Labor and Employment Issues [1 ed.] 9781616688837, 9781607412861

Citation preview

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved. Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved. Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

ECONOMIC ISSUES, PROBLEMS AND PERSPECTIVES

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

LABOR AND EMPLOYMENT ISSUES

No part of this digital document may be reproduced, stored in a retrieval system or transmitted in any form or by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in rendering legal, medical or any other professional services.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

ECONOMIC ISSUES, PROBLEMS AND PERSPECTIVES Additional books in this series can be found on Nova’s website at: https://www.novapublishers.com/catalog/index.php?cPath=23_29& seriesp=Economic%20Issues%2C%20Problems%20 and%20Perspectives&sort=2a&page=1

Additional E-books in this series can be found on Nova’s website at:

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

https://www.novapublishers.com/catalog/index.php?cPath=23_29& seriespe=Economic+Issues%2C+Problems+and+Perspectives

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

ECONOMIC ISSUES, PROBLEMS AND PERSPECTIVES

LABOR AND EMPLOYMENT ISSUES

NICKOLAS H. MULLEN

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

EDITOR

Nova

Nova Science Publishers, Inc. New York

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010 by Nova Science Publishers, Inc. All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical photocopying, recording or otherwise without the written permission of the Publisher. For permission to use material from this book please contact us: Telephone 631-231-7269; Fax 631-231-8175 Web Site: http://www.novapublishers.com NOTICE TO THE READER The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained in this book. The Publisher shall not be liable for any special, consequential, or exemplary damages resulting, in whole or in part, from the readers’ use of, or reliance upon, this material. Any parts of this book based on government reports are so indicated and copyright is claimed for those parts to the extent applicable to compilations of such works.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Independent verification should be sought for any data, advice or recommendations contained in this book. In addition, no responsibility is assumed by the publisher for any injury and/or damage to persons or property arising from any methods, products, instructions, ideas or otherwise contained in this publication. This publication is designed to provide accurate and authoritative information with regard to the subject matter covered herein. It is sold with the clear understanding that the Publisher is not engaged in rendering legal or any other professional services. If legal or any other expert assistance is required, the services of a competent person should be sought. FROM A DECLARATION OF PARTICIPANTS JOINTLY ADOPTED BY A COMMITTEE OF THE AMERICAN BAR ASSOCIATION AND A COMMITTEE OF PUBLISHERS. LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA Labor and employment issues / editor, Nickolas H. Mullen. p. cm. Includes index. ISBN  (HERRN) 1.Labor supply--United States. 2. Manpower policy--United States. 3. Unemployment--Government policy--United States. I. Mullen, Nickolas H. HD5706.L217 2010 331.10973--dc22 2010012226

Published by Nova Science Publishers, Inc.  New York

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

CONTENTS Preface Chapter 1

The Labor Market during the Great Depression and the Current Recession Linda Levine

1

Chapter 2

Long-Term Unemployment Congressional Budget Office

Chapter 3

How Slower Growth in the Labor Force Could Affect the Return on Capital Congressional Budget Office

57

Job Loss and Infrastructure Job Creation Spending during the Recession Linda Levine

67

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects of Underrepresenting Unauthorized Residents United States Government Accountability Office

81

Chapter 4

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

vii

Chapter 5

27

Chapter 6

Unemployment and Economic Recovery W. Brian Cashell

105

Chapter 7

Unemployment: Issues and Policies Jane G. Gravelle, Thomas L. Hungerford and Marc Labonte

113

Index

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

131

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved. Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

PREFACE In the short run, the relationship between economic growth and the unemployment rate may be a loose one. One reason that unemployment may not fall appreciably when economic growth first picks up is that some firms may have underutilized labor. Laying off workers when times are bad and rehiring them as conditions improve has costs. In other words, firms may be able to increase their output by raising the productivity of the labor on hand, which yields a temporary boost in measured labor productivity above its longer-run rate of growth. This book examines the relationship between economic growth and the unemployment outlook to anticipate possible future developments. This book consists of public documents which have been located, gathered, combined, reformatted, and enhanced with a subject index, selectively edited and bound to provide easy access. Chapter 1 - A good deal of commentary has addressed similarities between the recession that began in December 2007 and the Great Depression. Comparisons between the two have extended beyond conditions in financial markets to conditions in the labor market. The analogy appears to be fueled by projections that the unemployment rate could reach double digits in the coming months. Little if any comparative labor market research has been undertaken, however. To address the situation, this report analyzes the experiences of workers during the 193 0s, which encompassed the almost five years of the Great Depression. Because it was a period very distant and different from today, the report devotes considerable time to examining the employment and unemployment measures then available. The report ends by comparing the labor market conditions of the 1 930s with those encountered by workers thus far during the nation’s eleventh recession of the post-World War II period. A labor market analysis of the Great Depression finds that many workers were unemployed for much longer than one year. Of those fortunate to have jobs, many experienced cutbacks in hours (i.e., involuntary part-time employment). Men typically were more adversely affected than women. This was especially true for older and black men at a time when age- and race-based job discrimination were not unlawful and when occupational shifts in labor demand were operating against them. Higher-skilled workers fared better than lower-skilled workers. Those who toiled on farms and in factories were displaced in very large numbers. States whose economies were dependent on agriculture and manufacturing reported high unemployment rates. There are several similarities not only between the Great Depression and the recession that began in December 2007, but also between the Great Depression and other recent

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

viii

Nickolas H. Mullen

recessions. They include the greater impact of economic downturns on male blue-collar workers in the goods- producing sector (e.g., construction and manufacturing), lower-skilled workers, and older workers. Chapter 2 - Each year, millions of people become unemployed and find a job within a few weeks, while many others change jobs without any intervening unemployment. The flexibility of the labor market is generally considered a source of strength for the U.S. economy. Some people, however, remain jobless for many months. Those longterm unemployed workers account for a large portion of the total weeks of unemployment. This paper examines the extent to which unemployment is concentrated among workers who are unemployed for more than six months. It also examines the characteristics of those long-term unemployed workers as well as their sources of income and subsequent activities. The analysis is largely based on data from the most recently completed panel available of the Census Bureau’s Survey of Income and Program Participation, which covers people who experienced any unemployment during 2001, 2002, or 2003. Chapter 3 - The labor force in the United States is expected to grow at a significantly reduced rate in coming decades for several reasons: a long-term decline in fertility rates, the leveling off of a substantial rise in women’s labor force participation, and the aging and retirement of large numbers of people from the baby-boom generation. The Congressional Budget Office (CBO) projects that the growth of the labor force, which averaged 1.6 percent per year from 1950 to 2007, will slow to about half a percent per year over the next 20 years. Most mainstream economic models predict that the slowdown is likely to boost the ratio of capital to labor and thereby make the rate of return on capital—as reflected in the rate of interest on bonds or other borrowing, and the rate of return on stocks—lower than might otherwise be expected. Wages will be higher than would otherwise be the case for the same reason, although that effect will be much smaller than the increase in wages that is projected to result from productivity growth. Various economic models and simulations project that, other things being equal, the coming slowdown in labor force growth could subtract between 0.8 and 2.6 percentage points from the rate of return on capital within the next several decades; wages would rise by somewhat less than half the percentage decline in the rate of return on capital. Chapter 4 - After the long economic expansion that characterized much of the current decade, the nation entered its eleventh postwar recession in December 2007. The unemployment rate, which is a lagging economic indicator, did not start to rise until May 2008 when it jumped 0.5 percentage points to 5.5%. By December 2008, the unemployment rate exceeded 7.0% and well over 600,000 jobs were lost—the biggest monthly decrease since December 1974, when another deep recession was taking place. These labor market indicators and comments equating the latest recession to the Great Depression intensified congressional interest in passage of legislation early in 2009 aimed at encouraging creation of new jobs and warding off further loss of jobs. (See CRS Report R40655, The Labor Market During the Great Depression and the Current Recession.) To mitigate all but one recession since the 1960s, Congress chose to increase federal spending on infrastructure. (See CRS Report 92-939, Countercyclical Job Creation Programs.) But, there are a number of issues associated with using expenditures on public works to quickly create jobs in times of recession. (See CRS Report R40107, The Role of Public Works Infrastructure in Economic Stimulus.)

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Preface

ix

Public works expenditures traditionally have gone chiefly to construction activities (e.g., building highways and bridges, dams and flood control structures) which indirectly increase demand in industries that supply their products to construction firms (e.g., manufacturing). Today, the definition of infrastructure has been expanded to include green jobs, which include those in industries that utilize renewable resources (e.g., electricity generated by wind), produce energy- efficient goods and services (e.g., mass transit), and install energyconserving products (e.g., retrofitting buildings with thermal-pane windows). Chapter 5 features a letter written to the U. S. House of Representatives. Chapter 6 - Even though the economy seems to be growing again, it may be a while before the unemployment rate begins to decline, and it may even continue rising for some time after the resumption of sustained economic growth. The unemployment rate is generally a lagging indicator, meaning that its ups and downs happen some time after the ups and downs of other broad indicators of economic activity. Unemployment may not fall appreciably when economic growth first picks up because some firms may have underutilized labor. At the end of a recession as demand increases, some firms may initially be able to increase production without adding workers. Firms may be able to increase their output by raising the productivity of the labor on hand. Over the longer run, there tends to be a link between the rate of economic growth and changes in the unemployment rate. Estimates based on data since 1949 suggest that real economic growth of about 3.5% was associated with a stable unemployment rate. When economic growth was faster than 3.5%, the unemployment rate tended to fall, and when economic growth was below 3.5% the unemployment rate tended to rise. In the long run, a one percentage point difference in the economic growth rate has historically led to a change in the unemployment rate of about 0.4 percentage points. In other words, although economic growth of 3.5% was sufficient to maintain a stable unemployment rate, an annual increase in real output of 4.5% would result in a decline in the unemployment rate of 0.4 percentage points. Even if economic growth picks up and an expansion gets going, experience suggests it may be some time before there are significant declines in the unemployment rate. After the end of the two most recent contractions, it was well over a year before there was a clear downward trend in the unemployment rate. After nine of the past 10 contractions, it took at least eight months for the unemployment rate to fall by one full percentage point. Chapter 7 - The National Bureau of Economic Research (NBER) has declared the U.S. economy to be in recession since December 2007. The unemployment rate in December 2007 was 4.9%; by October 2009, the unemployment rate was above 10%. Although economic output began to grow in the third quarter of 2009, many economists expect that the labor market will remain weak into 2010. In response to high unemployment, some Members of Congress have proposed a job creation bill. This follows several policy steps taken since the economy entered the recession, including stimulus bills in 2008 (P.L. 110-185) and 2009 (P.L. 111-5), an unprecedented expansion in direct assistance to the financial sector by the Federal Reserve, and the creation of the Troubled Asset Relief Program (P.L. 110-343). President Obama, in a speech on December 8, 2009, proposed an additional stimulus package, which would include tax and other benefits for small business, infrastructure investments, incentives to promote energy efficiency, an extension of benefits for the unemployed, aid to state and local governments, and emergency assistance. The Jobs for Main Street Act of 2009 (H.R. 2847) passed the House on December 16, 2009, and included

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

x

Nickolas H. Mullen

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

an extension in unemployment insurance benefits and Consolidated Omnibus Budget Reconciliation Act (COBRA) health benefits, aid to troubled U.S. states and small businesses, and an increase in infrastructure spending. In addition, some policy analysts have proposed a small business hiring subsidy modeled on the 1977-1978 New Jobs Tax Credit. Most of the proposals discussed as part of a potential additional macroeconomic jobs bill are traditional fiscal stimulus policies. That is, their objective is to increase total spending (aggregate demand) either through direct spending on programs or by providing funds to others that they will spend (through tax cuts, transfer payments, and aid to state and local governments). Fiscal stimulus is only effective when the policy options actually increase aggregate demand.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

In: Labor and Employment Issues Editor: Nickolas H. Mullen

ISBN: 978-1-60741-286-1 © 2010 Nova Science Publishers, Inc.

Chapter 1

THE LABOR MARKET DURING THE GREAT DEPRESSION AND THE CURRENT RECESSION



Linda Levine

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

SUMMARY A good deal of commentary has addressed similarities between the recession that began in December 2007 and the Great Depression. Comparisons between the two have extended beyond conditions in financial markets to conditions in the labor market. The analogy appears to be fueled by projections that the unemployment rate could reach double digits in the coming months. Little if any comparative labor market research has been undertaken, however. To address the situation, this report analyzes the experiences of workers during the 193 0s, which encompassed the almost five years of the Great Depression. Because it was a period very distant and different from today, the report devotes considerable time to examining the employment and unemployment measures then available. The report ends by comparing the labor market conditions of the 1 930s with those encountered by workers thus far during the nation’s eleventh recession of the post-World War II period. A labor market analysis of the Great Depression finds that many workers were unemployed for much longer than one year. Of those fortunate to have jobs, many experienced cutbacks in hours (i.e., involuntary part-time employment). Men typically were more adversely affected than women. This was especially true for older and black men at a time when age- and race-based job discrimination were not unlawful and when occupational shifts in labor demand were operating against them. Higher-skilled workers fared better than lower-skilled workers. Those who toiled on farms and in factories were displaced in very large numbers. States whose economies were dependent on agriculture and manufacturing reported high unemployment rates.



This is an edited, reformatted and augmented version of a CRS Report for Congress publication dated June 2009.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

2

Linda Levine

There are several similarities not only between the Great Depression and the recession that began in December 2007, but also between the Great Depression and other recent recessions. They include the greater impact of economic downturns on male blue-collar workers in the goods- producing sector (e.g., construction and manufacturing), lower-skilled workers, and older workers. But, there remain substantial differences between the Great Depression and the current recession: 

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.



In 1933, at the depth of the Depression, one in four workers was unemployed. In contrast, the unemployment rate had risen to 9.4% by May 2009. The number of jobs on nonfarm payrolls fell 24.3% between 1929 and 1933. Thus far during the current recession, firms have cut nonfarm employment by 4.3%. The first 17 months of the ongoing recession compare favorably with the first two years of the Depression as well. In addition to the greater magnitude of unemployment and job loss during the early 1930s as compared with today, the implications of being unemployed have changed much in the intervening years. One reason for the altered situation facing today’s unemployed is the increased prevalence of families in which both spouses work. Another is the deeper drop in earnings and hours worked that occurred during the Depression. And, the social safety net that is now available to displaced workers and their families did not exist before the onset of the Great Depression.

References often have been made to similarities between the Great Depression and the recession that began in December 2007. Both were preceded by long periods of economic expansion during which consumers financed many of their purchases by taking on considerable debt. The substantial losses in savings that resulted from the bank failures of the 1930s made it all the more difficult for individual and commercial depositors to continue spending and investing to the degree they had previously. Today, credit also has become difficult for individuals and businesses to obtain—not because of bank failures, given the availability of federal deposit insurance, but rather because risky investments have led to deteriorated balance sheets at some banks and other financial institutions that have impaired their ability to extend credit. 1 Comparisons between the Great Depression and the eleventh recession of the post-World War II period have extended beyond conditions in the financial market to conditions in the labor market. Speculation that the unemployment rate could reach double digits in the coming months appears to have fueled the analogy. Little if any comparative research has been undertaken, however. This report analyzes the labor market experiences of workers during the 193 0s, which encompassed the almost five years of the Great Depression. Because it was a period very distant and different from today, considerable time is devoted to examining the employment and unemployment measures available at that time. The report ends by comparing the labor market conditions of the 1 930s with those encountered by workers thus far during the recession that began in December 2007.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

The Labor Market during the Great Depression and the Current Recession

3

THE LABOR MARKET DURING THE GREAT DEPRESSION Analysis of labor market conditions during the Great Depression is complicated by the fact that ―throughout the worst years of the Depression, no one knew how many unemployed persons there were, much less their characteristics,‖ because not until March 1940 did the federal government initiate a monthly survey of the labor force defined much as it is today. 2 In 1937, the 75th Congress passed Pub. Law No. 409, which required the President to conduct a census of unemployment.3 At least one of the 14-question Unemployment Report Cards was delivered by postal carriers to each dwelling in the United States and additional cards were made available at local post offices to the ―employable unemployed.‖ But, the 1940 census of the population was the first statistical undertaking to include questions on the labor force defined as persons who are employed or without jobs but actively seeking work within a prescribed period of time. Before then, the 1930 census of the population, the 1937 census of unemployment, and the occasional survey conducted in various states and cities utilized a very different concept—the ―gainful worker‖—that is, individuals who had at some time worked in an occupation in which they earned money or the equivalent, or in which they assisted in the production of marketable goods. Different occupational classification systems also were utilized in the 1930 and 1940 population censuses. For these reasons, this section of the report relies greatly on the U.S. Census Bureau’s adjustment of selected data from the 1930 census to make it as consistent as possible with the 1940 census.4 The Appendix discusses the evolution of labor force data over time. It is intended to supplement information in the detailed table notes in the body of the report about the specific Depression-era data presented in the tables.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Employment What commonly is referred to as the Great Depression comprised two downturns in the business cycle. The first recession began in August 1929, and lasted 43 months until March 1933. A sluggish comeback of the economy occurred during the next four years, before the business cycle peaked in May 1937. The subsequent 13-month decline in the economy’s performance ended in June 1938, but employment did not fully recover until the United States entered World War II.

Demographic and Occupational Characteristics The percentage of the working-age population employed fell substantially between 1930 and 1940, as shown in Table 1. Virtually all of the decrease occurred among men, with the proportion of the male population with jobs dropping 10 percentage points to 67.5%. Men continued to dominate the ranks of the employed during the 1930s despite the Great Depression’s very different impact by gender. The number of men employed fell by 898,000 over the decade, while the number of women employed rose by almost 1.3 million. Women’s share of employed persons consequently grew to one-fourth by 1940. Employers’ hiring decisions partly account for the differential employment effect of the Great Depression by age group. The substantial decrease in employment among teenagers (14- to 19- year-olds) likely was associated with the last-hired/first-fired approach that

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

4

Linda Levine

employers traditionally observe. Additionally, with passage of the Fair Labor Standards Act (FLSA) in 1938, the age limit for employment in manufacturing industries was raised to 16 years, which effectively reduced the number of job opportunities for young persons. This, in turn, might have prompted some teenagers to refrain from entering the workforce and instead, remain voluntarily in school after reaching age 14. School attendance requirements were increased over the decade in some geographic areas as well.5 The order in which employers usually conduct layoffs varies inversely with seniority, which would be expected to benefit older workers. But, length of service is of no help to older workers if entire plants are closed and all employees are let go. Once laid off, older workers found it particularly difficult to find new jobs. Because age discrimination was not unlawful at the time, employers could refuse with impunity to hire older workers if they generalized that younger workers were more adaptable and possessed more up-to-date skills, for example. ―During the 1930s many firms adopted a policy of not hiring anyone over some stated maximum age, the limit being placed sometimes as low as 45 years or even lower.‖6 Table 1. Employment by Gender and Age, 1930 and 1940 (numbers in thousands)

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Gender and Age

Number Employed

Employment as a Percent of Population 1930 1940 1930 1940 Total 44,953 45,338 50.5% 44.8% Men 34,997 34,099 77.6 67.5 14-19 years 2,575 1,752 36.9 23.7 20-24 years 4,409 3,961 82.5 69.6 25-44 years 16,652 16,456 91.2 83.6 45-64 years 9,645 10,256 86.2 76.7 65 years and over 1,716 1,674 51.5 38.0 Women 9,956 11,239 22.6 22.2 14-19 years 1,445 961 20.7 13.1 20-24 years 2,222 2,263 40.1 38.4 25-44 years 4,264 5,515 23.8 27.6 45-64 years 1,786 2,243 17.4 17.6 65 years and over 239 257 7.2 5.6 Source: U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: Estimates of Labor Force, Employment and Unemployment in the United States, 1940 and 1930, Washington, DC, 1944. Notes: In the 1930 census, individuals were asked to report whether they were at work yesterday (i.e., the day before the enumerator called) in which case they were counted as employed. If they responded otherwise, they were asked into which of seven classifications of unemployment they fell. Based on newly introduced definitions of employment and unemployment, the 1940 census would have considered some individuals in the seven classifications to be employed. The Census Bureau adjusted estimates in the 1930 and 1940 decennial censuses for this and other differences to make the data as consistent as possible. For further explanation of the adjusted data that appear in the table see the Appendix.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

5

The Labor Market during the Great Depression and the Current Recession

Table 2. Occupational Distribution of Gainful Workers in 1930 and of the Experienced Labor Force in 1940, by Gender (number of persons age 14 and older in thousands)

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Occupation and Gender Total Professionals Proprietors, managers, officials Farmers (owners and tenants) Clerical, sales, related workers Skilled workers and supervisors Semiskilled workers Unskilled workers Nonfarm laborers Farm laborers Servants Men Professionals Proprietors, managers, officials Farmers (owners and tenants) Clerical, sales, related workers Skilled workers and supervisors Semiskilled workers Unskilled workers Nonfarm laborers Farm laborers Servants Women Professionals Proprietors, managers, officials Farmers (owners and tenants) Clerical, sales, related workers Skilled workers and supervisors Semiskilled workers Unskilled workers Nonfarm laborers Farm laborers Servants

1930 48,595 2,946 9,665 6,012 7,936 6,283 7,973 13,792 6,273 4,187 3,332 37,916 1,498 9,160 5,749 4,865 6,202 5,444 10,747 6,116 3,607 1,024 10,679 1,448 506 263 3,072 81 2,528 3,045 156

Number in the Experienced Labor Forceb 1940 52,020 3,382 9,234 5,275 8,924 6,105 10,918 13,457 5,566 3,708 4,182 39,446 1,847 8,702 5,121 5,269 6,001 7,336 10,292 5,437 3,362 1,492 12,574 1,535 532 154 3,655 104 3,582 3,165 130

1930 100.0% 6.1 19.9 12.4 16.3 12.9 16.4 28.4 12.9 8.6 6.9 100.0% 4.0 24.2 15.2 12.8 16.4 14.4 28.3 16.1 9.5 2.7 1 00.0% 13.6 4.7 2.5 28.8 0.8 23.7 28.5 1.5

1940 100.0% 6.5 17.8 10.0 17.2 11.7 21.0 25.9 10.7 7.1 8.0 100.0% 4.7 22.1 13.0 13.4 15.1 18.6 26.1 13.8 8.5 3.8 1 00.0% 12.2 4.2 1.2 29.1 0.8 28.5 25.2 1.0

580 2,308

346 2,690

5.4 21.6

2.7 21.4

Number of Gainful Workersa

Percent Distribution

Source: U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: Comparative Occupation Statistics for the United States, 1870 to 1940, Washington, D.C., 1943. a. Gainful workers are employed and unemployed persons who reported having a gainful occupation, i.e., an occupation in which they earned money or a money equivalent, or in which they assisted in the production of marketable goods, regardless of their activity at the time the 1930 census was conducted. Although in 1930 and earlier censuses gainful workers between 10 and 13 years old were counted, the Census Bureau excluded them from the data presented in the above table because only persons age 14 and older were included as members of the labor force in the 1940 census. The Bureau made other adjustments to the 1930 count of gainful workers to make the data comparable with the labor force concept introduced in the 1940 census. b. The size of the experienced labor force is based on a complete count of persons age 14 and older employed for pay or profit (or at unpaid family work) by occupation at the time the 1940 decennial census was taken and a 5% cross-section sample count of the usual occupations of experienced workers seeking jobs and of persons on public emergency work at the time the 1940 census was taken.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

6

Linda Levine

In part because mechanization also could substitute for physical strength, the demand for unskilled workers fell as well. This, too, adversely affected men to a greater extent than women because men more often were unskilled farm and nonfarm (construction and factory) laborers whose jobs could be replaced by machinery. (See Table 2.) Black men in particular were susceptible to the shift in demand away from unskilled jobs because the occupations of farm and nonfarm laborers accounted for 43% of employed black men compared to 17% of employed white men.9 In contrast, unskilled women primarily were servants whose jobs could not as readily be supplanted by machinery. The varying impact of technological developments by occupation also helps to explain the differing pattern of job loss across gender and age groups. Increased mechanization and the advent of the assembly line permitted substitution of semiskilled workers for skilled workers, which operated to the advantage of women and younger men compared to older men.7 As shown in Table 2, the proportion of workers in skilled occupations fell from 12.9% to 11.7% between 1930 and 1940, with the entire decrease occurring among men. Older men in particular were displaced by mechanization because skilled workers more often were age 45 and older.8 In contrast, few women worked in skilled occupations (0.8% in 1930) while many more worked in the growing semiskilled occupations (23.7% in 1930).

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Table 3. Employees on Nonfarm Payrolls by Major Industry Group in Peak and Trough Years of the Great Depression (numbers in thousands) Industry 1929 1933 1937 1938 All nonfarm industries 31,324 23,699 31,011 29,194 Goods- producing sector 13,301 8,965 12,936 11,401 Mining 1,087 744 1,015 891 Construction 1,512 824 1,127 1,070 Manufacturing 10,702 7,397 10,794 9,440 Service- providing sector 18,023 14,734 18,075 17,793 Transportation and public utilities 3,916 2,672 3,134 2,863 Wholesale and retail trade 6,123 4,755 6,265 6,179 Finance, insurance, and real estate n.a. n.a. n.a. 1,447 Services n.a. n.a. n.a. 3,502 Government (excludes public emergency 3,065 3,166 3,756 3,883 workers) Source: U.S. Bureau of Labor Statistics, Employment, Hours, and Earnings, United States, 1909-90, Bulletin 2370, vol. I and II, Washington, D.C., 1991. Notes: The Current Employment Statistics (CES) program queries nonfarm employers about the number of jobs on their payrolls, and therefore, excludes self-employed persons and unpaid family workers who are counted in censuses and surveys of the population. However, employer surveys are considered a more accurate count of employment than can be gleaned from individuals through population censuses and surveys. n.a.=not available.

Women also fared better than men during the Depression due to the increased demand for workers in white-collar occupations (e.g., professional and clerical workers). Growth in white- collar occupations was especially pronounced among clerical and related workers, which, as seen in Table 2, was the single largest employment category for women. In

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

The Labor Market during the Great Depression and the Current Recession

7

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

contrast, men’s concentration in the farm operator category meant that they suffered most directly the consequences of the shift from an agrarian to an industrial economy. Black men were especially vulnerable: farm owners and tenants accounted for 21% of black men’s employment in 1940 compared to 14% of white men’s employment.10

Industry Characteristics No industry except government was immune from the Great Depression. According to the U.S. Bureau of Labor Statistics’ annual count of jobs on the payrolls of nonfarm employers, employment dropped by 7.6 million or 24.3% between 1929 and 1933.11 Almost three out of every five jobs lost were from the goods-producing sector, where employment fell by one-third from 13.3 million to less than 9 million. (See Table 3.) Within the goodsproducing sector, construction and manufacturing workers were the hardest hit: by 1933, there were 45.5% fewer construction jobs and 30.9% fewer manufacturing jobs than in 1929. In absolute terms, manufacturing accounted for over three-fourths of all job losses in the sector. Manufacturers of durable goods bore much of the decrease, with motor vehicle and equipment firms providing 45.5% fewer factory jobs in 1933 than in 1929 for example.12 During the 1937-1938 downturn, job losses were even more concentrated in the goodsproducing sector. Over three-fourths of the almost 2 million jobs lost in that year were in mining, construction and manufacturing industries. Manufacturing accounted for almost nine out of ten (1.4 out of 1.5 million) job losses in the sector. The milder second downturn produced relatively fewer job cutbacks than the first downturn at manufacturers (12.5%) and construction firms (5.1%). Companies in the service-producing sector also felt the Depression’s influence as laid off workers became unable to afford their former level of purchases at such establishments as grocers and clothing stores. Employment in the service-producing sector declined by 3.3 million jobs or 18.2% during the 1929-1933 downturn. Job losses in the sector numbered 282,000 or 1.6% during the second downturn of the decade. Agricultural employment fell as well during the 1930s. Droughts and dust storms led farm families to migrate from such states as Oklahoma and Arkansas, and the introduction of technological innovations (e.g., tractors, combines, mechanical cotton and corn pickers) displaced sharecropper families.13 The number of persons working in agriculture was estimated to have been about 10.5 million in 1929.14 Farm employment decreased steadily to about 10.0 million by 1933. It remained largely unchanged through 1937, after which the decline in farm employment resumed. Part-time Employment Of the 2.5 million workers on the payrolls of the 6,551 firms that responded to a federal government questionnaire sent to 25,000 firms in March 1932, 56% were employed part-time. Part-timers were even more prevalent in manufacturing industries in which they represented 63% of all workers at the companies that responded. The high incidence of part-time employment was involuntary; that is, people worked fewer hours than they would have preferred due to the low operating rates at the surveyed firms. Only 26% of the responding firms were operating on full- time schedules and only 28% were open five or more days a week. 15

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

8

Linda Levine

Part-time employment also was a deliberate strategy pursued under both the Hoover and Roosevelt Administrations. ―Work-sharing‖ was intended to spread across employees the scarcer work hours that resulted from greatly reduced demand for goods and services. But, [w]here President Hoover had tried to prevent the loss of some jobs by persuading industry leaders to cut hours, President Roosevelt tried, with some success, to reemploy many of those who had lost jobs by cutting hours still further and establishing minimum wages.16

Before the National Industrial Recovery Act was declared unconstitutional in 1935, ―business adopted voluntary codes, including minimum wages and maximum hours. These foreshadowed the Fair Labor Standards Act of 1938,‖ which ―contained a work sharing measure in the form of an overtime penalty for weekly hours more than 40.‖17

Unemployment The unemployment rate rose from 3.2% in 1929 to 24.9% in 1933 during the Great Depression’s more severe first downturn. While almost 1.6 million persons were unemployed in 1929, more than 12.8 million individuals lacked jobs in 1933. The unemployment rate declined during the brief recovery and measured 14.3% in 1937, when the second downturn began. Although the Great Depression came to an end in June 1938, the unemployment rate averaged 19.0% for the year. The number of unemployed persons jumped by almost 3 million (from 7.7 million to 10.4 million) between 1937 and 1938.18

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Demographic Characteristics Gender In relative terms, women experienced a greater increase in unemployment than men (3 04% and 208%, respectively) between 1930 and 1940. (See Table 4.) Contributing to the elevated joblessness of women was their entering the labor force at an accelerated pace during such poor economic times, which they did in part to compensate for the unemployment of their spouses. 19 Although the percentage of women unemployed rose during the 1930s, their unemployment rate remained below that of men at 13.6% and 15.4%, respectively, in 1940. The Census Bureau suggested that this might be explained by women being more likely to leave the labor force upon losing their jobs while men were more likely to search for another job and therefore continue to be counted as members of the labor force.20 Women’s lower unemployment rate also might have been associated with ―less unemployment in the clerical and service occupations in which women predominate, and because women tend to secure jobs formerly held by men.‖21 Age The increase in unemployment was greatest among young workers. As shown in Table 4, the number of unemployed 14 to 24 year olds rose by 251% between 1930 and 1940. Since the younger members of this group often are leaving school and looking for their initial jobs, one would expect them to encounter more unemployment than other labor force participants.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

The Labor Market during the Great Depression and the Current Recession

9

Additionally, their lack of experience could make it difficult to find another job once laid off. As previously mentioned, the then recently enacted Fair Labor Standards Act (which prohibited 14- and 15-year-olds from working for manufacturers) effectively limited the job options of the very youngest workers as well. Among workers age 25 and above, persons between 45 and 64 years old suffered the largest increase in unemployment. Although the Census Bureau did not calculate separate figures for workers age 55 to 64, it speculated that such data would reveal a greater proportion of unemployment in the older subset than among those age 45 to 54.22 This speculation was borne out by two special state censuses of unemployment. In Michigan in 1935, 16.8% of employable persons age 45 to 49 and 19.5% of those age 50 to 54 were unemployed in contrast with 23.0% of employables age 55 to 59 and 27.3% of those age 60 to 64.23 (Employables were defined as persons either employed or looking for work.) Similarly, in Massachusetts in 1934, between 22.0% and 23.3% of employables age 45 to 54 were jobless compared to between 26.0% and 30.0% of those age 55 to 64.24 The Census Bureau noted that the comparatively low unemployment rates of workers age 65 and older cannot be interpreted to mean that employment opportunities were most favorable for aged persons. Workers who lose their jobs at ages over 65 tend to retire from the labor force rather than undertake new enterprises or to search for jobs in competition with younger workers.25

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Race Although data on minority workers was very limited, it appears that black workers made up a disproportionate share of unemployed persons. In January 1934, for example, blacks comprised just 1.2% of all employables in Massachusetts, but 33.2% of employable blacks did not have jobs. In contrast, many fewer employable whites were unemployed (19.0%) despite their accounting for virtually all employables in the state (98.8%).26 Similarly, based on a 19-city survey conducted in 1931, black women were a larger proportion of all unemployed women in each city than they were of all gainful women workers in each city. The converse was true among white women.27 The rate of increase in unemployment among blacks also was greater than among whites. Between February 1929 and February 1930, for example, there was an 8% rise in joblessness among usually gainfully employed black men in Baltimore compared to a 2% rise among white men. Similarly, unemployment among black women in Baltimore climbed 13% compared to 7% among white women in the one-year period. 28 The greater impact of the Depression on older men appears to have been compounded for black men. Once laid off, they were ―at a disadvantage [in the labor market] on account of their age and on account of their race.‖29

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

10

Linda Levine

Table 4. Unemployment by Gender and Age, 1930 and 1940 (numbers in thousands) Gender and Age

Number Unemployed 1930 1940

Unemployment Rate 1930 1940

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Men 2,011 6,185 5.4% 15.4% 14-19 years 220 867 7.9 33.1 20-24 years 338 1,074 7.1 21.3 25-44 years 846 2,361 4.8 12.5 45-64 years 528 1,698 5.2 14.2 65 years and over 79 185 4.4 10.0 Women 440 1,776 4.2 13.6 14-19 years 146 434 9.2 31.1 20-24 years 94 425 4.1 15.8 25-44 years 140 592 3.2 9.7 45-64 years 56 307 3.0 12.0 65 years and over 4 18 1.6 6.5 Source: U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: Estimates of Labor Force, Employment and Unemployment in the United States, 1940 and 1930, Washington, DC, 1944. Notes: The Census Bureau adjusted the data from the 1930 census to accord with labor force definitions utilized in the 1940 census. The Bureau did not adjust estimates by race.

Occupational and Industry Characteristics Lower skilled workers were more adversely affected by the Depression than higher skilled workers. For example, while semiskilled workers represented 16.5% of gainful workers in 1930, they accounted for a larger share of all unemployed persons (25.3%) in 1937.30 Nonfarm laborers similarly were greatly overrepresented among the unemployed. In contrast, farmers and other proprietors, managers, and officials as well as professional and clerical workers were underrepresented among the unemployed compared to their presence among gainful workers. (See Table 5.) The manufacturing, services, and construction industries accounted for a majority (63%) of all unemployed persons in 1937.31 In 1937, 29% of all jobless persons formerly worked for manufacturers, with nearly one-fourth having been employed by iron and steel producers. Another 12% had worked in textile industries and 10% in clothing industries. Service-related firms (e.g., recreation and amusement, hotels, restaurants, and professional services) accounted for almost 18% of the unemployed. Some 16% of jobless persons were formerly employed in the construction industry. Duration of Unemployment Very long spells of unemployment were common during the 1930s. Among unemployed men in Massachusetts in January 1934, for example, 62% had been without jobs a year or longer while 45% had not worked in two or more years, 25% in three or more years, and 11% in the previous four or more years.32 Women in the state fared slightly better with 54%

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

The Labor Market during the Great Depression and the Current Recession

11

unemployed a year or longer; 35%, two or more years; 18%, three or more years; and 8%, four or more years. Long periods of unemployment remained prevalent later in the decade as well. Of the 4.7 million unemployed persons who reported the number of weeks they had worked during the preceding 12 months, 31% had been jobless for more than one year as of November 1937.33 Even in 1940, after the Great Depression had ended, workers still were experiencing extended periods of unemployment. Of unemployed workers who reported when they were last employed, one-third had not held a job in a year or more.34

Geographic Characteristics Although it is unlikely that any geographic area was not touched by the Depression, some areas felt it more than others. In 1940, the Middle Atlantic region reported the highest rate of unemployment (13.6%) and two states within it—Pennsylvania (14.4%) and New York (13.7%)—had the highest jobless rates in the nation. In New England, which had the next highest unemployment rate (10.6%), every state except Connecticut and Vermont reported double-digit rates. Both the Pacific and Mountain regions also reported over one-tenth of their workers unemployed. (See Table 6.) At the other end of the spectrum, the South Atlantic region had the lowest unemployment rate in the nation. On a state-by-state basis, South Carolina (4.0%) in the South Atlantic region and Mississippi (4.8%) in the East South Central region reported the lowest unemployment rates.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Table 5. Distribution of the Unemployed by Occupation in 1937 and the Distribution of Gainful Workers by Occupation in 1930 (15-74 year olds) Occupation Unemployed Workers Gainful Workers Total 100.0% 100.0% Professionals 3.4 6.1 Farmers (owners and tenants) 2.5 1 2.2 Other proprietors, managers, and officials 1.4 7.5 Clerical, sales and related workers 14.0 16.4 Skilled workers and supervisors 14.6 13.0 Semiskilled workers 25.3 16.5 Nonfarm laborers 19.9 12.9 Farm laborers 10.8 8.4 Servants 8.1 6.9 Source: U.S. Office of Administrator of the Census of Partial Employment, Unemployment, and Occupations, Final Report on Total and Partial Unemployment, 1937, vol. I, Washington, D.C., 1938. Notes: The 1930 decennial census provided the latest nationwide count of gainful workers to which the 1937 count of unemployed workers could be compared. (See the Appendix for the definition of gainful workers.) The two also used the same occupational classification system. To attain age comparability between the 1937 unemployment registration and the 1930 census, the U.S. Office of the Administrator adjusted the census data to remove persons under age 15 and over age 74. The number of unemployed persons in 1937 includes public emergency workers and excludes new workers and workers who did not report an occupation.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

12

Linda Levine

The number of workers laid off from manufacturing industries was large, the particular industries varying by region.35 In New England, for example, there were substantial layoffs in the shoe, textile, and steel manufacturing industries. In the Middle Atlantic region, job losses were high at clothing, textile, and steel manufacturers. Layoffs were prevalent in the steel and auto industries of the East North Central region. Farming, a major source of jobs at the time, accounted for much of the unemployment in other regions. They were the South Atlantic, East South Central, West South Central, West North Central, and Mountain regions.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Table 6. Unemployment Rates by Region and State, 1940 Region by State Unemployment Rate Region by State Unemployment Rate New England 10.6% Georgia 5.3% Connecticut 8.7 Maryland 7.4 Maine 11.8 North Carolina 5.4 Massachusetts 11.2 South Carolina 4.0 New Hampshire 10.3 Virginia 6.4 Rhode Island 12.3 West Virginia 11.2 Vermont 7.2 East South Central 7.2 Middle Atlantic 13.6 Alabama 6.6 New Jersey 11.4 Kentucky 9.6 New York 13.7 Mississippi 4.8 Pennsylvania 14.4 Tennessee 7.3 East North Central 9.0 West South Central 8.3 Illinois 9.1 Arkansas 6.9 Indiana 8.0 Louisiana 8.1 Michigan 9.4 Oklahoma 10.2 Ohio 9.5 Texas 8.1 Wisconsin 8.4 Mountain 10.2 West North Central 8.0 Arizona 11.0 Iowa 6.4 Colorado 9.5 Kansas 7.3 Idaho 10.2 Minnesota 10.0 Montana 10.3 Missouri 8.5 Nevada 8.9 Nebraska 6.8 New Mexico 12.1 North Dakota 7.2 Utah 10.1 South Dakota 6.4 Wyoming 8.7 South Atlantic 6.5 Pacific 10.3 Delaware 7.6 California 10.6 District of Columbia 7.2 Oregon 9.7 Florida 7.6 Washington 9.9 Source: U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: Comparative Occupation Statistics for the United States, 1870 to 1940, Washington, D.C., 1943.

In summary, about one in every four workers in the nation was unemployed in the worst year of the Great Depression. Many workers were unemployed for much longer than one Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

The Labor Market during the Great Depression and the Current Recession

13

year. Of those fortunate to have jobs, many experienced cutbacks in hours, and hence, earnings. Men— particularly those older and black—were among the most adversely affected during a time when age- and race-based employment discrimination were not unlawful and when occupational shifts in labor demand were operating against them. Lower skilled workers were at a greater disadvantage in the labor market than higher skilled workers. Those who toiled on farms and in factories were displaced in especially large numbers. And, states whose economies were dependent upon agriculture and manufacturing reported comparatively high unemployment rates.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

COMPARISON WITH THE RECESSION THAT BEGAN IN DECEMBER 2007 The following analysis compares the Great Depression with the eleventh recession of the post-World War II period. It utilizes, when possible, seasonally adjusted monthly data for December 2007 (the recession’s start) and the latest month in 2009 for which data are available. In those cases in which the U.S. Bureau of Labor Statistics (BLS) does not adjust monthly data for seasonal fluctuations, which prevents month-to-month comparisons, annual average data for 2008 are used. For the Depression period, data from tables earlier in the report are supplemented with data from the 1940 decennial census when it provides statistical series more akin to those available today. From a labor market perspective, the eleventh recession in the post-World War II period is reminiscent in several respects of the Great Depression. The similarities are not unique to the two economic downturns, however. They are shared with many of the recessions that occurred in the intervening years. Analogously, the Great Depression differs from the latest and intervening recessions by having had a much worse effect on workers based on labor force measures (e.g., unemployment rate) and on the lack of a pre-existing safety net of government programs.

Similarities There are a number of similarities between the characteristics of the unemployed during the Great Depression and the recession that began in December 2007. Three similarities are intertwined: 1. The cyclically sensitive goods-producing sector lays off workers during recessions in numbers that are out of proportion to the sector’s share of total employment. 36 2. The concentration of blue-collar occupations in the goods-producing sector means that workers in these jobs are particularly vulnerable to displacement during downturns in the business cycle. 3. Men, who dominate the ranks of workers in the goods-producing sector and in bluecollar occupations, typically are more adversely affected by recessions than women. Workers in the mining, construction, and manufacturing industries are overrepresented among those who lose jobs during economic downturns. About 40%

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

14

Linda Levine of nonfarm jobs during the Great Depression were in the goods-producing sector, but it accounted for 57% of all jobs cut from employer payrolls between 1929 and 1933 and 77% between 1937 and 1938.37 By 2008, the goods-producing sector’s share of nonfarm employment had fallen to 15.6%. Nonetheless, 50.4% of employees who lost jobs between December 2007 and May 2009 formerly worked in the mining, construction, and manufacturing industries.38 Workers in blue-collar occupations historically have accounted for a majority of employment in the cyclically sensitive goods-producing sector. In 1940, 66.1% of craft workers, operatives, and nonfarm laborers were employed in the mining, construction, and manufacturing industries.39 The same holds true, albeit to a lesser extent, today: 51.1% of persons employed in construction and extraction occupations (e.g., craft workers, laborers) and in production occupations (e.g., machine operators, assemblers) work in the mining, construction and manufacturing industries.40 As a result, workers in blue-collar occupations are at a higher than average risk of displacement during recessions.41 The greater concentration of men in cyclically sensitive industries and occupations contributes to the more adverse impact of recessions on male members of the labor force. One means of measuring this differential effect is to examine the trend in employment by gender. During the Great Depression, the number of employed women increased while the number of employed men decreased.42 In 1940, men accounted for 94.3% of persons employed in the mining industry, 98.3% of those in the construction industry, and 78.0% of workers in the manufacturing industry.43 In that year, men also composed 97.9% of craft workers, 96.8% of nonfarm laborers, and 75.2% of operatives and related workers.44 Seven decades later, men remain the dominant jobholders in the goods-producing sector: in 2008, men accounted for 87.2% of employment in mining, 90.3% in construction, and 70.7% in manufacturing.45 Last year, men also composed 97.5% of workers in construction and extraction occupations (e.g., craft workers, laborers) and 70.3% workers in production occupations (e.g., machine operators, assemblers).46 Partly as a result, men’s employment overall fell precipitously—by 5.4% (4.3 million)—from 78.3 million in December 2007 to 74.0 million in May 2009. Despite their somewhat increased presence over time in these industries and occupations, women experienced substantially less job loss than men during the current recession. The number of employed women fell 2.2% (1.5 million) from 68.0 million in December 2007 to 66.5 million in May 2009.47 The comparatively worse impact of recessions on male employment is not limited to the Great Depression and the recession that began in December 2007. According to an analysis conducted in 1993 of data from the Current Employment Statistics program, which began to collect data by gender in 1964, most of those who lost jobs in the five recessions that occurred between December 1969 and March 1991 were men. The researchers found that although women lost jobs in the last two of the five recessions covered by their analysis, men lost 9 to 19 times more jobs than women in the July 1990-March 1991 and July 1981-November 1982 recessions, respectively. They concluded that

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

The Labor Market during the Great Depression and the Current Recession

15

[t]he chief explanation for the vast differences in employment loss between women and men in recessions concerns the proportions of jobs held by women in the various industries.... [B]ecause the goods-producing industries bear most of the job loss during recessions and because employment in this sector is heavily male, men lose the great majority of jobs in recessions. The industry divisions that fare best during recessions, services and government, have a high concentration of women, partially accounting for women’s relative job stability.‖48

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

The more negative effect of the Great Depression and recent recessions on male members of the labor force also is discernible from unemployment statistics. As was the case during the Depression,49 unemployed men as a proportion of the male labor force exceeded unemployed women as a proportion of the female labor during the last three recessions. At the end of the 1981-1982 recession in November, the unemployment rate among men measured 11.1% compared to 10.2% among women; at the end of the 1990-1991 recession in March, the male unemployment rate was 7.2% compared to a female unemployment rate of 6.3%; and at the end of the 2001 recession in November, the unemployment rate for men measured 5.7% compared to 5.4% for women. Similarly, in May 2009, 17 months into the latest recession, men’s unemployment rate was 10.5% as opposed to women’s unemployment rate of 8.0%.50 Research also suggests that varying trends in the growth rates of male- and femaledominated industries contributed to the relatively worse unemployment experience of men starting with the 1981-82 recession.51 More specifically, it was estimated that faster employment growth in industries with large concentrations of women compared to industries with large concentrations of men narrowed the female-male unemployment differential to the point where men’s unemployment rate exceeded that of women during recent recessions—in stark contrast with the comparative level of their unemployment rates during most of the post-World War II period. 4. Lower-skilled workers are more susceptible to unemployment than higher-skilled workers. The incidence of unemployment tends to vary inversely with the skill level of a worker (regardless of the stage of the business cycle).52 In 1940, the unemployment rate among craft workers (13.7%) and nonfarm laborers (19.5%) was substantially higher than the unemployment rate among professionals (2.4%) and nonfarm proprietors, managers, and officials (2.0%), for example.53 Unemployment rates by occupation show the same pattern today. The unemployment rate of workers in management, professional, and related occupations (2.7%) was well below the rate of natural resources, construction, and maintenance occupations (8.8%) and production occupations (7.7%) in 2008.54 Looking at the data in terms of changes in employment reveals the same pattern: higher-skilled workers weather economic downturns better than lower-skilled workers on average. Between 1930 and 1940, employment among professionals and nonfarm proprietors, managers, and officials increased. 55 Similarly, between 2007 and 2008, employment of comparatively high- skilled workers increased by 973,000 among management, professional and related occupations.56 In contrast, employment

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

16

Linda Levine among craft workers and nonfarm laborers decreased during the Depression period. Employment of workers in construction and extraction occupations and in production occupations decreased as well between 2007 and 2008, falling by 934,000 and 422,000, respectively. 5. Once unemployed, older workers have a more difficult time becoming reemployed.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

The high proportion of workers under 25 years old who were unemployed at the time of the [1940] census were due to a comparatively rapid turnover in employment, with frequent but relatively short periods of idleness. Among workers 55 to 64 years old, on the other hand, unemployment apparently occurred less frequently, but those who lost their jobs experienced relatively great difficulty in finding another job, and tended to remain unemployed for comparatively long periods.57

In 1940, the median duration of unemployment was about seven months for job seekers up to age 35. The length of unemployment spells increased for each age group thereafter.58 The pattern of older workers having more difficulty finding new jobs has recurred. In 2008, the median duration of unemployment for all workers was 9.4 weeks as opposed to 12.2 weeks for 45 to 54 year olds and 11.8 weeks for 55 to 64 year olds. 59 Jobless workers age 45 and older also were overrepresented among the long-term unemployed (i.e., those without jobs for at least six months). While 28.5% of all unemployed workers were at least 45 years old, these baby-boom generation workers accounted for 37.8% of the long-term unemployed. As was the case during the Depression, some speculate that age discrimination plays a role in the reemployment problems of today’s older workers. For example, the publisher of a job listing website (Workforce50.com) recently asserted that older workers suffer from a misperception among employers of ―being overqualified, overpriced, technologically challenged and inflexible.‖60 ―Many out-of-work baby boomers have despaired as they wonder whether to trim their resumes to avoid giving away their decades of work experience, or to dye their hair.‖61 A report by the House Select Committee on Aging offered age discrimination as one explanation for the reemployment difficulties of older workers during the steep 1981-1982 recession, when the national unemployment rate last broke 10%.62 6. The number of people involuntarily working part-time increases during economic downturns. Just as involuntary part-time employment increased during the 1930s, so too has it grown today. The number of persons working part-time for economic reasons, that is, who would prefer full- time jobs if they were available, almost doubled in nonfarm industries between December 2007 and May 2009, from 4,548,000 to 8,928,000.63

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

The Labor Market during the Great Depression and the Current Recession

17

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Differences In Magnitude There are substantial differences in the extent of unemployment during the Great Depression and the current recession. The unemployment rate rose almost eight-fold between 1929 (3.2%) and 1933 (24.9%). In contrast, it almost doubled between December 2007 (4.9%) and May 2009 (9.4%). At the peak of unemployment during the Great Depression (1933), one in four workers was unemployed, in contrast with fewer than one in eleven today. To approximate the pervasiveness of unemployment at the depth of the Depression, the number of workers without jobs would have to have totaled 38.6 million in May 2009, which is 24 million more workers than were unemployed this May (14.5 million).64 Employers cut the total number of jobs on their payrolls much more deeply during the Great Depression than they have thus far in the latest recession. Between 1929 and 1933, employment on nonfarm payrolls fell by 24.3%, compared to 4.3% thus far in the recession. To approximate the relative extent of cutbacks that took place over the four-year period between 1929 and 1933, employers would have had to have shed 27.6 million more workers than they did between December 2007 and May 2009. In the goods-producing sector, 7.2 million rather than 3.0 million workers would have to have been laid off since the recession began to equal the relative impact of the four-year (1929-1933) decline. Within the goodsproducing sector, construction companies would have had to have pared payrolls by 2.2 million more jobs than the 1.2 million positions cut through May 2009. Manufacturers would have had to have let go 2.5 million workers beyond the 1.8 million they displaced since December 2007 if the industry was in as relatively bad shape as it was in 1933. 65 A comparison between these labor force measures two years into the Depression and thus far in the ongoing recession similarly shows the latter to be less severe than the former. By 1931, the unemployment rate had risen to 15.9% or five times higher than the 3.2% recorded in 1929.66 In contrast, the unemployment rate almost doubled 17 months into the latest recession (going from 4.9% in December 2007 to 9.4% in May 2009).67 To approximate the pervasiveness of unemployment in 1931, the number of workers without jobs would have to have totaled 24.7 million in May 2009, which is 10 million more workers than were unemployed in that month. Similarly, the 4.3% decrease in nonfarm payroll employment 17 months into the recession compares favorably with the 15.0% decrease recorded two years into the Depression.68 To approximate this difference in job loss, employers would have had to have let go 14.7 million more workers than they did between December 2007 and May 2009. In the goods-producing sector, 5.0 million rather than 3.0 million workers would have to have been laid off since the current recession began to equal the relative impact of the 1929-1931 decline. Within the goods- producing sector, construction companies would have had to have pared payrolls by 1.4 million jobs rather than the 1.2 million positions eliminated through May 2009. Manufacturers would have had to have shed almost 1.5 million workers beyond the 1.8 million displaced since December 2007 if the industry was performing as badly as it had two years into the Depression.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

18

Linda Levine

Geographically At least part of the different effect by geographic area of the Great Depression and current recession flows from the industry composition of states. As shown in Table 7, many of the states in the East North Central region had among the highest unemployment rates in May 2009 (the latest month for which state data are available).

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Table 7. Unemployment Rates by Region and State, May 2009 Region by State Unemployment Rate Region by State Unemployment Rate New England 8.3% Maryland 7.2 Connecticut 8.0 North Carolina 11.1 Maine 8.3 South Carolina 12.1 Massachusetts 8.2 Virginia 7.1 New Hampshire 6.5 West Virginia 8.6 Rhode Island 12.1 East South Central 10.3 Vermont 7.3 Alabama 9.8 Middle Atlantic 8.3 Kentucky 10.6 New Jersey 8.8 Mississippi 9.6 New York 8.2 Tennessee 10.7 Pennsylvania 8.2 West South Central 7.0 East North Central 11.0 Arkansas 7.0 Illinois 10.1 Louisiana 6.6 Indiana 10.6 Oklahoma 6.3 Michigan 14.1 Texas 7.1 Ohio 10.8 Mountain 7.8 Wisconsin 8.9 Arizona 8.2 West North Central 7.3 Colorado 7.6 Iowa 5.8 Idaho 7.8 Kansas 7.0 Montana 6.3 Minnesota 8.2 Nevada 11.3 Missouri 9.0 New Mexico 6.5 Nebraska 4.4 Utah 5.4 North Dakota 4.4 Wyoming 5.0 South Dakota 5.0 Pacific 11.1 South Atlantic 9.6 Alaska 8.4 Delaware 8.1 California 11.5 District of Columbia 10.7 Hawaii 7.4 Florida 10.2 Oregon 12.4 Georgia 9.7% Washington 9.4 Source: U.S. Bureau of Labor Statistics, data from the Local Area Unemployment Statistics program.

The difficulties of the area, whose employment base remains heavily dependent upon goods production,69 reflect the current recession’s intensification of the long-term problems of U.S. manufacturers—particularly those companies that directly and indirectly employ many of the region’s workers at auto assembly plants and parts suppliers.70 In contrast, the Middle

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

The Labor Market during the Great Depression and the Current Recession

19

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Atlantic and New England regions reported the highest unemployment rates in 1940.71 The comparatively better unemployment picture in these regions during the current recession likely is associated with diversification of their employment away from manufacturing industries (e.g., shoe, clothing, and textile production) over the past several decades.72 Another reason for the different geographic effect of the Great Depression and the current recession is the varying impact of the two downturns in nonmanufacturing industries. Many states that b Research Triangle.75

In the Impact of Being Unemployed The implications of being unemployed have changed substantially over time. One reason for the altered situation facing jobless persons is the increased prevalence of families in which both spouses work. In 1940, three-fifths of married-couple families reported the husband as sole breadwinner.76 In 2008, the husband was the lone earner in about one-fifth of marriedcouple families. Both spouses were employed in another 51.4% of married-couple families.77 Consequently, the financial impact of joblessness is likely to be less today than it was more than half a century ago, when unemployment more often meant the total absence of a paycheck for a larger proportion of families. Even for those persons who were employed during the Great Depression, the dismal economy created a greater hardship than is the case today because of the much deeper drop in earnings and hours worked. Average hourly earnings of factory workers in manufacturing industries fell by 21.4% between 1929 and 1933 and by 8.9% between 1929 and 1931. In addition, average weekly hours for these workers decreased by 13.8% between 1929 and 1933 and by 8.4% between 1929 and 1931.78 As of May 2009, in contrast, average hourly earnings of nonmanagerial workers in manufacturing industries decreased by 3.9% and their hours by 5.6%.79 Another important difference between the Great Depression and the current recession, in terms of the economic hardship inflicted by high unemployment, is the nationwide availability of public assistance programs in place today. Before the 1930s, public assistance was provided by private charities and local governments.80 State governments did not become involved in the provision of public relief until the latter part of 1931, as it became increasingly apparent that the traditional sources of assistance could not meet the rising needs of unemployed workers and their families. Seven states (New York, New Jersey, Wisconsin, Pennsylvania, Rhode Island, Illinois, and Ohio) appropriated and distributed unemployment relief between September 1931 and July 1932. Following passage of the Emergency Relief and Construction Act of 1932 by Congress, 33 more states became involved in providing public assistance. The act was the only major federal measure passed during the Hoover Administration explicitly to aid the jobless. It came almost four years after the Depression began. The Emergency Relief and Construction Act was signed reluctantly by President Hoover, who believed that the traditional sources of relief were the most appropriate agencies to handle the needs of the unemployed. The act was envisioned as a temporary measure, not to imply long-term commitment of federal resources in this area. The federal money was offered to the states in the form of loans that were to be repaid with interest or through deductions from future federal aid to states for highways. (This repayment feature was cancelled by Congress in 1938 during the Roosevelt Administration.)

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

20

Linda Levine

Not until enactment of the Federal Emergency Relief Act of 1933 (FERA) during the Roosevelt Administration did the federal government truly commit its own resources for provision of assistance to the unemployed. FERA, through matching grants and discretionary funds under Title I, aided employable persons and their dependents with cash and in-kind benefits between 1933 and 1935. Even under this program, however, direct public relief was considered a temporary measure (―the dole‖), while work relief (job creation) programs were emphasized. With passage of the Social Security Act in 1935—two years after the end of the Depression period’s first downturn—the federal government established a permanent presence in the area of benefit transfer programs (e.g., old-age and survivors insurance, unemployment insurance). Thus, persons unemployed during the current recession did not have to wait until controversy ended over whether aid should be administered by private or public agencies, until state governments and the federal government awoke to the magnitude of the jobless problem and intervened, or until an administrative system was established at the state and federal levels before they received benefits. In addition to already having safety net programs in place to cushion the impact of unemployment, today’s jobless also benefit more from the social legislation than did the Depression’s unemployed. A nationwide federal-state unemployment compensation program was part of the original Social Security Act, but when it was enacted, many groups were excluded from coverage. Among the excluded were workers in government and educational organizations as well as those laid off from small firms. As a result, the unemployment compensation program covered less than half of all employees in 1940.81 Over the years, the system has been expanded so that today almost all wage and salary workers are covered.82 Workers eligible for unemployment compensation (UC) during the current recession also collect benefits for a longer period of time than persons unemployed during the Depression. In December 1937, for example, 12 states provided benefits for a maximum of 12-14 weeks; 33 states, 15-17 weeks; and 4 states, 18-20 weeks.83 Most states now provide regular UC benefits for a maximum of 26 weeks. There is a permanently authorized extended benefit program in effect today as well that provides an additional 13 or 26 weeks of UC benefits in high-unemployment states. In addition, the 110th Congress authorized as part of P.L. 110-252 a temporary unemployment benefit program for workers in all states who exhausted their regular benefits. The 111th Congress extended the temporary program and expanded upon it in the American Recovery and Reinvestment Act of 2009 (P.L. 111-5). 84 Yet another liberalization of the law since the 1930s that helps today’s unemployed workers involves the waiting period before benefits can be collected. In the early years of the system, states had waiting periods ranging from two to four weeks.85 Currently, 25 states have no waiting period and the remaining states have a one-week requirement.86 Moreover, in the Emergency Unemployment Compensation Amendments of 1992 (P.L. 102-318), a work-sharing program was authorized permanently. It allows employers who temporarily cutback the hours of their employees rather than laying them off to develop shorttime compensation plans. If the plans are approved by the state in which the firms are located, employees involuntarily working reduced schedules can receive partial UC benefits.87 No program to compensate workers for the wage losses associated with pervasive involuntary part- time employment existed during the Great Depression. Finally, the UC system was of no value to jobless workers during the Depression period’s first economic downturn as the Social Security Act was not passed until 1935. In addition, no

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

The Labor Market during the Great Depression and the Current Recession

21

benefits were paid out for the first two years in order for reserve funds to accumulate. 88 It also was of no value to those who were jobless for so long during the 1930s that they did not meet the earnings or work time criteria for eligibility. Other federal programs available to assist those persons unemployed today also were in their infancy during the 1930s. In September 1933, it became national policy to provide aid to farmers and the unemployed by purchasing the farmers’ surplus produce and distributing it to the unemployed and their families. However, because of problems with the commodity distribution program, a new idea – food stamps – emerged in the late 1930s. The first food stamps were purchased in Rochester, New York in May 1939. Before the program ended in March 1943, it operated in half the counties of the nation. Thus, although food was available to the unemployed and their dependents in some geographic areas during the Depression, both the commodity distribution and food stamp programs did not start until after the economic downturns of the 1930s had ended.89

APPENDIX. THE EVOLUTION OF LABOR FORCE DATA During the past several decades, the definition and measurement of employment and unemployment have undergone many changes. The data available during the Great Depression are first described below. Next, the data available today are contrasted with the Depression-era statistics.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

The Great Depression Between the 1930 and 1940 censuses, the Census Bureau went from the gainful worker to labor force concept. Up to and including the 1930 census, individuals aged 10 and above reported themselves as gainful workers if they had at some time worked in an occupation in which they earned money or the equivalent, or in which they assisted in the production of marketable goods – regardless of their activity at the time of the census. Thus, persons who had never worked but were available for work were not counted as gainful workers. Beginning with the 1940 census, individuals aged 14 and above reported themselves as members of the labor force if they were employed for pay or profit, or at unpaid family work, during the week of March 24 to 30, 1940, or were seeking jobs or performing public emergency work during the specified period. The changeover from a gainful worker to labor force concept raised the age criteria, included new workers, and specified a time period in which labor market activity must have occurred. It also created discrepancies between the two census counts in the way other individuals were classified (e.g., seasonal workers, inmates of institutions, retired and disabled persons).90 The Census Bureau’s adjustment of the gainful worker concept to the labor force concept produced a net decrease of 1.2 million persons in the figure for gainful workers reported in the 1930 census. The second change was in the measurement of unemployment. The 1930 census was the first national effort to count the unemployed. Gainful workers were asked if they had been at work the day before the census-taker called. If they were not at work, the individual was asked

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

22

Linda Levine 

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

      

whether they had a job, whether they were able to work, whether they were looking for work, whether they had lost pay because they were not at work, and the reason for their idleness. On the basis of this information, gainful workers not at work ―yesterday‖ were classified into the following seven categories: 91 were jobless, able to work, and looking for work; were on layoff without pay, excluding the sick and voluntarily idle; were jobless and unable to work; had a job but were idle due to sickness or disability; were jobless and not looking for work; had a job but were voluntarily idle without pay; or had a job and were being paid although not at work.

According to the definition of the labor force introduced in the 1940 census, however, an individual was classified as either employed if at work within a given week, unemployed if seeking work or performing public emergency work within a given week, or not in the labor force if neither of the prior two classifications applied. The Census Bureau adjusted the much broader definition of unemployment in the 1930 census to comply with that of the 1940 census by estimating for each of the seven categories the number of individuals who would have been classified as unemployed had the 1940 census definition been utilized. The Bureau also estimated for those not included in the gainful worker definition (e.g., new workers) the number of persons who would have been counted as unemployed. Of the 2,451,000 workers the Census Bureau estimated to be unemployed in 1930 based on the 1940 census definition, 213,000 were new workers and 2,238,000 were classified as experienced workers seeking jobs. Several surveys of states and cities as well as the 1937 nationwide census of unemployment were conducted during the course of the Depression to gauge the pervasiveness of unemployment and the characteristics of the unemployed. The data sources were based on outdated (un)employment concepts and permit only a snapshot rather than time-series analysis as they were not conducted on a recurring basis. However, the sources were included in the preceding analysis because they provide data unavailable from the decennial censuses and data for intercensal years.

Today Compared to the Great Depression When comparisons are made in the report between unemployment rates during the course of the Depression and recession, the Depression-era data are retrospective estimates designed to be comparable with the Current Population Survey (CPS) rather than the 1930 and 1940 decennial censuses.92 Today, the CPS is the official source of timely labor force data. Its precursor was developed by the Works Progress Administration (WPA) in 1940. The Census Bureau assumed responsibility for the monthly unemployment report from the WPA in 1942.93 In the late 1950s, the Bureau of Labor Statistics assumed responsibility for the content, analysis and release of labor force data derived from CPS which continues to be conducted by the Census Bureau. Based on the CPS, the Bureau of Labor Statistics releases estimates each month on the labor force status of individuals.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

The Labor Market during the Great Depression and the Current Recession

23

End Notes

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

1

For more information, see CRS Report R40007, Financial Market Turmoil and U.S. Macroeconomic Performance, by Craig K. Elwell; and CRS Report R40198, U.S. Economy in Recession: Similarities To and Differences From the Past, by Marc Labonte. 2 John E. Bregger, ―The Current Population Survey: a Historical Perspective and BLS’ Role,‖ Monthly Labor Review, June 1984, p. 8. 3 U.S. Office of Administrator of the Census of Partial Employment, Unemployment, and Occupations, Final Report on Total and Partial Unemployment, 1937, vol. I, Washington, D.C., 1938, pp. VII-VIII. 4 For example, the Census Bureau’s adjustment of the 1930 gainful worker concept to the 1940 labor force concept reduced the former by 1.2 million persons. 5 U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: Comparative Occupation Statistics for the United States, 1870 to 1940, Washington, D.C., 1943. 6 John D. Durand, The Labor Force in the United States, 1890-1960 (New York: Social Science Research Council, 1948), pp. 114-115. 7 Ibid.. 8 Ibid. 9 U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: Comparative Occupation Statistics for the United Sates, 1870 to 1940, Washington, D.C., 1943. 10 Ibid. 11 Although the ―establishment survey‖ is limited to nonfarm employers, they have accounted for the majority of total U.S. employment since at least 1900. Annual estimates of employment that were developed retrospectively and appear in Historical Statistics of the United States, Colonial Times to 1970 (U.S. Bureau of the Census, Washington, D.C., 1975) show that nonfarm employees composed 77% of total U.S. employment in 1929 for example. 12 U.S. Bureau of Labor Statistics, Employment, Hours, and Earnings, United States, 1909-90, Bulletin 2370, vol. I, Washington, D.C., 1991. 13 Mitchell Broadus, ―Depression Decade: From New Era to New Deal, 1929-1942,‖ in Economic History of the United States, vol. IX (New York: Holt, Rhinehard, and Winston, 1962). 14 Annual estimates of the number of employed and unemployed persons nationally and in the farm sector and nonfarm sector appear in U.S. Bureau of the Census, Historical Statistics of the United States, Colonial Times to 1970 (Washington, D.C., 1975). These estimates are not available for the individual nonfarm industries shown in Table 3. They are not consistent with data from the decennial censuses and the establishment survey. The estimates were constructed retrospectively to be comparable with the Current Population Survey which was initiated in 1940. 15 William J. Barrett, ―Extent and Methods of Spreading Work,‖ Monthly Labor Review, September 1932. 16 Martin Nemirow, ―Work-sharing Approaches: Past and Present,‖ Monthly Labor Review, September 1984, p. 35. 17 Ibid., pp. 35-36. 18 U.S. Bureau of the Census, Historical Statistics of the United States, Colonial Times to 1970 (Washington, D.C., 1975). These retrospectively constructed estimates are not consistent with data from the decennial censuses. 19 Mitchell Broadus, ―Depression Decade: From New Era to New Deal, 1929-1941,‖ in Economic History of the United States, vol. IX (New York: Holt, Rinehart, and Winston, 1962). 20 U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: Estimates of the Labor Force, Employment, and Unemployment in the United States, 1940 and 1930, Washington, D.C., 1944. 21 Mitchell Broadus, ―Depression Decade: From New Era to New Deal, 1929-1941,‖ in Economic History of the United States, vol. IX (New York: Holt, Rinehart, and Winston, 1962), p. 98. 22 U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: Estimates of the Labor Force, Employment, and Unemployment in the United States, 1940 and 1930, Washington, D.C., 1944. 23 ―Michigan Population and Unemployment Census, 1935,‖ Monthly Labor Review, November 1936. 24 ―Census of Unemployment in Massachusetts,‖ Monthly Labor Review, December 1934. 25 U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: Estimates of the Labor Force, Employment, and Unemployment in the United States, 1940 and 1930, Washington, D.C., 1944, p. 3. 26 ―Census of Unemployment in Massachusetts,‖ Monthly Labor Review, December 1934. 27 ―Unemployment Among Women in the Early Years of the Depression,‖ Monthly Labor Review, April 1934. 28 U.S. Bureau of Labor Statistics, ―Surveys of Unemployment in Baltimore, 1928, 1929, and 1930,‖ Handbook of Labor Statistics, Washington, D.C., 1931. 29 John D. Durand, The Labor Force in the United States, 1890-1960 (New York: Social Science Research Council, 1948), p. 115. 30 The 1937 census of unemployment collected information from unemployed persons only. The Final Report on Total and Partial Unemployment, 1973 compared its unemployment estimates with data on gainful workers from the then most recent source of national data, namely, the 1930 census of population.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

24

Linda Levine

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

31

U.S. Office of Administrator of the Census of Partial Employment, Unemployment, and Occupations, Final Report on Total and Partial Unemployment, 1973, vol. I, Washington, D.C., 1938. 32 ―Census of Unemployment in Massachusetts,‖ Monthly Labor Review, December 1934. 33 U.S. Office of Administrator of the Census of Partial Employment, Unemployment, and Occupations, Final Report on Total and Partial Unemployment, vol., I, Washington, D.C., 1938. 34 U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: The Labor Force, Washington, D.C., 1943. 35 U.S. Office of Administrator of the Census of Partial Employment, Unemployment, and Occupations, Final Report on Total and Partial Unemployment,, vol. I, Washington, D.C., 1938. 36 The goods-producing sector is more sensitive to downturns in the business cycle than the service-providing sector because inventories of goods produced by the mining, construction and manufacturing industries build up due to reduced product demand during recessions. Firms in these industries consequently cut production and layoff workers as unsold homes and cars, for example, accumulate. Not until inventories diminish sufficiently do businesses increase production and not until they are confident that a sustained recovery is underway do firms (re)hire workers. 37 Calculated from the data in Table 3. 38 BLS, data from the Current Employment Statistics program, http://stats.bls.gov/ces. 39 U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: The Labor Force, Washington, D.C., 1943. 40 BLS, data from the Current Population Survey, http://stats.bls.gov/cps/cpsaat17.pdf. 41 During the 2001-2002 period, which encompassed the 2001 recession, the rate of job loss among long-tenured workers previously employed in production occupations was 8.7% while the average displacement rate was a much lower 4.8%. (This is the latest recessionary period for which data are available from BLS’ Displaced Worker Survey.) See the following report for additional information on the risk of displacement by occupation, industry, and other variables over time, CRS Report RL32292, Offshoring (a.k.a. Offshore Outsourcing) and Job Insecurity Among U.S. Workers, by Linda Levine. 42 See Table 1. 43 U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: The Labor Force, Washington, D.C., 1943. 44 Ibid. 45 BLS, data from the Current Population Survey, http://stats.bls.gov/cps/cpsaat17.pdf. 46 BLS, data from the Current Population Survey, http://stats.bls.gov/cps/cpsaat11.pdf. 47 BLS, data from the Current Population Survey, http://stats.bls.gov/cps. 48 William Goodman, Stephen Antczak, and Laura Freeman, ―Women and Jobs in Recessions: 1969-92,‖ Monthly Labor Review, July 1993, p. 28-29. 49 See Table 4 for 1930 and 1940 unemployment rates by gender. 50 BLS, data from the Current Population Survey, http://stats.bls.gov/cps. 51 Larry DeBoer and Michael Seeborg, ―The Female-Male Unemployment Differential: Effects of Changes in Industry Employment,‖ Monthly Labor Review, November 1984, pp. 8-15. 52 Often, occupational groups are used as proxies for skill level. An occupational group’s comparative skill level is based on the predominant number of years of schooling completed by persons employed in the group. In 1940, for example, 68.1% of professional and related workers had attended or completed college as had 20.3% of nonfarm proprietors, managers and officials (except farm) and 18.4% of clerical, sales and related workers. In contrast, 35.1% of nonfarm laborers, 38.6% of operatives and related workers, and 41.7% of craft and related workers had at most completed the first seven or eight years of elementary school. U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: Comparative Occupation Statistics for the United States, 1870 to 1940, Washington, D.C., 1943. 53 U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: The Labor Force, Washington, D.C., 1943. 54 BLS, data from the Current Population Survey, http://stats.bls.gov/cps/cpsaat25.pdf. 55 See Table 2. 56 BLS, data from the Current Population Survey, http://stats.bls.gov/cps. 57 U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: The Labor Force (Sample Statistics), Part I: General Characteristics, Washington, D.C., 1943, p. 15. 58 Ibid. 59 BLS, data from the Current Population Survey, http://stats.bls.gov/cps/cpsaat31.pdf. 60 Tiffany Hsu, ―Older Jobless Workers Face Tough Prospects,‖ Chicago Tribune, May 3, 2009. 61 Michael Luo, ―Longer Unemployment for Those 45 and Older,‖ The New York Times, April 13, 2009. 62 U.S. Congress, House Select Committee on Aging, The Unemployment Crisis Facing Older Americans, 97th Cong., 2nd sess., October 8, 1982, Comm. Pub. No. 97-367 (Washington: GPO, 1982). 63 BLS, data from the Current Population Survey, http://stats.bls.gov/cps.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

The Labor Market during the Great Depression and the Current Recession

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

64

25

The analysis in this paragraph is based on depression-era data from Historical Statistics of the United States, Colonial Times to 1970 (U.S. Bureau of the Census, Washington, D.C., 1975) and recent data from BLS, Current Population Survey. 65 The analysis in this paragraph is based on depression-era data in BLS, Employment, Hours, and Earnings, United States, 1909-90, Bulletin 2370, vol. I, Washington, D.C., 1991, and data for later years from BLS, Current Employment Statistics program. 66 U.S. Bureau of the Census, Historical Statistics of the United States, Colonial Times to 1970 (Washington, D.C., 1975). 67 BLS, data from the Current Population Survey. 68 BLS, Employment, Hours, and Earnings, United States, 1909-90, Bulletin 2370, vol. I, Washington, D.C., 1991, and BLS, Current Employment Statistics program. 69 In 2008, according to data from the Current Employment Statistics program, manufacturing jobs accounted for 11.1% of total nonfarm employment in Illinois, 17.7% in Indiana, 13.8% in Michigan and Ohio, and 17.2% in Wisconsin. The proportion in every state in the East North Central region was well above the average for the nation of 9.8% in 2008. 70 Some states in other regions that are comparatively dependent on motor vehicle and parts manufacturing similarly have comparatively high unemployment rates (e.g., Kentucky, North Carolina, South Carolina, and Tennessee) as shown in Table 7. See also ―Southeast Auto Parts Makers Feel Detroit’s Pain,‖ MSN Money, May 12, 2009. 71 See Table 6. 72 In 2008, according to data from the Current Employment Statistics program, manufacturing’s share of total nonfarm employment in the Middle Atlantic region (7.9%) was well below the national average (9.8%). Specifically, the proportion in New Jersey was 7.4%; in New York, 6.1%; and in Pennsylvania, 11.1%. Manufacturing’s share of total nonfarm employment in the New England region was equal to the national average of 9.8%. The shares of Maine (9.6%) and Massachusetts (8.7%) were below the average, while those of Connecticut (11.0%), New Hampshire (11.7%), Rhode Island (10.0%) and Vermont (11.4%) were above the average for the nation. 73 See Table 6. 74 Jennifer Robison, ―Nevada Economy,‖ Las Vegas Review-Journal, May 23, 2009; Peter Wong, ―Is This Recession Oregon’s Worst?,‖ Statesman Journal, May 17, 2009; and Bob Willis, ―California’s Unemployment Rate Rises to 26- Year High,‖ Bloomberg.com, March 20, 2009. 75 ―Recession Suddenly Humbles High-Tech Sector,‖ The Sacramento Bee, May 24, 2009. 76 U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: Families, Employment Status, Washington, D.C., 1943. 77 BLS, Employment Characteristics of Families in 2008, Washington, D.C., May 27, 2009. 78 BLS, Employment, Hours, and Earnings, 1909-1990, Bulletin 2370, vol. I, Washington, D.C., March 1991. 79 BLS, data from the Current Employment Statistics program, http://www.bls.gov/ces/. 80 Josephine C. Brown, Public Relief 1929-1939 (New York: Henry Holt and Company, 1940). 81 Tax Foundation Inc., Unemployment Insurance: Trends and Issues, Washington, D.C., 1982. 82 According to the U.S. Department of Labor, 62% of unemployed workers in May 2009 received UC benefits from all programs operating at the time. For information on benefit receipt see the Unemployment Compensation chapter of the Committee on Ways and Means’ 2008 Green Book at http://waysandmeans.house.gov/media/pdf/111/uc.pdf. 83 Paul A. Brinker, Economic Insecurity and Social Security (New York: Appleton-Century-Crofts, 1968). 84 CRS Report RL33362, Unemployment Insurance: Available Unemployment Benefits and Legislative Activity, by Alison M. Shelton and Julie M. Whittaker. 85 Tax Foundation, Unemployment Insurance: Trends and Issues, Washington, D.C., 1982. 86 CRS Report RL33362, Unemployment Insurance: Available Unemployment Benefits and Legislative Activity, by Alison M. Shelton and Julie M. Whittaker. 87 Short-time compensation programs exist in a minority of states today. 88 Broadus Mitchell, ―Depression Decade: From New Era to New Deal,‖ in Economic History of the United States (New York: Holt, Rinehart, and Winston, 1962). 89 U.S. Congress, House Committee on Agriculture, Food Stamp Act of 1977, 95th Cong., 1st sess., H.Rept. 95-464 (Washington: GPO, 1977). 90 U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: Estimates of the Labor Force, Employment, and Unemployment in the United States, 1940 and 1930, Washington, D.C., 1944. 91 U.S. Bureau of the Census, Sixteenth Census of the United States: 1940, Population: Estimates of the Labor Force, Employment, and Unemployment in the United States, 1940 and 1930, Washington, D.C., 1944, p. 2. 92 The methodology utilized and resultant data from 1900 to 1947 appears in U.S. Bureau of the Census, Historical Statistics of the United States, Colonial Times to 1970 (Washington, D.C., 1975). 93 John E. Bregger, ―The Current Population Survey: A Historical Perspective and BLS’ Role,‖ Monthly Labor Review, June 1984.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved. Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

In: Labor and Employment Issues Editor: Nickolas H. Mullen

ISBN: 978-1-60741-286-1 © 2010 Nova Science Publishers, Inc.

Chapter 2

LONG-TERM UNEMPLOYMENT



Congressional Budget Office

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

SUMMARY Each year, millions of people become unemployed and find a job within a few weeks, while many others change jobs without any intervening unemployment. The flexibility of the labor market is generally considered a source of strength for the U.S. economy. Some people, however, remain jobless for many months. Those longterm unemployed workers account for a large portion of the total weeks of unemployment. This paper examines the extent to which unemployment is concentrated among workers who are unemployed for more than six months. It also examines the characteristics of those long-term unemployed workers as well as their sources of income and subsequent activities. The analysis is largely based on data from the most recently completed panel available of the Census Bureau’s Survey of Income and Program Participation, which covers people who experienced any unemployment during 2001, 2002, or 2003.

Key findings include these: 





Although fewer than one-in-ten of the unemployment spells of adults that began in the 2001–2003 period lasted more than half a year, those long-term spells accounted for 37 percent of all weeks of work lost by adults due to unemployment during that period (see Summary Figure 1). The share jumps to 44 percent when shorter spells of unemployment experienced by those same people are included. Spells of unemployment often ended with the job seekers leaving the labor force rather than taking a job. About 70 percent of the unemployment spells begun by adults during the 2001–2003 period ended with the individuals taking a job; 25

This is an edited, reformatted and augmented version of a Congressional Budget Office publication dated October 2007.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

28

Congressional Budget Office





percent ended with them stopping their search; and the remaining spells were still in progress when the survey ended. Workers who never finished high school were not only more likely than other labor force participants to become unemployed, they also remained unemployed for a longer time. During the 2001–2003 period, although only one-in-ten adults in the labor force had not graduated from high school, one-in-six of the adults who experienced any unemployment and onein-five of the adults who experienced a longterm unemployment spell did not have a high school diploma. A majority of the long-term unemployed received unemployment insurance benefits for at least part of the period in which they were unemployed, offsetting a portion of their lost earnings. Also, although a majority of the long-term unemployed had health insurance coverage while unemployed, the percentage lacking insurance during their spell was higher than it had been before those workers became unemployed.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Introduction The frequent occurrence of short spells of unemployment by people entering the labor force or changing jobs is characteristic of a dynamic economy: As new opportunities for workers and firms arise, temporary periods of unemployment may occur. Even in a strong labor market, some people will be unemployed. (People are defined as unemployed if they are not working, are available to work, and have recently made specific efforts to find a job or are expected to be recalled from a temporary layoff.)1 Recent statistics from the monthly Current Population Survey (CPS) illustrate the dynamics of unemployment. In 2006, an average of 7.0 million people age 16 and over were unemployed. The number of unemployed people declined throughout the year, falling from 7.3 million at the end of 2005 to 6.8 million at the end of 2006. Each month in that year, an average of 2.6 million people became unemployed (see Table 1).2 The number of people leaving unemployment exceeded the number entering unemployment by an average of 40,000 per month (accounting for the overall drop of 500,000), which means that a slightly larger number of people who had been unemployed in the previous month must have either found a job or left the labor force. Estimates based on the CPS also suggest that, even in a period in which the labor market is generally healthy overall, a significant number of people take a long time to find a job. On average, about one-in-six unemployed people in 2006 (1.2 million) had already been unemployed for more than half a year. Few of the unemployed people under age 25, but one-in-five of the unemployed people age 25 or older, had been unemployed that long. Moreover, by the time they found a job or left the labor force, some of the individuals with shorter unemployment spells would have been unemployed that long as well. Long-term unemployment may result in serious problems for the unemployed individuals themselves as well as for the overall economy. Previous studies by the Congressional Budget Office (CBO) and others have documented substantial variability in workers’ earnings and in household income from one year to the next, some of which is associated with long-term unemployment.3 People who are taking a long time to find a job are losing the opportunity to

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Long-Term Unemployment

29

earn income, gain work experience, and, in some cases, receive health insurance and other nonwage benefits provided by employers.

Source: Congressional Budget Office based on data from the 2001 panel of the Survey of Income and Program Participation. Note: This figure is based on reports by respondents who began about 10,800 spells of unemployment during a three-year survey period that started in late 2000 or early 2001 and ended in late 2003.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Summary Figure 1. Duration of Unemployment Spells Begun by People Age 25 and Older, 2001 to 2003

Long-term unemployment can result from difficulties related to job seekers or potential employers or from the mechanisms that match supply to demand. For example, job seekers may not have the skills sought by employers in their labor market or may have unrealistic expectations; potential employers may not be hiring any new workers in their market; or unemployed workers may not know how to productively look for work. Long-term unemployment could also result from job seekers having less need to find a job quickly because of access to other sources of income, such as unemployment insurance (UI) benefits or income from a working spouse. For the overall economy, long spells of unemployment can indicate lost output. Taking the responses to the CPS at face value, the lost output of someone who says that he or she had been available for work and had been seeking a job for more than half a year could be substantial. Responses to a survey should not always be taken at face value, though. For example, some respondents might have told census interviewers that they were available to start work at any time during a certain period, when in fact they might not really have been available to do so or might not have been seriously looking. In addition, the end of a spell of unemployment does not always mean that an individual has found a job. Many spells end with the unemployed people no longer actively seeking work; that is, they leave the labor force.4 It is difficult to determine the extent to which that outcome indicates that the people had not been actively seeking a job and the extent to which it indicates that they stopped looking out of discouragement because suitable jobs were not available for them.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

30

Congressional Budget Office Table 1. Number and Distribution of Unemployed People, by Age and Duration of Spells in Progress, 2006 Age

Total

16 to 24 25 and Over Total, 16 years and over

2,350 4,650 7,000

16 to 24 25 and Over Total, 16 years and over

100 100 100

Duration of Spells Less Than 5 to 26 27 or More 5 Weeks Weeks Weeks Thousands of People 1,040 1,030 280 1,580 2,120 950 2,610 3,150 1,230 Percentage Distribution 44 44 12 34 46 21 37 45 18

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Source: Congressional Budget Office based on 2006 annual averages of monthly Current Population Survey statistics reported by the Bureau of Labor Statistics, available at www.bls.gov/cps/cpsa2006.pdf.

According to CPS data tabulated by the Bureau of Labor Statistics (BLS), in 2006 an average of 1.4 million people who were not in the labor force wanted to work, had searched for a job sometime in the previous year, and were available to work. However, only about 400,000 of those people indicated that they had stopped looking for work because of discouragement over job prospects; the rest had not looked for other reasons, such as problems with child care or transportation. Over the past several decades, long-term unemployment as a percentage of total unemployment has been highly cyclical and generally increasing, as measured by the CPS (see the upper panel of Figure 1).5 About one-in-six workers who were unemployed in 2006 had been unemployed for 27 weeks or longer, even though the overall unemployment rate was low. In several earlier periods with relatively low unemployment rates (such as in the early 1 970s), the percentage of unemployed people who had been unemployed for 27 weeks or longer was lower. Those statistics are incomplete in that they reflect the duration of unemployment spells that had not yet ended. That is, they are spells in progress, not completed spells. Nonetheless, it is likely that increases in the length of spells in progress were accompanied by increases in completed spells. The rise in long-term unemployment as a percentage of total unemployment reflects a decline in the percentage of the labor force that became unemployed as well as an increase in unemployment duration. In an average month in 2006, 1.7 percent of the labor force had been unemployed for less than five weeks—a rough measure of the percentage of the labor force that had become unemployed during the month. That short-term unemployment rate is the lowest it has been for half a century (see the lower panel of Figure 1). Those statistics indicate that, apart from variations stemming from the ups and downs of the business cycle, people are less likely to become unemployed than in the past, but those who do become unemployed are more likely to remain unemployed for more than half a year.6 Many workers who lose their job are eligible for unemployment insurance benefits, which offset a portion of an eligible worker’s lost wages. The duration of those benefits is generally limited to no more than 26 weeks, however, at which point recipients have exhausted their benefits.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Long-Term Unemployment

31

Source: Congressional Budget Office based on annual averages of monthly Current Population Survey statistics reported by the Bureau of Labor Statistics. Note: The duration of unemployment percentages in this figure depicts the duration of spells in progress, not completed spells. Vertical bars indicate periods of recession. Figure 1. Long-Term Unemployment and the Unemployment Rate, 1950 to 2006

The rate at which recipients exhaust their UI benefits has been gradually rising. In the years immediately after the 2001 recession, more than 40 percent of recipients of UI benefits had exhausted them—a rate that was higher than at any time in recent history. Since then, the exhaustion rate has fallen to about 35 percent, as the labor market has strengthened.7 Although it is not surprising that the exhaustion rate would climb as job opportunities decline and fall as opportunities increase, the gradual long-term rise in the rate is hard to explain. One part of the explanation for that rise, as well as for the rise in long-term unemployment, may be that an increasing share of job losses are permanent separations rather than temporary layoffs. According to research based on data from surveys conducted by the Census Bureau for the Department of Labor, the percentage of workers who are permanently separated from their job seems to have risen somewhat over the past two decades (adjusted for overall economic conditions).8

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

32

Congressional Budget Office

Moreover, research based on the same surveys indicates that the adverse consequences of losing a job because of slack work, a plant closing, or a position being abolished have increased. One study found that, on average, workers who lost a full-time job between 2001 and 2003 and found a new job by the time they were interviewed in 2004 earned about 17 percent less per week than they would have earned had they not been displaced.9 That amount was roughly double the average loss in weekly earnings incurred by workers who were displaced in the late 1990s. The increase in the size of the average loss in earnings was especially large for better-educated workers. Finally, as the author of that study points out, those estimates understate the total economic losses incurred by workers because the estimates do not take into account workers forgone earnings while they were unemployed and any losses in fringe benefits. A previous CBO study also underscores the difficulties associated with some job transitions.10 Examining unemployment insurance benefits provided to people who lost their job in the 2001 recession, the study found that the former recipients of UI benefits who went back to work within three months after their benefits ended were earning about 15 percent less than they had earned before they lost their job. Concerns about long-term unemployment have generated a search for new ways of identifying and aiding unemployed workers most likely to remain unemployed for a long period. The federal/state unemployment insurance system—the main public program for unemployed workers—is not designed to deal with unemployment spells lasting more than six months. Moreover, because the program temporarily replaces a portion of workers’ lost earnings, some job seekers have an incentive to stay unemployed longer than they would have otherwise. Early identification of individuals most likely to need help, job-search assistance, wage insurance, and retraining programs are among the approaches being used to shorten unemployment duration. The remainder of this paper provides information about how long it took for spells of unemployment to end and about whether those spells ended because the individuals found a job or because they left the labor force. The information about completed spells is used to show that unemployment is highly concentrated among a small number of people who experience long spells of unemployment. That finding underscores the potential payoff to the economy, as well as to the people themselves, of finding ways of reducing long-term unemployment.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.



The Extent to Which Unemployment Is Concentrated Millions of people enter the labor force, leave the labor force, or change jobs each month without experiencing any unemployment.11 Moreover, the extent of long-term unemployment in the United States appears relatively small in comparison with rates in many other countries.12 Nonetheless, unemployment in recent periods was quite heavily concentrated among a small portion of the workers who experienced long spells of unemployment. CBO estimated the degree of unemployment concentration by examining the experience of labor force participants during the three-year period covered by the most recently completed panel of the Survey of Income and Program Participation (SIPP) available, 2001 through 2003.13

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Long-Term Unemployment

33

BOX 1. MEASURES OF UNEMPLOYMENT The question of how often long-term unemployment occurs can be answered different ways, depending on precisely which measure is being used. As shown below, similarsounding measures can produce markedly different answers.

New Spells, Completed Spells, and Spells in Progress Measures based on new spells can be used to answer questions such as, How many of the people who become unemployed are starting spells that will last longer than six months? (This measure is analogous to the statistic often cited from unemployment insurance administrative data: the fraction of new unemployment insurance recipients who exhaust their entitlement to benefits.) Measures based on completed spells can answer questions such as, How many of the people unemployed at a point in time are in the midst of a spell that, when completed, will have lasted more than six months? Measures based on spells in progress can answer questions such as, How many people who are unemployed have been so for more than six months? This is the most common measure of unemployment duration and the one reported each month by the Bureau of Labor Statistics, based on data from the Current Population Survey (a monthly survey of about 60,000 households conducted by the Census Bureau for the Bureau of Labor Statistics). Because this measure uses information about how long people unemployed in a month had already been unemployed, rather than about the total length of unemployment spells, it can be provided on a timely basis.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

An Example In this example, suppose the question is, How common is an unemployment spell lasting more than one month? Assume that in January, two people became unemployed; one of them takes one month to find a job or leave the labor force, and the other takes two months. In February, the process is repeated, with two people becoming unemployed, one of whom is unemployed for one month and the other for two months. In every month thereafter, the process repeats.  



New spells: One of the two spells that began in February would last more than one month, although a survey taken in that month would not provide that information. Completed spells: Two of the three people showing up as unemployed in February were in spells that would last more than one month. (One was already in the second month and one had just begun a two-month spell.) Equivalently, two-thirds of the person-months (a unit of one month’s activity by one person) of unemployment in that month resulted from people in long spells. That is the measure emphasized in this paper. Spells in progress: A survey taken in February, for example, would show that three people were unemployed, only one of whom had been unemployed for more than one month.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

34

Congressional Budget Office 





Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.



SIPP is a large, nationally representative longitudinal survey that is conducted by the Census Bureau. Every four months for several years, respondents are asked detailed questions about the activities and sources of income for members of their household, including which weeks they were employed or unemployed. For this analysis of SIPP data, CBO classified respondents as unemployed in weeks in which they were not working, were available to work, and either spent time looking for work or were waiting to be recalled from a temporary layoff. A spell of unemployment could begin or end if, in adjacent weeks, a respondent was either employed or not in the labor force (that is, neither employed nor unemployed). The analysis answers two questions about long-term unemployment, both based on lengths of unemployment spells recorded during the 2001–2003 period: How many of the spells of unemployment that began during that period lasted at least 27 consecutive weeks? And what portion of all unemployment that was experienced during that period was accounted for by those long-term spells? Those questions differ from the ones that can be answered using estimates of unemployment duration reported each month by the BLS and cited in the previous section (see Box 1). By focusing on long-term unemployment, this analysis does not directly address related issues concerning non- participation in the labor market. Several studies have documented an increase in the percentage of the nonelderly adult male population who withdraw from the labor force for long periods of time. Although those people are not counted as unemployed (because they are not actively seeking work), their nonemployment may result in serious problems for them as well as for the overall economy.14 One source of that decrease in participation in the labor market appears to be an increase in the receipt of disability benefits.15 Most of the analysis reported in this paper is limited to respondents who were at least age 25 when the 2001 SIPP panel began and who worked in at least one week of the three-year period. The age restriction was imposed in order to focus on spells of unemployment that were not related to students seeking seasonal employment or transitioning from school to work. The employment restriction was imposed to screen out reported unemployment spells by a small number of people who may not really have been seeking employment, as was suggested by their lack of employment at any time in three years. The SIPP database and CBO s analysis of it have several limitations, none of which are likely to alter the qualitative findings of the analysis (see Appendix A). 

Even though the 2001–2003 period was a time of considerable upheaval in the labor market—including a recession and the aftermath of the September 11, 2001, terrorist attacks—about 75 percent of the people age 25 and older who worked during that period did not experience any unemployment. That is, they either were employed every week or, when they were not working, were not looking for a job. But some people among the 25 percent who did become unemployed during those three years were unemployed for substantial amounts of time. (The key results concerning the extent of unemployment concentration were found for the late 1990s as well, except that, in that period of falling unemployment, the duration of unemployment spells was generally shorter. See Appendix B for details.)

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Long-Term Unemployment

35

Source: Congressional Budget Office based on data from the 2001 panel of the Survey of Income and Program Participation. Notes: This figure is based on reports by respondents who began about 10,800 spells of unemployment during a three-year survey period that started in late 2000 or early 2001 and ended in late 2003. About 1 percent of those spells lasted more than 52 weeks; they are not depicted in the figures but are included in the calculations reported in the text. The relatively large percentages at weeks 17, 18, 34, and 35 are almost certainly the result of a shortcoming in SIPP responses known as ―seam bias‖ (see Appendix A). Figure 2. Duration of Unemployment Spells, 2001 to 2003

Estimates of Unemployment Spells During the three years covered by the 2001 SIPP panel, workers who were at least age 25 when the panel began experienced roughly 62 million spells of unemployment. That number is based on the nearly 11,000 unemployment spells recorded among participants in the survey and is adjusted using population weights provided by the Census Bureau. Although that large

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

36

Congressional Budget Office

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

a number of spells over just three years may seem high, it is consistent with data from a separate survey (see Appendix A).16

Duration A large portion of those nearly 11,000 unemployment spells ended quickly (see the upper panel of Figure 2). In particular, three-in-ten of the spells lasted only one week, and another one-in-ten was over in two weeks.17 CBO used information on the duration of each spell to estimate the cumulative distribution of spell completions—that is, the percentage of unemployment spells that had been completed within a specified time. About half of the unemployment spells were over in less than five weeks, as many of the newly unemployed found a job fairly quickly or left the labor force; more than 90 percent of the spells had ended by the twenty-sixth week, leaving only 8 percent lasting 27 weeks or more (see the lower panel of Figure 2). Even though only a small portion of the spells were long term, however, they accounted for a large portion of the total amount of unemployment recorded during the 2001–2003 period. CBO estimates that the 8 percent of spells that were long term accounted for 37 percent of all weeks of unemployment (see Table 2). By contrast, even though about half of all spells ended in less than five weeks, they accounted for only 10 percent of all weeks of unemployment. Those estimates mean that, on average, at any given point during the three-year period, 37 percent of the unemployed adults were in the midst of spells that would last 27 weeks or more. That number is much larger than what is found in the duration statistics reported each month by the BLS because those statistics reflect only the length of the unemployment spells in progress, not the completed spells.18 That finding is important because it underscores the extent to which the losses from unemployment—both to the individuals who become unemployed and to the overall economy, in terms of lost output—may be much more heavily concentrated among the longterm unemployed than is suggested by the statistics on the duration of spells in progress reported each month. By definition, longer spells must account for a disproportionate share of all weeks of unemployment. What is noteworthy is that such a high percentage of spells ended quickly and that the remainder lasted so long. The average spell duration of 9.3 weeks could have resulted from a constant exit rate of about 11 percent per week (rather than a much higher initial exit rate followed by a lower rate after the first few weeks). Had everyone had the same probability of exiting unemployment each week, only 5 percent of the spells would have lasted 27-plus weeks and those long-term spells would have accounted for only about 19 percent of all weeks of unemployment—about half their actual share. The fact that some people exit unemployment quickly and others take much longer stems from multiple causes. Some of the differences in spell duration among workers probably result from either the circumstances that led to their unemployment or the workers’ characteristics. For example, those who exit quickly include workers who become unemployed because their employers temporarily lay them off and then recall them or because they are in industries, such as construction, in which workers frequently change jobs with short intervening spells of unemployment. Likewise, some people enter the labor force to test the market and quickly leave. Those whose unemployment spells are much longer include job seekers who have characteristics that make them less desirable to potential

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

37

Long-Term Unemployment

employers (such as the lack of a high school diploma) or who are searching for work in labor markets with relatively few job openings. Table 2. Unemployment Spells of People Age 25 and Over, by Duration, 2001 to 2003

100 9.3

Less than 5 Weeks 53 1.8

Duration of Spells 5 to 26 Weeks 39 12.6

27 or More Weeks 8 42

100

10

53

37

Total Percentage of Spells Average Duration of Spells (Weeks) Percentage of Total Weeks of Unemployment

Source: Congressional Budget Office based on data from the 2001 panel of the Survey of Income and Program Participation. Note: The sample on which this table is based consists of about 10,800 unemployment spells that began during a three-year survey period that started in late 2000 or early 2001 and ended in late 2003. The respondents were age 25 or older at the beginning of the period. About 69 percent of those spells ended with the unemployed person becoming employed; 26 percent ended with the person leaving the labor force (that is, no longer looking for work); and the remainder ended with the person still unemployed in the final week of the survey period.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Table 3. Unemployment Spells of People Age 25 and Over, by Employment Status Immediately before and after the Spells, 2001 to 2003 Employment Status

Percentage of All Spells

Percentage of Spells Within Category

Employment Before Spell Spell ended with employment Spell ended with exit from labor force Spell ongoing Not in Labor Force Before Spell Spell ended with employment Spell ended with exit from labor force Spell ongoing All

72 57 11 5 28 12 15 2 100

100 78 15 6 100 42 52 5 n.a.

Source: Congressional Budget Office based on data from the 2001 panel of the Survey of Income and Program Participation. Notes: The sample on which this table is based consists of about 10,800 unemployment spells that began during a three-year survey period that started in late 2000 or early 2001 and ended in late 2003. The respondents were age 25 or older at the beginning of the period. n.a. = not applicable.

An additional explanation is that job seekers attractiveness to potential employers declines the longer they are unemployed.19 For example, in the absence of better information, employers could interpret the fact that someone has already been unemployed for over a month as an indication that there is something wrong with them that other potential employers had spotted. Finding a job could take workers longer if their skills or informal job networks have deteriorated over time or if they are searching less intensively because of discouragement. 

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

38

Congressional Budget Office

Although it is beyond the scope of this paper to estimate a causal model of why some people experience long-term spells of unemployment and others do not, the statistics provided in the remainder of the paper provide information about the circumstances leading to individuals becoming unemployed and the characteristics of the job seekers who remain unemployed for long periods.

Employment Status Not all spells of unemployment begin with a worker leaving a job, nor do spells always end because the unemployed person found a job. Almost 30 percent of the spells begun during the 2001–2003 period started after the individual had entered or reentered the labor force rather than after he or she had been working in the week before the spell began (see Table 3). Table 4. Unemployment Spells of People Age 25 and Over, by Duration and Employment Status Immediately before and after the Spells, 2001 to 2003

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Employment Status

Total

Less Than 5 Weeks

Duration of Spells 27 or 5 to More 26Weeks Weeks

Average Duration (Weeks)

Employed Before Spell Spell ended with employment 100 60 34 6 7.3 Spell ended with exit from labor 100 34 52 14 13.5 force Not in Labor Force Before Spell Spell ended with employment 100 38 53 8 10.9 Spell ended with exit from labor 100 64 30 7 7.8 force All 100 53 39 8 9.3 Source: Congressional Budget Office based on data from the 2001 panel of the Survey of Income and Program Participation. Note: The sample on which this table is based consists of about 10,800 unemployment spells that began during a three-year survey period that started in late 2000 or early 2001 and ended in late 2003. The respondents were age 25 or older at the beginning of the period. About 600 of those spells had not yet ended at the conclusion of the survey period; the duration of those spells up to the end of the survey period are included in the ―All‖ row but not shown separately.

Not surprisingly, unemployed people who recently had jobs were much more likely to end their unemployment spell with a job than were people who had entered (or reentered) the labor force. About 80 percent of the recently employed people had taken a job, compared with only about 40 percent of the labor force entrants and reentrants. That large difference could reflect the greater labor force attachment or employability of the people who recently worked. (Some of the difference could also reflect the sometimes thin line between being unemployed and being outside the labor force.)

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Long-Term Unemployment

39

The unemployment spells that ended with either an individual who had been working going back to work or with someone who had been out of the labor force withdrawing again were generally shorter than the other spells (see Table 4). The spells that began with a worker losing or leaving a job and ended with that worker leaving the labor force were the spells most likely to last 27 weeks or more. Most of the long-term unemployment spells in which the job seekers had worked immediately before becoming unemployed (rather than having been out of the labor force) appear to have resulted from job loss, rather than from the workers quitting their jobs. That finding is based on limited information provided by survey respondents who had not worked at all during the entire four- month period since they had previously been interviewed.20 Comparable information about the circumstances leading to shorter-term unemployment spells is not available from the SIPP. Much of the concern about unemployment in recent years has been focused on ―displaced‖ workers—those who lost their job permanently because their employer closed or moved, there was insufficient work for them to do, or their position was abolished. Every other year (in January or February) a set of questions about displaced workers is added to the Current Population Survey.21 The most recent displaced-worker survey took place in January 2006, asking respondents about jobs they lost in the preceding three years. Estimates from that survey provide additional evidence that unemployment is heavily concentrated among a relatively small group of job seekers (see Box 2).

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

BOX 2. THE DURATION OF JOBLESSNESS AMONG DISPLACED WORKERS In the January 2006 supplement to the Current Population Survey, displaced workers were defined as those who had become permanently separated from their job because their employer closed or moved, there was insufficient work for them to do, or their position was abolished. When displaced workers who subsequently found a new job were asked, ―After that job ended, how many weeks went by before [you] started working again at another job?,‖ the responses indicated that even though some displaced workers found another job quickly, many experienced a lengthy period of joblessness. (The term ―joblessness,‖ rather than ―unemployment,‖ is used here because the responses do not distinguish between periods in which the respondents were actively seeking work and periods in which they were not.) One-quarter of the reemployed workers who had lost their job in 2003 were jobless for at least 27 weeks. They accounted for 70 percent of all person-weeks (a unit of one week’s activity by one person) of joblessness incurred by reemployed workers (see the table, below). Moreover, the workers who had been jobless for the longest time were most likely to incur a large reduction in weekly earnings when they did return to work. Overall, about one-quarter of the reemployed workers who had lost a full-time job and were subsequently working in a full-time job had weekly earnings that were at least 20 percent below those of their old job. But about two-fifths of the workers who had been jobless for 27 or more weeks experienced that large a loss.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

40

Congressional Budget Office Joblessness and Reemployment of Displaced Workers, 2003 Duration of Joblessness (Weeks) Zero 1 to 4 5 to 26 27 or More Total

Percentage of Reemployed Displaced Workers 10 25 40 25 100

Average Duration of Joblessness (Weeks) Zero 2.5 14.5 57.2 20.9

Percentage of Person-Weeks of Joblessness Zero 3 27 70 100

Source: Congressional Budget Office based on data from the January 2006 Current Population Survey.

In sharp contrast, the remaining one-quarter of the unemployed people incurred a total of at least 27 weeks of unemployment during the entire period, accounting for almost two-thirds of all weeks of unemployment. Some of them had a number of relatively short spells that totaled up to 27 or more weeks. They accounted for 11 percent of all the people who had any unemployment and 20 percent of all weeks of unemployment Table 5. Total Unemployment Experience of People Age 25 and Over, 2001 to 2003

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Group

Percentage of Total Group

Average Number of Weeks Unemployed 20.4

Percentage of Total Weeks of Unemployment

Unemployed at Any Time During the 100 100 Three-Year Period Unemployed for up to 26 Weeks During 73 10.1 36 the Three-Year Period Unemployed for 27 or More Weeks 27 48.7 64 During the Three-Year Period Including a long-term spell 16 55.5 44 Not including a long-term spell 11 38.2 20 Source: Congressional Budget Office based on data from the 2001 panel of the Survey of Income and Program Participation. Note: The estimates presented in this table are for people who were at least age 25 at the beginning of a three-year survey period that started in late 2000 or early 2001 and ended in late 2003.

Particularly striking is the group whose members had experienced at least one of the long-term unemployment spells described in the preceding section. They accounted for 16 percent of all the people who experienced any unemployment (about 5 million of the 32 million people). Counting all of their weeks of unemployment—that is, the weeks from the shorter spells, as well as the longterm spell—they were responsible for 44 percent of all weeks of unemployment recorded during the three-year period. The next section of the paper takes a closer look at the characteristics of those people.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Long-Term Unemployment

41

CHARACTERISTICS OF THE LONG-TERM UNEMPLOYED AND SOURCES OF SUPPORT The preceding section showed that even though a small percentage of the adults who became unemployed during a recent period remained unemployed for more than half a year, they accounted for a large share of the number of weeks of work lost because of unemployment. This section briefly describes the people who experienced such long-term unemployment and their sources of support, including health insurance coverage, while they were not working. That information may be relevant to policy- makers as they develop and consider proposals for identifying unemployed workers most likely to take a long time to find a job, helping them find new jobs faster, or providing them with additional support while they remain unemployed.

Characteristics

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Unemployment rates are persistently higher among some groups than others. For example, the unemployment rate of adults who did not graduate from high school has long been more than double that of adults with a college degree. One group can have a higher unemployment rate than another group because its members become unemployed more frequently, remain unemployed longer, or both. Analysis of the data from the 2001 panel of the SIPP provides a way of identifying those differences.

Education As a person’s level of educational attainment increases, his or her likelihood of experiencing unemployment decreases. That negative relationship is clearly shown in the data. Among labor force participants who were age 25 or older in 2001, 37 percent of the individuals who had not graduated from high school were unemployed at some point, which was double the incidence of unemployment among college graduates (see Table 6). Moreover, among the people who experienced any unemployment, those with fewer years of education were much more likely to have experienced many weeks of unemployment, including a long-term spell. Marital Status Men and women who were married were less likely to have experienced any unemployment and, if they did, were less likely to have remained unemployed for a long time than were unmarried men and women. Only about one-in-five of the married men and women were ever unemployed during the three-year period of the survey, compared with almost onein-three of the men and women who were not married. Moreover, married men and women who experienced any unemployment were less likely to have had a long-term spell than were unmarried men and women. Married women had an especially small likelihood of having a long-term unemployment spell.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

42

Congressional Budget Office Table 6. Unemployment Experience of Labor Force Participants Age 25 and Over, by Characteristic, 2001 to 2003

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Characteristic

Percentage of Labor Force Participants Who Experienced Any Unemployment

Percentage of Unemployed People With At Least One Long-Term Spell

Educational Attainment Less than high school 37 20 High school graduate 27 17 Some postsecondary 25 14 College graduate 18 15 Sex Male 24 17 Female 26 15 Marital Status Married male 20 16 Unmarried male 31 20 Married female 23 12 Unmarried female 31 19 Race/Ethnicity Non-Hispanic White 23 14 Non-Hispanic Black 33 21 Hispanic 31 19 All 25 16 Source: Congressional Budget Office based on data from the 2001 panel of the Survey of Income and Program Participation. Note: The estimates presented in this table are for people who were at least age 25 at the beginning of a three-year survey period that started in late 2000 or early 2001 and ended in late 2003.

Ethnicity Black and Hispanic labor force participants were much more likely to have been unemployed than were non-Hispanic whites. Moreover, higher percentages of them had at least one long-term spell of unemployment. The preceding differences among groups in their likelihood of experiencing any unemployment or at least one long-term unemployment spell result in some striking differences between the characteristics of the long-term unemployed and the characteristics of the overall adult labor force (see Table 7). In particular, although individuals who did not graduate from high school constituted only 10 percent of the adult labor force, they accounted for 16 percent of the people who experienced any unemployment and 20 percent of the people who had at least one long-term unemployment spell. In contrast, there were almost three times as many college graduates in the labor force, yet they accounted for the same share of individuals with long-term unemployment.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

43

Long-Term Unemployment Table 7. Distribution of Labor Force Participants and Unemployed People Age 25 and Over, by Characteristic, 2001 to 2003

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Characteristic

Percentage of Labor Force

Percentage of Participants with Any Unemployment

Percentage of Participants with at Least One Long-Term Spell

Educational Attainment Less than high 10 16 20 school High school 30 33 34 graduate Some 31 31 27 postsecondary College graduate 29 21 20 Sex Male 51 49 52 Female 49 51 48 Marital Status Married male 36 29 28 Unmarried male 15 19 24 Married female 31 29 22 Unmarried female 18 23 26 Race/Ethnicity Non-Hispanic 74 67 59 White Non-Hispanic 11 14 19 Black Hispanic 10 13 15 All 100 100 100 Source: Congressional Budget Office based on data from the 2001 panel of the Survey of Income and Program Participation. Note: The estimates presented in this table are for people who were at least age 25 at the beginning of a three-year survey period that started in late 2000 or early 2001 and ended in late 2003. Race/ethnicity categories do not sum to 100 percent because of missing information and excluded categories.

Sources of Income Most workers in the midst of a long-term spell of unemployment in the 2001–2003 period had other sources of income. In the fourth month of their spell, about half of the unemployed workers were living in families in which at least one other person worked and about half were receiving unemployment insurance benefits (see Table 8).24 In addition, some of the unemployed workers who did not receive UI benefits in that month did receive benefits in other months. Separate tabulations (not shown in the table) indicate that more than 60 percent of the people who had long-term unemployment spells received UI benefits in at least one

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

44

Congressional Budget Office

month of their spell; that group included almost three-quarters of the people who had been employed immediately before their unemployment spell began. Much of the interest among policymakers in long-term unemployment has focused on unemployed workers who lost their jobs, rather than workers who voluntarily left them or recently entered (or reentered) the labor force. A previous CBO paper examined the role that UI benefits play in helping unemployed workers maintain their family income during periods without earnings and what happens to those workers and their families after their entitlement to UI benefits ends.25 That paper did not include people who were in long-term unemploy ment spells who did not receive UI benefits. As found in the previous paper, workers in the midst of a long-term spell of unemployment generally incurred a substantial reduction in their family income, compared with their income in the month before the spell began— both while they were unemployed and after they returned to work or left the labor force (see Figure 3). Following the end of their unemployment spell, those people who returned to work were, on average, earning roughly 20 percent less per month than they had been before the spell began.26 Consequently, their average family income was about 10 percent lower. For people who had not returned to work, their monthly income was, of course, much lower. Table 8. Major Sources of Family Income During a Long-Term Unemployment Spell, 2001 to 2003

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Source of Income

Percentage Receiving Income from Source

Average Monthly Amount (Dollars) For Recipients of For All Source Individuals 3,180 1,630 1,010 510

Earnings of Relatives 51 Unemployment Insurance 51 Benefits All Sources 91 2,910 2,640 Source: Congressional Budget Office based on data from the 2001 panel of the Survey of Income and Program Participation. Note: The sample on which this table is based consists of about 850 unemployment spells that began during a three-year survey period that started in late 2000 or early 2001 and ended in late 2003. Each spell lasted at least 27 consecutive weeks. The respondents were age 25 or older at the beginning of the period. Income reported in this table is for the fourth full month following the month in which the spell began.

Sources of Health Insurance Coverage The majority of people under age 65 obtain their health insurance through either their own employer or their spouse’s employer. For that reason, many policymakers have been concerned about unemployment jeopardizing health insurance coverage and have enacted legislation designed to help unemployed workers retain their coverage. The Consolidated Omnibus Budget Reconciliation Act of 1986 (COBRA) generally requires firms with 20 or more employees to continue offering coverage to certain terminated workers for a maximum

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

45

Long-Term Unemployment

of 18 months.27 Firms may charge their former employees the full group premium for that coverage plus 2 percent for administrative costs. The relatively high cost of those premiums, compared with the smaller amount workers generally pay while employed, probably deters many unemployed workers from participating, however.28 The people who experienced long-term unemployment between 2001 and 2003 were less likely to have had health insurance coverage than were other adults even before their longterm spell began. In the month prior to the beginning of their spell, 30 percent of those people lacked health insurance—roughly double the rate for nonelderly adults in the overall population. Most of the people with insurance obtained it through an employer (see Table 9). The share of people without health insurance rose to about 40 percent during their spell of unemployment, as many no longer received coverage through their employer; increases in other coverage offset only a small portion of the lost employer-sponsored coverage. In the month after their spell ended, the percentage of people without insurance continued to be higher than it had been before their spell began. Table 9. Sources of Health Insurance Coverage before, during, and after a Long-Term Spell of Unemployment, 2001 to 2003

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

(Percent) Source of Coverage

Month before Spell Began

During Spell

Month after Spell Ended

Employer Other Coverage Uninsured

54 15 30

40 18 41

42 20 37

Source: Congressional Budget Office based on data from the 2001 panel of the Survey of Income and Program Participation. Note: The sample on which this table is based consists of about 850 unemployment spells that began during a three-year survey period that started in late 2000 or early 2001 and ended in late 2003. Each spell lasted at least 27 consecutive weeks. The respondents were age 25 or older at the beginning of the period. Health insurance coverage during spell is for the fourth full month following the month in which the spell began. In a few cases, the health insurance coverage was not known; those spells are not included in the percentage distributions in this table.

Employer-sponsored health insurance was the main source of coverage for people in long-term unemployment spells who had been working immediately before they became unemployed (see Figure 4). Before becoming unemployed, about 25 percent of the people with jobs lacked health insurance.29 In the midst of their unemployment, nearly 40 percent were uninsured, as the percentage with employer-sponsored insurance fell. One month after their unemployment spell ended, coverage increased for the people who went back to work but remained lower for those who were still jobless. (The federal government also funds training and education programs that can help workers acquire new skills, thereby increasing participants employment opportunities and earnings.30 Although those programs can reduce the likelihood that their participants will incur future spells of long-term unemployment, that is usually not their immediate purpose.) 

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

46

Congressional Budget Office

Source: Congressional Budget Office based on data from the 2001 panel of the Survey of Income and Program Participation. Notes: This figure is based on reports by respondents who began long-term spells of unemployment during a three-year survey period that started in late 2000 or early 2001 and ended in late 2003 and who had earnings in the month before the spell began. UI = unemployment insurance. Figure 3. Average Monthly Family Income Before, During, and After a Long-Term Unemployment Spell, 2001 to 2003

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Long-Term Unemployment

47

Unemployment Insurance and Related Programs The UI program is by far the largest public program offering assistance to unemployed workers. Authorized by the Social Security Act of 1935, the program is intended to ease labor-market transitions by providing temporary income support for workers who lose their job and are looking for work. Because the amount of benefits provided by the program automatically increases in periods of high unemployment, the UI program also can help to dampen cyclical fluctuations in aggregate income. The federal government pays to administer the program, funds benefits for certain groups of unemployed workers, and provides general guidelines and some restrictions on how states may operate their UI programs. Each state sets eligibility requirements for UI benefits, determines the duration and amount of regular benefits, and specifies the payroll taxes that fund those programs. The outlays and revenues of the state programs are recorded in the federal budget. CBO estimates that outlays for UI benefits will total about $35 billion in fiscal year 2008. Nearly all of that amount will go to unemployed workers for benefits under the regular program, which generally limits the duration of benefits to no more than 26 weeks over a one-year period.31 A second level of UI benefits, jointly funded by the federal and state governments, is available in states with unemployment rates that exceed certain thresholds; CBO estimates that few workers will receive benefits through that program in 2008. Although the UI program provides partial earnings replacement to unemployed workers while they are looking for new jobs or awaiting recall from a temporary layoff, it is not designed to avert long-term unemployment or to help people after they are no longer eligible for benefits. In recent years, more than one-third of the unemployed workers who received UI benefits exhausted their entitlement to those benefits. (UI benefits are not available to unemployed individuals who are new entrants to the labor force or to previously employed workers who did not have a specified amount of employment and earnings in a recent year.32) The availability of cash benefits while unemployed can increase a recipient’s duration of unemployment, although the magnitude of that effect is uncertain.33 To ascertain whether UI recipients are actually seeking work—and to aid them in their job search—recipients (other than those who have been temporarily laid off) are generally required to register with public employment offices and to go on job interviews. Those offices provide free listings of job openings and job seekers. In addition, the offices may offer career counseling, job-search assistance, and referrals to job training and other employment-related services. Job seekers need not be UI recipients (or even have previously been employed) to use those services; the majority of participants are not. A small fraction of workers who lose their job are also eligible for benefits under the Trade Adjustment Assistance (TAA) program. That program is designed to help workers who lose their job as a result of increased imports or a shift in production out of the United States. Cash benefits are available to eligible workers who receive training, but only after they have exhausted their UI benefits. Legislation enacted in 2002 expanded eligibility for the program and provided displaced workers with a refundable tax credit for a portion of their health insurance premiums. (The 2002 amendments also provided a subsidy, described below, to eligible workers who accept jobs that pay less than the ones that they lost.)

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

48

Congressional Budget Office

Source: Congressional Budget Office based on data from the 2001 panel of the Survey of Income and Program Participation. Note: This figure is based on reports by respondents who began long-term spells of unemployment during a three-year survey period that started in late 2000 or early 2001 and ended in late 2003 and who had earnings in the month before the spell began. Figure 4. Health Insurance Coverage before, during, and after a Long-Term Unemployment Spell, 2001 to 2003

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Long-Term Unemployment

49

Worker Profiling and Reemployment Services Legislation enacted in 1993 requires states to screen new UI claimants, identify the ones most likely to exhaust their benefits, and refer them to reemployment services. If those profiled UI recipients do not participate in the services, they can be disqualified for benefits. The federal Department of Labor assisted states in developing statistical tools used to identify which claimants were most likely to exhaust their benefits. One widely cited study of profiling, as it was implemented in Kentucky, found a reduction in UI receipt resulting from the participation requirement itself.34 That is, some people who were called in to participate in reemployment services appear to have decided not to continue receiving UI benefits rather than participate. But whether profiling generally has been successful in providing services that have led to participants finding new jobs sooner than they otherwise would have is not yet known.35

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Reemployment Bonuses A second approach to reducing long-term unemployment that has received considerable attention is to provide financial inducements for newly unemployed workers to search for work more intensively or to accept job offers that they might otherwise have rejected. That approach was tested in several experiments conducted in the 1980s and 1990s in which UI recipients were provided with bonuses if they found new jobs within a specified period. Several evaluations found that such financial inducements did result in shorter durations of UI receipt. They also found that when those bonuses were combined with job-search assistance, their impact was larger.36 A subsequent analysis suggested that targeting eligibility for reemployment bonuses toward UI claimants with an above-average likelihood of exhausting their entitlement to UI benefits would further add to their effectiveness.37 Wage Insurance A third approach that has been available on a very limited basis is known as ―wage insurance.‖ Wage insurance subsidizes a fraction of the difference between the wage a worker earns in a new job and the wage he or she was earning in the old job for a limited time. The premise on which wage insurance proposals are based is that many of the workers who lose their job, especially workers who have been with the same employer for many years, are unlikely to find a new job that pays as well as the one that they lost. (Estimates of earnings losses discussed earlier in this paper support that premise.) Even though unemployment insurance provides workers with temporary income support while they search for a new job, it does not compensate them for the possibly permanent reduction in their earnings that resulted from the loss of their previous job. Wage insurance proposals aim to induce those workers to accept lower-paying jobs— which some workers might have been reluctant to take, perhaps in the hope that they would be able to get their old job back or that they would eventually be able to find a job at a similar pay level—as well as to compensate them for their loss in earnings. One wage insurance proposal would provide eligible displaced workers who found a lower-paying job within 26 weeks of displacement with half of the difference between the weekly earnings in their new and old jobs for up to two years.38 Unlike the experiments with bonuses, in which all unemployed workers who took a new job within a specified period would receive the money, job seekers under the wage insurance proposal would only qualify

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

50

Congressional Budget Office

for subsidies if their new job paid less than their old one. In that way, the subsidies would, in effect, compensate the unemployed workers for a portion of the financial loss they incurred when their old jobs were abolished. A version of wage insurance was tested in five cities in Canada in the mid-1990s. In that experiment, unemployment insurance claimants who found a new lower- paying full-time job within six months could receive an earnings supplement of 75 percent of their earnings loss (up to a cap) for up to two years. Researchers found that the earnings supplement appeared to have little impact on how quickly participants found new jobs. Its major effect was to partially compensate workers for the wage losses that they incurred.39 An amendment to the Trade Adjustment Assistance program enacted in 2002 offers a version of that plan for certain workers age 50 or older who are certified as eligible for TAA benefits. If those workers accept a new job that pays less than the one that they lost, the federal government will pay them half of the difference in wages for up to two years.40 Thus far, few people have taken the subsidy and no information is yet available about its effectiveness.41

APPENDIX.A: DATA AND METHODOLOGY Most of the estimates reported in this paper are based on data from the 2001 panel of the Survey of Income and Program Participation (SIPP). SIPP is a longitudinal survey of the population of the United States that has been conducted by the Bureau of the Census since the mid-1980s. Each panel comprises a nationally representative sample of households selected by the bureau and interviewed every four months for up to four years. The 2001 panel of the SIPP is the most recently completed panel available.42 That survey provided information about respondents activities during a period that began in one of the last three months of 2000 or in January 2001; the final month was 36 months later (that is, in September, October, November, or December 2003). Key findings were replicated using data from an earlier panel (see Appendix B). Responses to detailed questions about when each person worked or looked for work enabled census staff to determine each person s labor force status (employed, unemployed, or not in the labor force) week by week. The Congressional Budget Office (CBO) used those data to construct a string of weekly observations for each person who was age 16 or older when first interviewed. Demographic characteristics of respondents—such as their age, educational attainment, and marital status—were based on information provided in the initial interview. Census- provided longitudinal weights were applied to the data so that the respondents represented the overall U.S. civilian noninstitutionalized population.43 The analysis focused on respondents who were age 25 or older at the start of the survey. Of the approximately 23,000 respondents in that age group who were in the labor force, 5,600 experienced at least one spell of unemployment. In total, about 10,800 spells of unemployment were identified. That is, about one-quarter of the labor force participants during the 2001–2003 period experienced any unemployment; those who were unemployed had an average of almost two spells.44 Each spell was categorized by its length (number of consecutive weeks of unemployment), how it began (whether the respondent was employed or not in the labor force in the week immediately before the spell), and how it ended (whether

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.





Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Long-Term Unemployment

51

the respondent was employed or not in the labor force in the week immediately after the spell). For most of the analysis, the relevant unit is the spell, not the person. Thus, a person with two unemployment spells will show up twice in the data. The contribution of each new spell to the total number of person-weeks (a unit of one week’s activity by one person) of unemployment during the three-year observation period equals the length of that spell. (Because about 6 percent of the spells were still in progress in the final week of the three-year period, the length of those spells is understated.)

Limitations

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

The SIPP database and CBO’s analysis of it have several limitations, none of which are likely to alter the qualitative findings. Two limitations may be particularly important. First, the distinction between being unemployed and not being in the labor force is not always sharp. Some weeks in which respondents were not actively seeking a job might have been erroneously reported as weeks of unemployment, and some weeks in which respondents were looking for work might have been reported as weeks not in the labor force. Such errors could result in longer reported spells for some respondents and shorter spells for others. Second, changes in employment status are not always reported correctly. Some respondents, rather than accurately reporting the specific week in which a change occurred, evidently tell interviewers that they had no change in employment status during an entire four- month reference period and then report being in a different employment status at the beginning of the next four- month period. That ―seam‖ problem is evident in the relatively large numbers of unemployment spells reported to last exactly four or eight months.

Comparison with Estimates Based on the Current Population Survey As discussed in the main text, CBO estimates that workers who were age 25 or older at the start of the 2001 SIPP panel experienced about 62 million spells of unemployment over a three-year period (based on about 10,800 spells among respondents to the survey). That number of spells over a three-year period means that, on average, about 1.7 million people age 25 or older became unemployed each month, according to SIPP data. Two estimates derived from the monthly Current Population Survey (or CPS, a survey of about 60,000 households conducted by the Census Bureau for the Bureau of Labor Statistics) are consistent with at least that large a number. The first estimate is of the number of people who had been unemployed for less than five weeks. The Bureau of Labor Statistics (BLS) reported that during the 2001– 2003 period, an average of 1.7 million people age 25 or older had been unemployed for less than five weeks. Because that time frame roughly approximates one month, the measure provides an estimate of the average number of people who became unemployed each month. However, that CPSbased estimate would not include some very short-term spells that began and ended between the monthly interviews. Had those spells been included, the CPS would have indicated even more unemployment spells than were identified in the SIPP.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

52

Congressional Budget Office

A separate measure derived from the monthly CPS, though not regularly published, provides a further indication that the large number of unemployment spells estimated from the SIPP is credible. Because about three- quarters of the households in the CPS sample in one month are in the sample again in the following month, it is possible to link responses in pairs of successive months to produce ―gross flow‖ estimates. In principle, then, the CPS can be used to directly estimate monthly flows into and out of unemployment. In practice, however, a number of difficulties have prevented regular publication of those estimates. A particular concern has been that the net changes in employment and unemployment from one month to the next derived from the gross-flow data often do not match the net changes from the full CPS sample.45 Based on unpublished gross-flow data that were adjusted by BLS staff to be more consistent with the full CPS sample, a recent study estimated that in 2003 an average of 4.0 million people age 16 or older became unemployed each month, with about half of the inflows from employment and the rest from outside the labor force.46 Somewhat fewer people became unemployed in the two preceding years. The estimates provided in that study suggest that over the entire 2001–2003 period, roughly 130 million spells of unemployment occurred among people age 16 or older. That estimate is about 40 percent larger than the 92 million spells CBO estimates on the basis of the SIPP data (30 million spells for labor force participants who were ages 16 to 24 at the beginning of the panel and 62 million spells for participants age 25 or older). One reason that the gross-flow estimates are higher may be the tendency of classification errors in monthly series to magnify gross-flow estimates.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

APPENDIX B: LONG-TERM UNEMPLOYMENT IN THE LATE 1990S The estimates of the concentration of unemployment that were presented in the main body of this report were based on data from the 2001 panel of the Census Bureau’s Survey of Income and Program Participation (SIPP), which covers the 200 1–2003 period. In interpreting those estimates, one issue is whether the concentration of unemployment among a small group of long-term unemployed workers was the result of circumstances unique to that period. To examine the robustness of the estimates from the 2001 panel, the Congressional Budget Office compared the key findings with a similar analysis of data from the previous panel of the SIPP, which began in 1996 and ended three years later. During that period, the unemployment rate declined from about 5.5 percent (in early 1996) to 4.4 percent. By comparison, the respondents in the 2001 panel became unemployed during a period in which the unemployment rate was rising—from about 4.0 percent at the beginning of the period to nearly 6.0 percent in late 2003. The findings from the analyses reflect those differences: About 10 percent fewer adult workers experienced any unemployment in the earlier period (22 percent versus 25 percent), the average duration of the unemployment spells was shorter, and a much smaller percentage of spells lasted at least 27 weeks (see Table B-1).

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

53

Long-Term Unemployment

Table B-1. Comparison of Estimates of Unemployment Concentration Based on the 1996 and 2001 Panels of the Survey of Income and Program Participation 1996 Panel 22

2001 Panel 25

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Percentage of Workers Age 25 and Older Who Experienced Any Unemployment Average Duration of All Spells (Weeks) 7.4 9.3 Workers Who Experienced at Least One Spell of Unemployment That Lasted 11 16 at Least 27 Weeks, as a Percentage of Workers With Any Unemployment Percentage of Unemployment Spells That Lasted at Least 27 Weeks 4 8 Percentage of All Weeks of Unemployment Accounted for by Spells 23 37 That Lasted at Least 27 Weeks Percentage of All Weeks of Unemployment Accounted for by 34 44 Workers Who Had at Least One Spell of Unemployment That Lasted at Least 27 Weeks Source: Congressional Budget Office based on data from the 1996 and 2001 panels of the Survey of Income and Program Participation. Note: The estimates presented in this table are for people who were at least age 25 at the beginning of the survey period. The sample from the 1996 panel consists of people who provided sufficient information for nine consecutive four- month waves beginning in early 1996. The sample from the 2001 panel consists of respondents to nine consecutive four-month waves beginning in late 2000 or early 2001.

Nonetheless, the qualitative results concerning the concentration of unemployment are similar for both periods, which shows that the main findings reported in this paper were not unique to the 2001–2003 period. In the earlier period (like in the later period), the small percentage of unemployment spells that lasted at least 27 weeks accounted for a large percentage of all of the weeks of unemployment. Only 4 percent of the spells that began in 1996, 1997, and 1998 were long term, yet they accounted for almost one-quarter of total unemployment. Moreover, taking into account the additional unemployment some of those workers experienced from other spells, the workers with long-term spells incurred about onethird of all unemployment.

End Notes 1

2

3

The Bureau of Labor Statistics uses those criteria to determine whether respondents to the Current Population Survey should be classified as unemployed. In that survey (in which the Census Bureau collects data on the labor force every month), respondents are asked about their activities during the week preceding the interview, and respondents who are unemployed provide information about how long they have been unemployed thus far. Respondents who did any paid work during the reference week (or were self-employed or worked in a family business) are classified as employed. Those not counted as employed or unemployed are classified as not in the labor force. About 60,000 households participate in the survey. That estimate is based on the observation that, on average, 2.6 million people age 16 and over had been unemployed for less than five weeks, which is approximately one month. It understates the number of people who actually became unemployed to the extent that it misses people with very short-term spells who entered and left unemployment between the monthly surveys. See Congressional Budget Office, Trends in Earnings Variability Over the Past 20 Years (April 17, 2007); Peter Gottschalk and Robert Moffitt, ―Trends in the Transitory Variance of Earnings in the United States,‖

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

54

Congressional Budget Office

Economic Journal, vol. 112, no. 478 (2002), pp. C68–C73; and Karen E. Dyan, Douglas W. Elmendorf, and Daniel E. Sichel, ―The Evolution of Household Income Volatility‖ (draft, Brookings Institution, Washington, D.C., June 2007). 4 Estimates from the CPS indicate that almost half of the unemployment spells completed in 2003 ended with the individuals leaving the labor force rather than becoming employed. See Randy Ilg, ―Analyzing CPS Data Using Gross Flows,‖ Monthly Labor Review (September 2005), pp. 10–18. 5 A major redesign of the CPS in 1994 may have reduced the estimated amount of short-term unemployment somewhat, but it is unlikely to have resulted in a substantial change in the series depicted in Figure 1. See Katharine G. Abraham and Robert Shimer, ―Changes in Unemployment Duration and Labor Force Attachment,‖ in Alan B. Krueger and Robert M. Solow, eds., The Roaring Nineties: Can Full Employment Be Sustained? (New York: Russell Sage Foundation, 2001), pp. 367–420. 6 Statistics from a separate set of questions in the CPS show that 9.3 percent of all people who worked or looked for work in 2005 experienced any unemployment in that year—a smaller percentage than in most of the previous half century. For the latest statistics, see Department of Labor, Bureau of Labor Statistics, Work Experience of the Population in 2005, USDL 07-0 199 (February 9, 2007). A paper by Lawrence F. Katz and Alan B. Krueger that examined the strong performance of the labor market in the 1990s discussed the secular decline in the short-term unemployment rate and in the percentage of the labor force that had experienced any unemployment; see ―The High-Pressure U.S. Labor Market of the 1990s,‖ Brookings Papers on Economic Activity, vol. 1999, no. 1 (1999), pp. 1–87. 7 In the 12-month period ending in September 2007, 7.5 million workers began receiving unemployment insurance payments and 2.6 million recipients exhausted their benefits. Statistics on unemployment insurance are available from the Department of Labor at http://workforcesecurity.doleta.gov/unemploy/finance.asp. 8 Henry S. Farber, ―What Do We Know About Job Loss in the United States? Evidence from the Displaced Workers Survey, 1984–2004,‖ Economic Perspectives, Federal Reserve Bank of Chicago, vol. 29, no. 2 (2005), pp. 13– 28. See also Erica L. Groshen and Simon Potter, ―Has Structural Change Contributed to a Jobless Recovery?‖ Current Issues in Economics and Finance, Federal Reserve Bank of New York, vol. 9, no. 8 (August 2003). Women’s increased attachment to the labor force may also have contributed to the overall increase in longterm unemployment; see Abraham and Shimer, ―Changes in Unemployment Duration and Labor Force Attachment.‖ 9 Farber, ―What Do We Know About Job Loss in the United States?‖ 10 Congressional Budget Office, Family Income of Unemployment Insurance Recipients (March 2004). The people included in that analysis had received UI benefits for at least four consecutive months in 2001 or early 2002. 11 A survey of nonfarm employers conducted each month by the Bureau of Labor Statistics indicates a considerable amount of turnover. In 2006, for example, employers reported that they hired about 59 million workers and that about 55 million workers left. That is, more than 40 percent of their workforce turned over during the year. See Department of Labor, Bureau of Labor Statistics, Job Openings and Labor Turnover: January 2007, USDL 07- 0373 (March 13, 2007), Tables 11 to 14. 12 Each year, the Organization for Economic Cooperation and Development (OECD) publishes estimates of longterm unemployment for each of its 30 member countries. The OECD measures long-term unemployment as the percentage of unemployed people who have been unemployed for 12 months or more. In the most recent year for which data are available (2005 for most of the countries), the long-term unemployment rate was about 12 percent in the United States, compared with 33 percent for the OECD as a whole. Only 6 of the 30 countries in the OECD had a lower long-term unemployment rate than that of the United States. See OECD Factbook 2007 (Paris: Organization for Economic Cooperation and Development, 2007), pp. 140–141. 13 The Census Bureau staggered the start of the three-year period so that the first month for about one-quarter of the respondents was January 2001; for the other respondents, the initial month was in late 2000. The final month (also staggered) was September, October, November, or December 2003. 14 See Chinhui Juhn, Kevin M. Murphy, and Robert H. Topel, ―Current Unemployment, Historically Contemplated,‖ Brookings Papers on Economic Activity, vol. 2002, no. 1 (2002), pp. 79–116. 15 See David H. Autor and Mark G. Duggan, ―The Rise in the Disability Rolls and the Decline in Unemployment,‖ Quarterly Journal of Economics, vol. 118, no. 1 (February 2003), pp. 157–205. 16 A previous paper based on the 2001 SIPP panel estimated about half that number of spells (29.4 million) for people age 25 or older, but that estimate was based on a monthly variable that did not include many short unemployment spells. To be included in the previous paper’s estimate, an individual had to be jobless for an entire month and had to have spent at least one week laid off or looking for work. Based on the weekly variable, roughly half of the spells lasted less than one month. See Alfred O. Gottschalck, Dynamics of Economic Well-Being: Spells of Unemployment, 2001– 2003, Current Population Reports, P70-105 (Bureau of the Census, March 2006). 17 The apparent exceptions—the relatively large percentages of unemployment spells lasting 17, 18, 34, or 35 weeks—are almost certainly the result of a shortcoming in SIPP responses known as ―seam bias.‖ In brief, because respondents are interviewed every four months, changes in employment status tend to be disproportionately reported as having occurred between four-month (17- or 18-week) intervals.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Long-Term Unemployment

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

18

55

Based on data from the CPS, about 22 percent of the spells in progress by unemployed people age 25 and over in 2002 had lasted 27 or more weeks. Estimates based on the SIPP indicate that a smaller percentage of the spells in progress in that year had lasted that long. 19 In practice, it is extremely difficult to distinguish between true state dependence and unmeasured heterogeneity. See Theresa J. Devine and Nicholas M. Kiefer, Empirical Labor Economics: The Search Approach (New York: Oxford University Press, 1991). 20 Respondents were asked the main reason they had not worked. Some of the options probably reflected job loss (for example, respondents were unable to find work or had been laid off). Other options appeared to reflect voluntary decisions (for example, respondents were not interested in working, were retired, or were caring for children). 21 The displaced-worker survey began in 1984. See Department of Labor, Bureau of Labor Statistics, Worker Displacement, 2003– 2005, USDL 06-1454 (August 17, 2006). 22 As reported in the previous section, about 62 million distinct spells of unemployment were identified during the three-year period covered by the 2001 SIPP panel. 23 The sample for Table 5 includes spells of unemployment that were already under way in the first week of the survey. As a result, it includes more weeks of unemployment, especially long-term unemployment, than do the tables on spells. It might also include some people who never worked. 24 The information in this section refers to income reported for the fourth full month following the month in which the long-term spell began. 25 Congressional Budget Office, Family Income of Unemployment Insurance Recipients. 26 The lower monthly earnings do not necessarily indicate a lower wage rate. Another explanation could be that the individuals may not have worked as many hours during the month after their spell ended. 27 To qualify for coverage through COBRA, workers must have voluntarily or involuntarily terminated their employment for reasons other than gross misconduct. Additional information about eligibility and terms is provided by the Department of Labor at www.dol.gov/ebsa/faqs/faq_consumer_cobra.html. 28 A recent survey of employers estimated the average annual premiums for employer-sponsored health insurance in 2007. For family coverage, the premium is $12,106, with workers paying $3,281of that amount; for individual coverage, the premium is $4,479, with workers paying $694. See the Kaiser Family Foundation and Health Research and Educational Trust, Employer Health Benefits 2007 (Menlo Park, Calif.: The Henry J. Kaiser Family Foundation, 2007). 29 About 40 percent of the people who had not been working imme-diately before becoming unemployed lacked health insurance. 30 Those programs include training authorized by the Workforce Investment Act of 1998 and educational assistance through Pell grants and subsidized student loans. For an examination of federally funded training programs, see Christopher J. O’Leary, Robert A. Straits, and Stephen A. Wandner, eds., Job Training Policy in the United States (Kalamazoo, Mich.: W. E. Upjohn Institute for Employment Research, 2004). The impact of community college schooling on the subsequent earnings of displaced workers is examined by Louis Jacobson, Robert LaLonde, and Daniel G. Sullivan in ―Estimating the Returns to Community College Schooling for Displaced Workers,‖ Journal of Econometrics, vol. 125 (2005), pp. 271–304. 31 CBO estimates that about 8 million workers will begin receiving regular UI benefits in 2008. On average, they will receive $287 per week for 15.4 weeks. 32 Under each state’s UI laws, eligibility for benefits depends on a worker’s employment experience in a preceding ―base period,‖ which usually consists of four consecutive calendar quarters. To qualify for benefits, claimants must have earned a specified amount of wages, worked a certain number of weeks, or met some combination of earnings and employment requirements during the base period. Additional requirements concerning how workers became unemployed and their availability to work also apply. 33 See, for example, Stephen A. Woodbury and Murray A. Rubin, ―The Duration of Benefits,‖ and Paul T. Decker, ―Work Incentives and Disincentives,‖ in Christopher J. O’Leary and Stephen A. Wandner, eds., Unemployment Insurance in the United States: Analysis of Policy Issues (Kalamazoo, Mich.: W. E. Upjohn Institute for Employment Research, 1997), pp. 211–320. 34 Dan A. Black and others, ―Is the Threat of Reemployment Services More Effective than the Services Themselves? Evidence from Random Assignment in the UI System,‖ American Economic Review, vol. 93, no. 4 (September 2003), pp.1313–1327. 35 Government Accountability Office, Unemployment Insurance: More Guidance and Evaluation of WorkerProfiling Initiative Could Help Improve State Efforts, GAO-07-680 (June 2007). 36 For a summary of the experiments and their results, see Bruce D. Meyer, ―Lessons from the U.S. Unemployment Insurance Experiments,‖ Journal of Economic Literature, vol. 33, no. 1 (March 1995), pp. 91–131. For a recent review of other techniques for reducing the duration of UI receipt, see Christopher J. O’Leary, ―State UI Job Search Rules and Reemployment Services,‖ Monthly Labor Review (June 2006), pp. 27–37. 37 See Christopher J. O’Leary, Paul T. Decker, and Stephen A. Wandner, ―Cost-Effectiveness of Targeted Reemployment Bonuses,‖ Journal of Human Resources, vol. 40, no. 1 (Winter 2005), pp. 270–279.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

56

Congressional Budget Office

38

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

See Lori G. Kletzer and Robert E. Litan, A Prescription to Relieve Worker Anxiety, International Economics Policy Brief 01-02 (Washington, D.C.: Institute for International Economics, 2001). See also a recent proposal by Jeffrey R. Kling to substantially restructure the unemployment insurance system in Fundamental Restructuring of Unemployment Insurance: Wage-Loss Insurance and Temporary Earnings Replacement Accounts, Hamilton Project Discussion Paper 2006-05 (Washington, D.C.: Brookings Institution, September 2006). 39 Howard Bloom and others, ―Testing a Financial Incentive to Promote Re-employment Among Displaced Workers: The Canadian Earnings Supplement Project (ESP),‖ Journal of Policy Analysis and Management, vol. 20, no. 3 (2001), pp. 505–523. 40 Eligible workers age 50 or older who obtain new full-time employment at wages of less than $50,000 within 26 weeks of their separation may receive half of the difference between the old and new wages (up to $10,000 paid over a period of up to two years). 41 Government Accountability Office, Trade Adjustment Assistance: Changes to Funding Allocation and Eligibility Requirements Could Enhance States’ Ability to Provide Benefits and Services, GAO-07- 701 (May 2007). 42 The subsequent SIPP panel was begun in 2004. Thus far, the Census Bureau has released detailed data from the first four waves of interviews. (Each wave covers four months.) See www.bls. census.gov/ sipp_ftp. html#sipp04. 43 Almost half of the respondents who provided information in the initial interview did not provide data in at least one subsequent period and therefore were excluded from CBO’s analysis. The Census Bureau’s longitudinal weights are intended to adjust for attrition. 44 About 53 percent of those who were unemployed had one spell of unemployment, 23 percent had two spells, 10 percent had three spells, and most of the others had four, five, or six spells. 45 BLS staff and other researchers have been working on improving the gross-flow data. See, for example, Harley J. Frazis and others, ―Estimating Gross Flows Consistent with Stocks in the CPS,‖ Monthly Labor Review (September 2005), pp. 3–9. In 1979, a national commission recommended that priority be given to improving the gross-flow data so that they could be published on a regular basis; see Final Report of the National Commission on Employment and Unemployment Statistics, Counting the Labor Force (1979), p. 5. 46 Randy Ilg, ―Analyzing CPS Data Using Gross Flows,‖ Monthly Labor Review (September 2005), pp. 10–18.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

In: Labor and Employment Issues Editor: Nickolas H. Mullen

ISBN: 978-1-60741-286-1 © 2010 Nova Science Publishers, Inc.

Chapter 3

HOW SLOWER GROWTH IN THE LABOR FORCE COULD AFFECT THE RETURN ON CAPITAL 

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Congressional Budget Office The labor force in the United States is expected to grow at a significantly reduced rate in coming decades for several reasons: a long-term decline in fertility rates, the leveling off of a substantial rise in women’s labor force participation, and the aging and retirement of large numbers of people from the baby-boom generation. The Congressional Budget Office (CBO) projects that the growth of the labor force, which averaged 1.6 percent per year from 1950 to 2007, will slow to about half a percent per year over the next 20 years. Most mainstream economic models predict that the slowdown is likely to boost the ratio of capital to labor and thereby make the rate of return on capital—as reflected in the rate of interest on bonds or other borrowing, and the rate of return on stocks—lower than might otherwise be expected. Wages will be higher than would otherwise be the case for the same reason, although that effect will be much smaller than the increase in wages that is projected to result from productivity growth. Various economic models and simulations project that, other things being equal, the coming slowdown in labor force growth could subtract between 0.8 and 2.6 percentage points from the rate of return on capital within the next several decades; wages would rise by somewhat less than half the percentage decline in the rate of return on capital. Shifts in wages and rates of return on capital would have budgetary implications, both for the overall federal budget and for individual programs such as Social Security. The lower interest rates that would result from a slowdown in labor force growth would tend to reduce federal interest payments and slow the growth of debt relative to output. Lower interest rates would also reduce the interest income of the Social Security trust fund and increase the actuarial imbalance of the system (because future deficits in the system would be ―discounted‖ at a lower rate of interest), although the accompanying higher wages would tend to increase payroll tax revenues somewhat.1 The rate at which the labor supply grows is, however, only one factor influencing wages and rates of return on capital. Therefore, it is not clear whether, on balance, future wages and 

This is an edited, reformatted and augmented version of a Congressional Budget Office publication dated October 2009.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

58

Linda Levine

rates of return will be higher or lower than today’s. For example, budget deficits reduce national saving—the sum of private and government saving— and tend to crowd out investment in productive capital (office buildings, factories, machines, computers, and other equipment, which support future production and consumption). That crowding out implies a smaller capital stock and, other things being equal, lower wages and a higher rate of return on capital. If current laws remained unchanged, rising health care costs and the aging of the population would lead to substantial increases in budget deficits in the longer term, which would tend to offset the effect of slower growth in the labor force on the rate of return on capital. Increases in capital outflows—the net amount of savings that flows to other countries—would also tend to increase the domestic rate of return on capital. CBO’s 10-year economic projections illustrate some of those offsetting effects. They incorporate lower rates of return from a slackening in labor force growth, but that effect is outweighed by the impact of budget deficits and economic recovery, resulting in a projection that real (inflation-adjusted) interest rates will increase over the 10-year period. (CBO’s 10year projections also suggest that wages will rise, but that development stems from rising productivity rather than changes in the relative size of the capital stock.)

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

HOW THE GROWTH RATE OF THE LABOR FORCE AFFECTS RATES OF RETURN The rate of growth of labor supplied in the economy can affect wages and rates of return on capital by influencing the long-run ratio of the stock of productive capital to the quantity of labor.2 That effect stems from a simple mathematical relationship: If the economy begins in a stable long-run equilibrium, a slowdown in the growth of the supply of labor, together with the same rate of saving and investment, implies a rising ratio of capital to labor. Fewer additional productive capital goods would be required to preserve the same ratio because those incoming cohorts of workers would be smaller than they would be if the labor supply grew more quickly. Standard economic theory predicts that, all else being equal, a larger quantity of capital per worker translates into a lower marginal product of capital—the amount added to production by a small additional amount of capital—and a higher wage rate. As there would be more machines for each worker to utilize, the productivity of workers (and therefore wages) would go up, but the contribution to output provided by each additional piece of machinery (and therefore the marginal product of capital) would fall.3 Rates of return in the economy—such as the return on equity (in the form of dividends or capital gains) and the interest rate on bonds or other borrowing—are related, in varying degrees, to the contribution to output that is attributable to capital and therefore tend to move in the same direction as the marginal product of capital. Equities represent an ownership stake in the capital and other assets of a corporation, so when the marginal product of capital declines, the return on equities tends to fall as well. The return earned on equities is tied to long-term interest rates because people with savings can choose whether to invest in equities or interest-bearing assets such as bonds. Pricing in financial markets adjusts so that equities provide a higher return, on average, than relatively safe assets such as bonds because the return on equities is more uncertain. The difference between the return on equities and the

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Job Loss and Infrastructure Job Creation Spending during the Recession

59

interest rate paid on government bonds is referred to as the equity premium; that premium compensates investors for the additional risk associated with holding equities. Given a constant equity premium, the interest rate on government bonds would change by the same number of percentage points as the return on equities. Of course, the equity premium might not remain constant because changes in the return on equities could also alter the perceived riskiness of equities relative to bonds; in most cases, though, a change in the rate of return on equities would tend to move the interest rate in the same direction. The preceding discussion assumes that, as the growth rate of the labor supply slows, the rate of saving remains unchanged. In fact, saving rates could fall with lower growth in the labor force as the ratio of working-age people (who tend to save) to people of retirement age (who tend to spend from their savings) falls. However, models incorporating the effect of changes in demographics on saving rates suggest that the net effect would still be an increase in the relative size of the capital stock and a decrease in the rate of return on capital. In addition, a variety of evidence suggests that the elderly do not draw down their savings as much as some economic models imply, which suggests that the impact of an aging population on the rate of saving would not be large.4

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

HOW RATES OF RETURN AFFECT THE FEDERAL BUDGET Changes in rates of return (on both debt and equity instruments) and wages affect the federal budget. The most significant effect would occur because changes in interest rates have a direct effect on the amount of interest paid on federal debt, which will rise much more slowly for a given path of the primary (or noninterest) deficit if interest rates are lower. Similarly, the balances of the Social Security trust fund, or privately held savings accounts, will rise much more slowly given lower rates of return.5 Calculations involving ―present values,‖ which discount future flows of income and expenses by interest rates, are also quite sensitive to interest rates. That type of calculation includes the actuarial imbalance of the Social Security system. A lower interest rate means that future outlays and receipts are discounted by less and, therefore, figure more prominently in present-value calculations. In cases such as Social Security, where outlays are projected to rise more than revenues, a lower interest rate thus implies a larger actuarial imbalance.6 By contrast, the impact of changes in rates of return on primary (or noninterest) budget deficits would probably be small. Income earned on capital investments is taxed at a lower effective rate, on average, than income earned from participating in the labor force.7 Therefore, government revenues would be higher if a decline in the rate of return on capital shifted income shares so that the share of labor compensation rose and that of capital income fell. However, typical estimates suggest that changes in rates of return do not affect the balance between the amount of total income that is paid in returns on capital versus the amount that is paid in labor compensation. When an increase in the relative abundance of capital reduces the return on capital, there are two opposing effects on the fraction of overall income that is paid as capital income: Although each increment added to the capital stock generates less income, there are proportionately more units of capital earning income. (There are corresponding, but opposite, effects on the fraction of output paid as labor earnings.)

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

60

Linda Levine

Many economic models assume that the two effects exactly balance, leaving the total fraction of output paid as capital income (and labor earnings) unchanged.8

ESTIMATED IMPACT USING SIMPLE THEORETICAL MODELS One standard, simple model of economic growth assumes that a fixed fraction, s, of output is saved; that the population grows at rate n; that the stock of productive capital depreciates at the rate ö; and that a fraction, a, of output is received by owners of capital. Under those assumptions, the rate of return on capital (r) in the long run is

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

r = a(n + ö)/s - ö According to this equation, for every percentage point that the rate of the growth of the labor force falls—that rate is assumed to be the same as the rate of population growth, n, in the long run—the rate of return on capital falls by a/s percentage points. In the United States today, the share (a) of income that goes to capital is roughly 0.3, and the gross rate of saving, s (the rate of saving before subtracting the share needed to cover the depreciation of existing capital), is roughly 20 percent of output, or 0.2; so a/s is about 0.3/0.2, or 1.5. That result implies that a decline of 1 percent in the growth rate of labor—roughly the amount that the growth rate is projected to decline, relative to its postwar average, over the next few decades—would lead to a long-run decline in the rate of return on capital of about 1.5 percentage points, or about 20 percent of the pretax rate of return on capital. (The pretax rate, as measured by profit and interest income divided by the value of the capital stock, has averaged roughly 7 percent over the past 40 years.) The percentage increase in the wage rate would be approximately /(1- ), or 0.4 times as large as the decrease in the rate of return; or, compared with the level if the growth rate of labor stayed constant, roughly 8 percent.9 CBO uses a similar model to project the real rate of interest on Treasury notes over the 10-year budget horizon, although the projections incorporate productivity growth and, therefore, wages that rise over time even if the rate of return on capital is constant.10 However, the implied impact of changes in the growth rate of labor on the rate of return on capital and wages is the same as those described above. In the example described above, the rate of saving, s, is simply assumed. However, more complex models that incorporate explicit motives for saving imply similar results. For example, another class of economic models, overlapping generations models, incorporate successive generations that each save during their working years to finance retirement (and, in some cases, to provide insurance against unforeseen fluctuations in income). Those models also tend to predict that slower growth in labor inputs implies a lower rate of return on capital. For example, one textbook presentation of an overlapping generations model finds the steadystate rate of return is given by R = (/(1- ))(1+n)(2+) -  where the variables , n, and  are, as in the previous example, the share of income that goes to capital, the rate of growth of the labor force, and the rate of depreciation; and  is a Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Job Loss and Infrastructure Job Creation Spending during the Recession

61

measure of people’s impatience, or the rate at which they would trade consumption today for consumption tomorrow.11 In this model, too, a higher rate of population growth, n, which translates into faster growth of the supply of labor, implies a higher rate of return, r. Under reasonable assumptions of a = 0.3 and  = .03, a fall in the rate of growth of the labor force of 1 percentage point implies a decrease in the rate of return on capital of about 0.9 percentage points (or roughly 15 percent). The same assumptions imply a wage rate that is 6 percent higher than it would be without the slowdown in labor growth. The results are similar to those from the simpler example presented earlier because the principle is the same: With a slower rate of growth in the labor force, a stable rate of saving would imply a rising ratio of capital to labor. The decrease in the rate of return is smaller in this case, in part, because the drop in the rate of return reduces the saving rate, offsetting the initial effect somewhat. There is one type of economic model in which the rate of growth of the labor force does not affect the long-run rate of return. In this ―dynastic‖ model, there is inter-generational altruism: People’s well-being depends, in part, on the well-being of their children and succeeding descendants. In the strongest version of this model, what people care about is the sum of all their children’s well-being, and that well-being matters as much to them as their own. Under those assumptions, a slower rate of labor force growth—which indicates fewer children per parent, on average—would induce people to save less because there would be fewer heirs to receive bequests. In fact, that reduced saving would exactly equal the reduction in saving that is required to maintain the same ratio of capital to labor with lower population growth, leaving the rate of return unchanged. However, there are reasons to doubt that altruism affects the relationship between the growth rate of the labor force and rates of return as profoundly as this model implies. First, the strongest version of the altruism model has economic implications that do not appear to be supported by the data.12 Second, only the strongest forms of ―altruistic linkage‖ imply that rates of return are unrelated to the growth rate of labor. For example, if parents cared about the average welfare of each generation of their descendants, rather than the sum of all welfares in each generation, faster population growth would once again imply higher rates of return in much the same way that it does in the other theories described above. Moreover, if changes to children’s wellbeing at a given time were not as important to parents as changes to their own wellbeing at that same time, saving would not fully adjust to compensate for changes in the growth rate of labor, and rates of return would still be affected (albeit by a smaller amount than they would be otherwise).

POTENTIAL OFFSETTING EFFECTS FROM DEFICITS AND CAPITAL OUTFLOWS It is important to note that the implication that slower rates of labor force growth will result in lower rates of return and higher wages assumes that other factors are unchanged. However, when considering the coming decades, that ―ceteris paribus‖ (other things being equal) assumption may not be valid. In particular, under current law, federal spending is projected to grow relative to gross domestic product as health care costs continue to rise and increasing numbers of the baby-boom generation become eligible for Social Security and Medicare benefits. In the absence of policies that reduced spending or increased revenues,

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

62

Linda Levine

federal budget deficits would grow substantially. Increased deficits would reduce the amount of national saving available for investment, just as decreased private saving would. Lower rates of investment would reduce the size of the capital stock, which would in turn tend to increase rates of return and decrease wages, moving in the opposite direction of the effect of slowing labor force growth. (In the context of the simple model presented earlier, higher deficits reduce the rate of national saving, s, thereby increasing the rate of return.) The impact on rates of return would depend on the size of the deficits and the degree to which they crowded out investment. Under some budgetary assumptions, such as those of CBO’s alternative fiscal scenario, together with CBO’s typical assumption for the crowding out of investment, the effect of deficits would easily outweigh the effect of lower labor force growth, implying increases in rates of return in the coming decades.13 Under other assumptions, such as those of CBO’s extended baseline scenario, deficits would be much smaller but still would eventually outweigh the effect of slower labor force growth. The impact of declining labor force growth on rates of return might also be offset by net outflows of capital from the United States to the rest of the world. The theories discussed above are built on the assumption of a closed economy—that is, one with no economic relations with other countries. That assumption precludes the possibility that, in the event of a slowdown in the growth of labor supply, capital outflows from the United States to the rest of the world could offset the tendency for the relative prevalence of capital in the domestic economy to rise. Clearly, capital can flow relatively freely to and from the United States, so international capital flows are likely to reduce, but not eliminate, the effect of slowing labor force growth on rates of return on capital. Falling rates of return in the United States could lead to outflows of capital from investors seeking higher rates of return elsewhere. However, the economy of the United States is large, and its interest rates influence interest rates around the world. Therefore, if those outflows were large enough, they could raise the capital stock and depress rates of return in the rest of the world as well, which would temper the outflow from the United States. Furthermore, studies have found that despite the relatively free flow of capital across national borders, rates of national saving and domestic investment tend to be correlated— countries with higher rates of saving tend to have higher rates of investment.14 That correlation suggests, for whatever reason, that the impact of domestic saving is not entirely undone by capital flows. Explanations for this phenomenon differ, but there appears to be a significant degree of home bias in investment—other things being equal, investors are more likely to invest their savings in assets based in their own country. Finally, population growth in other developed countries and, especially, in the developing world is also projected to slow in the coming decades. For example, the United Nations projects that world population growth will slow from about 1.2 percent per year from 2005 through 2010 to about 0.3 percent per year between 2045 and 2050, which is similar to the slowdown expected for the United States over the same period.15 If the slowdown in population growth and resulting fall in rates of return turned out to be a worldwide phenomenon, the changes in the United States would probably not be offset by capital flows because foreign investments would not become relatively more attractive. Indeed, some estimates indicate that accounting for capital flows could actually result in larger projected declines in rates of return because, on net, capital would flow into rather than out of the United States.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Job Loss and Infrastructure Job Creation Spending during the Recession

63

SIMULATION RESULTS Economic simulations have led a number of researchers to conclude that the coming slowdown in labor force growth will reduce rates of return in the United States and other developed countries. Those estimates are based on economic models that are more complex than the simple estimates discussed above. However, they incorporate the simpler models’ logical connections and key assumptions, so it is not surprising that they produce broadly similar estimates. Robin Brooks, for example, estimates that a shift in population growth mirroring that of the baby-boom generation would result in a decline in rates of return in the United States of about 0.8 percentage points between 2000 and 2020, rebounding to a net decline of about 0.3 percentage points in the longer run.16 David Domeij and Martin Floden project that interest rates in countries in the Organisation for Economic Co-operation and Development will fall by about 1.5 percentage points between 2000 and 2050.17 Dirk Krueger and Alexander Ludwig project a decline in the rate of return in the United States of 0.8 percentage points (assuming no capital flows) or 0.9 percentage points (assuming capital flows) between 2005 and 2080; under either assumption, wages would rise by about 4 percent.18 Orazio Attanasio, Sagiri Kitao, and Giovanni L. Violante project a decline in interest rates of 1.9 percentage points in the developed world (assuming no capital flows) or 2.6 percentage points (assuming capital flows) between 2005 and 2040; wages would rise by 10 percent or 13 percent, respectively, under the same assumptions.19 Although those estimates vary, in each case the rate of return on capital is projected to decline by a substantial amount. (Wages would also rise in each case, although those results were not published in some of the studies.)

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

EMPIRICAL EVIDENCE There is some empirical evidence in the historical record to support the idea that the rate of growth of labor affects rates of return. For example, equations predicting the equilibrium rate of return that include the growth rate of the labor force, such as the one presented on page 5, have been found to be broadly consistent with actual rates of return in both the United States and Japan.20 The evidence, however, is not definitive. One possible reason for that ambiguity is that there have not been enough major, sustained changes in the rate of growth of the labor force over the period for which reliable data are available to identify the effects. Also, over shorter time periods, other factors, such as cyclical fluctuations and monetary policy, may have a greater impact on rates of return than long-run factors such as the growth rate of labor supply. In the absence of clear evidence, substantial weight should be placed on the results of economic models and simulations in projecting the effects of labor force growth on rates of return.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

64

Linda Levine

End Notes

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

1

The actuarial imbalance in the system is a measure of the gap between Social Security’s projected tax revenues and expenditures over some future period, with both taxes and expenditures adjusted by a discount factor to account for the time value of money. The discount factor has the effect 2 For simplicity, this discussion refers to the ratio of labor to capital. When technological progress raises productivity, it is actually the ratio of capital to output that must be stable in the long run to be consistent with a stable rate of return. 3 The ratio of capital to labor would not rise indefinitely because, eventually, the rising depreciation that resulted from a larger stock of capital would increase faster than the additional capital’s contribution to output. As the ratio of capital to labor rose, capital would continue to depreciate at the same rate, but each increment added to capital would contribute less and less to output. Eventually, an increasing share of saving would be required to replace depreciating capital, offsetting some of the effect of slower labor growth. In the long run, the ratio of capital to labor would rise to a point at which the rate of saving was just sufficient to maintain it. 4 See Congressional Budget Office, Will the Demand for Assets Fall When the Baby Boomers Retire? Background Paper (September 2009). 5 The higher wages that would result from slow growth in the labor force would improve Social Security’s finances by boosting payroll tax revenues. However, that effect would be relatively minor because it affects the level, rather than the growth rate, of wages and because higher wages eventually imply higher benefit payments. 6 Although changes in projected interest rates alter present-value calculations, they do not necessarily affect the proper weight to place on current versus future costs and benefits, especially when the trade-offs span different generations. 7 Although income taxes are levied on both capital income and labor income, capital income faces lower taxes, on average, because much of labor income is subject to payroll taxes and a substantial share of capital income (including returns earned in tax-free savings accounts or by pension funds, charities, colleges, and other nonprofit institutions) is not taxable. In addition, depreciation of fixed capital—a sizable portion of capital income—is not taxed. 8 Technically, the assumption that economic output is based on a ―Cobb-Douglas production function‖ implies that capital and labor income shares are constant. Many economic projections assume that type of production function. Such an assumption is supported by the fact that income shares in the United States have been fairly constant, even as the ratio of capital to labor has risen over time. However, some studies have suggested that the effect on the rate of return outweighs an increase in relative abundance of an input. See Eric Miller, An Assessment of CES and Cobb-Douglas Production Functions, Congressional Budget Office Working Paper 200 8-5 (June 2008). That paper evaluates which is more consistent with historical economic data, the CobbDouglas function or a more generalized formulation called the constant elasticity of substitution (CES) function. The paper finds that the data are more consistent with a specification of the CES type in which projected decreases in the rate of return on capital would lead to a somewhat smaller share of capital income in output and a larger share of labor income. 9 More precisely, the multiplicative effect on wages should be the multiplicative effect on rates of return raised to the /(1- ) power, which for relatively small effects implies a percentage increase roughly  /(1- ) times as large. 10 See Congressional Budget Office, How CBO Projects the Real Rate of Interest on 10-Year Treasury Notes, Background Paper (December 2007). 11 See Olivier Jean Blanchard and Stanley Fischer, Lectures on Macroeconomics (Cambridge, Mass.: MIT Press, March 1989), p.103. 12 See Joseph G. Altonji, Fumio Hayashi, and Laurence J. Kotlikoff, ―Is the Extended Family Altruistically Linked? Direct Tests Using Micro Data,‖ American Economic Review, vol. 82, no. 5 (December 1992), pp. 1171–1198. 13 See Congressional Budget Office, The Long-Term Budget Outlook (June 2009). The alternative fiscal scenario deviates from CBO’s normal baseline projections, beginning in 2010, by incorporating some changes in policy that are widely expected to occur and that policymakers have regularly made in the past. The extended baseline scenario adheres closely to current law, applying the concepts underlying CBO’s 10-year baseline budget projections. 14 George Georgopoulos and Walid Hejazi, ―The Feldstein-Horioka Puzzle Revisited: Is the Home- Bias Much Smaller?‖ International Review of Economics and Finance, vol. 18, no. 2 (March 2009), pp. 341–350. 15 See United Nations Population Division, ―World Population Prospects: The 2008 Revision Population Database,‖ available at www.esa.un.org/unpp/index.asp. 16 See Robin Brooks, ―Asset-Market Effects of the Baby Boom and Social-Security Reform,‖ American Economic Review, vol. 92, no. 2 (May 2002), Figure 1, p. 404. 17 David Domeij and Martin Floden, Population Aging and International Capital Flows, Stockholm School of Economics Working Paper No. 539 (November 2003), Figure 1.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Job Loss and Infrastructure Job Creation Spending during the Recession 18

19

Dirk Krueger and Alexander Ludwig, ―On the Consequences of Demographic Change for Rates of Return to Capital, and the Distribution of Wealth and Welfare,‖ Journal of Monetary Economics, vol. 54, no. 1 (2007), Table 7, p. 79. Orazio Attanasio, Sagiri Kitao, and Giovanni L. Violante, ―Global Demographic Trends and Social Security Reform,‖ Journal of Monetary Economics, vol. 54, no. 1 (2007), Table 2, p. 174. See Kaiji Chen, Ayse Imrohoroglu, and Selahattin Imrohoroglu, ―The Japanese Saving Rate Between 1960 and 2000: Productivity, Policy Changes, and Demographics,‖ Economic Theory, vol. 32, no. 1 (July 2007), pp. 87– 1 04; Kaiji Chen, Ayse Imrohoroglu, and Selahattin Imrohoroglu, A Quantitative Assessment of the Decline in the U.S. Saving Rate and the Current Account Balance (May 2008), available at www.sef.hku. hk/~kaijic/USeconomy.pdf; and Congressional Budget Office, How CBO Projects the Real Rate of Interest on 10-Year Treasury Notes, Figure 3, p. 6.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

20

65

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved. Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

In: Labor and Employment Issues Editor: Nickolas H. Mullen

ISBN: 978-1-60741-286-1 © 2010 Nova Science Publishers, Inc.

Chapter 4

JOB LOSS AND INFRASTRUCTURE JOB CREATION SPENDING DURING THE RECESSION 

Linda Levine

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

SUMMARY After the long economic expansion that characterized much of the current decade, the nation entered its eleventh postwar recession in December 2007. The unemployment rate, which is a lagging economic indicator, did not start to rise until May 2008 when it jumped 0.5 percentage points to 5.5%. By December 2008, the unemployment rate exceeded 7.0% and well over 600,000 jobs were lost—the biggest monthly decrease since December 1974, when another deep recession was taking place. These labor market indicators and comments equating the latest recession to the Great Depression intensified congressional interest in passage of legislation early in 2009 aimed at encouraging creation of new jobs and warding off further loss of jobs. (See CRS Report R40655, The Labor Market During the Great Depression and the Current Recession.) To mitigate all but one recession since the 1960s, Congress chose to increase federal spending on infrastructure. (See CRS Report 92-939, Countercyclical Job Creation Programs.) But, there are a number of issues associated with using expenditures on public works to quickly create jobs in times of recession. (See CRS Report R40107, The Role of Public Works Infrastructure in Economic Stimulus.) Public works expenditures traditionally have gone chiefly to construction activities (e.g., building highways and bridges, dams and flood control structures) which indirectly increase demand in industries that supply their products to construction firms (e.g., manufacturing). Today, the definition of infrastructure has been expanded to include green jobs, which include those in industries that utilize renewable resources (e.g., electricity generated by wind),



This is an edited, reformatted and augmented version of a CRS Report for Congress publication dated October 2009.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

68

United States Government Accountability Office

produce energy- efficient goods and services (e.g., mass transit), and install energyconserving products (e.g., retrofitting buildings with thermal-pane windows). A question that typically arises during congressional consideration of economic stimulus legislation is which approach produces the most bang for the buck. In the instant case, this means how many jobs might be supported by federal expenditures on traditional and green infrastructure projects. Once stimulus legislation is signed into law, the focus of Congress customarily turns to estimates of the number of jobs that result as federal funds are allocated to specific activities. Therefore, after briefly examining the trend in employment and unemployment since the recession’s onset, the report turns to an in-depth look at estimates of job creation, including the limitations of the methodology often used to derive them and the difficulties associated with developing job estimates for green infrastructure in particular. The report closes with a review of what is known to date about the number of jobs supported by infrastructure spending among other provisions in the American Recovery and Reinvestment Act (ARRA, P.L. 111-5). Section 1512 requires entities that receive ARRA appropriations from federal agencies, totaling approximately $271 billion, to include in quarterly reports the number of jobs created or maintained as a result. Section 1513 requires the Council of Economic Advisors to report quarterly on the effect of ARRA provisions on employment and other economic indicators. After the long economic expansion that characterized much of the current decade, the nation entered its eleventh postwar recession in December 2007. The unemployment rate, which is a lagging economic indicator, did not start to rise until May 2008 when it jumped 0.5 percentage points to 5.5%. By December 2008, it exceeded 7.0% according to data from the U.S. Bureau of Labor Statistics (BLS). Well over 600,000 jobs were lost in December 2008— the biggest monthly decrease recorded by the BLS Current Employment Statistics program (CES) since December 1974, when another deep recession was taking place. The Business Cycle Dating Committee of the National Bureau of Economic Research, the official arbiter of peaks and troughs in the business cycle, announced at the end of November 2008 that a substantial and widespread decline in economic activity had begun a year earlier. December 2007 marks both the end of the 73-month economic expansion that began in March 2001, and the beginning of the latest recession. As part of its announcement, the committee noted that it ―views the payroll employment measure, which is based on a large survey of employers, as the most reliable comprehensive estimate of employment. This series [the CES] reached a peak in December 2007 and has declined every month since then.‖ The committee’s announcement intensified congressional interest in passage of legislation aimed at encouraging creation of new jobs and warding off further loss of jobs. So, too, did comments equating the recession to the Great Depression. (See CRS Report R40655, The Labor Market During the Great Depression and the Current Recession.) To mitigate all but one recession since the 1960s, Congress chose to increase federal expenditures on infrastructure (public works), thereby directly raising demand for goods and services to offset the reduced demand of consumers. (See CRS Report 92-93 9, Countercyclical Job Creation Programs.) But, there are a number of issues associated with using spending on public works to quickly create jobs during a recession. (See CRS Report R40107, The Role of Public Works Infrastructure in Economic Stimulus.) When Congress considers spending on infrastructure to help stimulate a flagging economy, ―how many jobs are created‖ is a commonly asked question. After first briefly examining trends in job loss since the latest recession began, this report focuses on job

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects…

69

creation estimates associated with increased spending on traditional and so-called green infrastructure, placing a heavy emphasis on explaining the methodology often used to derive them and the difficulties associated with developing estimates for green infrastructure in particular. Once stimulus legislation is signed into law, the focus of Congress customarily turns to estimates of the number of jobs that result as federal funds are allocated to specific activities. In the case of the American Recovery and Reinvestment Act (ARRA, P.L. 111-5), Congress included language requiring entities that receive ARRA appropriations from federal agencies to report the number of jobs created or maintained as a result and requiring the Council of Economic Advisors to report on the employment and other economic effects of ARRA provisions. The report closes with a review of what is known to date about the number of jobs associated with the stimulus act.

EMPLOYMENT AND UNEMPLOYMENT THROUGH JOB LOSS As shown in Table 1, employment on nonfarm payrolls has steadily declined since December 2007.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Table 1. Payroll Jobs at Nonfarm Employers (seasonally adjusted employment in thousands) Year by Month 2007 December 2008 January February March April May June July August September October November December 2009 January February March April May June July August September

Total Employment

Private Sector Employment

138,152

115,783

138,080 137,936 137,814 137,654 137,517 137,356 137,228 137,053 136,732 136,352 135,755 135,074

115,689 115,515 115,373 115,203 114,029 114,834 115,691 114,497 114,197 113,813 113,212 112,542

134,333 133,652 133,000 132,481 132,178 131,715 131,411 13 1,210(p) 1 30,947(p)

111,793 111,105 110,457 109,865 109,573 109,182 108,936 108,754 108,544

Source: U.S. Bureau of Labor Statistics, data from the Current Employment Statistics program. Notes: (p) = preliminary. Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

70

United States Government Accountability Office

The number of job cutbacks intensified starting in late 2008. Of the 7.2 million jobs lost since the recession’s onset, the majority have disappeared since November 2008. As is typical during economic downturns, employees in the goods-producing sector have been the most adversely affected. They saw their ranks shrink by almost 3.6 million between December 2007 and September 2009. (See Table 2.) Workers in the sector’s construction industry began experiencing job losses before the economy-wide downturn began. Nonetheless, between the recession’s onset and September 2009, construction firms cut almost 1.5 million jobs. Across all manufacturing industries, employment fell by 2.1 million as well. Table 2. Number of Payroll Jobs by Industry (seasonally adjusted employment in thousands) Employment, September 2009 (p) Goods-producing sector 22,043 18,465 Mining and logging 743 708 Construction 7,523 6,038 Manufacturing 13,777 11,719 Service-providing sector 116,109 112,482 Trade, transportation and utilities 26,725 25,092 Wholesale trade 6,045 5,649 Retail trade 15,568 14,700 Transportation and warehousing 4,555 4,178 Utilities 557 565 Information 3,025 2,826 Financial activities 8,243 7,702 Professional and business services 18,109 16,597 Education and health services 18,570 19,311 Leisure and hospitality 13,551 13,154 Other services 5,517 5,397 Government 22,369 22,403 Source: U.S. Bureau of Labor Statistics, data from the Current Employment Statistics program. Notes: (p)=preliminary.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Industry by Sector

Employment, December 2007

Although manufacturing job losses have been widespread, two industries that produce durable goods—fabricated metal products (e.g., hardware, wire, and screws) and transportation equipment (e.g., motor vehicles and parts)—have been particularly hard hit. Employment in the service-providing sector most recently peaked in December 2007, when the recession began. Although some service-providing industries have continued to grow—utilities, education and health services—cutbacks elsewhere have far outweighed their gains. As shown in Table 2, these industries reported higher employment in September 2009 than at the start of the recession. In contrast, the financial activities industry began to lose jobs before the advent of the economy-wide downturn. This mirrors the above-mentioned trend in construction employment in part because real estate is a component of financial activities and

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects…

71

it, like construction, has been hurt by the collapse of the housing market. Other components of financial activities, such as brokerage firms that packaged high-risk mortgages and the investors (e.g., banks) that purchased them, have been negatively affected by the housing market downturn as well. Despite a marked slowdown in the pace of job loss in recent months and the widely expressed belief that the recession ended this summer, the prospect of steady job growth beginning in the near term looks dim according to a CRS analysis of employment trends after the end of the prior ten recessions. In all but one instance, ―the number of jobs on employer payrolls fluctuated for months.... Sustained job growth occurred within three to five months of the start of seven recoveries. In sharp contrast, steady job growth did not commence until March 1992—12 months after the July 1990-March 1991 recession ended—and not until September 2003—22 months after the March-November 2001 recession ended.‖1

The unemployment rate in September 2009 rose to 9.8% from 5.0% in December 2007, according to BLS data derived from the Current Population Survey.2 The Blue Chip Economic Indicators reported that the latest consensus forecast among the nation’s leading business economists is that the unemployment rate will continue to rise to 10% or higher and ―recede from that level only grudgingly‖ during the last six months of 2010. Most of these economists think it will not ―be until the second half of 2012, or later, before the unemployment rate falls below 7 percent on a sustained basis.‖3

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

INFRASTRUCTURE SPENDING AND JOB CREATION ESTIMATES When in response to a recession Congress has acted to create jobs by raising demand for goods and services through increased federal spending, it often has chosen to direct the funds to infrastructure (public works) activities. Other means of direct countercyclical job creation— employment tax credits, state revenue-sharing, and public service employment— have been relied on much less often.4 A more expansive definition of infrastructure than was used in the past is now under consideration. Historically, public works has been synonymous with heavy and civil construction activities (e.g., road and bridge building, flood control structures and dam building). Today, it includes so-called green jobs. Although numerous studies on the emerging green economy have been released in the last several years, no consistent definition of green jobs exists at present. Green jobs seemingly are those in and related to industries that utilize renewable resources to produce their outputs (e.g., energy generated by wind, solar, and geothermal technologies) and jobs in and related to industries that produce energyefficient goods (e.g., Energy Star appliances and equipment) and services (e.g., intra- and inter-city mass transit).5 For this reason, the following discussion focuses on what is known about the job-generating impact of infrastructure spending broadly defined. The section below begins with an in-depth examination of how job creation estimates usually are developed. The focus then narrows to look at two models that can be used to calculate the number of jobs nationwide dependent upon demand in the construction industry among other industries, and one model that can be used to calculate the number of jobs by

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

72

United States Government Accountability Office

state dependent on the construction industry among other industries. The section ends by reviewing the difficulties that researchers encounter in estimating the number of jobs supported by expenditures on green infrastructure and the consequent caution that should be taken when utilizing these estimates in particular.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Job Creation Estimates: What Are They? Interest in how many jobs are created by a particular type of economic activity has surfaced when the economy is in a downturn and policymakers seek to compare the relative advantages of different stimulus options. It also has arisen when policymakers want to know the impact of shifting expenditures from one federal budget category to another (e.g., away from defense and towards social services programs). Unless there is an increase in total spending, however, the number of jobs in the labor market would remain largely unchanged.6 Although there are other bases upon which to develop estimates of the number of jobs created by a given economic activity, an input-output (I-O) model of the economy often is utilized due to its cost-effectiveness.7 An I-O model describes the interrelationships between industries in the production process, showing how the dollar value of a sale is distributed across industries at a particular point in time. It thus reflects how much of the purchased product comes from final and supplier industries. An I-O table might show, for example, the dollar value of roof trusses produced by the veneer, plywood, and engineered wood products manufacturing industry and the dollar value of bricks produced by the clay product and refractory manufacturing industry used by the construction industry to erect residential buildings. The output requirements from each industry must then be converted to employment requirements. Employment requirements are derived from productivity estimates for each industry at a particular point in time. The total employment requirement associated with a given type of final demand (e.g., a water reuse program) is the employment in the industry producing the final product or service and in the supplier industries. In other words, it is an approximation of both the direct and indirect employment dependent upon/supported by the economic activity. It commonly is expressed as the number of jobs per billion dollars of expenditures valued in a particular year’s dollars. Like an I-O table, an employment requirements table is a matrix of hundreds of columns and rows. Each column displays the number of jobs supported in each of the industry rows by an expenditure of one billion dollars in the column industry. For example, one billion dollars spent in the construction industry supports (direct) employment in the various components of that industry (e.g., residential and commercial building, highway and bridge building) and (indirect) employment in the many industries that supply their goods and services to the construction industry (e.g., asphalt shingle manufacturing, fabricated metal bridge section manufacturing). An employment requirements table thus permits estimation of the varying impact of an expenditure on different industries and the varying impact of different kinds of expenditures.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects…

73

Some Caveats I-O models freeze technology and productivity at a particular point in time. Thus, the jobgenerating potential of an economic activity undertaken today could differ from that of an earlier period if there were technological and productivity improvements in the intervening years. Similarly, the estimates often are stated in terms of the number of jobs created for every billion dollars of expenditures, but a billion dollars spent in one year could buy less (more) than a billion dollars spent in another year depending on changes in price levels over time. There also could be differences in estimated versus actual job creation because I-O models assume that resources are unlimited. If, for example, the economy was performing at a fairly high level with plants operating near full capacity and with fairly few workers unemployed, the actual number of new jobs might fall short of the estimate due to capital and labor constraints. This is less likely to matter during a broad-based economic downturn. Further, I-O tables do not necessarily differentiate between imported and domestically produced goods. As a consequence, the domestic employment impact of expenditures might be overstated to the extent that inputs are imported. Similarly, I-O tables typically do not express employment in terms of full-time equivalents (i.e., both full-time and part-time jobs are counted equally). Thus, programs which draw upon industries that rely relatively more on part-time workers (e.g., retail trade) might appear to create more jobs than programs that draw to a greater extent on industries employing relatively more full-time workers (e.g., manufacturing).

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

The Multiplier Effect A more comprehensive estimate of the number of jobs created by a particular type of economic activity has three components, namely,   

the number of jobs directly attributable to the activity, the number of jobs indirectly attributable to the activity, and the number of jobs induced throughout the economy as a result of the activity.

Induced jobs are those dependent upon the purchases of persons in direct and indirect jobs. For example, workers who are directly or indirectly employed as the result of a highway construction program might spend some portion of their wages in their communities at grocery stores, auto repair shops, and movie theaters. Estimates of induced jobs or the multiplier are considered tenuous. To calculate the multiplier effect, one must estimate how much of the additional money earned by directly and indirectly employed workers will likely be spent versus saved. The actual number of jobs created by this added spending will further depend on economic conditions (e.g., the availability of labor, the inflation rate). As a result, there are widely varying estimates of the multiplier effect and those job creation studies that include induced employment utilize different multipliers.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

74

United States Government Accountability Office

Job Estimates and Construction Spending The Federal Highway Administration Perhaps the most widely known estimate of the employment impact of federal spending on our nation’s roads comes from the Federal Highway Administration (FHWA). Although the FHWA twice updated its 1997 analysis, which estimated that $1 billion of federal-aid highway expenditures plus a $250 million state match supported 47,575 jobs, some proponents of stimulating job growth through increased federal spending on infrastructure continue to use this figure. The most recent update by the FHWA to 2007 indicates that a $1.25 billion expenditure on highway construction consisting of $1 billion from the federal government and $250 million from state government could support 34,779 jobs. If a state match is not required, ―then $1 billion in Federal funds supports 27,800 jobs.‖8 The jobs number has decreased over time in part because of increases in the price of inputs, such as asphalt and diesel fuel. The FHWA breaks down the estimate of 27,822 jobs per billion dollars of federal spending on highways as follows: 



Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.



9,536 construction-oriented jobs (i.e., jobs at construction companies working on the projects and at businesses that provide direct inputs to the projects such as asphalt, concrete, and guard rails); 4,324 jobs in supporting industries (i.e., employment at firms that provide inputs to the industries directly providing the materials and equipment utilized in highway construction such as producers of sheet metal who supply the manufacturers of guard rails); and 13,962 induced jobs (i.e., jobs throughout the economy dependent upon consumer expenditures from the wages of workers in ―construction-oriented‖ and ―industrysupporting‖ jobs).

Thus, the multiplier effect accounts for one-half of the total estimate. The FHWA notes one caveat about I-O analysis in addition to those mentioned above, that is, the job estimate ―utilizes the national average mix of construction materials and labor inputs. Specific projects and local utilization ratios will alter the estimated number of jobs supported.‖9 For example, a different combination of materials and number of workers might be required for road resurfacing projects compared to bridge building or commuter rail projects.

The FHWA also states that [t]he employment figures have recently been used as a justification for including highway spending in an economic stimulus package. But with the exception of short-term resurfacing and preservation projects, highway funds spend out slowly, with only 27% of a project, on average, outlaying in the first year.10

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects…

75

BLS Employment Requirements Table In recognition of the fact that ―people want to assess the impact on employment of different policies or actions,‖ the U.S. Bureau of Labor Statistics (BLS) makes available electronically free-of-charge to the public the employment requirements tables it develops as part of its employment projections program.11 I-O and employment requirements tables developed and utilized by others often are proprietary and not made widely available. The employment requirements tables are based on the official I-O tables for the nation that the U.S. Bureau of Economic Analysis (BEA) develops every five years. BLS takes the latest national I-O table available from BEA – in this case, 1997 – and updates it to reflect more recent production and distribution technologies. It then utilizes the updated I-O table and recent labor productivity data to develop an employment requirements table. Because the base year for the most recently published employment projections is 2006, the latest employment requirements table reflects 2006 technologies of production and distribution as well as labor productivity. The BLS employment requirements table provides information for the construction industry as a whole. The construction industry, according to the North American Industry Classification System, is composed of three major subdivisions:  

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.



construction of buildings (residential and nonresidential), heavy and civil engineering construction (highway, street, and bridge construction; utility system construction; construction of flood control structures, dams, and hydroelectric power generation facilities), and specialty trade contractors (foundation, structure, and building exterior contractors; building equipment contractors; building finishing contractors).

The BLS employment requirements table shows 11,768 jobs directly and indirectly dependent upon $1 billion of spending on construction. A majority of the jobs are in the construction industry itself (i.e., 6,925 direct jobs). The figure from the BLS employment requirements table for construction expenditures (11,768) is somewhat lower than the direct and indirect jobs figure for highway expenditures from the FHWA (13,860). Potential explanations for the disparity include differences in industry definition, data sources, method of updating the model, and time period. The employment requirements available from BLS do not break out other types of construction that have been discussed as part of a federal job creation package (e.g., public school construction). BLS formerly conducted surveys to estimate full-time year-long employment associated with a variety of different construction activities, including new schools, hospitals, water and sewer facilities, roads, mass transit, and maintenance and repair construction. The survey information was last updated a few decades ago, however.

BEA’s Regional Input-Output Modeling System (RIMS II) From its Regional Input-Output Modeling System (RIMS II), the BEA produces estimates by geographic area of the employment, earnings, and output dependent on additional spending in hundreds of different industries.12 For a fee to most parties, BEA

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

76

United States Government Accountability Office

currently utilizes either the 1997 benchmark I-O for the nation or the 2006 annual I-O for the nation adjusted by 2006 data from its regional economic accounts to provide these estimates at the subnational level. 13 As shown in Table 3, the number of jobs directly and indirectly supported by an expenditure of $1 billion in the construction industry in a given state ranges widely. The main reason for the disparity in job creation estimates is that each state has a different mix of industries within its borders. As a consequence, one state varies from the next in its capacity to supply all the intermediate goods needed to carry out construction projects. A secondary explanation is that earnings vary by state.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Table 3. Number of Direct and Indirect Jobs by State Dependent on an Expenditure of $1 Billion in the Construction Industry State Alabama Alaska Arizona Arkansas California Colorado Connecticut Delaware District of Columbia Florida Georgia Hawaii Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Minnesota Mississippi Missouri

Number of Jobs 15,851 11,009 12,238 15,306 12,289 12,575 10,709 9,518 1,874 13,127 14,224 11,614 15,860 11,916 13,747 14,330 13,625 15,039 13,731 15,988 10,687 10,714 13,354 12,998 15,357 13,241

State Montana Nebraska Nevada New Hampshire New Jersey New Mexico New York North Carolina North Dakota Ohio Oklahoma Oregon Pennsylvania Rhode Island South Carolina South Dakota Tennessee Texas Utah Vermont Virginia Washington West Virginia Wisconsin Wyoming United States

Number of Jobs 16,127 13,946 11,459 12,374 11,118 14,279 10,106 15,555 13,500 14,391 16,232 13,184 12,390 10,767 15,319 15,316 14,556 12,985 14,692 14,883 12,085 12,171 13,834 13,673 13,091 14,315

Source: Prepared by CRS from RIMS II estimates supplied by the BEA Regional Product Division.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects…

77

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Job Estimates and Green Infrastructure Spending Estimating the number of jobs dependent upon green infrastructure activities presents a greater challenge than estimates related to infrastructure projects as traditionally defined. The basis for most data collection by U.S. statistical agencies is the North American Industry Classification System (NAICS). It currently does not identify separately so-called green industries (e.g., those that utilize renewable resources to produce their outputs, those that manufacture goods which minimize energy use). For example, the NAICS disaggregates the electric utility industry into hydroelectric, fossil fuel, nuclear, and other power generation, transmission, and distribution. Such renewable sources of energy production as wind, solar, and biomass are not uniquely recognized; they are included in the ―other‖ category. If harnessing the wind to produce electricity and plant material to produce biofuel requires a substantially different mix of inputs than relying on coal and gasoline, for example, the conventional I-O model does not seem well- suited as a basis for estimating the number of jobs supported by these green activities. Similarly, within NAICS, the construction industry does not have a unique category for retrofitting (e.g., installing additional insulation, fluorescent lighting, or energy-efficient heating and air-conditioning systems). Retrofitting likely requires a combination of inputs from supplier industries that differs from the mix for the top-to-bottom construction of buildings, once again making use of conventional I-O models problematic. This recognized difficulty generally is either not mentioned, or how it is dealt with is not described, in the analyses of green job creation. One study, commissioned by the Center for American Progress that is discussed in more detail below, does address the problem. The researchers explain that because ―the U.S. government surveys and accounts that are used to construct the input-output tables do not specifically recognize wind, solar, biomass, building retrofitting, or new mass transit as industries in their own right,‖ they created synthetic industries by combining parts of industries for which data are available. The researchers provided an example in the case of the biomass ―industry:‖ they constructed it by combining farming, forestry, wood products, and refining industries; then they ―assigned relative weights to each of these industries in terms of their contributions to producing biomass products.‖14 Further complicating the matter is the context and manner in which estimates of green jobs generally are presented. Studies often develop employment projections based on differing sets of assumptions and time horizons. For example, the number of direct and indirect jobs some 10 or more years in the future supported by an assumed increase in the demand for energy that is met by an assumed shift during the projection period from coal to wind and geothermal power generation. Some reports also include induced employment, but this is not always made clear. In addition, some analyses relate to a particular state. Their results may not be generalizeable to other areas because state economy’s have different mixes of industries and may not be able to provide any or all of the inputs for a particular green output. Additionally, the assumptions and methodologies underlying the job creation estimates often are not clearly articulated, which makes thoughtful review of the results very difficult. It should be noted that many of the studies by green economy proponents that were available when Congress was crafting stimulus legislation had not been conceived for the purpose of quickly stabilizing or increasing the number of jobs in the nation or in industries particularly hard hit by the recession. Job creation estimates from two organizations that

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

78

United States Government Accountability Office

proposed broad-based green economy strategies intended in part to stimulate the deteriorating labor market are briefly described below. 

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.



The September 2008 report, Green Recovery: A Program to Create Jobs and Start Building a Low-Carbon Economy, was commissioned by the Center for American Progress (a research and educational institute). It represents an acceleration of a 10year program included in a 2007 report (Capturing the Energy Opportunity: Creating a Low-Carbon Economy). The 2008 report’s authors at the Department of Economics and Political Economy Research Institute (University of Massachusetts – Amherst), who relied on I-O analysis, estimate that almost 2 million jobs (935,200 direct jobs, 586,000 indirect jobs, and 496,000 induced jobs) could be created or preserved by a two-year $100 billion ―green economic recovery program.‖ Of the $100 billion total, $46 billion would be in the form of federal spending for such activities as public building retrofits, mass transit and freight rail expansion, and smart electrical grid development. Much of the remainder would be in the form of tax credits to encourage businesses and homeowners to retrofit commercial and residential buildings. The authors acknowledge that not all of the green activities can contribute equally to a short-term green economic recovery program. Some ... strategies are clearly capable of delivering within a year, while others will require as long as two years to be implemented.15 In December 2008, the Apollo Alliance (a coalition of labor, environmental, business and community leaders) proposed The Apollo Economic Recovery Act.

It is an initial step toward achievement of a 10-year $500 billion program to create 5 million green-collar jobs, which had been released in September 2008. The new initiative calls for federal spending of about $50 billion to create or maintain more than 650,000 direct jobs and 1.3 million indirect jobs. The derivation of these job creation figures is not always clear, appearing to rely much of the time on spending-to-jobs relationships estimated by other organizations. The proposed allocation of federal funds and associated job estimates include $10 billion to improve the efficiency and reliability of the electric transmission grid (131,000 direct and indirect jobs), $8 billion to repair roads and bridges (278,000 direct and indirect jobs), and $8 billion to encourage localities to replace aging buses and trains with U.S.-made clean-energy vehicles (37,600 direct jobs in vehicle manufacturing and 167,000 indirect jobs).

MEASURING JOBS SUPPORTED BY SPENDING PROVISIONS IN THE AMERICAN RECOVERY AND REINVESTMENT ACT While crafting the American Recovery and Reinvestment Act (ARRA), Congress was concerned about timely tracking of the number of jobs whose creation or maintenance results from the legislation. The 111th Congress therefore addressed this matter in bill language much more than prior Congresses had in countercyclical job creation legislation. At Title XV— Accountability and Transparency of Division A—Appropriations Provisions, P.L. 111-5 requires entities that receive ARRA appropriations from federal agencies (e.g., grant, loan, or contract recipients; states) to include in their quarterly reports to the agencies estimates of the

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects…

79

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

number of direct jobs created and retained by infrastructure projects, for example. Recipients of recovery funds must make their first submission of the required information in October 2009, and the agencies must post the contents of the reports on websites 30 days after the end of each calendar quarter. The Office of Management and Budget (OMB) is directed to provide guidance to help recipients prepare the reports, including the development of job estimates.16 The act further charges the Congressional Budget Office and the Government Accountability Office with commenting on the job estimates contained in the reports within 45 days after they have been submitted to federal agencies. P.L. 111-5 additionally tasks the Council on Economic Advisers (CEA) with submitting quarterly reports to the Committees on Appropriations on the effect of ARRA-provisions on employment and other economic indicators. The CEA’s mandate thus extends well beyond the above-described reporting requirements, which apply only to $271 billion in direct government investment spending out of a total of $787 billion.17 Accordingly, as well as utilizing the direct job estimates provided by recipients of investment spending under ARRA, the CEA anticipates incorporating into its macroeconomic model the data the Treasury and the OMB are collecting weekly on tax cuts and other spending. The first quarterly report of the CEA was issued in September 2009. Based on two different estimating procedures, it found that ARRA might have added some one million jobs to employer payrolls in August 2009 compared to what employment would have been in the absence of the legislation. The CEA focused on the impact of the state fiscal relief contained in ARRA because $38.4 billion had been provided to the states by August, which ―represents almost half of outlays and one-quarter of total ARRA stimulus (that is, outlays plus tax reductions).‖18 A positive relationship was estimated between the fiscal relief thus far provided to states and employment in state and local government as well as in the education and health care industries.

End Notes 1

CRS Report R40798, Unemployment and Employment Trends Before and After the End of Recessions, by Linda Levine. 2 Data from the Current Population Survey of households is available at http://stats.bls.gov/cps. 3 ―Recession Finished, Say Blue Chip Analysts,‖ Daily Labor Report, September 10, 2009. 4 CRS Report 92-939, Countercyclical Job Creation Programs, by Linda Levine. 5 Related jobs include, for example, those in industries that manufacture wind turbines and install thermal-pane windows. 6 Small differences in the total number of jobs could occur at the same spending levels if the economic activities to (from) which funds were being shifted were more (less) capital-intensive, for example. 7 Another basis for estimating the impact of policy and other changes on the economy is conducting surveys. According to the U.S. Bureau of Economic Analysis (BEA), the advantage of the I-O approach to making impact estimates is the accessibility of the data sources required to develop the I-O model. 8 U.S. Department of Transportation, Federal Highway Administration, Employment Impacts of Highway Infrastructure Investment, p. 1, http://www.fhwa.dot.gov/policy/otps/publications.htm. 9 U.S. Department of Transportation, Federal Highway Administration, Employment Impacts of Highway Infrastructure Investment, p. 2, http://www.fhwa.dot.gov/policy/otps/publications.htm. 10 U.S., Employment Impacts of Highway Infrastructure Investment, U.S. Department of Transportation, Federal Highway Administration, p. 2, http://www.fhwa.dot.gov/policy/otps/publications.htm. 11 U.S. Bureau of Labor Statistics, Layout and Description for 201-order Employment Requirements Tables, Washington, D.C., December 2007, p. 3, http://stats.bls.gov/emp/empind4.htm. 12 For additional information on RIMS II see BEA, Regional Multipliers: A User Handbook for the Regional InputOutput Modeling System, http://www.bea.gov/scb/pdf/regional/perinc/meth/rims2.pdf.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

80

United States Government Accountability Office

13

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

More specific detail by industry is available from the 1997 benchmark I-O than from the annual I-O. Therefore, Table 1 (Number of Direct and Indirect Jobs Per $1 Million of Output Produced by the Water, Sewage and Other Systems Industry) in CRS Report R40 107, The Role of Public Works Infrastructure in Economic Stimulus, was drawn from the 1997 benchmark I-O because the 2006 annual I-O provides data only for the utilities industry as a whole. 14 Robert Pollin, Heidi Garrett-Peltier, and James Heintz, et al., Green Recovery: A Program to Create Good Jobs and Start Building a Low-Carbon Economy, Center for American Progress, Washington, D.C., September 2008, p. 20, http://www.americanprogress.org. 15 Robert Pollin, Heidi Garrett-Peltier, and James Heintz, et al., Green Recovery: A Program to Create Good Jobs and Start Building a Low-Carbon Economy, Center for American Progress, Washington, D.C., September 2008, p. 5, http://www.americanprogress.org. 16 Office of Management and Budget, Implementing Guidance for the Reports on Use of Funds Pursuant to the American Recovery and Reinvestment Act of 2009, M-09-21, Washington, D.C., June 22, 2009. 17 The remaining ARRA funds fall into five categories, as the Council of Economic Advisors states in Estimates of Job Creation from the American Recovery and Reinvestment Act of 2009 (May 2009, p. 2): ―individual income tax cuts; a two-year patch to the alternative minimum tax; investment incentives; aid to people directly hurt by the recession; and state fiscal relief.‖ 18 Council of Economic Advisers, The Economic Impact of the American Recovery and Reinvestment Act of 2009, First Quarterly Report, Washington, DC, September 10, 2009, p. 33.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

In: Labor and Employment Issues Editor: Nickolas H. Mullen

ISBN: 978-1-60741-286-1 © 2010 Nova Science Publishers, Inc.

Chapter 5

U.S. LABOR FORCE STATISTICS: ILLUSTRATIVE SIMULATIONS OF THE LIKELY EFFECTS OF UNDERREPRESENTING UNAUTHORIZED RESIDENTS 

United States Government Accountability Office

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

ABBREVIATIONS ACS BLS CPS DHS

American Community Survey Bureau of Labor Statistics Current Population Survey Department of Homeland Security

November 30, 2009 The Honorable Jim McDermott Chairman Subcommittee on Income Security and Family Support Committee on Ways and Means House of Representatives Dear Mr. Chairman: In times of economic uncertainty as well as in times of stability, policymakers and the public rely on labor force statistics, such as the unemployment rate, to provide important information on the current state of the economy. These statistics, published by the Department of Labor’s Bureau of Labor Statistics (BLS), include key figures that are based on data obtained from the Current Population Survey (CPS). The CPS, a household interview survey 

This is an edited, reformatted and augmented version of a U. S. Government Accountability Office publication dated November 2009.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

82

United States Government Accountability Office

administered by the Department of Commerce’s U.S. Census Bureau, is designed to represent the entire United States civilian noninstitutional population. However, certain U.S. residents— specifically, foreign-born persons who are not authorized to live here, to whom we refer as unauthorized residents in this report—may not be represented in CPS data to the same extent as the general population. Little research has been done on whether potential underrepresentation of the unauthorized population might noticeably affect labor statistics, but policy efforts that depend on valid and reliable labor force statistics would benefit from such information. We agreed with your office to examine issues concerning unauthorized residents’ impact on labor force statistics. We addressed the following key questions: 1. Extent of underrepresentation: What is known about the extent of any underrepresentation of unauthorized residents in CPS data used to compile labor force statistics? 2. Labor force status: What is known about the likely labor force status of unauthorized residents?

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

3. Possible effects: How might CPS underrepresentation of unauthorized residents affect key labor force statistics? To answer the first two questions, we examined documents and data from and interviewed officials from the Census Bureau, the Department of Homeland Security, and BLS. We also interviewed and obtained data from immigration and statistical experts outside these agencies (see slide 44 in app. I for a listing of these experts and their roles). Additionally, we reviewed previous GAO work and other studies on unauthorized residents, government surveys, and labor force statistics and interviewed immigration advocates about these topics. To answer the third question, we conducted simulations to illustrate how undercounting unauthorized residents may have affected the unemployment rate, size of the U.S. labor force, and number of employed persons in March 2008. We determined that March 2008 was the most recent time period for which data were available that were sufficiently reliable for the purposes of our main simulations. We created the simulated statistics by first using combinations of assumptions obtained from experts about the extent of underrepresentation of unauthorized residents and their level of unemployment in March 2008. Then, because the estimates obtained from experts were uncertain, we used much broader assumptions to test the robustness of our results. We compared all of the simulated statistics with the margins of error for the original statistics. We also updated these simulations using June 2009 national labor force statistics as a test of whether changing economic conditions might affect our results.1 We conducted our work from November 2008 to November 2009 in accordance with all sections of GAO’s Quality Assurance Framework that are relevant to our objectives. The framework requires that we plan and perform the engagement to obtain sufficient and appropriate evidence to meet our stated objectives and to discuss any limitations in our work. We believe that the information and data obtained, and the analysis conducted, provide a reasonable basis for any findings and conclusions.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects…

83

RESULTS IN BRIEF On October 2, 2009, we briefed your staff on the results of our work. This report formally conveys the information provided during that briefing (see app. I for the briefing slides). In general, we found the following: 





The extent of CPS underrepresentation of unauthorized residents is unknown, but experts we consulted told us that the CPS data might not represent approximately 10 to 15 percent of unauthorized residents. Little information is available about the labor force status of unauthorized residents, but experts we consulted suggested that their approximate unemployment rate in March 2008 may have been 6.5 to 8.5 percent, compared with 5.2 percent nationally. Little is known about the effect of underrepresentation of unauthorized residents on labor force statistics. However, using the above information provided by experts, we simulated the likely effects of adding unauthorized residents assumed not represented in March 2008 labor force statistics. Because the expert assumptions were uncertain, we tested our results using a broader range of assumptions. Our simulations showed that adding unauthorized residents not represented in CPS data would likely have a minimal effect on the overall U.S. unemployment rate. The simulations did not, however, rule out the possibility that adding such residents would increase the estimated size of the national labor force and the estimated number of persons employed nationally. Updating these simulations using national labor force statistics for June 2009 revealed similar effects.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

AGENCY COMMENTS AND OUR EVALUATION Prior to our October 2, 2009, briefing, we provided the Departments of Labor, Homeland Security, and Commerce with a draft of the briefing document and incorporated technical comments provided by the Bureau of Labor Statistics and the Census Bureau as appropriate. Homeland Security had no comments on the slides. Since then, Labor, Homeland Security, and Commerce have reviewed a draft of this report. Homeland Security had no comments, and BLS informally remarked positively on the way that our draft accounted for uncertainty related to unauthorized workers. Commerce provided written comments, reproduced in appendix II, which did not dispute any of our three main findings concerning the possible impacts of underrepresentation of unauthorized residents on national labor force statistics. However, Commerce suggested that we (1) further explicate how survey data are weighted to improve statistical representation and how we conducted our simulations, as well as providing—in this letter—more specifics about the experts we consulted; (2) remove information on the two-card (or ―grouped answers‖) approach for estimating the number of illegal immigrants represented in a survey, because including it implies both that this method is appropriate for use in the CPS and that it can be used to estimate underrepresentation;2 and (3) reduce the range of underrepresentation in our illustrative simulations, which we vary from 0 to 50 percent, because Commerce stated that using this range gives the range more plausibility than is supported by research.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

84

United States Government Accountability Office

In response to Commerce’s first suggestion, we added some additional information to the slides. One new footnote cites a Census Bureau publication that provides detailed information on how CPS data are weighted, and another new footnote provides a numerical illustration of how we conducted our simulations. We also added a parenthetical note in this letter, indicating the slide that lists the experts we consulted and their roles, and clarified that we consulted experts in both immigration and statistics. In response to Commerce’s second suggestion, we did not change our presentation of information on the grouped answers method for the following reasons: 

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.



Our slide presentation clearly states that the grouped answers method is appropriate for a survey conducted by a private-sector organization. Consistent with Commerce’s comments, we concluded in our September 2006 report3 on estimating the unauthorized population that the grouped answers method is not appropriate for any existing government-conducted survey, including the CPS; the 2006 report therefore raised the possibility of a new survey designed for the foreign-born population. If the grouped answers method were validated and included in a privatesector survey of the foreign-born, the resulting data logically could be used to help assess statistics based on a government-conducted general population survey such as the CPS, as well as serve various other policy and evaluation purposes.4 While the grouped answers method is not intended to estimate underrepresentation, if validated and used in a new private-sector survey, it could provide information on specific groups within the foreign-born population (such as the number of employed unauthorized workers and the number of unauthorized workers in the labor force). Such data would be useful to the task addressed in this report in at least two ways. First, in the illustrative simulations we conducted, we made assumptions about unemployment levels within the unauthorized worker population because there was no survey estimate of unemployment for this group. The grouped answers method, applied in a new private-sector survey, might provide such an estimate. Second, our illustrative simulations used an indirect estimate of the unauthorized worker population represented in the CPS; this indirect estimate had been produced by a researcher using a subtraction method (see slide 12 of app. I) and various assumptions regarding labor force status in this group. A private-sector survey using the grouped answers method could help researchers using the subtraction method or other indirect methods by, for example, providing a survey estimate of the number of unauthorized workers represented in a private- sector survey.5

Regarding Commerce’s third suggestion, we did not reduce the range of the test assumptions for underrepresentation of unauthorized workers in the CPS. Our reasons are that this report (1) already presents analyses that use fairly narrow ranges of assumptions deemed most plausible by experts (for both underrepresentation and unemployment among unauthorized workers) and (2) uses much broader ranges of 0 to 50 percent to test results produced using the fairly narrow ranges. We conducted the test simulations because of uncertainty about the narrower expert-based ranges. Specifically, using alternative assumptions of 0 to 50 percent, we tested how far ―off the mark‖ the expert assumptions would have to be for our findings to change. Our tests show that the initial expert-based

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

.

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects…

85

narrow-range results would hold up even if the expert-based assumptions were rather far off the mark. Finally, Commerce provided additional information about correspondence between GAO and the Census Bureau concerning an open GAO recommendation.6 This recommendation dates from 1998 and concerns devising a plan of research for evaluating the quality of census and survey data on foreign-born persons. There have been various communications in addition to the letter referenced by the Census Bureau,7 but the Census Bureau has not proposed a strategy or plan of research for evaluating the quality of census and survey data on foreign-born persons. Innovative thinking and collaboration may be required to make progress on this longterm recommendation. For example, one approach may be for the Census Bureau to design research and evaluation options that it might pursue in partnership or coordination with others (such as other federal agencies, universities, or private-sector organizations). As agreed with your office, unless you publicly announce its contents earlier, we plan no further distribution of this report until 7 days from the report date. We will then send copies of this report to relevant congressional committees, the Secretary of Labor, the Secretary of Commerce, the Secretary of Homeland Security, and other interested parties. In addition, this report will be available at no charge on GAO’s Web site at http://www.gao.gov. If you or your staff have any questions about this report, please contact either Nancy Kingsbury at (202) 512-2700 or [email protected] or Cornelia Ashby at (202) 512-7215 or [email protected]. Contact points for our Offices of Congressional Relations and Public Affairs may be found on the last page of this report. GAO staff who made key contributions to this report are listed on slide 47 of appendix I.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Sincerely yours,

Nancy Kingsbury Managing Director, Applied Research and Methods

Cornelia M. Ashby Director, Education, Workforce, and Income Security Issues

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

86

United States Government Accountability Office

APPENDIX I: BRIEFING SLIDES U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects of Underrepresenting Unauthorized Residents Briefing for Staff of Chairman Jim McDermott Subcommittee on Income Security and Family Support House Committee on Ways and Means October 2, 2009 Prepared by GAO’s Applied Research and Methods and Education, Workforce, and Income Security Teams

Introduction Labor force statistics are economic benchmarks for policymakers and the general public. 





The Labor Department’s Bureau of Labor Statistics (BLS) produces key statistics, such as the unemployment rate, the size of the labor force, and the number of employed persons. BLS compiles these statistics using data from the Current Population Survey (CPS), a monthly household interview survey that is designed and weighted to represent the U.S. population as measured by updated census counts. The CPS is intended to represent the entire civilian noninstitutional population, including those foreign-born persons who are not authorized to reside in the United States.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

CPS data may represent many—but likely not all—unauthorized foreign-born residents.8 

 

CPS data differentiate between native and foreign-born persons, but not between the foreign-born who reside here legally and those who reside here illegally (referred to as unauthorized residents in this briefing). Indirect analyses suggest that millions of unauthorized residents may be represented in CPS data.9 Some additional unauthorized residents may not be represented in these data.

Introduction: Effects of Underrepresentation The effect that underrepresenting unauthorized residents may have on CPS-based labor force statistics depends on two key factors: 

The number of unauthorized residents who are not represented in CPS data. For example, if this number is small compared with the general population, then omitting data on them would probably have a small effect.



Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects… 

87

The labor force status of unauthorized residents who are not represented in CPS data. For example, if this group’s employment status is similar proportionately to that of the general population, then omitting data on them would have little effect on some statistics, such as the unemployment rate.

Key Questions For this briefing, we examined these questions: 1. EXTENT OF UNDERREPRESENTATION: What is known about the extent of any underrepresentation of unauthorized residents in CPS data used to compile labor force statistics? 2. LABOR FORCE STATUS: What is known about the likely labor force status of unauthorized residents? 3. POSSIBLE EFFECTS: How might CPS underrepresentation of unauthorized residents affect key labor force statistics?

Scope and Methodology To obtain background information and answer questions 1 and 2, we:

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

    

Examined documents and data from the Department of Commerce’s Census Bureau, the Department of Homeland Security (DHS), and BLS. Reviewed existing GAO work. Interviewed federal officials, immigration experts, and immigrant advocates about unauthorized residents and employment. Examined selected academic and policy institute studies of unauthorized residents, government surveys, and labor statistics.10 To answer our third question:  We conducted simulations to illustrate how underrepresenting unauthorized residents might have affected three BLS statistics in March 2008:11 unemployment rate; size of the U.S. labor force; and number of employed persons.

 



We updated this analysis using the more limited data available for June 2009. For these analyses, we obtained ranges of possible effects by varying our assumptions about the  extent of underrepresentation of unauthorized residents and  level of unemployment for unauthorized residents who were not represented in CPS data.12 See appendix A for more information on our methods.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

88

United States Government Accountability Office 





We did not focus on  statistics for local areas, specific occupations, or economic sectors;  trends in labor force statistics over time; or  other data sources for labor statistics, besides the CPS. (See app. B.) We provided a draft of this briefing to the Departments of Labor, Commerce, and Homeland Security for review. Within these departments, BLS and the Census Bureau provided technical comments that we incorporated as appropriate. DHS had no comments. We conducted our work from November 2008 to October 2009 in accordance with all sections of GAO’s Quality Assurance Framework that are relevant to our objectives. The framework requires that we plan and perform the engagement to obtain sufficient and appropriate evidence to meet our stated objectives and to discuss any limitations in our work. We believe that the information and data obtained, and the analysis conducted, provide a reasonable basis for any findings and conclusions.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Summary of Findings 1. UNDERREPRESENTATION: The extent to which unauthorized residents are underrepresented in CPS data is unknown, but experts we consulted suggested this might range from 10 to 15 percent. 2. LABOR FORCE STATUS: Little is known about unauthorized residents’ labor force status, but experts we consulted suggested that their unemployment rate may have been between 6.5 and 8.5 percent in March 2008, versus a national rate of about 5.2 percent. 3. POSSIBLE EFFECTS: Our simulations for March 2008 and update for June 2009 indicated that, despite uncertainties,  adding unauthorized residents not represented in CPS data would likely have a minimal effect on the U.S. unemployment rate, but  increases to the estimated size of the U.S. labor force or the estimated number of employed persons cannot be ruled out.

Background Labor Force Statistics 

BLS is the source of government labor force statistics, which include but are not limited to the following:

 

Unemployment Rate: Percentage of the labor force that is not working and is looking for work.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects… 

89

Labor Force Size: Number of persons who are working or looking for work.

Employed Persons: Number of persons who are working.13 BLS bases these three statistics on CPS data. BLS reports statistics at the national and state levels. 

  

The Current Population Survey (CPS)  





The Census Bureau administers the CPS monthly to approximately 60,000 households. The CPS asks a household member to report characteristics of each individual in the household, including country of birth, as well as overall household characteristics. BLS bases its labor force statistics on CPS data for those aged 16 and older. The CPS does not ask about authorization to reside or work in the United States. BLS officials told us that asking about legal status could discourage cooperation with the CPS. The Census Bureau takes several steps to minimize underrepresentation in the CPS and increase its accuracy. See appendix C for more information.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Almost 12 Million Unauthorized Residents Are Estimated to Reside in the United States DHS and the Pew Hispanic Center have estimated the number of unauthorized residents by using existing data on foreign-born U.S. residents. Essentially, DHS and Pew have 

  

begun with an estimate of the number of all foreign-born persons residing here (including both those who are and those who are not authorized to live in the United States);14 subtracted the number who are authorized to reside here;15 made various assumptions about, for example, the effects of underrepresentation; and concluded that in early 2008, almost 12 million residents— approximately one-third of all foreign-born—were unauthorized.

Previous GAO Work In 1998, we reported that policymakers need reliable information on foreign-born persons for policy-related activities and decision making, but such data were not available (GAO/GGD-98-164).

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

90

United States Government Accountability Office   





We recommended that the Census Bureau and DHS publish a plan of research to evaluate census and survey data on the foreign-born. Given continued interest by Members of Congress in related issues, we continue to follow up on agency efforts in this area. We are working with Census Bureau officials to plan a meeting to discuss letters we exchanged in 2008 and 2009 concerning difficulties that the Census Bureau anticipates in implementing some approaches to evaluating representation of the foreign- born. In 2004, we reported that lack of data precluded us from estimating the cost of educating children who are not authorized to reside in the United States (GAO-04733). In 2006, we reported that foreign-born respondents appear to accept an innovative method of asking about legal status in personal interviews conducted by a nongovernment entity (GAO- 06-775).16

Finding 1: Underrepresentation The Extent of Underrepresentation Is Unknown, but Some Experts Suggest 10 to 15 Percent  

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.







Information on the number of unauthorized residents who are not represented in CPS data is unavailable. Three experts told us that 10 to 15 percent of unauthorized residents might not be represented in CPS data, but these figures are uncertain.17 One of the experts, Jeffrey S. Passel of the Pew Hispanic Center, used analytic procedures to estimate that the underrepresentation of unauthorized residents in the March 2008 CPS was approximately 12.5 percent.  Passel applied Census Bureau estimates of undercounts for the overall U.S. population in 2000, by race, age, and gender, to foreign-born CPS respondents. He assumed higher undercount rates for recent and unauthorized immigrants.18 The figures for the underrepresentation of unauthorized residents cited in this briefing are uncertain, because they rely on inferences based on demographic characteristics and/or a single study with limited scope. For example, the three experts we consulted on underrepresentation all referred to a key study that attempted to measure the size of the census undercount of unauthorized Mexicans in Los Angeles County.19 In our analyses, we first simulated underrepresentation as ranging from 10 to 15 percent. Then, because the expert figures were uncertain, we tested the robustness of our results by broadening this range to 0 to 50 percent.  Note: The experts estimated the 10 to 15 percent range for the entire unauthorized population. To conduct our simulations, we applied this range to unauthorized residents in the labor force.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects…

91

Finding 2: Labor Force Status Little Is Known about the Labor Force Status of Unauthorized Residents 



 

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.



Labor force status includes  having a job (employed);  having no job and looking for work (unemployed); and  having no job and not looking for work (out of the labor force), such as homemakers, retirees, and ―discouraged workers.‖ Limited information is available on whether the labor force status of unauthorized residents who are not represented in CPS data differs from that of  unauthorized residents represented in the CPS or  the overall, general population of the United States. Passel estimated that in March 2008, there may have been about 8.3 million unauthorized residents in the U.S. labor force.20 Passel and two other experts who provided input on unemployment among unauthorized residents told us the following:  The unemployment rate of unauthorized residents is likely to vary over time and be affected by changes in the economy and government policies.  The unemployment rate for unauthorized residents might have been between 6.5 and 8.5 percent in March 2008, higher than BLS’s reported rate of about 5.2 percent for the general population. However, these figures are uncertain.21 In our simulations of the unemployment rate and number employed, we first assumed March 2008 unemployment rates for unauthorized residents of 6.5 to 8.5 percent. Then, because the expert figures were uncertain, we broadened this range to 0 to 50 percent to test the robustness of our results.

Finding 3: Possible Effects Experts Do Not Know the Effect of Under-representation; Some Suggest It May Be Minimal 





Some experts suggested to us that the effect of underrepresenting unauthorized residents on overall statistics would likely be small. However, due to a lack of data, experts ultimately do not know the effect of underrepresentation on labor force statistics. Our simulations focused primarily on three March 2008 statistics:22  unemployment rate (U.S. reported figure: 5.2 percent),  size of the labor force (U.S. reported figure: 153.1 million), and  number of employed persons (U.S. reported figure: 145.1 million). We also conducted rough simulations to determine whether similar patterns of results might hold for June 2009.23

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

92

United States Government Accountability Office 

To test whether including unauthorized residents who were not represented in CPS data would change the three statistics listed above, we recalculated them using various assumptions about (1) the underrepresentation of unauthorized residents and (2) the unemployment rate of those who were not represented.

Finding 3: Possible Effects (March 2008) Unemployment Rate: Likely Effect of Underrepresentation Was Minimal  



Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.



According to BLS data, the March 2008 national unemployment rate was 5.24 percent. The margin of error for this statistic is +/- 0.17 percent.24 We conducted two sets of simulations for March 2008:  In the first set of simulations, we varied underrepresentation from 10 to 15 percent and varied unemployment rates for the unauthorized residents not represented in the data from 6.5 to 8.5 percent.  In the next set of simulations, we used a broader range for underrepresentation (0 to 50 percent) and unemployment rates (0 to 50 percent) to test the robustness of the initial results. The first set produced simulated unemployment rates that fell within the margin of error of BLS’s reported rate (5.24 percent, +/-0.17 percent), which we regard as a minimal effect. In the second set, only assumptions considerably more extreme than the ranges suggested by experts produced simulated unemployment rates that fell outside BLS’s margin of error. See figure 1.

Figure 1. Simulated U.S. Unemployment Rates after Including Unauthorized Residents Assumed Not Represented in CPS Data, March 2008 Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects…

93

Labor Force: Potential Increases Cannot Be Ruled Out   





BLS estimated that the March 2008 national labor force size was 153 million. The margin of error for this figure is about +/- 506,000. For our simulations for March 2008, we assumed that the BLS estimate of 153 million includes 7.2 million unauthorized members of the labor force. For our first set of simulations for this time period, we further assumed that the 7.2 million figure underrepresented unauthorized members of the labor force by 10 to 15 percent. The resulting simulations added 800,000 to 1.3 million more unauthorized members of the labor force to the BLS figure of 153 million, for a simulated total of roughly 154 million. Our second set of simulations for this time period explored a broader range of assumptions for underrepresentation (0 to 50 percent). These simulations added 0 to 7.2 million to the BLS figure of 153 million, for simulated totals ranging from 153 to 160 million. Most simulations produced simulated national labor force sizes that were above the margin of error of BLS’s reported labor force size.25

Employed Persons: Potential Increases Cannot Be Ruled Out  

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.







In March 2008, BLS estimated that there were 145 million employed persons. The margin of error was about +/- 541,000. Our simulations for March 2008 assumed, as on the previous slide, that the BLS labor force estimate of 153 million persons included 7.2 million unauthorized residents. In the first set of simulations:  As on the previous slide, we varied underrepresentation from 10 to 15 percent, adding from 800,000 to 1.3 million unauthorized residents to the labor force.  We then assumed that from 6.5 to 8.5 percent of the added unauthorized residents were unemployed, meaning that 91.5 to 93.5 percent were employed; we applied these percentages to the 800,000 to 1.3 million added to the labor force.  Resulting simulations added 700,000 to 1.2 million employed persons to the BLS figure of 145 million, for a simulated total of roughly 146 million. In the next set of simulations:  As before, we used a broader range of assumptions for underrepresentation (0 to 50 percent), adding from 0 to 7.2 million to the labor force.  We then assumed that 0 to 50 percent of the added unauthorized residents were unemployed, or that 50 to 100 percent were employed, and applied these employed percentages to the 0 to 7.2 million added to the labor force.  These simulations added from 0 to 7.2 million employed persons to the BLS figure, for simulated totals ranging from 145 to 152 million. Most simulations produced simulated numbers of employed persons that were above the margin of error of BLS’s reported national figure.26 See figure 2.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

94

United States Government Accountability Office

Figure 2. Simulated U.S. Employed Persons after Including Unauthorized Residents Assumed Not Represented in CPS Data, March 2008

Finding 3: Possible Effects (June 2009)

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Effects of Underrepresentation May Be Similar under More Recent Economic Conditions 







After March 2008, the month of our detailed simulations, the U.S. unemployment rate rose, reaching 9.5 percent in June 2009, and two experts told us that the rate for unauthorized residents may have risen more steeply over this time period, perhaps to 11 or even 15 percent. While more recent estimates of size of the overall unauthorized population are not available, one expert believes this population may have decreased due to declining economic conditions, and another thinks its size has stayed roughly the same.27 We performed simulations of the effects of underrepresentation on June 2009 labor force statistics for which we assumed no change since March 2008 in the size of the unauthorized labor force represented in CPS data.  As a check, we also performed sensitivity analyses in which we increased and decreased the size of the unauthorized labor force (represented in the CPS data) by up to 30 percent. The sensitivity analyses indicated that changes of this magnitude would not change our simulation findings. Our simulations for June 2009 were approximate, or rough, in that we used the broad assumptions of 0 to 50 percent underrepresentation of unauthorized residents and 0 to 50 percent unemployment for this group without identifying a narrower range of potentially more likely values based on expert opinion.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects…  

95

We found patterns in our June 2009 simulations similar to our 2008 results. Barring extreme assumptions or other differences we were not able to measure, underrepresentation may have had similar effects on 2008 and 2009 statistics.28

Effects of Underrepresentation May Be Larger in Some States We did not simulate state-level labor force statistics, but our research indicates the following:    

Unauthorized residents probably represent a higher percentage of the population in certain states. In some cases (e.g., Arizona, California, and Nevada), the percentage may be about twice as high as the national figure. Additionally, some experts believe that the underrepresentation of unauthorized residents is greater in certain states. Effects on state labor force statistics could be greater in states with more unauthorized residents or higher underrepresentation, especially if, for example, the unemployment rate of unauthorized residents who are not represented in CPS data differs from estimates for the rest of the state population.

Implications

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.





Although the overall unemployment rate appears not to be sensitive to possible underrepresentation of unauthorized residents, other labor force statistics (such as the size of the labor force) could be. Therefore:  Fuller representation of unauthorized residents in the CPS and other surveys would improve the accuracy of some labor force statistics.  Fuller representation might be especially important in states with high concentrations of unauthorized residents, where underrepresentation may produce pronounced effects on labor force statistics.29 In general, better information about the unauthorized population would help policymakers implement laws and evaluate policies that apply differently to foreignborn persons who reside here legally and to unauthorized residents. We are continuing to monitor federal efforts related to possible approaches for evaluating representation of the foreign-born population in data collected.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

96

United States Government Accountability Office

APPENDIX A: METHODS FOR SIMULATING NATIONAL LABOR FORCE STATISTICS

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.



For our March 2008 detailed simulations, we used publicly available CPS estimates of the U.S. unemployment rate, size of the labor force, and number of employed persons aged 16 and older.30  March 2008 was the most recent time for which we were able to obtain published estimates of the number of unauthorized residents in the labor force (Passel and Cohn, 2009(a)).  We also used Jeffrey S. Passel’s unpublished estimate that 7.2 million unauthorized residents aged 16 and older in the U.S. labor force were represented in the March 2008 CPS.31  Passel made this estimate before he adjusted CPS data for u nde rrepresentation.  While we did not vary this figure in our simulations, we performed sensitivity analyses and determined that our findings would not change if we increased or decreased Passel’s figure by up to 20 percent.  We initially simulated national labor force statistics using a range of assumptions based on the opinions of one government and three private sector experts, who suggested that  the underrepresentation of unauthorized residents in CPS data might be between 10 and 15 percent and  unemployment among unauthorized residents, as of March 2008, might range from 6.5 to 8.5 percent.32  Using these values, we simulated revised versions of  the national unemployment rate,  the number of persons in the U.S. labor force, and  the number employed in the United States.  To test whether more extreme assumptions would change our results, we also calculated additional simulated values for March 2008 using values of 0 to 50 percent for both underrepresentation and the unemployment rate of unauthorized residents. We examined whether the resulting simulated labor force statistics fell within the margins of error for reported BLS statistics.33 

 



To determine whether our results would likely be similar under more recent economic conditions, we updated our analysis using June 2009 CPS data on the unemployment rate (8.7%), labor force (155 million), and employed persons(141 million). We assumed that the number of unauthorized residents in the labor force represented in the CPS was unchanged from March 2008 (7.2 million). We used the same broad range of assumptions for underrepresentation (0 to 50 percent) and unemployment rate (0 to 50 percent) and compared the results with the 90 percent confidence interval for the June 2009 statistics. While this study focused on possible underrepresentation of one specific group, unauthorized residents, it is also possible that CPS data under- or overrepresent the

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects…





97

general population by a small amount (based on factors such as net undercounting and overcounting in the 2000 census).34 Our simulations did not adjust for any overall under or overrepresentation in the CPS, but we tested our results to determine whether they would be affected by an overall error of less than one-half of 1 percent in either direction. This sensitivity analysis showed that adjusting for a small overall error would not substantially change our results. Table 1. March 2008 Labor Force Statistics and Margins of Error (Civilian Noninstitutional Population Aged 16 and Older)

Lower 90% Confidence Interval 5.24% 5.08% Unemployment Rate 153,135,000 152,629,000 Labor Force 145,108,000 144,567,000 Employed Persons Source: Calculated by GAO using published CPS data. Statistic

CPS Estimate

Upper 90% Confidence Interval 5.41% 153,641,000 145,649,000

APPENDIX B: DATA SOURCES FOR LABOR FORCE STATISTICS WE EXAMINED 

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.





The Current Population Survey (CPS), which is the primary data source for the statistics examined here, is a nationally representative monthly survey of approximately 60,000 households administered by the Census Bureau.  BLS uses CPS data to calculate statistics such as unemployment rates, the size of the labor force, and the number of employed persons. The decennial census (which is not a survey, but a comprehensive household-based count) and the American Community Survey (ACS) are not used directly in calculating official labor force statistics.  The Census Bureau uses them in calculating the national population estimates, which are used to weight CPS data.35 None of these data sources collects information on legal status.

Data Sources for Labor Force Statistics We Did Not Examine 



The Current Employment Statistics survey is a monthly survey of approximately 400,000 employers. The results are used in calculating the number of nonfarm jobs, known as payroll employment. Those surveyed are drawn from businesses making unemployment insurance payments.  This survey does not collect data on legal status. Unemployment Insurance claims data are collected by states and used by BLS as supplementary information in calculating some state and local labor force statistics.  Unemployment insurance claims data may exclude most unauthorized residents, who are ineligible to collect unemployment benefits.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

98

United States Government Accountability Office

APPENDIX C: APPENDIX C: CPS DATA, NATIONAL POPULATION ESTIMATES, AND U NDERREPRESENTATION 







Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.





The Census Bureau weights CPS data to make them consistent with national population estimates.  The Census Bureau produces the national population estimates (1) using counts from the decennial census and (2) updating these counts based on birth and death records as well as data on net migration from the ACS, a large survey that it conducts every year.36  According to the Census Bureau, the purpose of weighting CPS data is to improve accuracy and help minimize underrepresentation.37 CPS data distributions—such as Hispanics in defined age groups—thus match the distributions from the national population estimates. However, problems affecting the national population estimates logically ―carry over‖ to the weighted CPS data, including any underrepresentation of unauthorized residents. National population estimates may underrepresent a group if, for example, that group is undercounted in the census or partially composed of foreign-born residents who may be underrepresented in the ACS. Undercounting in the decennial census is thought to be higher for unauthorized residents than for others.  Undercounting in the decennial census occurs when a household is missed or an individual is omitted from the list of persons residing in a covered household.38  To illustrate, if a subpopulation (such as the unauthorized population) numbers 12 million and the census fails to count 1.2 million of them, the undercount would be 1.2 million or 10 percent. While the Census Bureau uses census data and the national population estimates to improve the accuracy of CPS data, it does not make statistical adjustments to correct undercounting in census or underrepresentation in the national population estimates. Regarding the issue of statistical adjustment of the census,39 a topic of discussion among policymakers over the years, the director of the Census Bureau said in a May 15, 2009, testimony before the Senate Committee on Homeland Security and Governmental Affairs:

―I agree fully... that statistical adjustment of the census is eliminated as an option for reapportionment... [or] redistricting. The 2003 decision of director... assured that no implementation infrastructure for adjustment was put in place for [the] 2010 [census].‖

Underrepresentation of unauthorized residents in CPS data may be caused by factors additional to undercounting in the census, including 



possible underrepresentation of unauthorized residents in the ACS or mis categorization of some members of key groups (e.g., Hispanics)—which would affect how CPS data are weighted, and nonresponse in the CPS, which occurs, for example, when no one in a selected household is available to take the survey.40

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects…

99

APPENDIX D: GROUPED ANSWERS APPROACH TO ESTIMATING THE UNAUTHORIZED RESIDENT POPULATION 





Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.



The "grouped answers" approach is a questionnaire method that might be used in private-sector, personal-interview surveys to obtain information about the overall population of unauthorized residents, while ensuring the anonymity of individual survey responses.41 This personal-interview approach  shows each respondent a flash card with answers grouped in three boxes,  asks the respondent to "just pick the box" that contains his or her current legal status (along with other statuses), and thus  avoids questioning respondents about whether they are unauthorized. Estimation for an overall population is possible because the grouped answers approach (two card version)  divides respondents into two subsamples; 42  shows subsamples 1 and 2 slightly different cards (see the following slide); and  compares the percentages that picked key boxes, as indicated by the arrow on the following slide. When the grouped answers approach is applied in a personal interview survey conducted by a university or private-sector organization, it appears to be acceptable to foreign-born respondents and, with certain provisions, to immigrant advocates.

Figure 3. Legal Status Cards 1 and 2 Compared

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

100

United States Government Accountability Office

APPENDIX E: LIST OF EXPERTS CONSULTED 



Experts who suggested parameters for our simulations:43  Steven A. Camarota, Center for Immigration Studies  Randy Capps, Migration Policy Institute  Michael D. Hoefer, U.S. Department of Homeland Security  Jeffrey S. Passel, Pew Hispanic Center Experts who reviewed our work:  Douglas S. Massey, Office of Population Research, Princeton University  Fritz J. Scheuren, National Opinion Research Center, University of Chicago  Robert E. Warren, former Director, U.S. Immigration Statistics Division, Immigration and Naturalization Service

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Selected References Camarota, Steven A. & Karen Jensenius. ―A Shifting Tide: Recent Trends in the Illegal Immigrant Population.‖ Washington, D.C.: Center for Immigration Studies, July 2009. Groves, Robert M. et al. Survey Methodology, 2nd edition. Hoboken, N.J.: John Wiley & Sons, Inc, 2009. Hoefer, Michael, Nancy Rytina, & Bryan C. Baker. ―Estimates of the Unauthorized Immigrant Population Residing in the United States: January 2008.‖ Office of Immigration Statistics, U.S. Department of Homeland Security. Marcelli, Enrico A. & Paul M. Ong. ―Estimating the Sources of the 2000 Census Undercount among Foreign-born Mexicans in Los Angeles County.‖ Prepared for the Annual Population Association of America Meetings. Atlanta, Georgia: May 10, 2002. Mulry, Mary. ―Summary of Accuracy and Coverage Evaluation for Census 2000.‖ Research Report Series, Statistics #2006-3. Washington, D.C.: U.S. Census Bureau, Feb. 28, 2006. Passel, Jeffrey S., & D’Vera Cohn. ―A Portrait of Unauthorized Immigrants in the United States.‖ Washington, D.C.: Pew Hispanic Center, April 14, 2009 (a). Passel, Jeffrey S., & D’Vera Cohn, ―Mexican Immigrants: How Many Come? How Many Leave?‖ Washington, D.C.: Pew Hispanic Center, July 22, 2009 (b). Slater, Courtenay M. ―The Impact of Census Undercoverage on Federal Programs.‖ Conference on Census Undercount: Proceedings of the 1980 Conference. Washington, D.C.: U.S. Bureau of the Census, July 1980. Pages 107-111. U.S. Census Bureau. Current Population Survey: Design and Methodology. Technical Paper 66. Washington, D.C.: October 2006.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects…

101

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

APPENDIX II: COMMENTS FROM THE DEPARTMENT OF COMMERCE

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

United States Government Accountability Office

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

102

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

U.S. Labor Force Statistics: Illustrative Simulations of the Likely Effects…

103

End Notes

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

1

After March 2008, the month of our detailed simulations, the U.S. unemployment rate rose, reaching 9.5 percent in June 2009. Declining economic conditions may affect the number of unauthorized residents in the United States, although recent estimates of the size of this population are not available. With this in mind, we performed sensitivity analyses that included decreasing the size of the unauthorized labor force (represented in the CPS data) by up to 30 percent. The sensitivity analyses indicated that a decrease of this magnitude would not change our simulation findings. 2 The grouped answers method is described in slides 41 to 43 in appendix I of this report. 3 GAO, Estimating the Undocumented Population: A “Grouped Answers” Approach to Surveying Foreign-Born Respondents, GAO-06-775 (Washington, D.C.: Sept. 29, 2006). 4 For example, data on a respondent’s legal status (such as legal permanent resident or refugee) are not collected by government-conducted general-purpose surveys such as the CPS—but potentially could be collected by a private-sector survey using an approach such as the grouped answers method. Such data are relevant to implementing various laws and evaluating various immigration policies. 5 Additionally, because the subtraction method estimates the unauthorized population by subtracting estimates of immigrants in specific legal status categories from an overall estimate of the total foreign-born population, researchers using this approach would be helped by grouped answers estimates of foreign-born persons in specific legal status categories (such as the number of legal permanent residents). The grouped answers method could provide direct estimates of these categories. 6 GAO, Immigration Statistics: Information Gaps, Quality Issues Limit Utility of Federal Data to Policymakers, GAO/GGD-98-164 (Washington, D.C.: July 31, 1998). 7 Perhaps most notably, in addition to the letter mentioned by the Census Bureau, we outlined various possible approaches that Census might take in estimating undercoverage of foreign-born persons, including a new record linkage (data-matching) approach. We shared this information with Census Bureau staff at a meeting and subsequently discussed this and other possible methods with them during 2008. We also made additional suggestions in a February 2009 letter to the Census Bureau. 8 Foreign-born residents include naturalized citizens, other authorized persons, and unauthorized residents. 9 See, for example, Jeffrey S. Passel and D’Vera Cohn, A Portrait of Unauthorized Immigrants in the United States, (Washington, D.C.: Pew Hispanic Center, Apr. 14, 2009). 10 We selected studies on the basis of their relevance to our work. 11 March 2008 is the most recent time period for which an expert has published estimates of not only the unauthorized population, but also the number of unauthorized workers. 12 We determined that the data we used were sufficiently reliable for the purposes of our simulations. 13 BLS publishes two estimates of employment: (1) the number of persons employed, measured by the CPS, and (2) the number of jobs, measured by the Current Employment Statistics survey. We focus here on the CPS estimate only. 14 The Pew Hispanic Center estimates this using data from the CPS, while DHS uses data from another national survey, the American Community Survey. 15 This number, estimated using primarily DHS data, includes persons such as legal permanent residents and refugees. 16 This survey approach groups legal statuses in boxes on a flash card. Each foreign-born respondent chooses the box containing his or her legal status, along with other statuses. No one specifically chooses the unauthorized status. See appendix D for more details. 17 These experts were Steven A. Camarota, Michael D. Hoefer, and Jeffrey S. Passel. See appendix E. 18 We use the term "undercount" when referring to census counts of the population and the term "underrepresentation" when referring to statistics based on data from a sample survey, such as CPS. 19 See Enrico A. Marcelli and Paul M. Ong, Estimating the Sources of the 2000 Census Undercount among Foreignborn Mexicans in Los Angeles County (prepared for the Annual Population Association of America meetings, Atlanta, Georgia, May 10, 2002). 20 This estimate includes an upward adjustment for assumed underrepresentation. Passel’s figure before adjusting for underrepresentation was 7.2 million. 21 These experts were Steven A. Camarota, Randy Capps, and Jeffrey S. Passel. The figures they provided are marked by varying degrees of uncertainty because they rely on assumptions and inferences, such as using CPS data on all immigrants from Mexico and Central America as a proxy population for unauthorized residents. 22 In this briefing, we use labor force statistics that are not seasonally adjusted. 23 Our simulations cannot be generalized to other time frames without further analysis. 24 BLS' published estimate is 5.2 percent, but we are providing a more precise figure for purposes of comparison with simulated values and margins of error. The margins of errorused in this briefing correspond to 90 percent confidence intervals around BLS’s reported statistics.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

104

United States Government Accountability Office

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

25

The simulated labor force size fell within the margin of error only in simulations with assumed underrepresentation of less than 7 percent. We provide this information as an observation; we have not conducted formal tests of significance. 26 The simulated number of employed persons fell within the margin of error of the reported statistic for a fairly narrow range of assumptions, including (1) underrepresentation of less than 8 percent or (2) underrepresentation of less than 9 percent and an unemployment rate of more than 13 percent. This is provided as an observation; we did not conduct formal tests of significance. 27 See Jeffrey S. Passel and D’Vera Cohn, Mexican Immigrants: How Many Come? How Many Leave? (Washington, D.C.: Pew Hispanic Center, July 22, 2009), and Steven A. Camarota and Karen Jensenius, A Shifting Tide: Recent Trends in the Illegal Immigrant Population (Washington, D.C.: Center for Immigration Studies, July 2009). 28 Our findings for both years are consistent with much earlier analyses that adjusted for census undercoveragefor the overall population. See Courtenay M. Slater, ―The Impact of Census Undercoverageon Federal Programs,‖in Conference on Census Undercount: Proceedings of the 1980 Conference (Washington, D.C.: U.S. Bureau of the Census, 1980), 107-111). 29 The same may be true for certain occupational categories, industrial sectors, or local areas, but examinations of the effects for these groups are beyond the scope of this briefing. 30 CPS data, which we and Passel used as starting points in our analyses, are weighted to reflect the Census Bureau’s national population estimates and to correct for survey nonresponse. 31 Passel estimated the March 2008 labor force participation rate of represented unauthorized residents to be 76.5 percent. 32 For example, if 7.2 million of the unauthorized resident labor force are represented in the CPS but 10 percent of this population are not represented, the total numbers 8 million, with 800,000 not represented. If the 800,000 have an unemployment rate of 8 percent, 64,000 are unemployed. To recalculate the existing unemployment rate, we add 64,000 to itsnumerator (number unemployed) and 800,000 to its denominator (total in labor force). 33 The margins of error correspond to 90 percent confidence intervals around the reported statistics. We use labor force statistics that are not seasonally adjusted. 34 For more information on possible overall counting error in the 2000 census, see Mary Mulry, ―Summary of Accuracy and Coverage Evaluation for Census 2000,‖ Research Report Series, Statistics #2006-3 (Washington, D.C.: U.S. Census Bureau, Feb. 28, 2006). 35 See appendix C. 36 For more information on national population estimates, see U.S. Bureau of the Census, Methodology for the United States Resident Population Estimates by Age, Sex, Race, and Hispanic Origin (Vintage 2008): April 1, 2000, to July 1, 2008, http://www.census.gov/popest/topics/methodology/2008-nat-meth.pdf.. 37 U.S. Census Bureau, Current Population Survey: Design and Methodology, Technical Paper 66, October 2006. See chapter 10, especially the section on national coverage adjustment. 38 For a discussion of undercoverage and related issues, see Groves et al., Survey Methodology, 2nd Edition (Hoboken, N.J.: John Wiley & Sons, Inc., 2009). 39 The U.S. Supreme Court ruled in 1999 that the use of statisticalsampling was statutorily prohibited for purposes of apportioning seats in the House of Representatives, but did not specifically address its use for other purposes. Dep’t of Commerce v. U.S. House of Representatives, 525 U.S. 316 (1999). 40 To correct CPS data for nonresponse, the Census Bureau increases the weights of the answers of respondents who might be similar to nonrespondents, but this may not fully address nonresponseon the part of unauthorized residents. 41 See GAO, Estimating the Undocumented Population: A “Grouped Answers” Approach to Surveying ForeignBorn Respondents, GAO-06-775 (Washington, D.C.: Sept. 29, 2006). 42 No one is asked to respond to both cards. This method (1) has thus far not been developed for mail surveys and (2) is not appropriate for telephone interviewing because respondents must view the flash cards. 43 We chose these four experts on the basis of their recent experience in estimating the unauthorized population. Camarota, Hoefer, and Passel suggested possible underrepresentation levels, and Camarota, Capps, and Passel commented on unemployment levels.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

In: Labor and Employment Issues Editor: Nickolas H. Mullen

ISBN: 978-1-60741-286-1 © 2010 Nova Science Publishers, Inc.

Chapter 6

UNEMPLOYMENT AND ECONOMIC RECOVERY



W. Brian Cashell

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

SUMMARY Even though the economy seems to be growing again, it may be a while before the unemployment rate begins to decline, and it may even continue rising for some time after the resumption of sustained economic growth. The unemployment rate is generally a lagging indicator, meaning that its ups and downs happen some time after the ups and downs of other broad indicators of economic activity. Unemployment may not fall appreciably when economic growth first picks up because some firms may have underutilized labor. At the end of a recession as demand increases, some firms may initially be able to increase production without adding workers. Firms may be able to increase their output by raising the productivity of the labor on hand. Over the longer run, there tends to be a link between the rate of economic growth and changes in the unemployment rate. Estimates based on data since 1949 suggest that real economic growth of about 3.5% was associated with a stable unemployment rate. When economic growth was faster than 3.5%, the unemployment rate tended to fall, and when economic growth was below 3.5% the unemployment rate tended to rise. In the long run, a one percentage point difference in the economic growth rate has historically led to a change in the unemployment rate of about 0.4 percentage points. In other words, although economic growth of 3.5% was sufficient to maintain a stable unemployment rate, an annual increase in real output of 4.5% would result in a decline in the unemployment rate of 0.4 percentage points. Even if economic growth picks up and an expansion gets going, experience suggests it may be some time before there are significant declines in the unemployment rate. After the end of the two most recent contractions, it was well over a year before there was a clear



This is an edited, reformatted and augmented version of a CRS Report for Congress publication dated November 2009.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

106

W. Brian Cashell

downward trend in the unemployment rate. After nine of the past 10 contractions, it took at least eight months for the unemployment rate to fall by one full percentage point. This report examines the relationship between economic growth and the unemployment rate to anticipate possible future developments.

INTRODUCTION By many accounts, the economic contraction that began in December 2007 is already, or is about to be, over.1 The increase in real gross domestic product (GDP) in the third quarter of 2009 lends support to that view. What may concern policymakers as much or even more, however, is the outlook for the unemployment rate. Even though the economy seems to be growing again, it may be a while before the unemployment rate begins to decline, and it may even continue rising for some time after the beginning of any expansion. The unemployment rate is generally a lagging indicator, meaning that its ups and downs happen some time after the ups and downs of other broad indicators of economic activity.2 Not only might the unemployment rate be slow to fall at the beginning of an economic expansion, its rate of decline, or whether it declines at all, is likely to depend on the rate of growth. It is possible for there to be above-zero real economic growth that is insufficient to prevent continued increases in the unemployment rate. This report examines the relationship between economic growth and the unemployment rate to anticipate possible future developments.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

ECONOMIC GROWTH AND THE UNEMPLOYMENT RATE In the short run, the relationship between economic growth and the unemployment rate may be a loose one. One reason that unemployment may not fall appreciably when economic growth first picks up is that some firms may have underutilized labor. Laying off workers when times are bad and rehiring them as conditions improve has costs. Up to a point, firms may be willing to maintain more workers than they need to satisfy the depressed demand for their goods and services in order to avoid those costs. As a result, at the end of a recession as demand increases, some firms may, at first, be able to increase production without adding workers. In other words, firms may be able to increase their output by raising the productivity of the labor on hand, which yields a temporary boost in measured labor productivity above its longer-run trend rate of growth. Once the labor on hand is fully utilized, output will grow at the rate of growth of labor productivity until the firm begins to add workers. As the economic expansion progresses, output growth will be determined by the combined rates of growth of labor and the productivity of labor. As long as growth in total output exceeds the rate of growth of labor productivity, then employment will rise. If employment growth is more rapid than growth in the labor force, then the unemployment rate will fall. That there is a stable long-run relationship between the rate of economic growth and changes in the unemployment rate was most famously pointed out by economist Arthur

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Unemployment and Economic Recovery

107

Okun, which is why it is sometimes referred to as ―Okun’s Law.‖ It has been included in a list of ―core ideas‖ that are widely accepted in the economics profession.3 The key to the relationship between the rate of economic growth and the unemployment rate is the rate of growth of what economists refer to as ―potential output.‖ In brief, potential output is a measure of the capacity of the economy to produce goods and services given the available resources, such as labor and capital. The rate of growth of potential output is a function of the rate of growth of productivity and the rate of increase of the contribution of the labor force in the production of goods and services. When the unemployment rate is high, as it is now, then actual output falls short of potential. Labor’s contribution to output is determined by the hours worked by those who are employed. Ultimately, labor input is measured in terms of hours. This, in turn, depends on the size of the population, the share of the population that is in the labor force, and the share of the labor force that is actually employed. In the absence of productivity growth, as long as each new addition to the labor force is employed, growth in total output will just equal the growth in the labor force. If growth in output falls below the rate of growth of the labor force, then there will not be enough new jobs created to accommodate all of the new job seekers. The proportion of the labor force that is employed will fall, and the unemployment rate will rise. If growth in output exceeds the rate of growth in the labor force, some of the new jobs opening up to satisfy rising demand will be filled by drawing down the pool of unemployed labor.4 If productivity is rising, over time it will take fewer and fewer workers to produce a given quantity of goods and services. If growth in output just equals the growth rate of the labor force, then more people will be entering the labor force than are needed to produce the higher levels of output. The share of the labor force that is employed will fall, and the unemployment rate will rise. Only as long as the growth in output exceeds the combined growth rates of the labor force and productivity will the unemployment rate fall in the long run. Knowing what that rate is might be useful to policymakers. Even when output is growing rapidly enough to create jobs fast enough to push the unemployment rate down, it may still take time to match jobs and job seekers. If the new jobs being created require substantially different skills from those jobs that have disappeared, then it may be difficult for displaced workers to get rehired. Some of those job seekers may have skills that are easily transferred from one job to another and thus may not experience longterm unemployment. Those with skills that have become outmoded or are less applicable in those industries that are expanding may have more difficulty finding new work. The more of a mismatch in skills there is between new jobs and job seekers, the longer it will take for displaced workers to find new jobs. Limits to geographical mobility may also account for some of the mismatches in the labor market. There are costs, both monetary and emotional, associated with pulling up stakes and moving to another part of the country to get a new job. It may also be the case that the further removed a job prospect is, geographically, the less likely it is that a job seeker will even hear about it. This may be especially relevant in areas where job losses are permanent, such as seems to be happening in the automotive industry as well as those industries that are in decline because of foreign competition. Figure 1 shows the relationship between economic growth and changes in the unemployment rate since 1949. Each point in the graph refers to a calendar quarter and indicates a pair of observations. The first observation of each pair is the percentage change in

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

108

W. Brian Cashell

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

real GDP over the previous four quarters (shown on the horizontal scale). The second observation of each pair is the percentage point change in the civilian unemployment rate over the same period (shown on the vertical scale). The downward sloping line indicates the estimated statistical relationship between the two series over the interval shown.5 Where it crosses the zero line indicates, approximately, the rate of economic growth necessary to keep the unemployment rate from either rising or falling. Above that growth rate, the unemployment rate tended to fall, and below that growth rate, the unemployment rate tended to rise. This analysis of the relationship between economic growth and changes in the unemployment rate is relatively simple. There are certainly other economic factors to take into account, in addition to more sophisticated statistical techniques. Nonetheless, there is evidently a strong link between the rate of economic growth and changes in the unemployment rate. Based on this estimate of the relationship between the two variables over the entire period, real economic growth of about 3.5% was associated with a stable unemployment rate. When economic growth was faster than 3.5%, the unemployment rate tended to fall, and when economic growth was below 3.5% the unemployment rate tended to rise. There are times, however, when the relationship temporarily breaks down. Changes in productivity growth, for example, tend not to be correlated with changes in unemployment. In the short run, a change in the rate of growth of productivity can cause the economic growth rate and the unemployment rate to change in the same direction. For example, in 1993, the economic growth rate fell to 2.7% from 3% in 1992 and the unemployment rate fell to 6.9% from 7.5%. The reason was a decline in productivity growth in 1993 after a brief surge in the previous year.6 There may also be small variations in average hours worked that may cause changes in the growth rate that are independent of changes in unemployment.

Source: Department of Commerce, Bureau of Economic Analysis; Department of Labor, Bureau of Labor Statistics Figure 1. Economic Growth Rate (%) and Percentage Point Changes in the Unemployment Rate, 1949 – 2009

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Unemployment and Economic Recovery

109

In the long run, based on the relationship estimated in Figure 1, a one percentage point difference in the economic growth rate has historically led to a change in the unemployment rate of about 0.4 percentage points. In other words, although economic growth of 3.5% was sufficient to maintain a stable unemployment rate, an annual increase in real output of 4.5% would result in a decline in the unemployment rate of 0.4 percentage points. Depending on possible changes in productivity, economic growth might not have to be as rapid as 3.5% now to push down the unemployment rate. Between 1949 and 2000, the civilian labor force grew at an average annual rate of 1.6%. More recently, however, that growth rate has slowed. Between 2000 and 2008, the annual rate of growth of the labor force was 1.0% and its growth is expected to continue to slow down. The Bureau of Labor Statistics projects that the labor force will grow at a 0.7% rate between 2009 and 2016.7 Although the near-term growth rate of the labor force may be known with some degree of confidence, predicting productivity growth presents substantial difficulties. In the short run, over the course of the business cycle, productivity growth tends to vary in predictable ways.8 In the long run, projecting productivity growth is difficult. Economists have identified three time periods that correspond with three different trend rates of growth in productivity.9 Between 1949 and 1973, output per hour of labor grew by 2.8% at an annual rate. Between 1973 and 1995, labor productivity grew at a 1.4% rate. Between 1995 and 2008, labor productivity grew at a 2.6% annual rate.10 Since 2000, the labor force has grown at an annual rate of 1%. To get the unemployment rate to fall, economic growth will likely have to exceed the sum of the 1% rate of growth of the labor force and the recent 2.6% rate of growth of labor productivity. This would seem to indicate that, if recent trends continue, sustained economic growth above 3.5% may be necessary to push down the unemployment rate. Even if economic growth picks up and an expansion gets going, experience suggests it may be some time before there are significant declines in the unemployment rate. Suppose that two successive monthly declines is taken as the beginning of a meaningful downward trend in the unemployment rate.11 Table 1 shows how long it has taken following the end of each of the past 10 economic contractions for that trend to begin. For example, following the end of the contraction in October 1949 that downward trend did not start until January 1950. After the two most recent contractions, well over a year had passed before there was a clear downward trend in the unemployment rate. Just as the time it takes for the unemployment rate to begin coming down has varied, the rate at which it falls has varied as well. Each of the lines in Figure 2 shows the percentage point change in the unemployment rate following the end of each of the past 10 economic contractions. In nine of those 10 instances, it took at least eight months for the unemployment rate to fall by one full percentage point. The line showing the steepest decline was the expansion that began in October 1949 when the unemployment rate was 7.9%. Of those 10, the expansion that began with the highest unemployment rate was the one that began in November 1982 when the unemployment rate was 10.8%. In that case, it took eight months for the unemployment rate to fall below 10%.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

110

W. Brian Cashell Table 1. Time in Months between the End of Economic Contractions and Two Successive Monthly Declines in the Unemployment Rate

Months Following End of Contraction Until There Were Two Successive Declines October 1949 4 May 1954 6 April 1958 5 February 1961 9 November 1970 11 March 1975 4 July 1980 2 November 1982 5 March 1991 17 November 2001 21 Source: National Bureau of Economic Research; Department of Labor, Bureau of Labor Statistics. Date of End of Contraction

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

The two thickest lines in Figure 2 show a wide range of possible outcomes. The top line represents the two years following the July 1980 business cycle trough, which ended with the unemployment rate two percentage points higher than it was at the trough. In that case, the expansion only lasted a year and the another contraction began in July 1981. The bottom line shows the two years following the October 1949 business cycle trough, which ended with the unemployment rate 4.4 percentage points lower than it had been at the trough.

Source: National Bureau of Economic Research; Department of Labor, Bureau of Labor Statistics. Figure 2. Percentage Point Change in the Unemployment Rate in the First Two Years of Each of the Past 1 0 Economic Recoveries

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Unemployment and Economic Recovery

111

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

THE OUTLOOK From a policy perspective, what matters for the sustainable reduction of the unemployment rate is what the growth rate of potential output will be in the future. According to estimates by Economist Robert J. Gordon, potential output has grown at an average annual rate of 3.4% since 1875.12 Gordon doubts, however, that growth in potential output will be that rapid over the next 20 years. He argues that the acceleration in productivity growth of the late 1990s was temporary. He examines economy-wide productivity rather than just that for the private business sector and finds that productivity growth slowed between 2004 and 2008 because the gains from information technology investments were beginning to diminish. His assumption of slower productivity growth along with expected declines in the growth rate of the labor force lead him to project a 2.4% rate of growth in potential output over the next 20 years. If that view is correct, then over the long run, real economic growth in excess of 2.4% would be likely to yield a declining rate of unemployment. Economics Susanto Basu and John G. Fernald also examined the current outlook for growth in potential output.13 They point out that there has been a significant decline in household net worth during the most recent contraction. That drop in wealth will likely make it more difficult to afford leisure time or to make down payments on purchases of durable goods like automobiles. Consequently, the supply of labor may be larger in the near term than it might otherwise have been, and that would tend temporarily to raise growth in the labor force and potential output. At the same time, they expect that disruptions in financial markets will tend to constrain growth in potential output over the near term because of higher risks associated with investment spending. Those factors tend to offset and mainly serve to emphasize how uncertain estimates of growth in potential output can be. The Congressional Budget Office (CBO) publishes projections of growth in potential output. In its August 2009 economic outlook, CBO projected that potential output will grow at an average annual rate of 2.0% between 2009 and 2013, rising to a 2.4% rate of increase between 2014 and 2019.14 CBO expects the unemployment rate to average 10.2% in 2010 and 9.1% in 2011. The Organisation for Economic Co-operation and Development (OECD) projects that the U.S. unemployment rate will remain above 10% through the end of 2010.15 In conclusion, if the current rate of growth in potential output is as low as 2%, then real economic growth would not have to be much above that to eventually end the rise in the unemployment rate. Nonetheless, history suggests that even after the contraction is over, the unemployment rate may continue to rise or at least remain steady at relatively high rates for some time.

End Notes 1

For example, in October 2009, the National Association for Business Economics released the results of a survey of 44 professional forecasters, the ―vast majority‖ of whom felt that the recession was already over. See http://www.nabe.com/publib/macsum.html. 2 See also CRS Report R40798, Unemployment and Employment Trends Before and After the End of Recessions, by Linda Levine. 3 Alan Blinder, ―Is There A Core of Practical Macroeconomics That We Should All Believe?,‖ American Economic Review, vol. 87, no. 2, May 1997.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

112

W. Brian Cashell

4

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

If there is considerable slack in the economy, as is now the case, this does not pose a problem, but once unemployment reaches relatively low levels then the increased demand for labor is more likely to be satisfied by rising wages than by higher levels of employment and there may be a risk of accelerating inflation. CBO estimated, in 2008, that the rate close to which that becomes a risk (which CBO refers to as the nonaccelerating inflation rate of unemployment or NAIRU) may have been as low as 5%. At current levels of unemployment, the risk of accelerating wages and inflation would seem low. See Robert Arnold, ―Reestimating the Phillips and the NAIRU,‖ Congressional Budget Office, Working Paper 2008-06, August 2008. 5 Using ordinary least squares. 6 David Altig, Terry Fitzgerald, and Peter Rupert, ―Okun’s Law Revisited: Should We Worry about Low Unemployment?,‖ Federal Reserve Bank of Cleveland, Economic Commentary, May 15, 1997, available at http://www.clevelandfed.org/research/commentary/1997/0515.htm. 7 Data available at ftp://ftp.bls.gov/pub/special.requests/ep/labor.force/clfa0616.txt. 8 For example, in the very early stages of an economic expansion, productivity growth tends to be relatively rapid, as the most productive of the idle labor force are put back to work first. 9 See CRS Report RL34677, Productivity Growth: Trends and Prospects, by Brian W. Cashell. 10 As measured by changes in output per hour of the nonfarm private business sector. Data published by the Department of Labor, Bureau of Labor Statistics. 11 This may not be a rigorous definition, other standards could be used. See Figure 2 to appreciate the time it takes for the unemployment rate to decline significantly after the end of an economic contraction. 12 Robert J. Gordon, ―The Slowest Potential Output Growth in U.S. History: Measurement and Interpretation,‖ presented at the Center for the Study of Innovation and Productivity at the Federal Reserve Bank of San Francisco, November 2008, available at http://www.frbsf.org/csip/research/200811_Gordon.pdf. 13 Susanto Basu and John G. Fernald, ―What Do We Know and Not Know About Potential Output?,‖ Federal Reserve Bank of San Francisco, Working Paper, March 2009, available at http://www.frbsf.org/ publications/ economics/papers/ 2009/wp09-05bk.pdf. 14 Congressional Budget Office, The Budget and Economic Outlook: An Update, August 2009, available at http://www.cbo.gov/ftpdocs/105xx/doc10521/08-25-BudgetUpdate.pdf. 15 Organisation for Economic Co-operation and Development, ―OECD Employment Outlook: Tackling the Jobs Crisis,‖ 2009.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

In: Labor and Employment Issues Editor: Nickolas H. Mullen

ISBN: 978-1-60741-286-1 © 2010 Nova Science Publishers, Inc.

Chapter 7

UNEMPLOYMENT: ISSUES AND POLICIES



Jane G. Gravelle1, Thomas L. Hungerford2 and Marc Labonte3 1

Economic Policy, 2Public Finance, 3Macroeconomic Policy

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

SUMMARY The National Bureau of Economic Research (NBER) has declared the U.S. economy to be in recession since December 2007. The unemployment rate in December 2007 was 4.9%; by October 2009, the unemployment rate was above 10%. Although economic output began to grow in the third quarter of 2009, many economists expect that the labor market will remain weak into 2010. In response to high unemployment, some Members of Congress have proposed a job creation bill. This follows several policy steps taken since the economy entered the recession, including stimulus bills in 2008 (P.L. 110-185) and 2009 (P.L. 111-5), an unprecedented expansion in direct assistance to the financial sector by the Federal Reserve, and the creation of the Troubled Asset Relief Program (P.L. 110-343). President Obama, in a speech on December 8, 2009, proposed an additional stimulus package, which would include tax and other benefits for small business, infrastructure investments, incentives to promote energy efficiency, an extension of benefits for the unemployed, aid to state and local governments, and emergency assistance. The Jobs for Main Street Act of 2009 (H.R. 2847) passed the House on December 16, 2009, and included an extension in unemployment insurance benefits and Consolidated Omnibus Budget Reconciliation Act (COBRA) health benefits, aid to troubled U.S. states and small businesses, and an increase in infrastructure spending. In addition, some policy analysts have proposed a small business hiring subsidy modeled on the 1977-1978 New Jobs Tax Credit. Most of the proposals discussed as part of a potential additional macroeconomic jobs bill are traditional fiscal stimulus policies. That is, their objective is to increase total spending (aggregate demand) either through direct spending on programs or by providing funds to others that they will spend (through tax cuts, transfer payments, and aid to state and local 

This is an edited, reformatted and augmented version of a CRS Report for Congress publication dated January 2010.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

114

Jane G. Gravelle, Thomas L. Hungerford and Marc Labonte

governments). Fiscal stimulus is only effective when the policy options actually increase aggregate demand. Some argue that the job tax credit proposal is different from traditional fiscal policies in that its objective is to directly increase employment through a subsidy to labor costs. Studies that examined the 1977-1978 jobs tax credit found mixed results—some conclude that the tax credit was responsible for creating a significant number of jobs, while others conclude that it was ineffective. The choice of financing affects both the macroeconomic impact and the cost-benefit tradeoff of the policy proposal. Policy measures can be financed by cutting other spending, raising other taxes, or increasing the budget deficit. Economic theory indicates that a deficitfinanced policy proposal would have the maximum impact on employment in the short term. Policy changes that increase the deficit, however, move the budget further from long-term sustainability. Some policymakers have proposed redirecting funds under the Troubled Asset Relief Program (TARP) to finance job creation proposals. Proposals to redirect TARP funds to finance job creation proposals in essence pay for those proposals by reducing the amount that the Treasury Secretary is authorized to purchase under TARP by the cost of the proposal. Since TARP is not near that ceiling today, any proposal that reduces TARP authority by less than $150 billion would not force TARP to be reduced from its currently planned size. Therefore, it would cause the actual budget deficit to increase from the current deficit by the size of the job creation proposal. In response to high unemployment, some Members of Congress have proposed a job creation bill. This follows several policy steps taken since the economy entered a recession in December 2007, including stimulus bills in 2008 (P.L. 110-185) and 2009 (P.L. 111-5), an unprecedented expansion in direct assistance to the financial sector by the Federal Reserve, and the creation of the Troubled Asset Relief Program (P.L. 110-343). President Obama, in a speech on December 8, 2009, proposed an additional stimulus package. This package would include tax and other benefits for small business, infrastructure investments, incentives to promote energy efficiency, and an extension of measures provided in the 2009 legislation with respect to benefits for the unemployed, aid to state and local governments, and emergency assistance. H.R. 2847, the ―Jobs for Main Street Act of 2009,‖ passed the House on December 16, 2009. It included an extension in unemployment insurance benefits and COBRA health benefits, aid to troubled U.S. states and small businesses, and an increase in infrastructure spending. This report discusses the current unemployment outlook, and analyzes policy proposals to address the issue and options for financing proposals.

THE UNEMPLOYMENT PROBLEM At the onset of the recession in December 2007, the unemployment rate climbed steadily from a rate of 4.9%, topping 10% in October 2009. This marked the first time since 1982 that unemployment topped 10%; prior to that, unemployment had not topped 10% since the Great Depression. This recession has been characterized by the biggest percentage point increase in the unemployment rate of any recession since World War II. The duration of unemployment

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Unemployment: Issues and Policies

115

has also been rising. Since mid-2009, over half of those unemployed had been so for 15 weeks or longer. While long-term unemployment has consistently trended upward in past recessions since 1970, the number of long-term unemployed workers has been greater in absolute terms in this recession. The rise in unemployment has been driven by a steady decline in employment. The decline in employment during this recession can be separated into three phases. From January to August 2008, employment fell between 72,000 and 175,000 per month. From September 2008 to July 2009, employment fell between 304,000 and 741,000 per month. In August 2009, employment declined by 154,000, and each subsequent month has seen a smaller decline than the month before.1 Nevertheless, employment has continued to fall in every month since December 2007. Economic output tends to reaches its trough and begin growing again before employment. Figure 1 illustrates that employment has typically begun a sustained increase within four months after the recession has ended, but in the last two ―jobless recoveries,‖ employment did not start rising for 12 and 22 months, respectively. The atypical performance of employment in the last two recoveries has at least two possible explanations. First, it may point to fundamental changes in the labor market that could have bearing on job creation in this recovery. Alternatively, it may point to the fact that short and shallow recessions such as the previous two are more likely to have initially weak recoveries, which suggests that a jobless recovery is unlikely to be repeated this time around. A ―hands off‖ policy approach would counsel for patience—the fall in unemployment will be gradual, but it is also inevitable. Every recession since World War II except the 1980 recession was followed by a period of sustained job creation.2 As Figure 1 indicates, although sustained job creation usually happens quickly, employment growth occurred with a long lag after the last two recessions. In the third quarter of 2009, economic output began to grow again. If the economy continues to grow, unemployment will eventually begin to decline. Many economists are projecting that unemployment will peak in mid-2010, and then begin to slowly fall.3 Historical experience confirms that strong economic growth is the most important factor for reducing unemployment after a recession.4 Nevertheless, because the unemployment rate is so high, even if the economy grew at a healthy rate, it would take a significant amount of time for unemployment to reach more normal levels. For example, after the unemployment rate peaked at 10.8% in December 1982, the unemployment rate was 8.3% one year later, and it took almost five years for the rate to fall by half. This gradual decline happened over a five-year period when economic growth averaged an unusually high rate of 4.5% annually. Another argument in favor of patience is that the government has already taken extraordinary steps to stabilize the economy through the creation of the Troubled Asset Relief Program, the Federal Reserve’s emergency lending facilities, and fiscal stimulus in 2008 and 2009, the latter of which contains significant outlays through 2011. These programs will be discussed in greater detail in the following section, entitled ―Policy Steps Taken Through 2009.‖ The full economic effects of these previous policy steps will take time to be felt. Proponents of this approach are likely to argue that, with these policies already in place, it is unlikely that further policy steps could sharply hasten the anticipated decline in unemployment.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

116

Jane G. Gravelle, Thomas L. Hungerford and Marc Labonte

Source: CRS Report R40798, Unemployment and Employment Trends Before and After the End of Recessions, by Linda Levine, based on data from the Bureau of Labor Statistics jobs data and Business Cycle Dating Committee for end date of recessions.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Figure 1. Number of Months Into Recovery When Sustained Job Growth Began

A more interventionist policy approach could be justified on at least two grounds. First, the loss in output caused by high unemployment is very costly in economic and noneconomic terms in the short run. If policy steps to reduce unemployment can be taken at relatively low costs, then the cost-benefit tradeoff would be favorable. A major policy debate at this time, discussed below in the section entitled ―Issues in Financing Job Creation Proposals,‖ focuses on just how costly financing these additional policy steps would be. The second rationale depends on whether high unemployment has any permanent effects. Mainstream economic theory suggests that the business cycle has no lasting effect on the natural rate of unemployment—booms and busts temporarily move the unemployment rate, but it always gravitates back toward its long-term equilibrium rate. In this view, policy steps could hasten the return to the natural rate, but market forces would eventually have caused unemployment to return to the same long-run level on its own. In other words, policy steps would result in temporary (but not permanent) improvements in well-being. Some economists have offered a competing theory called ―hysteresis,‖ however. In this view, bouts of high unemployment can lead to permanent increases in the natural rate of unemployment, so that unemployment never falls as low in the subsequent recovery as it had been at the previous peak. This theory was developed to explain the failure of high unemployment to fall in Western Europe in the 1980s expansion.5 Hysteresis could result from imperfections in labor markets or from workers losing some of their skills in long bouts of unemployment that reduce their subsequent employability. If hysteresis effects are significant, then policy steps that successfully reduce unemployment now could avoid some permanent loss in well-being.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Unemployment: Issues and Policies

117

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

POLICY STEPS TAKEN THROUGH 2009 The National Bureau of Economic Research (NBER) declared the U.S. economy to be in recession since December 2007. Numerous actions have already been taken to contain damages spilling over from housing and financial markets to the broader economy. These policies include traditional monetary and fiscal policy, as well as federal interventions into the financial sector. In February 2008, in response to weaker economic growth, an economic stimulus package of approximately $150 billion was adopted.6 A provision that was considered (but not enacted) in the February stimulus bill was a 26-week extension of unemployment benefits; this extension was eventually enacted.7 With the worsening performance of the economy, congressional leaders and President Obama proposed much larger stimulus packages at the beginning of 2009. The American Recovery and Reinvestment Act of 2009 (ARRA), an $820 billion package with $275 billion in tax cuts (offset by a $7 billion gain from the treatment of built-in losses) and the remainder in spending, was passed by the House on January 28 (H.R. 1). It was a wide-ranging package that included infrastructure spending, revenue sharing with the states, middle class tax cuts, business tax cuts, unemployment benefits, and food stamps. Similar legislation was passed in the Senate on February 10 (an amendment in the nature of a substitute for H.R. 1) and would cost $838 billion, with $292 billion in tax cuts. The version of the bill signed into law on February 17, 2009 (P.L. 111-5), was a $787 billion package with $286 billion in tax cuts and the remainder in spending.8 The largest budgetary effects of P.L. 111-5 occur in FY20 10 (equaling 2.2% of GDP, compared to 1.3% in 2009), so this legislation has not yet had its peak effect on the economy. A number of financial-sector interventions have also been undertaken, before and after financial market conditions worsened significantly in September 2008. The Federal Reserve (Fed) has reduced short-term interest rates to zero and introduced a number of facilities, providing direct assistance to the financial system that would eventually surpass $1 trillion. In October 2008, legislation was enacted granting the Treasury Department authority to purchase up to $700 billion in assets through the Troubled Assets Relief Program (TARP).9 A number of programs have been created under TARP, including programs to inject capital into banks, aid automakers and troubled financial firms, and modify mortgages. The Federal Reserve is close to completing purchases of $1.25 billion of mortgage-backed securities, $175 billion in Government Sponsored Enterprise (GSE) debt, and $300 billion of long-term Treasury debt. On March 23, 2009, the Treasury announced a plan for a public-private partnership to purchase troubled assets, including one part that uses the Federal Deposit Insurance Corporation (FDIC) to insure loans and another part that would allow access to the Federal Reserve’s Term Asset-Backed Securities Loan Facility (TALF).10 Other policies enacted in response to the financial crisis include an FDIC guarantee of debt issued by banks, a Treasury guarantee of money market mutual funds, and Treasury support of the GSEs.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

118

Jane G. Gravelle, Thomas L. Hungerford and Marc Labonte

POLICY PROPOSALS The House has passed H.R. 2847, which extends unemployment benefits, food stamps, and Medicare and COBRA health benefits, and provides additional funds for infrastructure and aid to state and local governments. The Senate Leadership has indicated that they will consider a jobs bill early in 2010. President Obama, in a speech on December 8, 2009, proposed an additional stimulus package to include tax and other benefits for small business (including incentives for hiring), infrastructure investments, incentives to promote energy efficiency, and an extension of relief provided in the 2009 legislation for benefits for the unemployed, state aid, and emergency assistance. Proposals have also been circulating for some time to provide an incremental jobs tax credit.

House Proposal (H.R. 2847)

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

On December 16, the House passed the Jobs for Main Street Act of 2010, H.R. 2847. This bill provided for $154 billion in spending, allocated to three categories: infrastructure spending, public service jobs, and emergency funding (see Figure 2). The last category includes a tax cut. The proposal redirects TARP funds to finance the spending.11 The bill also provides for an extension of the surface transportation authorization through the fiscal year and an extension of the statute of limitations for claims of discrimination in USDA’s credit programs and provides funding for remedies.

Source: CRS. Figure 2. Distribution of Spending under The Jobs for Main Street Act (H.R. 2847)

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Unemployment: Issues and Policies

119

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Infrastructure Spending The proposal allots $48.3 billion (31% of the total) for additional infrastructure spending. About three quarters of the infrastructure spending is for surface transportation (57% or $27.5 billion for highways, 17% or $8.4 billion for mass transit, and 1.7% or $800 million for Amtrak). The transit funds include $6.15 billion for urban and rural formula grants, $500 million for capital investment grants for new or expanded fixed guideway projects, and $1.75 billion in formula grants for repair. The Amtrak funds are capital grants. Other transportation projects include $500 million for airports improvement and $100 million for the Maritime Guaranteed Loan program. Slightly over 8% of the infrastructure total ($4.1 billion) is allocated to school renovation grants. About 6% ($2.8 billion) is allocated to environmental, flood, and similar spending: $2 billion for clean water ($1 billion for the Clean Water State Revolving Fund and $1 billion for the Safe Water Drinking Fund), $100 million for the Bureau of Reclamation to provide water in rural areas, and $0.7 15 billion for Corps of Engineers projects. The proposal also includes $2 billion for the Department of Energy Innovation Technology Loan Guarantee program for renewable energy, $1 billion for the National Housing Trust Fund to provide low-income rental housing, along with $65 million in vouchers, and $1 billion for repair and rehabilitation of public housing. Public Service Jobs The bill provides for $26.7 billion in funding (17% of the total) for public service jobs, with the bulk of funds ($23 billion, or 86% of the total in this category) for education. About 95% of these funds would be allocated to school districts and public institutions of higher learning, to retain or create jobs for teachers and other educational services and to modernize and repair facilities. The remaining 5% will be allocated to the states. The plan also provides funds to support law enforcement jobs ($1.18 billion) and firefighter jobs ($0.5 billion). The bill also provides the following amounts to support other types of jobs: $200 million for Americorps, $500 million for summer youth employment, $300 million for the college work study program, $270 million for parts and forestry workers, and $750 million for competitive grants to support job training in high-growth and emerging industry, particularly health care and green industries. Emergency Funding Over half of the total funds, $79 billion, would be directed at relief provisions. Almost all of this category (97%) is directed at three programs: a six-month extension of expanded unemployment benefits ($41 billion, or 52%), a six-month extension of the higher Medicaid match for doctors payments ($23.5 billion, or 3 0%), and an extension from nine to 15 months of the subsidy for health insurance for unemployed workers ($12.3 billion or 16%) under COBRA. The package also includes $2.3 billion to increase the eligibility for refundability of the child tax credit. This latter provision would remove the $3,000 floor that occurs before the child credit can become refundable. The plan also provides $354 million to continue Small Business Administration temporary loan authority guarantees through FY2010 and $305 million to freeze Department of Health and Human Services (HHS) poverty guidelines at 2009 levels to prevent a reduction in a variety of means-tested benefits. The bill also contains two provisions with negligible revenue cost: a provision that individuals may exclude tax

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

120

Jane G. Gravelle, Thomas L. Hungerford and Marc Labonte

refunds for assessing eligibility for means- tested programs for a year, and a provision for legal assistance to Social Security and Supplemental Security Income disability claimants.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

President Obama’s Proposal In a speech at the Brookings Institution on October 8, President Obama proposed a fourpart stimulus plan. 12 The plan includes the general elements in the House plan and some additional proposals. The first part of the plan would provide tax benefits and loans to small businesses. The proposal includes an unspecified tax incentive to keep and add employees. Administration officials have, according to news reports, suggested that tax credits for creating jobs or payroll tax holidays are under discussion. The plan would also extend a provision to allow increased expensing for equipment purchases by small businesses, enacted in the 2009 stimulus bill for 2009 through 2010. Under the increased expensing provision, firms may deduct up to $250,000 of the cost of equipment, with a phaseout beginning at $800,000.13 The plan would also completely exempt capital gains on small business stock. This provision has a relatively small revenue effect.14 The small business proposals also include plans to use remaining TARP funding for small business loans, which could be done without congressional action. The second part of the proposal is an increase in infrastructure spending, to extend further into the future than the spending authorized in the 2009 stimulus. The third part of the proposal is to provide incentives for individuals to retrofit homes for energy efficiency and to expand other initiatives in the 2009 stimulus to promote energy efficiency and clean-energy jobs. The final part of the proposal is to extend relief in the 2009 stimulus, including emergency assistance to seniors, unemployment insurance benefits, COBRA health benefits for those who have lost their health insurance, and relief to state and local governments to prevent layoffs. The President also indicated support of the House bill, H.R. 2847, as a complement to his plans, in a statement released on December 16, 2009.15

Incremental Jobs Tax Credit A proposal that has been circulating for some time, and that might be considered as a small business hiring incentive, is an incremental jobs tax credit. This type of credit would provide benefits for hiring employees in excess of a base amount. The United States had one historical experience with this type of credit in 1977 and 1978 (the New Jobs Tax Credit).16 The tax code has contained for some time permanent tax credits targeted at certain types of workers, and the target groups were expanded somewhat in 2009.17 These credits are applicable to newly hired workers from the targeted groups without requiring an increase in a firm’s total employment, however.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Unemployment: Issues and Policies

121

Discussion

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Both monetary and fiscal policy can be used to stimulate the economy. The proposals addressed above include direct spending by the government, transfers to state and local governments (for either infrastructure spending, Medicare, or other purposes), direct transfers to individuals (such as unemployment compensation), tax cuts for individuals, and tax incentives aimed at businesses, including jobs tax credits. Jobs subsidies differ from policies aimed at increasing aggregate demand, in that they are supply-side subsidies. That is, the initial effect is not aimed at inducing spending that will then encourage firms to expand output and hire workers (although it may do so), but is aimed at reducing the cost of hiring workers, so as to induce more hires. The first section below discusses traditional fiscal policies, the second discusses incentives aimed at jobs.

Spending, Transfers, and Tax Cuts Most of the proposals discussed as part of a potential additional macroeconomic jobs bill are traditional fiscal policies similar to those enacted in 2009. That is, their objective is to increase total spending (aggregate demand) either through direct spending on programs or by providing funds to others that will spend (through transfer payments, tax cuts, and aid to state and local governments). The House bill is composed almost entirely of spending and transfers (with a small tax cut, the increased child credit refundability). President Obama’s proposal does not contain details, although it appears that, while much of the proposal involves spending and transfers, part of the proposal will be aimed at tax cuts for small businesses. (This section of the proposal would also include a subsidy for labor, discussed in the next subsection). The issues surrounding these fiscal instruments are the same as those relating to the previous stimulus, except that it is later in the business cycle and there is a greater possibility that the provisions may come later than is desirable.18 Many economists view fiscal policy as less effective than monetary policy in an open economy. 19 Fiscal stimulus is only effective from a macroeconomic perspective when it increases aggregate demand. The size of the proposal and financing are the most important determinants of its effect on aggregate demand. For example, the House bill is about 1% of output, about the size of the February 2008 stimulus, but considerably smaller than the February 2009 stimulus (which was estimated to be 1.3% of output in 2009 and 2.2% in 2010). As discussed in the section entitled ―Issues in Financing Job Creation Proposals,‖ in the standard macroeconomic model, only deficit- financed proposals would have a significant effect on aggregate demand. Fiscal stimulus can involve tax cuts, spending increases, or a combination of both. Tax cuts may be less effective than spending because some of the tax cut may be saved, which diminishes the effectiveness of the stimulus.20 Some argue that tax cuts that are temporary, appear in a lump sum rather than in withholding, or are aimed at higher-income individuals are more likely to be saved.21 Transfer payments are similar to tax cuts, but tend to be received by lower-income individuals who are more likely to spend them. Evidence generally suggests that tax subsidies for business tax cuts are not very effective.22 Mark Zandi, for example, has suggested that the multipliers (the dollar increase in real output for each dollar of stimulus) for personal tax cuts range between 1 and 1.3, the multipliers for spending and

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

122

Jane G. Gravelle, Thomas L. Hungerford and Marc Labonte

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

transfer payments range over 1.5, and multipliers for transfers to state and local governments range under 1.5.23 The challenge to spending programs is that there may be a lag time for planning and administration before the money is spent. Some analysts suggest that aid to state and local governments may be spent more quickly because these governments are likely to cut back on spending in downturns due to balanced budget requirements, and the aid may forestall these cuts. Extending ARRA infrastructure spending is unlikely to be spent quickly since all the spending under ARRA has not occurred yet. The receipt of tax cuts can also be delayed, if they are delivered through changes to withholding or through a delayed refund. If a stimulus is considered or enacted as the economy is beginning to recover, its benefits may be limited given these lags. Subsidies to business investment are, like other policies, aimed at increasing aggregate demand (through increased investment spending). The small business investment subsidies suffer from the same problems confronting business subsidies for investment in general, namely, although in theory a temporary subsidy should be the most effective investment stimulus, evidence from prior investment subsidies does not suggest that such subsidies are very effective.24 The lack of effectiveness may occur in part because businesses with losses cannot take advantage of the provision and in part because firms may already have excess capacity. The extension of the expensing provision for small business, however, has mixed effects because firms in the phaseout range have a marginal disincentive to invest. In any case, the potential effect on spending is limited by the fact that these provisions are relatively small in revenue effect.

Job Credits Some argue the job credit provisions are different from traditional fiscal policies in that their objective is to directly increase employment through a subsidy to labor costs. A general subsidy to labor (such as a forgiveness of the employer’s share of payroll taxes) would be relatively costly. (In the short run, a forgiveness of the employee’s share would be similar to an individual income tax cut while forgiveness of the employer’s share would be similar to a job credit.) More research has been done on an incremental job tax credit, which was a policy adopted in 1977 and 1978. Two proponents of this policy, Bartik and Bishop, have argued that the proposal will be successful in creating a significant number of jobs.25 Their estimates were done by assuming a labor demand elasticity of 0.3, which indicates that a 10% reduction in the cost of labor would increase employment by 3%. Note that their estimates did not rest on a study of the 1977-78 credit, but rather predicted the effect on jobs based on a central tendency labor demand elasticity. 26 Note that this estimate is a general demand elasticity, and might not necessarily be as high during a recession, when business is slack. Studies that examined the 1977-1978 credit found mixed results. Bishop studied the construction, retailing, and wholesaling industries, accounting for the effect of the jobs credit, and found that the credit was responsible for 150,000 to 600,000 of the 1 million increase in employment during that period.27 Perloff and Wachter compared firms who knew about the credit with those who did not and found employment growth to be greater among the former group, although they caution that this is not a random selection and there may be characteristics about firms with more knowledge that could independently affect growth. Overall, they seem to conclude that the credit did not work very well because many firms

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Unemployment: Issues and Policies

123

were not aware of it, and many firms did not have enough employment growth.28 Tannenwald surveyed Wisconsin and New England firms.29 He found that the effect was smaller than predicted. He indicated that most estimates of the labor demand response to a change in wages indicate that a 10% change in wages lead to labor demand increases of 2%. These estimates are general estimates, not associated with a downturn. He found an increase of only 0.4%, less than a quarter of the projected effects. The major reason was the lack of product demand. For example, one quote from his survey was: ―Orders determine levels of hiring, not tax gimmicks.‖ The main reservation about a jobs tax credit is that it might not be effective in those industries that are experiencing slack demand, causing the labor demand elasticity, already low in normal times, to approach a very low level.30 While an incremental credit can have a larger ―bang for the buck‖ by only providing subsidies for additional hiring, it is also much more complicated and can possibly be evaded (for example, firms may hire their contractors temporarily). The 1977-1978 credit was made incremental in Congress (presumably to increase bang for the buck), but an incremental subsidy was opposed by the Carter Administration because of complexity and unfairness. Sunley discusses a variety of distortions that arise from an incremental credit, depending on the design, such as hiring part-time workers instead of full-time, reducing overtime, and hiring and replacing workers. Also, it automatically favors firms that are growing anyway, which leads to geographic differentials.31

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

ISSUES IN FINANCING JOB CREATION PROPOSALS Another consideration is how to finance any proposed policy measure. The choice of financing affects both the macroeconomic impact and the cost-benefit tradeoff of the policy proposal. Policy measures can be financed by cutting other spending, raising other taxes, or increasing the budget deficit. Economic theory indicates that a deficit-financed policy proposal would have the maximum impact on employment in the short term. In a deep recession, total spending (aggregate demand) in the economy is inadequate to fully employ labor and capital resources. In other words, lack of aggregate demand is the main cause of high unemployment. Increasing the budget deficit can increase total spending in the economy and bring some of those idle resources back into use. Deficit-neutral proposals would tend to neutralize the effects of job creation provisions on total spending in the economy by cutting other spending or lowering the spending of those whose taxes are raised. Deficit-neutral policies can be used to lower the cost of labor, but without any increase in demand for their products, employers may be unresponsive to incentives to increase their labor force. In the context of a deep recession, the short-term economic cost of increasing the budget deficit may be quite low. The main economic costs of increasing the deficit comes from its tendency to ―crowd out‖ private investment spending or increase the trade deficit.32 Deficits crowd out private investment spending because their financing requires scarce private saving. Increasing the demands on this private saving raises interest rates, making private investment spending less attractive. In the current context, investment spending has been greatly reduced by the recession, so there is less chance of it being crowded out by the larger deficit in the short run. Unusually low Treasury bond rates are evidence that the crowding out factor is not

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

124

Jane G. Gravelle, Thomas L. Hungerford and Marc Labonte

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

significant at present.33 Deficits and domestic private investment spending can also be financed through foreign capital flows, however. An increase in net foreign capital inflows must be matched by an equal increase in the trade deficit.34 With perfect capital mobility, the stimulus to total spending caused by the larger deficit could be entirely offset by the decline in total spending resulting from a larger trade deficit. Since the trade deficit has fallen significantly since the beginning of 2007, this drawback to increasing the deficit may also be less important at present. While an economic argument can be made that increasing the deficit could have shortterm benefits, that argument may presuppose that the increase in the deficit would be reversed when economic conditions return to normal. Political constraints may make that difficult, and could lead one to conclude that the short-term benefits of higher deficits would be outweighed by the long-term costs—namely, that if deficits are not reduced when unemployment falls, the effects on investment spending and the trade deficit would become greater. Indeed, any proposal to increase the deficit can be viewed in the broader context of an overall deficit that in 2009 was larger relative to the size of the economy than all but a handful of previous wartime years. The 2009 deficit is not sustainable in the long run in the sense that deficits of that size would cause the national debt to continually rise relative to output. A deficit of this size cannot be maintained indefinitely without eventually resulting in a fiscal crisis where investors refuse to continue financing it because they no longer believe that the government would be capable of servicing it. While there is no sign of investor unwillingness to hold federal debt at the present (since borrowing rates are so low), it is also difficult to predict at what point investors would refuse to hold more debt. Essentially, investors are willing to hold federal debt as long as they believe that the government will eventually reduce the deficit to the point where it becomes sustainable. Policy changes that increase the deficit place the deficit further from sustainability.35

Redirecting TARP Funds to Finance Proposals Some policymakers have proposed redirecting funds under the Troubled Asset Relief Program (TARP) to finance job creation proposals. H.R. 2847, which passed the House on December 16, 2009, redirected $150 billion of TARP funds to finance the bill’s other provisions. Under the Emergency Economic Stabilization Act (Division A of H.R. 1424/P.L. 110-343), Congress authorized up to $700 billion to be outstanding for the life of the program to purchase ―troubled assets,‖ as defined by the Treasury Secretary. In its November 2009 monthly report to Congress, Treasury reported that $550 billion of TARP expenditures had been planned and $476 billion had been committed under signed contracts, signifying that Treasury has no plans at this time to use $150 billion of the funds authorized.36 Treasury’s planned use of available TARP funds can change at any time, however. When TARP was created, the Treasury did not collect and set aside $700 billion of revenue to finance the program—the Treasury Secretary was simply given legal authority to purchase $700 billion of assets. Therefore, Treasury holds no unused money under TARP that can be redirected toward new policy proposals. Like most spending programs, TARP expenditures are financed from general revenues. When the budget is in deficit, additional expenditures are financed by additional borrowing. If the Treasury Secretary wished to

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Unemployment: Issues and Policies

125

purchase more TARP assets, it would be necessary to first issue federal debt (thereby increasing the budget deficit) to do so. The size of the actual deficit today is based on the cost of TARP expenditures taken during this fiscal year (the method for measuring the cost is described below); the projected size of the future deficit in budget projections is based on some assumption about how much the Treasury Secretary plans to increase the size of TARP in the future and how much these future expenditures will cost. Proposals to redirect TARP funds to finance job creation policies rely, in essence, on a reduction in the amount that the Treasury Secretary is authorized to purchase under TARP. Since TARP is not near its ceiling today, any proposal that reduces TARP authority by less than $150 billion would not force TARP asset holdings to be reduced from the currently planned size. Thus, reducing the authorized size of TARP by less than $150 billion does not increase the revenues flowing to the Treasury because it does not force Treasury to sell any of the assets TARP currently holds. In effect, a new policy proposal that increases spending or reduces revenues would be deficit-financed if it included a reduction in TARP authority of less than $150 billion under Treasury’s current plan (since it would not result in any increase in revenues via a reduction in TARP assets outstanding). A separate issue is whether redirecting TARP funds to finance job creation proposals would increase the expected budget deficit in the future compared to projections of current policy. This measure would be relevant for scoring the proposal and for measuring the proposal’s macroeconomic effects relative to a baseline of current policy. The answer to this question depends on what assumption is made about the future size of the TARP program under current policy baselines, which need not be the same as Treasury’s announced plans since those plans are not binding. Through March 2009, the Congressional Budget Office (CBO) baseline assumed that the amount outstanding under TARP would reach $700 billion. In August 2009, CBO modified its baseline and assumed that TARP would peak at $600 billion, or $50 billion higher than Treasury’s current plans. That implies that reducing TARP by up to $100 billion would have no impact on future projections of the baseline deficit (because the baseline already assumed that the redirected money would not be spent by TARP), so redirecting TARP by up to that amount to finance a new proposal would be fully matched by an increase in the deficit.37 From a macroeconomic perspective, a job creation proposal to be financed by a decrease in TARP authority of up to $100 billion would have the same effect on economic projections compared to the August baseline as if it were deficitfinanced. For official scoring purposes in 2009, the 2009 budget resolution (S.Con.Res. 13) instructs CBO to use the March baseline, even though a more recent baseline is now available. (The same baseline is used so that scoring will be consistent throughout the session.) Therefore, a bill financed by redirecting any TARP funds would be officially scored as being offset by a decline in overall anticipated federal spending via a smaller TARP program since the March baseline assumed all $700 billion of TARP authority would be used in the future. The offset would not be one-for-one, however. Under Section 123 of EESA, the cost of asset purchases are scored as the net present value of the subsidy in the loan, modified for risk, and are not scored on a cash flow basis. For future TARP spending, CBO assumes a subsidy rate of 50%. Therefore, a dollar reduction in TARP authority is scored as reducing the official budget deficit by only 50 cents. 38

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

126

Jane G. Gravelle, Thomas L. Hungerford and Marc Labonte

End Notes

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

1

The data described here are net figures from the Bureau of Labor Statistics’ establishment survey. Despite the recession, gross job gains at firms averaged 7 million per quarter in 2008, down from 7.6 million per quarter in 2007. These gross job gains have been more than offset by gross job losses, however. Gross job losses at firms averaged 7.9 million per quarter in 2008, up from 7.4 million per quarter in 2007. 2 In the case of the 1980 expansion, the economy slid back into recession in 1981. 3 See Blue Chip Economic Indicators, Vol. 34, No. 12, December 10, 2009. 4 See CRS Report R40925, Unemployment and Economic Recovery, by Brian W. Cashell. 5 See Olivier Blanchard and Lawrence Summers, ―Hysteresis and the European Unemployment Problem,‖ in Stanley Fischer, ed., NBER Macroeconomics Annual, vol. 1 (Cambridge: MIT Press, 1986), p. 15. 6 A second stimulus plan (H.R. 7110) passed the House on September 26, 2008, but was not passed by the Senate before the 110th Congress ended. It included $36.9 billion on infrastructure ($12.8 billion highway and bridge, $7.5 billion water and sewer, $5 billion Corps of Engineers); $6.5 billion in extended unemployment compensation; $14.5 billion in Medicaid; and $2.7 billion in food stamp and nutrition programs. 7 For a discussion of the tax, housing, and unemployment legislation adopted in the 110th Congress, see CRS Report RS22850, Tax Provisions of the 2008 Economic Stimulus Package, by Jane G. Gravelle; CRS Report RS22 172, The Conforming Loan Limit, by N. Eric Weiss and Mark Jickling; and CRS Report RS22915, Temporary Extension of Unemployment Benefits: Emergency Unemployment Compensation (EUC08), by Julie M. Whittaker and Alison M. Shelton. 8 For a discussion of the American Recovery and Reinvestment Act of 2009, see CRS Report R40104, Economic Stimulus: Issues and Policies, by Jane G. Gravelle, Thomas L. Hungerford, and Marc Labonte. 9 See CRS Report RL34427, Financial Turmoil: Federal Reserve Policy Responses, by Marc Labonte for a discussion of Federal Reserve Policy, and CRS Report RL34730, Troubled Asset Relief Program: Legislation and Treasury Implementation, by Baird Webel and Edward V. Murphy. 10 For further discussion, see CRS Report RL34730, Troubled Asset Relief Program: Legislation and Treasury Implementation, by Baird Webel and Edward V. Murphy. 11 The estimates presented in this section are taken from Jobs for Mainstreet, posted on Speaker Pelosi’s website at http://www.speaker.gov/newsroom/legislation?id=0351, visited December 22, 2009. 12 The proposal is discussed at http://www.whitehouse.gov/the-press-office/president-obama-announces-proposalsaccelerate-job-growth-and-lay-foundation-robust. 13 Currently the allowance is scheduled to drop to $125,000 for 2010, with a phaseout beginning at $500,000. This higher level is itself a temporary provision as part of the 200 1-2003 tax cuts, so that the allowance will fall to $25,000, with a phaseout beginning at $200,000 in 2011. 14 A provision exempting 50% of the gain on small business stock and taxing the remaining gain at 28% has been in the tax law since 1993, but its relative benefit largely disappeared when capital gains tax rates were reduced to 15%. The 2009 legislation increased the exemption to 75% and the President’s proposal would increase it to 100%. See CRS Report R40728, Small Business Tax Benefits and the American Recovery and Reinvestment Act of 2009, by Gary Guenther for a discussion. 15 See http://www.whitehouse.gov/the-press-office/statement-president-house-passage-jobs-bill. 16 See CRS Report 92-939, Countercyclical Job Creation Programs, by Linda Levine for a discussion. 17 For more information, see CRS Report RL3 0089, The Work Opportunity Tax Credit (WOTC), by Linda Levine. 18 See the discussions below addressing the issues of financing the jobs programs from TARP funds, which indicate that this assumption does not change the characterization of stimulus proposals. See CRS Report R40104, Economic Stimulus: Issues and Policies, by Jane G. Gravelle, Thomas L. Hungerford, and Marc Labonte for a more extensive discussion of the issues surrounding the 2009 stimulus. See CRS Report RS22790, Tax Cuts for Short-Run Economic Stimulus: Recent Experiences, coordinated by Jane G. Gravelle for evidence on the effectiveness of recent policy options. 19 When fiscal expansion raises the deficit and drives up interest rates, capital is attracted from abroad. The purchase of U.S. dollars by foreigners to buy U.S. assets drives up the price of the dollar, causing export demand to decline. This reduction in the demand for exports offsets in part (perhaps in large part) the initial increase in demand induced by the stimulus. The more mobile international capital flows are the larger offsetting effect. 20 See CRS Report RS21136, Government Spending or Tax Reduction: Which Might Add More Stimulus to the Economy?, by Marc Labonte. 21 CRS Report RS21126, Tax Cuts and Economic Stimulus: How Effective Are the Alternatives?, by Jane G. Gravelle. 22 See CRS Report RL31134, Using Business Tax Cuts to Stimulate the Economy, by Jane G. Gravelle. 23 Mark Zandi, Moody’s Economy.com. Also see CRS Report R40104, Economic Stimulus: Issues and Policies, by Jane G. Gravelle, Thomas L. Hungerford, and Marc Labonte, p. 16, for a list of multipliers. This report also discusses the effects of alternative tax and spending policies in more detail. 24 See CRS Report RL3 1134, Using Business Tax Cuts to Stimulate the Economy, by Jane G. Gravelle.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Unemployment: Issues and Policies 25

127

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Timothy J. Bartik and John H. Bishop, The Job Creation Tax Credit, Economic Policy Institute Briefing Paper, October 20, 2009, http://www.epi.org/publications/entry/bp248/. 26 See Daniel L. Hamermesh, Labor Demand (Princeton University Press: Princeton, NJ, 1993), for a survey: Hamermesh suggests a midpoint elasticity of 0.3 on p. 92. 27 John Bishop, ―Employment in Construction and Distribution Industries: The Impact of the New Jobs Tax Credit,‖ in Studies in Labor Markets, University of Chicago Press, 1981, pp. 209-246. 28 Jeffrey J. Perloff and Michael L. Walchter, ―The New Jobs Tax Credit,‖ American Economic Review, vol. 69, May 1989, pp. 173-179. 29 Robert Tannenwald, ―Are Wage and Training Subsidies Cost-Effective? Some Evidence from the New Jobs Tax Credit,‖ New England Economic Review, September/October 1982, pp. 25-34. 30 Although the issues are somewhat different, studies of permanent targeted jobs tax credits that are aimed at disadvantaged workers have generally found limited effects. Daniel L. Hammermesh, Labor Demand (Princeton University Press: Princeton, NJ, 1993) reviews the evidence on the effects of several earlier jobs subsidies. For studies of the current work opportunity credit, see CRS Report RL3 0089, The Work Opportunity Tax Credit (WOTC), by Linda Levine. 31 These positions, as well as problems with the credit, were discussed by a Carter Administration official, Emil Sunley, in ―Legislative History of the New Jobs Tax Credit,‖ The Economics of Taxation, Edited by Henry J. Aaron and Michael J. Boskin, Washington, DC, The Brookings Institution, 1980. See also James Leigh Griffith, ―A Critical View of the Complexity and Effect of the New Jobs Credit,‖ The Tax Lawyer, vol. 32, Fall 1978, pp. 157-179, and Roland L. Hjorth, ―New Jobs Credit,‖ Taxes: The Tax Magazine, vol. 55, November 1977, pp. 707-714, for further discussions of complexity issues. 32 For more information, see CRS Report RL3 1775, Do Budget Deficits Push Up Interest Rates and Is This the Relevant Question?, by Marc Labonte. 33 Although the credit crunch has increased the risk premium on borrowing rates and cut off access to credit for some risky borrowers, it has led to a general decline in interest rates. 34 This relationship is due to the balance of payments accounting identity (i.e., dollars sold equals dollars bought). 35 For more information, see CRS Report R40770, Economic Effects of a Budget Deficit Exceeding $1 Trillion, by Marc Labonte. 36 U.S. Department of Treasury, Troubled Asset Relief Program – Monthly Section 105(a) Report – November 2009, December 10, 2009, available at http://financialstability.gov/latest/reportsanddocs.html. 37 Congressional Budget Office, ―Budget Savings from Legislated Reductions in Amounts Outstanding under the TARP,‖ Director’s Blog, December 10, 2009, http://cboblog.cbo.gov/?p=440. 38 For example, CBO scored the $20.8 billion reduction in TARP authority in H.R. 4173 as a $10.4 billion reduction in the deficit in 2010. Congressional Budget Office, Letter to Honorable Barney Frank, December 9, 2009, http://www.cbo.gov/ftpdocs/108xx/doc10844/hr4173asreported.pdf. Since other provisions of this bill increased the deficit, the overall effect of the bill is to reduce the deficit by $7.3 billion over 10 years.

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved. Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

INDEX

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

A accountability, 110 accounting, 10, 15, 30, 42, 66, 132, 137 achievement, 82 actual output, 117 adjustment, 3, 23, 25, 103, 112, 113 age, vii, 1, 3, 4, 5, 6, 9, 10, 12, 16, 17, 21, 23, 30, 35, 36, 38, 39, 40, 41, 42, 43, 44, 46, 47, 49, 53, 54, 55, 57, 58, 59, 60, 63, 94, 103 aggregate demand, x, 123, 131, 132, 133 aging population, 63 agriculture, vii, 1, 8, 12 airports, 129 Alaska, 19, 80 altruism, 65 ambiguity, 67 American Recovery and Reinvestment Act, 21, 72, 73, 82, 84, 127, 136 annual rate, 119, 121 ash, 51 assets, 62, 66, 127, 134, 135, 136 assumptions, 64, 65, 66, 67, 81, 86, 87, 88, 91, 93, 96, 97, 98, 99, 100, 112, 113 attachment, 41, 58 attacks, 36 attractiveness, 40 authority, 124, 127, 129, 134, 135, 137 automobiles, 121 availability, 2, 20, 51, 60, 78 B baby boomers, 17

background, 91 background information, 91 balance of payments, 137 balanced budget, 132 banks, 2, 75, 127 base year, 79 BEA, 79, 80, 83, 84 benchmarks, 90 bias, 37, 59, 66 binding, 135 biofuel, 81 biomass, 81 birth, 93, 103 black women, 10 blue-collar workers, viii, 2 bonds, viii, 61, 62 borrowers, 137 borrowing, viii, 61, 62, 134, 137 budget deficit, 62, 63, 66, 124, 133, 135 budget resolution, 135 business cycle, 3, 14, 16, 26, 32, 72, 119, 120, 126, 131 C Canada, 53 capital flows, 66, 67, 136 capital gains, 62, 130, 136 capital inflow, 134 capital outflow, 62, 66 career counseling, 51 cash flow, 135 charities, 20, 68 children, 59, 65, 94

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

130

Index

classification, 3, 12, 56 closed economy, 66 coal, 81 codes, 8 collaboration, 89 colleges, 68 commodity, 21 community, 60, 82 compensation, 21, 27, 63, 131, 136 competition, 10, 117 complement, 130 complexity, 133, 137 components, 75, 77 composition, 18 concentration, 7, 14, 15, 34, 37, 57, 58 concrete, 78 conditioning, 81 confidence, 100, 112, 113, 119 confidence interval, 100, 112, 113 Congress, iv, viii, ix, 1, 3, 20, 21, 26, 27, 71, 72, 73, 75, 81, 82, 94, 110, 115, 123, 124, 133, 134, 136 Congressional Budget Office, v, viii, 29, 30, 31, 32, 33, 37, 39, 40, 42, 43, 44, 46, 47, 48, 49, 50, 54, 57, 58, 59, 61, 68, 69, 83, 121, 122, 135, 137 consensus, 75 construction, viii, ix, 2, 5, 7, 8, 11, 14, 15, 16, 18, 26, 39, 71, 74, 75, 76, 77, 78, 79, 80, 81, 132 consumer expenditure, 78 consumers, 2, 72 consumption, 62, 65 control, ix, 71, 75, 79 corn, 8 correlation, 66 costs, vii, 47, 68, 116, 117, 124, 126, 132, 133, 134 cotton, 8 counsel, 125 counseling, 49 credit, 2, 124, 128, 129, 130, 131, 132, 133, 137 crowding out, 62, 66, 133

D data collection, 81 data distribution, 103 database, 35, 54 death, 103 debt, 2, 61, 63, 127, 134, 135 decisions, 3, 59, 111 defense, 76 deficit, 63, 124, 131, 133, 134, 135, 136, 137 definition, ix, 12, 23, 24, 38, 71, 75, 79, 122 demographic characteristics, 94 demographics, 63 Department of Commerce, 86, 91, 108, 118 Department of Energy, 129 Department of Health and Human Services, 129 Department of Homeland Security, 85, 86, 91, 106 depreciation, 64, 68 developed countries, 66, 67 diesel fuel, 78 disability, 23, 35, 130 discrimination, vii, 1, 4, 12, 17, 128 displacement, 14, 26, 52 distortions, 133 distribution, 21, 38, 79, 81, 89, 111 District of Columbia, 13, 19, 80 diversification, 18 doctors, 129 domestic investment, 66 draft, 58, 87, 92 drawing, 117 duration, 16, 17, 32, 33, 34, 35, 36, 37, 38, 40, 51, 57, 60, 124 Duration, 11, 31, 37, 38, 39, 40, 42, 57, 58, 60 E earnings, 2, 12, 20, 21, 30, 34, 42, 45, 48, 49, 50, 51, 52, 53, 59, 60, 63, 80 economic activity, ix, 72, 76, 77, 115, 116 economic downturn, viii, 2, 14, 16, 17, 21, 22, 74, 77 economic growth, vii, ix, 64, 115, 116, 117, 118, 119, 121, 125, 127

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

131

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Index economic indicator, viii, 71, 72, 83 economic theory, 62, 126 economics, 117, 122 Education, 44, 74, 89, 90, 107 educational attainment, 44, 54 educational services, 129 elasticity, 68, 132, 133, 137 elderly, 63 electricity, ix, 71, 81 elementary school, 26 e-mail, 111 employability, 41, 126 employees, 4, 8, 14, 21, 25, 47, 74, 130 employment growth, 16, 116, 125, 132 employment status, 55, 59, 91 energy, ix, 72, 76, 81, 82, 123, 124, 128, 130 energy efficiency, ix, 123, 124, 128, 130 engagement, 86, 92 equating, viii, 71, 72 equilibrium, 62, 67, 126 equities, 62 equity, 62, 63 estimating, 24, 76, 81, 83, 87, 88, 94, 112, 113 ethnicity, 46 evolution, 3 examinations, 113 expenditures, viii, ix, 68, 71, 72, 76, 77, 78, 79, 134 exports, 136 extraction, 14, 15, 16 F failure, 126 family, 6, 7, 23, 45, 46, 58, 59 farmers, 11, 18, 21 farms, vii, 1, 12 federal funds, 72, 73, 82 fertility, viii, 61 fertility rate, viii, 61 finance, 58, 64, 124, 128, 133, 134, 135 financial crisis, 127 financial institutions, 2 financial markets, vii, 1, 62, 121, 127 financial sector, ix, 123, 124, 127 financial system, 127

financing, 124, 126, 131, 133, 134, 136 firms, vii, ix, 2, 4, 7, 8, 11, 21, 26, 30, 47, 71, 74, 75, 78, 115, 116, 127, 130, 131, 132, 133, 136 fiscal policy, 127, 131 flexibility, viii, 29 flood, ix, 71, 75, 79, 129 fluctuations, 13, 50, 64, 67 focusing, 35 food, 21, 127, 128, 136 foreign capital flows, 134 foreign investment, 66 foreign-born population, 88, 99, 112 forgiveness, 132 fossil, 81 fringe benefits, 34 fuel, 81 full capacity, 77 funding, 111, 128, 129, 130 funds, x, 20, 21, 49, 51, 68, 75, 78, 79, 83, 84, 110, 123, 124, 127, 128, 129, 131, 134, 135, 136 G gasoline, 81 GDP, 116, 118, 127 gender, 3, 5, 15, 26, 94 generation, viii, 17, 61, 65, 67, 79, 81 geographical mobility, 117 Georgia, 13, 19, 80, 106, 112 goods and services, ix, 8, 72, 75, 77, 116, 117 government revenues, 63 grants, 20, 60, 129 graph, 117 Great Depression, v, vii, viii, 1, 2, 3, 7, 8, 11, 12, 13, 14, 15, 17, 18, 20, 21, 23, 24, 71, 72, 124 gross domestic product, 65, 116 groups, 5, 21, 43, 45, 51, 88, 103, 112, 113, 130 growth, vii, viii, ix, 16, 18, 61, 62, 63, 64, 65, 66, 67, 68, 75, 78, 115, 116, 117, 118, 119, 121, 125, 129, 132, 136 guidance, 83 guidelines, 51, 129

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

132

Index H

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

hands, 125 Hawaii, 19, 80 health, x, 30, 31, 43, 47, 49, 51, 59, 60, 62, 65, 74, 75, 83, 123, 124, 128, 129, 130 health care, 62, 65, 83, 129 health insurance, 30, 31, 43, 47, 49, 51, 59, 60, 129, 130 health services, 74, 75 heating, 81 heterogeneity, 59 high school, 30, 39, 43, 44, 45, 46 highways, ix, 20, 71, 78, 129 hiring, x, 3, 4, 31, 123, 128, 130, 131, 133 Hispanics, 103 homeowners, 82 hospitality, 74 hospitals, 79 hotels, 11 House, ix, 17, 26, 27, 85, 90, 113, 123, 124, 127, 128, 130, 131, 134, 136 households, 36, 54, 55, 58, 83, 93, 102 housing, 75, 127, 129, 136 husband, 20 hydroelectric power, 79 hysteresis, 126 I identification, 34 identity, 137 immigrants, 87, 94, 112 immigration, 86, 88, 91, 112 implementation, 103 imports, 51 incentives, ix, 49, 123, 124, 128, 130, 131, 133 incidence, 8, 16, 44 income, viii, 29, 31, 35, 45, 46, 49, 52, 59, 61, 63, 64, 68, 84, 129, 131, 132 income support, 49, 52 income tax, 68, 84, 132 indication, 40, 55 indicators, viii, ix, 71, 115, 116 industrial sectors, 113

industry, 7, 8, 11, 15, 18, 26, 74, 75, 76, 78, 79, 80, 81, 84, 117, 129 infancy, 21 inferences, 94, 112 inflation, 62, 78, 122 information technology, 121 infrastructure, viii, ix, 71, 72, 75, 76, 78, 81, 83, 103, 123, 124, 127, 128, 129, 130, 131, 132, 136 inmates, 23 institutions, 23, 68, 129 instruments, 63, 131 insulation, 81 insurance, 2, 7, 21, 30, 34, 36, 47, 49, 52, 53, 58, 64, 102 integrity, 111 interest rates, 61, 62, 63, 66, 67, 68, 133, 136, 137 interrelationships, 76 interview, 54, 58, 60, 85, 90, 105 investment, 62, 66, 83, 84, 121, 129, 132, 133, 134 investment incentive, 84 investors, 63, 66, 75, 134 iron, 11 J Japan, 67 job creation, ix, 20, 72, 73, 75, 76, 77, 78, 79, 80, 81, 82, 123, 124, 125, 133, 134, 135 job training, 51, 129 jobless, viii, 9, 11, 20, 21, 23, 29, 42, 48, 59, 125 job-search, 34, 51, 52 justification, 79 L labor force participation, viii, 61, 113 labor markets, 39, 126 language, 73, 82 laws, 60, 62, 99, 112 layoffs, 4, 12, 33, 130 learning, 129 legislation, viii, 21, 47, 71, 72, 73, 81, 82, 83, 124, 127, 128, 136 leisure time, 121

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

133

Index lending, 125 liberalization, 21 likelihood, 44, 45, 49, 52 line, 5, 41, 118, 119, 120 linkage, 65, 112 loans, 20, 60, 127, 130 local government, ix, x, 20, 83, 123, 124, 128, 130, 131, 132 logging, 74 Louisiana, 13, 19, 80

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

M machinery, 5, 62 maintenance, 16, 79, 82 manufacturing, vii, viii, ix, 1, 2, 4, 7, 8, 11, 12, 14, 15, 18, 20, 26, 27, 71, 74, 76, 77, 82 marginal product, 62 market, vii, viii, ix, 1, 2, 3, 10, 12, 14, 19, 23, 29, 30, 31, 33, 35, 36, 39, 49, 58, 71, 75, 76, 82, 117, 123, 125, 126, 127 matrix, 76 measurement, 23 measures, vii, 1, 2, 14, 18, 35, 59, 124, 133 media, 27 median, 16, 17 Medicaid, 129, 136 Medicare, 65, 128, 131 men, vii, 1, 3, 5, 7, 9, 10, 11, 14, 15, 16, 45 Mexico, 13, 18, 19, 80, 112 middle class, 127 migrants, 18 minimum wage, 8 mining, 8, 14, 15, 26 minority, 10, 27 model, 40, 64, 65, 66, 76, 79, 81, 83, 131 models, viii, 61, 63, 64, 67, 76, 77, 81 monetary policy, 67, 131 money, 3, 6, 20, 23, 52, 68, 77, 111, 127, 132, 134, 135 Montana, 13, 19, 80 mortgage-backed securities, 127 motives, 64

N nation, vii, viii, 1, 11, 12, 22, 27, 71, 72, 75, 78, 79, 80, 81 national borders, 66 national debt, 134 national policy, 21 natural rate of unemployment, 126 natural resources, 16 negative relation, 44 net migration, 103 New England, 11, 12, 13, 18, 19, 27, 133, 137 North America, 79, 81 nutrition, 136 O objectives, 86, 92 observations, 54, 117 occupational groups, 26 OECD, 59, 121, 122 Office of Management and Budget, 83, 84 Oklahoma, 8, 13, 18, 19, 80 open economy, 131 operator, 7 order, 4, 21, 35, 84, 111, 116 Organization for Economic Cooperation and Development, 59 oversight, 111 overtime, 8, 133 ownership, 62 P Pacific, 11, 13, 19 parameters, 106 parents, 65 partnership, 89, 127 payroll, 18, 51, 61, 68, 72, 102, 130, 132 permit, 24 planning, 132 plants, 4, 18, 77 plausibility, 87 poor, 9, 19 population, 3, 7, 25, 35, 38, 47, 54, 62, 64, 65, 66, 67, 86, 88, 90, 91, 94, 95, 98, 99, 101, 102, 103, 105, 112, 113, 117 population growth, 64, 65, 66, 67

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

134

Index

positive relation, 83 positive relationship, 83 potential output, 117, 121 poverty, 129 power, 68, 81 premiums, 47, 51, 59 present value, 63, 135 president, 136 primary data, 102 private investment, 133 private sector, 100 probability, 38 producers, 11, 78 production, ix, 3, 6, 14, 15, 16, 18, 23, 26, 51, 62, 68, 76, 79, 81, 111, 115, 116, 117 production function, 68 productivity, vii, viii, ix, 61, 62, 64, 68, 76, 77, 79, 115, 116, 117, 118, 119, 121, 122 productivity growth, viii, 61, 64, 117, 118, 119, 121, 122 profit, 6, 23, 64 program, 7, 15, 19, 20, 21, 26, 27, 34, 49, 51, 53, 72, 74, 75, 76, 77, 79, 82, 129, 134, 135 public service, 75, 128, 129

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Q questioning, 105 R race, vii, 1, 10, 12, 94 range, 87, 88, 92, 94, 95, 96, 97, 98, 100, 113, 120, 131, 132 rate of return, viii, 61, 62, 63, 64, 65, 66, 67, 68 real estate, 7, 18, 75 real rate of interest, 64 reason, vii, viii, 2, 18, 20, 23, 47, 56, 59, 61, 66, 67, 76, 80, 116, 118, 133 recall, 39, 51 recognition, 79 recovery, 9, 26, 62, 82, 83, 125, 126 recreation, 11 redistricting, 103 refining, 81 refugees, 112

region, 11, 12, 18, 27 rehabilitation, 129 relationship, vii, 62, 65, 116, 117, 118, 119, 137 relative size, 62, 63 relevance, 112 reliability, 82, 111 relief, 20, 83, 84, 128, 129, 130 renewable energy, 129 repair, 77, 79, 82, 129 Requirements, 60, 79, 84 resources, ix, 20, 71, 76, 77, 81, 117, 133 restaurants, 11 retail, 7, 77 retirement, viii, 61, 63, 64 retirement age, 63 returns, 63, 68 revenue, 75, 127, 129, 130, 132, 134 risk, 14, 26, 63, 75, 122, 135, 137 robustness, 57, 86, 94, 95, 96 rural areas, 129 S safety, 2, 14, 21 sales, 5, 6, 12, 26 sample survey, 112 savings, 2, 62, 63, 66, 68 savings account, 63, 68 school, 4, 9, 30, 35, 44, 46, 79, 129 schooling, 26, 60 search, 9, 10, 30, 34, 49, 51, 52 searching, 39, 40 Secretary of Commerce, 89 Secretary of Homeland Security, 89 self-employed, 7, 58 Senate, 103, 127, 128, 136 sensitivity, 98, 100, 101, 112 separation, 60 September 11, 36 shape, 18 shares, 27, 63, 68 sharing, 8, 21, 25, 75, 127 short run, vii, 116, 118, 119, 126, 132, 133 short-term interest rate, 127 Silicon Valley, 19 simulation, 98, 112

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

135

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Index skills, 4, 31, 40, 49, 117, 126 small firms, 21 Social Security, 20, 21, 27, 49, 61, 63, 65, 68, 69, 130 social services, 76 South Dakota, 13, 19, 80 spectrum, 11 speculation, 9 speech, ix, 123, 124, 128, 130 stability, 15, 85 standards, 122 statistics, 15, 23, 30, 32, 33, 38, 40, 58, 85, 86, 87, 88, 90, 91, 92, 93, 95, 96, 98, 99, 100, 102, 112, 113 statute of limitations, 128 steel, 11, 12 stimulus, ix, x, 72, 73, 76, 79, 81, 83, 123, 124, 125, 127, 128, 130, 131, 132, 134, 136 stock, 62, 63, 64, 66, 68, 130, 136 strategies, 49, 51, 82 strength, viii, 5, 29 students, 35 subsidy, x, 51, 53, 123, 124, 129, 131, 132, 133, 135 substitution, 5, 68 subtraction, 88, 112 summer, 75, 129 supervisors, 5, 6, 12 suppliers, 18 supply, ix, 31, 61, 62, 63, 65, 66, 67, 71, 77, 78, 80, 121, 131 Supreme Court, 113 surplus, 21 survey design, 88 survivors, 21 sustainability, 124, 134

technological progress, 68 teenagers, 3 telephone, 113 tenants, 5, 6, 7, 12 thinking, 89 thresholds, 51 time frame, 55, 112 time periods, 67, 119 Title I, 20 tracking, 82 trade, 7, 65, 68, 74, 77, 79, 133, 134 trade deficit, 133, 134 trade-off, 68 training, 49, 51, 60 training programs, 60 transfer payments, x, 123, 131, 132 transitions, 34, 49 transmission, 81, 82 transportation, 32, 74, 128, 129 trust, 61, 63 turnover, 16, 59 U U.S. economy, viii, ix, 29, 123, 127 uncertainty, 85, 87, 88, 112 unemployment insurance, x, 21, 30, 31, 32, 34, 36, 45, 48, 49, 52, 53, 58, 60, 102, 123, 124, 130 uninsured, 48 United Nations, 66, 68 United States, viii, 3, 4, 6, 7, 10, 13, 25, 26, 27, 34, 51, 54, 58, 59, 60, 61, 64, 66, 67, 68, 80, 86, 90, 93, 94, 95, 100, 106, 112, 113, 130 universities, 89 updating, 79, 103 USDA, 128 V

T tax credit, 51, 75, 82, 124, 128, 129, 130, 131, 132, 133, 137 tax cuts, x, 83, 123, 127, 131, 132, 136 tax incentive, 130, 131 tax rates, 136 teachers, 129 technological developments, 5

vacancies, 49 variability, 30 variables, 26, 64, 118 vehicles, 74, 82 vouchers, 129 W wage rate, 59, 62, 64, 65

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

136

Index windows, ix, 72, 83 women, vii, viii, 1, 3, 5, 7, 9, 10, 14, 15, 16, 45, 61 wood, 76, 81 work study, 129 Workforce Investment Act, 60 World War I, vii, 1, 2, 3, 13, 14, 16, 124, 125

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

wages, viii, 32, 53, 60, 61, 62, 63, 64, 65, 67, 68, 77, 78, 122, 133 wealth, 121 websites, 83 welfare, 65 Western Europe, 126 wind, ix, 71, 76, 81, 83

Labor and Employment Issues, edited by Nickolas H. Mullen, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,