. The success of the Kerala Model of migration surveys has the made the conception of an India Migration Survey possible. Emigration out of India has attained large proportions over the years and has changed drastically over that time. A case in point being that former states that had contributed majorly to the total emigrant population from India like Kerala and Tamil Nadu have been replaced by Uttar Pradesh, Bihar and West Bengal, states with much larger populations not only in total numbers, but also in working-age populations (Kumar and Rajan, 2015). With the only reliable source of on these trends being the emigration clearance records maintained by the government of India, the extent and significance of this change has not been fully examined, despite its persistence for almost a decade. It is in this regard that the India Migration Survey attains even more significance. A look at the proposed sampling of the India Migration Survey 2024 makes the aims and objectives of the survey clearer. The India Migration Survey 2024 thus proposes to have a sample size of 6,63,000 households spread over the 640 districts in India. The India Migration Survey thus aims to be the foundation for a coherent and broadbased migration policy, which is what India has been lacking so far. Migration still remains one of the main survival strategies for a number of households all over the country. Also, with the extent of internal migration happening all over the country, it is imperative that migration focused surveys be launched all over the country. The twenty-year experience of the Kerala Migration Survey, along with the implementation of the KMS model of surveys across different states, perhaps offers a path ahead for this huge and challenging, but necessary, undertaking.

Notes 1 The Latin American Migration Project collected a variety of samples among the various countries it represented. A better look at the survey methodology can be gathered at: https://lamp.opr.princeton.edu/research/methodologyen.htm 2 Official definition as given by the Census of India: http://censusindia.gov.in/Cen sus_And_You/migrations.aspx 3 As given in the RBI Inward Remittances Survey of 2016–17 4 http://cds.edu/research/ru/migrationresearch/migrationsurveydata/ 5 As given in point 12 (ii) and point 13 (v) of Chapter 2 of the Bill respectively

References Beauchemin, C. (2012). Migrations between Africa and Europe: Rationale for a survey design. Paris, France: Institut National d’Etudes Demographiques. Beauchemin, C. (2014). A manifesto for quantitative multi-sited approaches to international migration. International Migration Review, 48(4), 921–938.

20  S. Irudaya Rajan et al. Bhagat, R. B., Das, K. C., Prasad, R., and Roy, T. K. (2017). International outmigration from Gujarat, India: The magnitude, process and consequences. Migration and Development, 6(3), 448–459. Bilsborrow, R. E. (2013). Sample design for surveys on international migration: MEDHIMS methods and applications in Egypt and Jordan. Paris: ADETEF MEDSTAT III, for European Commission, World Bank, etc. Donato, K. M., Hiskey, J., Durand, J., and Massey, D. S. (2010). Migration in the Americas: Mexico and Latin America in comparative context. The Annals of the American Academy of Political and Social Science, 630, 6–17. Kumar, S. K., and Rajan, S. I. (2015). Emigration in 21st-century India: Governance, legislation, institutions. New Delhi: Routledge. MAFE Project. (2010). European policy briefing: MAFE project policy briefing no. 1. European Research Area. Massey, D. S. (2004). Social and economic aspects of immigration. Annals of the New York Academy of Sciences, 1038(1), 206–212. Nanda, A. K., Veron, J., and Rajan, S. I. (2020). Passages of fortune: Exploring dynamics of international migration from Punjab. New Delhi: Routledge. Plaza, S., Navarrete, M., and Ratha, D. (2011). Migration and remittances household surveys in Sub-Saharan Africa: Methodological aspects and main findings. Washington, DC: The World Bank. Plaza, S., and Ratha, D. (eds.). (2012). Diaspora for development in Africa. Washington, DC: The World Bank. Rajan, S. I. (2015). Migration and development: The Indian experience. In G. P. Freeman, and N. Mirilovic (eds.), Handbook on migration and social policy (pp. 137–161). Northampton, MA: Edward Elgar Publishers. Rajan, S. I., D’Sami, B., and Raj, S. S. A. (2017a). Tamil Nadu migration survey 2015. Working Paper No. 472. Thiruvananthapuram: Centre for Development Studies. Rajan, S. I., D’Sami, B., and Raj, S. S. A. (2017b). Tamil Nadu migration survey 2015. Economic & Political Weekly, 52(21), 85–94. Rajan, S. I., D’Sami, B., and Raj, S. S. A. (2018). International migration in Tamil Nadu: Results from the Tamil Nadu migration survey 2015. In S. I. Rajan (ed.), India migration report 2017 (pp. 249–260). New Delhi: Routledge. Rajan, S. I., and Nair, A. (2013). Children ‘Left Behind’ negotiating parental migration: Preliminary results from a state-wide survey. In S. I. Rajan (ed.), India migration report 2013: Social cost of migration (Chapter  1, pp.  1–16). New Delhi: Routledge. Rajan, S. I., and Sumeetha, M. (eds.). (2019). Handbook of internal migration in India. New Delhi: Sage Publications. Rajan, S. I., and Zachariah, K. C. (2011). Impact of emigration and remittances on Goan economy. In S. I. Rajan (ed.), India migration report 2011 (pp. 295–316). New Delhi: Routledge. Rajan, S. I., and Zachariah, K. C. (2017). Kerala migration survey 2016: New evidences. In S. I. Rajan (ed.), India migration report 2017: Forced migration (Chapter 8, pp. 289–305). New Delhi: Routledge. Rajan, S. I., and Zachariah, K. C. (2019).  Emigration and remittances: New evidences from the Kerala migration survey 2018. Working Paper No. 483. Thiruvananthapuram: Centre for Development Studies. Rajan, S. I., and Zachariah, K. C. (2020). New evidences from the Kerala migration survey, 2018. Economic & Political Weekly, 55(4), 41–49.

Kerala model of migration surveys 21 Santo Tomas, P. A., Summers, L. H., and Clemens, M. (2009). Migrants count: Five steps toward better migration data: Report of the commission on international migration data for development research and policy. Washington, DC: Center for Global Development. Schoorl, J., Heering, L., Esveldt, I., Groenewold, G., Van der Erf, R., Bosch, A., . . . De Bruijn, B. (2000). Push and pull factors of international migration:  A comparative report. Luxembourg: Office for Official Publications of the European Communities. Zachariah, K. C., Kannan, K. P., and Rajan, S. I. (eds.). (2002). Kerala’s Gulf connection: CDS studies on international labour migration from Kerala state in India. Thiruvananthapuram: Centre for Development Studies. Zachariah, K. C., Mathew, E. T., and Rajan, S. I. (1999). Impact of migration on Kerala’s economy and society. Working Paper No. 297. Thiruvananthapuram: Centre for Development Studies. Zachariah, K. C., Mathew, E. T., and Rajan, S. I. (2000). Socio-economic and demographic consequences of migration in Kerala. Working Paper No. 303. Thiruvananthapuram: Centre for Development Studies. Zachariah, K. C., Mathew, E. T., and Rajan, S. I. (2001). Impact of migration on Kerala’s economy and society. International Migration, 39(1), 63–87. Zachariah, K. C., Mathew, E. T., and Rajan, S. I. (2003).  Dynamics of migration in Kerala: Dimensions: Differentials and consequences. New Delhi: Orient Blackswan. Zachariah, K. C., Nair, P. R. G., and Rajan, S. I. (2001). Return emigrants in Kerala: Rehabilitation problems and development potential. Working Paper No. 319. Thiruvananthapuram: Centre for Development Studies. Zachariah, K. C., Nair, P. R. G., and Rajan, S. I. (2006). Return emigrants in Kerala: Welfare, rehabilitation, and development. New Delhi: Manohar Publishers. Zachariah, K. C., Prakash, B. A., and Rajan, S. I. (2002). Gulf migration study: Employment, wages and working conditions of Kerala emigrants in the United Arab Emirates. Working Paper No. 326. Thiruvananthapuram: Centre for Development Studies. Zachariah, K. C., and Rajan, S. I. (2004). Gulf revisited: Economic consequences of emigration from Kerala: Emigration and unemployment. Working Paper No.363. Thiruvananthapuram: Centre for Development Studies. Zachariah, K. C., and Rajan, S. I. (2005). Unemployment in Kerala at the turn of the century: Insights from CDS Gulf migration studies. Working Paper No.374. Thiruvananthapuram: Centre for Development Studies. Zachariah, K. C., and Rajan, S. I. (2007). Migration, remittances and employment: Short-term trends and long-term implications. Working Paper No. 395. Thiruvananthapuram: Centre for Development Studies. Zachariah, K. C., and Rajan, S. I. (2009). Migration and development: The Kerala experience. New Delhi: Daanish Books. Zachariah, K. C., and Rajan, S. I. (2010a). Migration monitoring study, 2008: Emigration and remittances in the context of surge in oil prices. Working Paper No. 424. Thiruvananthapuram: Centre for Development Studies. Zachariah, K. C., and Rajan, S. I. (2010b). Impact of the global recession on migration and remittances in Kerala: New evidences from the return migration study (RMS) 2009. Working Paper No. 432. Thiruvananthapuram: Centre for Development Studies.

22  S. Irudaya Rajan et al. Zachariah, K. C., and Rajan, S. I. (2011a). From Kerala via Kerala via the Gulf: Emigration experiences of return emigrants. Working Paper No. 443. Thiruvananthapuram: Centre for Development Studies. Zachariah, K. C., and Rajan, S. I. (2011b). Inflexion in Kerala’s Gulf connection: Report on the Kerala migration survey 2011. Working Paper No. 450. Thiruvananthapuram: Centre for Development Studies. Zachariah, K. C., and Rajan, S. I. (2012).  Kerala’s Gulf connection, 1998–2011: Economic and social impact of migration. New Delhi: Orient Blackswan. Zachariah, K. C., and Rajan, S. I. (2015a). Dynamics of emigration and remittances in Kerala: Results from the Kerala migration survey 2014. Working Paper No. 463. Thiruvananthapuram: Centre for Development Studies. Zachariah, K. C., and Rajan, S. I. (2015b).  Researching international migration: Lessons from the Kerala experience. New Delhi: Routledge. Zachariah, K. C., and Rajan, S. I. (2016). Kerala migration study 2014. Economic and Political Weekly, 51(6), 66–71. Zachariah, K. C., and Rajan, S. I. (2018). Emigration from Kerala: End of an era. Kochi, India: RedInk an Imprint of Nalanda Books.

2 Keeping up with Kerala’s Joneses Balasubramanyam Pattath

1 Introduction The first Malayali migrants crossed the Arabian sea on a dhow without communication aids and landed on Gulf shores, ignorant of both the beckoning demand for their labour and the emigration boom they would trigger. The fruits of their labour transformed not just their families, but also contributed to building a new Kerala through macrosocial and macroeconomic gains and spill-overs. However, at its core, migration remains an individual and household-level stimulus that elicits a change. While it is true that emigration moved a significant amount of Kerala’s poor out of state-defined poverty, the rising prices of land and labour were a double-edged sword that had adverse effects on non-migrant households with no Gulf connection (Prakash, 1998). After 50 years of sustained migration, it has been established as acredible opportunity for Kerala’s youth, who see in their neighbours, and cascading levels of social organisation, the fruits of migration ‘conspicuously consumed’. This phrase is gaining mileage among researchers in the context of Malayalis’ consumption patterns and will be discussed in the following section (Dominic, 2017). This phenomenon also informs the questions in this chapter, namely: In the presence of consumption patterns that can be deemed conspicuous, what role do perceptions of relative deprivation (if any) play among households as a driver of emigration? How do the multiple social identities of these households affect these perceptions? What is the role of absolute deprivation, both in terms of money and in terms of possessions, as drivers of emigration? To this end, this chapter uses data from the Kerala Migration Survey (KMS) 2011 and 2016 to investigate the implied hypothesis in my questions: that relative deprivation has a role to play in the emigration decision. I employ the relative deprivation variable defined by Stark and Taylor and the Yitzhaki index in the empirical analysis. To identify the influence of visible markers

24  Balasubramanyam Pattath of conspicuous consumption and the possession of status-endowing goods, I use various controls ranging from the value of assets held to an asset index to raw consumption expenditures, in order to tease out relative deprivation in various levels of social identification and consumption preference. Since 89.3% emigrants from Kerala are located in the Gulf Cooperation Countries (GCC) based on KMS 2016 reports, this chapter assumes that ‘emigration’ means ‘international migration to any GCC country’. The rest of this chapter will expand as follows: section 2 reviews the theoretical and empirical literature on Kerala migration, conspicuous consumption and relative deprivation; section 3 presents results of the qualitative analysis and establishes the hypothesis that informs the quantitative analysis. Section 4 quantifies relative deprivation and presents the data; section 5 describes the empirical models; section 6 discusses the results and outlines the robustness checks; and section 7 concludes and addresses the scope by acknowledging the limitations.

2  Context and literature review 2.1 Context: Kerala and international migration Kerala’s society is characterised by paradoxes unprecedented among Indian states: co-existence of poor economic growth with human development indices rivalling advanced economies, stunted industrial growth with high standards of living, low per-capita income with high consumption, and low rates of poverty with high rural and urban income inequalities (M.K, 2016; Zachariah et al., 2003). The positive halves of these contradictions are owed in large part to the 50-year Gulf Migration phenomenon, which burgeoned after the oil price boom of 1973, as a result of which Kerala has been among the largest sources of emigrants for the last three decades. Remittances travelling across the seas are crucial to the money order stability in the state, constituting 36.3% of state net domestic product (Zachariah and Rajan, 2012, 2014, 2015). It is from this unsustainable morass that this chapter teases out a cycle and disentangles the elements that be; namely, the cycle wherein migration and conspicuous consumption among a few leads to relative deprivation among a set of related others, which increases their propensity for emigration and so on. The theoretical migration literature regarding the determinants of migration is rich. Reviews of the drivers of migration, provided by Massey et al. (1993, 2005), Ghatak et  al. (1996), Greenwood (1997), and King and Skeldon (2010) were considered before crystallising the set of insights this chapter benefits from, namely: the role of the household in migration and consumption decisions, the influence of migrant households on their society of origin through return migration, the role of perceptions of relative deprivation and aspiration on migration decisions, and the multiple identities of migrants in their source and destination.

Keeping up with Kerala’s Joneses 25 2.2 Conspicuous consumption Conspicuous consumption was presented in ‘The Theory of The Leisure Class’ by Veblen, who proposed that such consumption is undertaken in the pursuit of status-promotion, after the fulfilment of one’s basic needs, to distinguish the upper class from the lower, undertaken in order to impress others, especially the social groups an individual identifies with (Veblen, 1899; Chen et  al., 2005; Riquelme et  al., 2011). Chaudhuri et  al. (2011) have acknowledged the correlation between the status-seeking behaviour of consumers and their intention to conspicuously consume. The tendency to enhance status and emulate consumption patterns for upward social mobility is a result of susceptibility to interpersonal influence, which has been defined as a construct that fetters an individual to a desire to please significant others. This susceptibility exists among consumers in emerging and less-developed contexts where extended family structures exist (Childers and Rao, 1992). This leads to attempts to bridge structural differences through consumption, which can be seen at the macroeconomic level as well, where demonstration effects push less-developed nations to emulate the consumption patterns of advanced ones (Mai and Tambyah, 2011; Ger and Belk, 1996). 2.3 Relative deprivation and migration Each one began to consider the rest, and to wish to be considered in turn; and thus, a value came to be attached to public esteem. Whoever sang or danced best, whoever was the handsomest, the strongest, the most dexterous, or the most eloquent, came to be of most consideration; and this was the first step towards inequality.

Jean Jacques Rousseau saw the gaze of others as a preeminent driver of the actions of humans, later picked up by Marx, who outlined society’s role as the rating agency for an individual’s satisfaction, postulating its absolute form against its relative cousin (Tischler, 2011). In the context of satisfaction and happiness, Easterlin (1974) found an unequivocal positive relation between individual income and happiness within 19 countries, but found no positive relationship using US time-series data between 1946–70. The fact is that the proportion of people who answered ‘very happy’ remained largely the same in the data, despite the doubling of US GDP in the same span of time. Runciman (1966) defines relative deprivation as the outcome of social comparisons between individual and relative well-being within defined reference groups. In addition to this, studies highlighting the spatiotemporal nature of the frame of reference have been done by Ferrer-i-Carbonell (2005), who used German panel data to demonstrate that relative income is just as important to the individual as absolute income to derive happiness. This is seen even when controlling for family income in the work of Oshio et al. (2010).

26  Balasubramanyam Pattath Mobility of individuals, if unrestricted, can be brought about by economic constraints acting as forces that motivate a ‘flight’ decision. Stark’s identification of individual relative deprivation as a determinant for human migration can even explain migration in the absence of significant spatial income differentials, which are essential for neoclassical migration models (Harris and Todaro, 1970; Sjaastad, 1962). Bhandari (2004) found a significant effect of relative deprivation on the propensity to migrate using the size of land holding as a proxy. 2.4 Keeping up with the Joneses in Kerala Both strands of literature can be married with relative ease to the Kerala context by looking at this from the supply side and the demand side. On the supply side, there exist households that have engaged in conspicuous consumption patterns because of an interplay of influences. First, the nature of family structures in Kerala allows for susceptibilities to interpersonal influence. Because of the migrant’s sense of dual residence, the currency arbitrage and cultural contrasts that migrants confront upon return makes differentiating between necessities and luxuries difficult (Dominic, 2017). In the context of Maslow’s hierarchy of needs, the higher wage differential that motivated the migrant’s emigration satisfies the subsistence needs of the household, advancing the migrant’s motivation towards higher level needs like love and esteem (Maslow, 1943). Allied with the greater propensity to consume from windfall earnings than from money earned through normal work, the Kerala migrant’s devotion to luxury cars, palatial houses and non-routine consumer goods find reason (Arkes et al., 1994). A study of 640 emigrants from six districts in Kerala confirmed previous findings by Batra et al. that there is a link between exposure to global standards and conspicuous consumption, the latter being positively related to materialistic tendencies (Daly and Varghese, 2016). Consumption has become the yardstick to measure socio-economic status in Kerala (Zachariah and Rajan, 2015). Housing and shopping trends that Kerala society pursues are indicative of the consumerist culture existing in the state (Moorthy, 1997; Nair, 1986; Zachariah, 2003). Apart from houses, a common object of emulation and of ostentation was found to be vehicles. 97% of the households surveyed had at least one vehicle, 50% had two personal vehicles and 33% had more than two. With regard to more aspirational products such as cars, 62% of households had one car and 13% had two. This is in sharp contrast to the national average of 13 cars per 1,000 people (Ghate and Sundar, 2014). It is in this societal setting, where migrants constitute about 10% of the state population, return migrants are aplenty, and households with a ‘Gulf Connection’ through a household member or close friend who act as points of reference and influence which Moutinho (1987) defines as reference

Keeping up with Kerala’s Joneses 27 groups, that the pressure to emulate and outdo emanates, thus constituting the demand side of the hypothesis. The numbers resulting in Kerala’s income inequality thus fall under the canopy of Sousa-Poza’s and Liebig’s (2004) work which found empirical evidence for the positive link between higher income distributions and higher migration rates. It would be remiss to exclude Stark (2006), who supplied the behavioural foundation for this structural relationship. In the presence of this felt pressure, the household members decide to migrate not necessarily to increase their expected income, but also to improve their relative position with respect to a specific reference group (Stark and Taylor, 1989). When individuals or households face a mismatch between their current living standards and their desired living standards, between what they get and what they think they deserve, there is discontentment and a call for individual or collective action (Brown, 1988). This mismatch or discrepancy is central to relative deprivation theory, as the influence of visible others is persistent on the aspirational psyche. Additionally, increasing availability of economic resources can not only enhance migration capabilities by loosening economic constraints, but can also increase the overall individual capacity for aspirations (Appadurai, 1988). Considering the positive relationship between aspirations and education levels which Kerala has, the latter being a social indicator which Kerala leads among the Indian states, feelings of dissatisfaction are exacerbated by perceptions of relative deprivation due to the visible consumption on display. In the analyses that follow, I will investigate the initial questions posed to disentangle some of these roles and effects through an initial qualitative analysis followed by a quantitative analysis.

3  Qualitative analysis 3.1 Design In addition to analysing the pre-departure dataset, my exposure to the field benefits this chapter from the doors it opened to semi-structured in-depth interviews. These interviews allowed me to address questions stemming from the hypothesis in this chapter, as well as to touch upon issues that could not be quantified and were not present in any of the previous KMS datasets. This qualitative research, as well as the following data analyses, was conducted during December 2017 to June 2018. I chose my respondents in the following manner: I randomly selected 15 answered pre-departure modules from the district of Thrissur, contacted the emigrant-aspirant who had answered them and requested him/her to introduce me to a friend who also had an interest in finding a job outside India. The result was a pool of 15 individuals who were all the result of one degree of separation, each found through this technique of snowball sampling (Biernacki and Waldorf, 1981). This method was beneficial as it

28  Balasubramanyam Pattath (1) did not come as a favour from a friend or acquaintance, (2) reduced the possibility of biased interviews (Hermanowicz, 2002), and (3) snowballing from different starting points reduced the tendency for strong selection bias (Lopes et al., 1996). Finally, this method of identifying subjects also guaranteed informed consent and confidentiality. As a result, none of the subjects will be named and instead will be called P1–P15. Among the pool, 13/15 interviewees were male while two were female. The range of age was 19–31 with a mean of 24.4 and standard deviation of 3.027. Interestingly, the pre-departure respondents and the subjects they led me to matched 100% in gender. In contrast to ‘focused interviews’ with a predefined problem statement, the method employed was of ‘elite interviews’, where I was willing – and often eager – to let the interviewee teach me what the problem, the question and the situation was (Dexter, 2006). This relaxed the interviewees from the preconceived notions formed around their first interaction with a stranger such as myself and the stimulus that is my own positionality as an interviewer and researcher; the ways in which my values and subjectivity are part of the construction of knowledge. A failure to consider this could lead to misrepresentation of the potential results (Vargas-Silva, 2012). To avoid potential digressions inherent in elite interviews, Lilleker (2003) cautions interviewers of elites to tiptoe carefully around the issue of recording, inquire deferentially about how much time the interviewee can spare and listen attentively. Hence, the interview transcripts that were created are not able to be shared. In the appendix, I instead attach a spreadsheet of sociodemographic and religious characteristics, with y/n questions answered by all 15 respondents (see Table 2A.4). Each interview follows a certain order of thematic questions after eliciting basic characteristics. However, as stated earlier, a dynamic paradigm that ensured better answers were allowed as long as the following themes were covered. 1 Have you already secured a job or are you planning to search for one? 2 What motivates your desire to migrate? Are you happy with your current living situation? 3 Do your family and/or friends motivate your desire to migrate? Which country do you wish to migrate to? What profession do you wish to pursue? 4 Do you wish to return? How far ahead have you planned? Is stepmigration to a richer country and more rewarding job in your plans? 5 Do you plan to remit to your family? 6 How strong are your networks abroad? Who influences you more and do you find their information reliable? Have you done research on your own? 7 Are you worried about any risks in the receiving country? Have you heard stories?

Keeping up with Kerala’s Joneses 29 8 In your estimate, what percentage of people in the age group 21–29 from Kerala wish to migrate? Why do you think so? What factors affect your estimate? Is this good or bad? 9 Does your household possess luxurious consumer items? 10 Have you noticed a difference in households/people you knew before and after their migration? How has that motivated your decision to migrate? Mixed analysis of the pre-departure module and the interviews The following subsection has been constructed in this manner: statistics derived from the pre-departure dataset are contextualised with paraphrased excerpts from the interviews and presented in a stylised way. This triangulation is simply being mindful of alternative sources of knowledge, seeing as how important they are in the field of development (Lewis et al., 2008). Findings i) Of the 529 respondents, 68% had thought about getting a job abroad while 25% had begun the search Only subjects P5 and P14 had found jobs and were preparing their departure after settling affairs at home. P5 had secured a job in the UAE, as an accountant, that paid three times the amount he was earning in Kerala. He wished to return in order to start his own auditing firm with his colleagues who were also considering migration to raise capital for their future venture. P14, however, was moving towards a desk job in a Dubai-based Malayali’s company – one that he had obtained through an acquaintance in the same firm and who had been motivating him with the promise of a job. P1’s student status (bachelor in engineering) has not stopped him from planning for the next 20 years, as he wished to work in the shipping industry in the Gulf and to return after a decade of higher savings in order to pursue the life of a farmer. P6, who is 31, had almost given up the search because of changes in his professional life in the previous three years: a promotion from accountant to manager, a revamp of the company he was once dissatisfied with enough to consider leaving and a series of misfortune in the marriage market because of his aspirant migrant disposition. This opinion was heard from P3 and P10 as potential drawbacks of their aspirations. However, they assumed that they would get the required years and savings before having to enter the marriage market with location-based constraints. Nonetheless, the perception of Gulf migration and its effect on the marriage market requires further inquiry, as there has been an inflexion in the migration trend and a decrease in the absolute number of emigrants (Zachariah and Rajan, 2015). While there might be negative perceptions of

30  Balasubramanyam Pattath emigrants constrained by poor working conditions and lack of proximity, there are some positive connotations associated with the emigrant status. None of the respondents reported that they had contacted agents. Based on the dataset, only 3% had spoken to agents. P8, P10 and P13 explicitly added that they distrusted agents due to the number of scams they had heard about, while P7 claimed that these days, everyone knows someone in the Gulf, thereby crowding out the need to even find an agent. From the dataset, 31% of those who had contacted agents had found out about them by advertisement. Only 6/15 of my subjects recalled seeing a print or television ad in the last five years about migration to the Gulf. Overall, the sense was that migration is just a few phone calls away for those who were slightly more connected than others. An overwhelming 89% were considering Gulf countries, consistent with the KMS 2016 figure of 89.3% emigrants residing in countries in the Gulf Cooperation Council. 57.17% wish to migrate to the UAE, while 13% wish to move to Saudi Arabia. Outside the GCC, 2.95% wish to emigrate to Canada and 1.57% to the USA. Does this imply a strength of networks, a predisposition to the familiar or a certain hierarchy among receiving migrant countries in terms of the skillsets of the migrants? I tried to pose this question to P5 and P6 because of their opinionated stances on the migration phenomenon. As it was a difficult question to pose in the aforementioned form, I asked them why the Gulf was a more popular destination than others, apart from the strength of networks. Now while this was a thinking question, P5 drew the comparison of jobs to countries, where the metric was skillsets possessed. He said that even with the degrees and superior education status that a vast majority of Malayalis possessed, they still entered the global market against superior European nations. He said that akin to the replacement migration of North Indian workers in Kerala, workers from Kerala were mostly adequate to supply the labour needed for the GCC countries’ infrastructure development, whereas they would not be suitable for jobs needing superior skillsets. While this may not be the exact situation, his comparison scratches the surface of what the data says: 10% of the pre-departure respondents were going to be an accountant, cashier, clerk, etc. 12% were going to be engineers and 11% were heading to be salesmen. There was not a single doctor. Nurses were the next category, with 3% of the respondents. That a large proportion of the emigrants from Kerala fall into the skill category required for blue collar jobs provides some credence to this opinion (Heller, 2015). P6 tried to trace the timeline of perceptions towards Gulf migration with an anchor placed on marriage as an eventual event or milestone that it supports. He said that initially you would be ‘hot property’ on the market if you had a job in the Gulf, whereas now it isn’t as valued anymore. He said that people have started considering the difficult aspects of it against the

Keeping up with Kerala’s Joneses 31 alternative of finding a stable job in Kerala. He argued that business is picking up and that the start-up culture allows graduates to discover niche areas of success within established and successful Kerala industries, especially tourism and hospitality, provided you had a knack for it. In the presence of visible and nearby examples of success, families had started becoming resistant to the Gulf-karan suitor for their daughters, he felt. His opinion is based on self-selection bias, but it does account for what the literature says about Gulf wives, the left-behind women of Kerala society who battle loneliness, depression, social stigma and difficult equations with their in-laws because of the confusion in social dynamics brought about by their newfound freedom (Zachariah and Rajan, 2001; Rajan, 2013). They face pressures in the household to manage its activities and act as the bridge between the in-laws and the husband, especially when the constraints faced by the parents are as practical as an inability to make international calls (Zachariah et al., 2001a, 2001b). The Gulf wife also has to act as the household’s figurehead and perpetuate the image and uphold its status in society while at the same time engaging in the market in a way that reflects positively on the household. Indeed, while the consumption pattern is conspicuous, another form of subtler inconspicuous consumption is visible in Kerala society today where female household members, empowered by absentee husbands, participate in reputation-building activities to enhance the status of the family (CurridHalkett, 2017). However, none of the male subjects (with whom the topic came up) wanted to face the situation of creating a Gulf wife. All of them had ambitiously envisioned a few successful years of savings and timed returns to their homes and the marriage market. Being with their family and friends was a strong connection for everyone except P5, who wanted to step-migrate to a better country. While the desire seems to be in place, none of them expected a predicament where the margin of savings wasn’t high enough, keeping the increasing living costs in Kerala in mind, which would eat into their savings by increasing the remittance amounts. This lack of fear could again be a strength of network influence, where both concerns related to fears of migration are allayed by the knowledge of a successful migrant whose lived reality prior to migration was relatable for the aspirant. However, the strength of networks gets an additional boost while considering the following statistic. ii) 87% knew at least one person who had migrated abroad, of which 12% knew 20 or more people working outside India Of the total number of people known, only less than 2% did not know anyone in their preferred country of destination. 80% of the people known in the receiving country were either friends or close family, suggesting that

32  Balasubramanyam Pattath the presence of a bankable contact increases familiarity with that country as well as increases chances of finding a job in that country. iii) 63% of the respondents felt that their networks had had fairly good experience working abroad, while less than 3% felt it was slightly bad This statistic is riddled with conjecture because of various issues. The aspirant could be aloof to the actual pressures faced by the migrant friend. The aspirant could also be convincing themselves to believe differently owing to an inherent disposition to go abroad. This was mentioned by P6 in the context that migrating still holds that charm connected to being an NRK, holding a VISA and being away from home. He said that the media’s influence, and movies especially, have shaped the mindsets related to migration influence. He considered the expected gains from migration to always be worth the costs. P8 and P4 mentioned that there was a certain quality of life related to the freedom they would get in Dubai and mentioned an interest in enjoying their young years. Migrants enjoy a certain status back home, as their return is celebrated and gifts are exchanged. Regarding the intention of going abroad, 60% had consulted their parents, while only 18% had spoken to their friends alone, meaning that migration is by and large a household decision. While around 64% of those who spoke to any family member found their advice to be the most valuable, the levels of appreciation for friends’ advice was around 78%. Both the female respondents, P11 and P12, considered their family’s decision to be secondary in their migration decision. Both of them were required to raise funds to manage their households and arrange for their siblings’ (females) weddings, sacrificing their young ages. The recent movie Take Off (2017) captures the tale of female nurses who migrate to Iraq, despite security issues, because of their need for money. They get trapped in a hostage situation involving a terrorist organisation which was unwilling to negotiate with any government. All the interview respondents were migrating in order to remit amounts to their families at home. 14 respondents wished to build a new house or renovate their existing house with money earned abroad. For some, purchasing new land was a decision to be made in the future. It is interesting to note that several post-materialist activities, such as NGO activities and free health camps in Kerala, are funded by Muslim emigrants who have made a fortune abroad (Osella and Osella, 2009). 12 out of 15 interview respondents answered that they had felt behavioural changes in one or more of their friends/family who had had an emigration experience. In each case, it was related to how ‘posh’ they had become in their language, appearance and consumption habits. Ten out of 15 admitted that seeing these changes made them want to have similar change of fortunes

Keeping up with Kerala’s Joneses 33 of their own, just so that they felt less deprived because they didn’t seize an opportunity that was easy. P10 and P14 mentioned that one of their first gifts following their eventual migration would be an LED TV for their parents. P11 and P12 wanted to purchase a washing machine, induction cooker and microwave oven for their houses, among other goods. iv) All the respondents felt that more than at least 60% of the population in the age group 21–29 wished to migrate (Table 2.D in the appendix) The mean of the rates suggested by each respondent suggest their perception of the general trend, based on their influences, their contacts and their own bias on the matter. With respect to the last sub-question, P6 and P5 felt that it was not good, as it left Kerala’s economy unstable due to a heavy dependence on remittances. P6, especially, argued that there were enough opportunities at home if people were innovative and entrepreneurial enough to take risks in their own backyard. Zachariah and Rajan (2015) write about the irony in people who do not take enough risks at home then emigrating and starting businesses in Gulf countries where more unstable elements can unsettle your prospects. 3.2 Discussion The analysis, with its real-life examples as well as examples from movies, bring to light some of the realities this chapter deals with. Although it is not generalisable by any mathematical extent, the analysis was enriching as an addition to the quantitative analysis, as it allowed me to tweak some of my specifications. Additionally, the case study method implicit in the interview profiles put the real lives of real people right at the centre of explanation and allows the exceptional and the peculiar to shine through, but not eclipse, whatever passes for normality (Bennett and ShurmerSmith, 2002). What I have conducted is an ethnographic approach, in part, and what it offers is a holistic and in-depth glimpse into the lives of a migrant aspirant: what motivates their aspirations, what they see themselves doing, how others’ actions influence them and how they feel about what they want, what they have and what others have. This forms the platform and also informs the quantitative section, which is the next part of the chapter.

4  Quantitative analysis 4.1 Data source The chapter relied the data base of the 2011 and 2016 Kerala Migration Surveys (KMS). The 2011 and 2016 surveys include a household module which recorded socio-demographic factors of the household and its members, and

34  Balasubramanyam Pattath employed additional modules for emigration, out-migration, return emigration and return migration.1 4.2 Quantifying relative deprivation The more an individual is deprived relative to a pre-defined reference group, the higher the propensity to migrate, is how Stark and Yitzhaki (1988) and Stark and Taylor (1989) first outlined the relationship. This relationship is also conditional on the fact that absolute deprivation does not restrict mobility. This makes Kerala fertile ground for this research due to the interconnected networks that facilitate Gulf migration, even at the expense of debts incurred to raise the capital required for emigration. Individuals choose to either physically leave the context that makes them absolutely and relatively deprived, or they may deal with it through adjusting their set of held norms and standards, changing their reference groups, or by mentally disengaging with their sense of deprivation (Carver et al., 1989). In a context where migration is largely a household decision, which may either benefit from familial connections (an existing emigrant) or whether the decision-making regarding departure is decided collectively, the family’s deprivation as a whole (absolute and relative), proxied by a legitimate marker, is important (De Jong, 2000). Hence, I measure the relative deprivation using the household as the unit of analysis. While Stark and Taylor (1989) see migration and deprivation as interchangeable and inversely related, the relationship between relative deprivation and utility is not as simple.2 The practical measure they proposed to measure, relative deprivation, assumes that relative deprivation can be measured by income which has a continuous distribution. In this chapter, I choose to employ two distinct measures of relative deprivation, both measured using consumption expenditure instead of household income. My reasons being: a) 30% of the values in the income variable from KMS 2011 and 43% of the values from KMS 2016 are unreported, thereby constraining a lot of the variation in income as well as the dependent variable corresponding to the households; and b) Household consumer expenditure, being a function of income, is calculated based on summations of more than 20 items, divided into nonfood expenses and food expenses, allowing for more accurate data capture and reporting. This makes the variable more robust. However, there are, in principle, problems of reverse causality for households reporting remittances received from an out-migrant. I use monthly consumption expenditure as opposed to yearly, which spreads the effect of the remittances, except for cases where the remittances feed the bulk of the expenditure.

Keeping up with Kerala’s Joneses

35

The following are the two distinct measures of relative deprivation which are used in this chapter: For the first approach, I draw on the work of Czaika (2012) and focus primarily on the importance of reference groups (Walker and Smith, 2002). People are assumed to compare themselves to various social groups, and then to further compare their social groups to other social groups within the same reference category. Using KMS data, I stratified the households into reference categories based on ‘consumption expenditure quartiles’ and religious group. While the latter is a more explicit social grouping, the former is a reasonable comparison for households to field. The first explanatory variable in this approach is Individual Relative Deprivation (IRD), which measures relative deprivation of households with respect to other households in the same reference group, defined as: c max

IRDi,r (ci ) = ∫ cir ir

[1 −F (z)]dz

In this measure of IRD, ci denotes annual consumption expenditures of household i, with F(z) representing the cumulative distribution of household consumption levels within a social reference group r. This variable was used to measure: IRD with respect to every other household in the state; IRD within consumption expenditure quartiles; and IRD with respect to three religious groups (Hindu, Muslim, Christian). The second explanatory variable in this approach is GRD, or Group Relative Deprivation, which is defined as: GRDir (cr ) =

cr−max cr

[1 −F (z)]dz with cr =

1 nr

∑c i εr

i ,r

In this measure, each household i that belongs to a social group r identifies with the economic well-being of the entire group, in this case religion. The well-being of the group is represented (proxy) by the annual household consumption cr and compares with the average consumption levels of all wealthier social groups within the same societal reference category (Czaika, 2012). However, I only use this measure for religion as a social category and exclude consumption quartiles because households may not identify as strongly with a consumption-based group, let alone the lack of explicit identification in such a classification.3 Further, I  also use a measure of Multiple Relative Deprivation (MRD), which is a principal component of the IRD state and IRD religious group measures to test the hypothesis that households may feel deprived across multiple social groups.4

36

Balasubramanyam Pattath

Thus, the propensity of an individual to migrate is a function of overall aspirations, which are motivated by the deprivation felt individually, as a unit, collectively across various reference groups, as a result of a combination of the two, and fundamentally as a function of the absolute deprivation. The second approach draws on a more popular strand of literature and employs the Yitzhaki index as a proxy for relative deprivation. Two variants of the index are used. The original Yitzhaki index for household i is calculated in the following manner: RDi =

1 N

∑ (y

j

−yi ), ∀y j > yi

The modified Yitzhaki index is the second measure, calculated by:

modified RD RDi i = modified

1 N

∑ (ln(y

i

−lnyi )), ∀y j > yi

Both compare an individual i’s income to that of every other individual j, whose income is higher than i and belongs to the same reference group. In my analysis, I use household consumption expenditure to calculate the indices. The modified Yitzhaki index draws on the suggestions made by Eibner and Evans to use the log of income instead of the nominal income in order to capture relative deprivation as an increasing function of yj and a decreasing function of yi, and concave with respect to yj, such that a transfer of consumption expenditure from a rich man to a poorer man, both having more income than person i, would increase person i’s relative deprivation (Lee, 2012). In addition to the cross-section analysis, I use the same Yitzhaki index as explanatory variable(s) on the 2011–2016 panel. 4.3 Constructing the 2011–2016 panel I construct the panel of KMS 2011–2016 in order to study the characteristics of households prior to migration. This helps reduce some of the potential endogeneity problems seen in cross-sectional studies (Phuong et  al., 2008). The panel contains 13,199 households  – the number of households which were resurveyed in KMS 2016, carried forward from the original 15,000 that were surveyed in KMS 2011. The unique ‘ScheduleNo’ variable was used to append both datasets in a long format, with the addition of a year variable. Unlike the income variable, the consumption expenditure variable did not have missing values, so it was merged without dropping any observations. The KMS 2016 dataset had outliers on the higher tail, so it was winsorised at 0.5% to bring parity to the KMS 2011 consumption variable without compromising the log-normal shape or variance.

Keeping up with Kerala’s Joneses 37 4.4 Reference group substitution Sherif et al. (1955) defined reference groups as those to which the individual relates himself as a part or to which he aspires to relate himself psychologically. However, the concept of a defined reference group is tricky in empirical studies (Podder, 1996). Most work in the literature on relative deprivation defines points of comparison towards individuals with proximate characteristics and similar opportunities (Deaton, 1997; Eibner & Evans, 2005; Ferrer-iCarbonell, 2005; Kuegler, 2009; Stark and Taylor, 1991). Even geographical proximity is assumed as a reference group, as are sociodemographic factors (Deaton, 1997; Eibner and Evans, 2005; Ferrer-i-Carbonell, 2005; Stark and Taylor, 1991). Further studies on KMS data can exploit individual-level data to tease out the effect of these characteristics within and across households. In fact, doing so gains credence from Mangyo and Park’s China Inequality and Distributive Justice Survey project, which says that classmates and relatives are reference groups for urban residents, while neighbours are points of comparison for rural residents (Mangyo and Park, 2011). Reference group substitutions are frequent causes of concerns in migration studies involving utility and deprivation. However, in the Kerala case, the reference group is stable for the emigrants because 89.3% of the emigrants identified by KMS 2016 are located in the Gulf Cooperation Council (GCC) countries where permanent residence is close to impossible (Guardian, 2017). Additionally, the increasing pressures of nationalisation, uncertainties and exposure to global shocks (as was seen in the aftermath of the 2008 global financial crisis) and the influx of migrants from other states of India crowding out new emigration from Kerala (according to KMS 2016 which saw a drop in the absolute number of emigrants), lends credence to a stable reference group (Roche, 2017). Emigrants from Kerala frequently complain about poor living and working conditions in the Gulf, citing their inability to adjust to the completely different social milieu and their desire to return home once they have accumulated the required amount of savings (Rahman, 2009). A large part of remittance money goes into buying a plot of land and/or building/renovating a house, a key reason why a household quality index has been a staple in the KMS for the last few rounds. Both the proclivity to return and the desire to build a new house was confirmed by the majority of the respondents interviewed in this chapter’s qualitative analysis. Feelings of homesickness are also rife among Kerala emigrants (Chakraborty and Mandal, 2016). This is owed in part to their temporary uprooting from the rest of the family, which stays back, and to the high incidence of festivals and community-based activities in the state. 4.5 Identification strategy The KMS 2011 cross-section has a variable named HHLDTYPE (Household Type) which categorises households into EMI, non-migrant, OMI, REM and ROM (see Table 2.1).

38  Balasubramanyam Pattath Table 2.1  Breakup of households based on migration status Household Type

Frequency

Percentage

EMI Non -Migrant OMI REM ROM Total

2,736 9,695 792 1,324 453 15,000

18.24 64.63 5.28 8.83 3.02 100.00

Source: KMS 2011.

Table 2.2 Change of household type by migration status between 2011 and 2016 2016 2011

EMI

REM

OMI

ROM

Non Migrant

Total

EMI Non-Migrant OMI REM ROM Total

64.8 4.8 15.8 18.8 5.0 17.6

17.5 2.0 1.9 35.8 2.5 7.9

1.5 2.0 32.2 1.8 5.5 3.6

0.6 0.7 10.6 0.6 14.9 1.6

15.5 90.5 39.6 43.1 72.0 69.3

100.0 100.0 100.0 100.0 100.0 100.0

Source: KMS 2016.

The KMS 2016 cross-section has a variable denoting whether Schedule 2 was used. This is a supplementary module that is used by field investigators who conduct the survey when they are informed of the presence of an emigrant or an out-migrant.5 Another extension of the analysis in this chapter is the interplay between short- and long-distance migration, and between out-migration and emigration. However, for the time being, I only concern myself with emigrants, as the context so heavily depends on them alone. The variable named out-migrant is used as dependent variable in the empirical models based on the 2016 cross-section. For the panel dataset, the dependent variable is a binary variable that equals one if a household in the 2011 survey with no emigrants had at least one emigrant during the 2016 survey (see Table 2.2). Socio-economic status of the household prior to the migration decision is measured in this setup, although the results are not telling for all reference groups.

5  Empirical models The following empirical analysis is based on the KMS 2016 cross-section as well as the 2011–2016 panel. The key explanatory variable is relative deprivation calculated according to the two approaches outlined in section 3.2.

Keeping up with Kerala’s Joneses

39

Every specification in the data analysis includes all households, irrespective of whether they have at least one emigrant or no emigrants at the time of the survey. For the first approach, which is based solely on the KMS 2016 crosssection, I  compute two measures of IRD (IRD state, IRD religion), one measure of GRD (GRD religion) and one measure of MRD (MRD religion and state). The MRD variable was calculated as a composite index based on eigenvalues generated by principal component analysis. In this regression, emigration of a former household member is the binary dependent variable M, which is set to one if the household has at least one emigrant, and zero otherwise, predicted by the following probit: M = 1 if ( 1RDir + X i

0

+

1

> 0) & 0, otherwise

Xi is a variable that includes absolute deprivation as well as other controls. Since the decision to emigrate is largely influenced by the household and is based on a lot of factors, I use the variables available from the 2016 crosssection to control for several socio-economic and demographic factors. First, household consumption expenditure along with a squared term is included because migration aspirations are assumed to be non-linearly related to the level of absolute deprivation (Taylor, 1996). Second, household size, including the number of out-migrants, are used. This effect may be because a larger household size can spur the need for emigration due to economic intra-household risk diversification strategy (Stark and Levhari, 1982). Third, household assets are included in mixed ways. I use a dummy for land possession over an acre, as it may signify farm holding or a plot for house construction or rent. I include a dummy for value of land, construction and gold possessed. I also include a dummy for the kind of fuel used at home and the quality of house owned. I differentiate the fuel dummy into good fuel and bad fuel, the latter being cheaper and possibly correlated with worse health outcomes. For the good house dummy, there is a 1–5 scale included in the survey instrument which is to be filled in by the field investigator. Since 50% of households received the median ranking, I exclude them and use ratings 4 and 5 for very good and excellent housing to generate the good house dummy. For the consumer household items which are broadly luxurious goods and the centrepiece of what constitutes conspicuous consumption, I  generate asset indices using the eigenvalues of the principal component analysis using World Bank guidelines (Filmer and Scott, 2008). For approach two, which is based on the KMS 2016 cross-section as well as the 2011–2016 panel, I generated four measures of the Yitzhaki index for the former and five measures for the latter. The Yitzhaki indices used as explanatory variables in the KMS 2016 cross-section are based on religion as well as quartiles based on household consumption expenditure. For the panel dataset, I generated Yitzhaki indices for the consumption expenditure

40  Balasubramanyam Pattath quartiles separately (2, 3, 4, 5), while also slicing the data based on consumption expenditure quartiles within the total number of households as a robustness check (6, 7, 8, 9). For the cross-section analysis, I used the same dependent variable as in the earlier approach, with different specifications. In this, I  included the household size, the consumption expenditure, a rural dummy for the administrative classification of the household, as well as a dummy for Rashtriya Swasthya Bima Yojana (RSBY) insurance, which provides free social security schemes and medical insurance to eligible families (India, 2012). RSBY dummy is important in the context of studies which examine the motivation to migrate depending on the availability of insurance vis a vis risk preference of migrants (J. Greenwood, 1997). Specification 2 used the modified Yitzhaki index, while the original Yitzhaki index was used in the others. Clustered standard errors are used. Additionally, an LPM was also run on the same specification using the Yitzhaki index as a robustness check (see Table 2.5). For the panel dataset, the dependent variable is a binary variable that equals one if a household in the 2011 survey with zero emigrants had at least one emigrant during the 2016 survey. Consumption expenditure of the households, household size and the rural dummy were used in each of the specifications which differed by the Yitzhaki explanatory variable. For specifications 2–5, the Yitzhaki index was generated within income quartiles separately, which reduced the number of observations. As a robustness check, the data was sliced into the consumption expenditure quartiles in specifications 6–9, while retaining the Yitzhakis generated for the whole state, to run the model. Clustered standard errors were used while controlling for urban-rural fixed effects. Additionally, the conditional logistic model was used on the same specification with clustered standard errors.

6 Results 6.1 IRD, GRD and MRD Table  2.3 reports the results on the probability of out-migration of the household members. I  find positive and significant effect for the relative deprivation variable across individual and group comparisons for the religious category (specifications 2 and 3). The positive sign of the IRD religion is maintained in (specification 4), but the GRD assumes a negative sign in the same. The effect of relative deprivation when the comparison group is the whole state is found to be negative on emigration, possibly because comparisons within a large group are less meaningful because of the lack of visibility of wealthier entities, preventing direct comparisons, or because lack of knowledge about one’s standing over such a large scale leads to overestimation of one’s position. However, I find a positive and statistically significant effect for MRD in (specification 5), which implies that

Table 2.3 Out-migration of household member  

1

2

3

4

5

Dependent variable

all state

IRD religion

GRD religion

IRD & GRD

MRD

  IRD state   cons_exp   cons_expsq   landpossacre   hindu   muslim   christian   goodhouse   badfuel   pc1   pc2   pc3   buildings   constructions   landvalue   goldvalue   IRD religion   GRD religion   MRD   no. of observations pseudo R-sq

  −0.193*** −0.046 −0.24*** −0.06 0.000*** 0 −0.247*** −0.05 0.154 −0.562 0.91 −0.563 0.266 −0.563 0.225*** −0.043 −0.057 −0.029 0.096*** −0.016 0.024 −0.012 0.108** −0.037 7.55e−08*** −1.16E−08 −0.000* −8.15E−08 −1.47e−08*** −4.29E−09 2.41E−07 −1.97E−07             13,199

      0.026*** −0.003 −0.000*** 0 −0.386*** −0.049             0.493*** −0.028 −0.046 −0.028                

0.119

  religion         0.028*** 0.023*** −0.003 −0.003 −0.000*** −0.000*** 0 0 −0.381*** −0.306*** −0.049 −0.05                         0.498*** 0.497*** −0.027 −0.028 −0.063* −0.06* −0.028 −0.028                                                                 0.182***   0.431*** −0.01   −0.019   0.082*** −0.303***   −0.01 −0.021             13,199 13,199 13,199 0.077

0.057

0.095

Note: Standard errors in parentheses; * p