Current Unemployment, Historically Contemplated

Current Unemployment, Historically Contemplated Chinhui Juhn University of Houston Kevin M Murphy University of Chicago and NBER Robert Topel University of Chicago and NBER March 2002 This Version: May 2002 Prepared for the Brookings Panel on Economic Activity, April 4-5, 2002 We thank Panel participants for helpful comments Earlier versions were presented at the NBER Labor Studies meeting and at the University of Missouri Murphy and Topel are also Senior Fellows at the Milken Institute We acknowledge financial support from the Center for the Study of the Economy and the State and the Milken Foundation I Introduction Twelve years ago, our Brookings paper “Why has Natural Rate of Unemployment Increased Over Time?” analyzed long term changes in joblessness among American men We documented the dramatic rise in unemployment and non-participation for prime aged males that had taken place between 1967 and 1989 Our main conclusion was that a steep and sustained decline in the demand for low skilled workers had reduced the returns to work for the less skilled This led to high rates of unemployment, labor force withdrawal, and long spells of joblessness for these men We found that the long-term growth in joblessness was accounted for a roughly equal increase in time spent out of the labor force and time spent unemployed We concluded that structural factors, primarily the decline in the demand for low skilled labor, had dramatically changed the prospects for a return to low rates of joblessness any time soon After our paper was published in BPEA in 1991, things appeared to change The 1990s opened with a brief recession that was followed by the longest sustained decline in unemployment in modern U.S history By the end of the expansion, the unemployment rate had reached its lowest level since the late 1960s, falling below percent for the first time since 1969 Some macroeconomists argued that the “natural rate,” had permanently shifted to a level at or below 5.0 percent (Stiglitz (1997), Gordon (1998), Staiger, Stock, and Watson (2001)) Because we had emphasized changes in the structure of labor demand that made a return to low rates of joblessness unlikely, these facts present a challenge to our 1991 framework Maybe we were just wrong – maybe the demand and supply framework of our previous work is inconsistent with rates of joblessness in the post-1990 period If so, we would join a distinguished group of social scientists who have drawn attention a significant empirical phenomenon, only to watch that phenomenon disappear immediately thereafter.1 As it turns out, however, the framework that we developed for thinking about the pre-1990 evolution of joblessness also does fairly well in helping to understand jobless time in the post-1990 period In this paper we look in some detail at employment data from the 1990s, revisiting issues raised in our earlier work Specifically, we ask: (1) Have the trends we identified in our earlier paper – the concentration of non-employment among the less skilled, the growth of non-participation in the labor force, and the increased duration of joblessness – been reversed with the fall in aggregate unemployment? (2) Did the ‘90s expansion really “return” the U.S labor market to conditions of the late 1960s, as unemployment statistics seem to indicate? (3) Does the economic framework of supply and demand we utilized a decade ago still help in understanding long-term developments in unemployment, non-employment, and labor force participation? Our answers are surprising First, the basic trends toward longer spells of joblessness and rising non-employment have continued in spite of the prolonged expansion of national output and the concomitant fall in unemployment rates Long spells and labor force withdrawal were more important in the 1990s than ever before Second, the fall in unemployment to levels close to historical lows is very misleading Malthus founded the club His theory that the forces of endogenous population growth doomed the common people to perpetual poverty “explained” why incomes had failed to increase over the period of his data Publication of Malthus’s theory was followed by centuries of almost continuous progress More recently, when the returns to college were at a record low in 1979, Richard Freeman offered a supply-based theory in “The Over Educated American”, only to have the returns to college increase steadily over the next 15 years, reaching a record high To Freeman’s credit his model did predict a rebound, even if not so large and sustained as the one that actually occurred Broader measures of joblessness show that the labor market of the late 1990s is more like the relatively “slack” labor market of the late 1980s than the booming labor market of the late 1960s Finally, the basic forces of supply and demand identified in our previous paper continue to have explanatory power The theory does a reasonable job of explaining the trends that have continued, as well as those that have changed Recent data also provide considerable insight into what has happened in the labor market over the past decade Over the 1990s, while unemployment was falling, time spent out of the labor force was rising In fact, the increase in time spent out of the labor force was so large that total joblessness – which combines the unemployed with those who have withdrawn from the labor force – at the business cycle peak in 2000 was as high as it was at the previous cyclical peak of 1989, in spite of the fact that the unemployment rate was roughly percentage points lower in 2000 In terms of total joblessness, the oft-praised boom of the 1990s really represented little in the way of employment progress for American males While the growth in the amount of time American males spent out of the labor force continues a trend found in our earlier research, other features of the data changed in the 1990s The real wages of less-skilled men, which had been falling steadily since the early 1970s, stabilized in the 1990s and even rebounded slightly in the second half of the decade It appears that the 30 year trend toward greater wage inequality has run its course, at least at the bottom of the wage distribution The data on joblessness reflect the impact of the changing wage trends The long-term divergence in employment rates between low-wage workers and those with higher wages – which was so pronounced in our earlier work – has stopped, while unemployment and wage gaps across skill groups have narrowed The congruence betweens patterns of change in wages and employment comports with our previous work, which stressed demand-driven wage changes as the dominant factor driving secular changes in employment rates We are not the first to study declining unemployment in the 1990s Others have emphasized changes in the composition of the labor force as a source of declining unemployment Shimer (1998) finds that aging of the labor force is important in explaining the decline in the unemployment, particularly compared to the late 1970s Katz and Krueger (1999) investigate to what extent the withdrawal of the incarcerated population from the labor force lead to a drop in the aggregate unemployment rate Other papers explore explore the role of improvements in job search technology For example, Autor (2000) argues that temporary-help agencies may have facilitated more efficient employer-employee matches, thus bringing about a decline in frictional unemployment The internet may have also reduced search costs, although its impact is less certain (Autor (2000), Kuhn and Skulerud (2000)) We show below that lower unemployment rates in the 1990s are accounted for by a sharp decline in the incidence of spells, while durations of spells have remained high This fact is inconsistent with a theory built on declining search costs, which would imply shorter unemployment spells A related line of research compares divergent employment outcomes between the United States and Europe While these countries may have experienced patterns of labor demand during the 1970s and the 1980s, there is widespread belief that more flexible labor markets and wages kept American unemployment rates relatively low, while European rates rose Along these lines, several papers (Bertola and Ichino (1995), Blanchard and Wolfers (2000), Ljungqvist and Sargent (1998), and Bertola, Blau and Kahn (2001)) emphasize the importance of interactions between macroeconomic shocks and labor market institutions These papers find that while macroeconomic variables (oil prices, real interest rates, total factor productivity, and labor share of income) or labor market institution variables (unemployment benefits and duration, union coverage, collective bargaining, employment protection policies) alone cannot explain the differences across U.S and Europe, a model that allows for interaction effects fits the data well While this shocks-institutions framework is useful for explaining crosscountry differences, it is less successful in understanding recent changes in U.S unemployment For example, Bertola, Blau and Kahn (2001) report that their model significantly under-predicts the decline in U.S unemployment in the late 1990s We revisit the evolution of joblessness in the United States, using 34 years (19672000) of micro-data from the Current Populations Surveys Our main conclusions are: Falling unemployment rates over the 1990s greatly exaggerate the improvement in labor market conditions for prime-age males While unemployment rates were falling, rates of joblessness – which include time out of the labor force – were roughly the same in the late 1990s as they had been in the late 1980s Rising non-participation among prime aged men largely offset declining unemployment, so that the employment to population ratio was constant Trends towards longer durations in both unemployment and non-employment have continued in the 1990s, in spite of declining unemployment rates The probability of entering unemployment (or non-employment) fell dramatically during the 1990s The decline in incidence of spells was so large that the likelihood of experiencing a spell reached its lowest level in the 34 years of our data But there was no decline in durations of unemployment spells – spells were about 2.8 weeks longer in 1999-2000 than they had been a decade earlier – and the durations of non-employment spells increased by over months during the 1990s Broadly speaking, all of the long-term growth in joblessness is the product of longer durations of jobless spells While non-employment continues to be concentrated among less skilled men, the trend toward rising joblessness of the least skilled reversed somewhat in the 1990s The largest declines in unemployment occurred among men in the lowest skill categories Unemployment among men from the first decile of the wage distribution fell by 4.6 percentage points between the cyclical peaks of 1989-90 and 1999-2000, while the decline in unemployment at the median of the wage distribution was about one percentage point In contrast, over the longer term the growth in nonemployment is heavily weighted toward less skilled men Among men at the bottom of the wage distribution the non-employment rate increased by 13.5 points between the late 1960’s and 2000, but by less than one percentage point for men with wages above the median of the distribution The long term decline in real wages of less-skilled men stopped in the early 1990s, and actually reversed slightly in the later part of the decade While wages of highly skilled men grew most rapidly of all during the 1990s – continuing past patterns of relative growth inequality between men at the bottom of the wage distribution and men at the median contracted slightly over the decade Overall, the trend toward greater wage inequality appears to have stopped for males in the bottom half of the wage distribution Joblessness of less-skilled men has shown up increasingly as time spent out of the labor force rather than time spent unemployed Consistent with our earlier work, we believe this continued trend toward labor force withdrawal reflects two factors: (i) the relatively low returns to work (real wages of the least skilled remain substantially below where they were in the past) and (ii) increasingly attractive outside opportunities, such as collecting disability payments, that have shifted labor supply among the least skilled We find that more than 40 percent of the growth in OLF status is associated with an increase in individuals claiming to be ill or disabled Despite rising wages and rates of labor force participation for women, high joblessness of less-skilled men is not the outcome of improved labor market opportunities for their working wives Non-employment rates and rates of labor force withdrawal increased most among men who did not have a working wife Looking across the male wage distribution, the proportion of men with a working wife actually fell among low-skilled men – whose wages and employment rates were falling – and rose among men in the top four deciles of the wage distribution – where wages rose and employment rates were stable We conclude that long term changes in joblessness were the result of adverse shifts in labor demand, perhaps coupled with policy-driven shifts in labor supply, among low-skilled men The paper is organized as follows Section II describes the methods we use to analyze the March CPS data that form the basis for our empirical analysis, and reports results on changes in the distribution of male wages since 1967 Section III provides empirical evidence on the evolution of unemployment and participation for 1967-2000 We disaggregate by reported reason for non-employment in order to shed light on the divergent trends in unemployment and non-participation We also decompose changes in unemployment and non-employment into inflows and outflows, documenting the increased importance of very long jobless spells, especially in the 1990s Section IV presents our analysis of unemployment, non-employment and wages over the recent period, disaggregated by wage percentile category Building on our earlier work, we examine whether labor supply elasticities estimated from cross-sectional data, together with the changes in the distribution of wages, are consistent with recent changes in male joblessness Section V summarizes and concludes II Data Our data are drawn from the 1968-2001 Annual Demographic Files that supplement the March Current Population Surveys (CPS) The CPS is a monthly random sample of approximately 50,000 rotating U.S households It forms the basis for published government statistics on earnings, employment, unemployment, and labor force participation, among other measures While published labor market statistics rely on questions about each survey respondent’s employment status in the reference week of the survey (usually the third week of the month), we study retrospective information on labor market outcomes in the previous calendar year that is collected each March – hence our data cover the 34 calendar years 1967-2000 In addition to personal and household characteristics for each respondent, the retrospective data in the March survey record the respondent’s annual weeks worked, unemployed, and out of the labor force, as well as the number of unemployment spells We measure time spent unemployed (U), out of the labor force (O) and non-employed (N=U+O) as the percentage of the year spent in each state – for example, for the ith individual Ui is weeks unemployed divided by 52 This differs from the usual method of measuring unemployment, which divides unemployed weeks by weeks in the labor force Our method better summarizes the allocation of time across the three states, and naturally aggregates across individuals.2 Using methods described below, we use information on weeks worked, unemployed, and out of the labor force to calculate both the incidence and the duration of spells The survey also records a respondent’s annual earnings and usual weekly hours worked from all jobs as well as occupation, industry, and other characteristics for the longest job held during the prior year We use the earnings, weeks worked and hours worked information to calculate average hourly wages and to assign individuals a percentile position in the overall wage distribution, as described below This allows us to track changes in employment outcomes (U, O, and N) for persons in different parts of the wage distribution We focus our analysis on males since they were the focus of our earlier work and because labor force participation issues for women are significantly more complex In order to avoid issues associated with early retirement, Social Security, and pensions, we focus on men who have to 30 years of potential labor market experience For high school graduates this cutoff yields men who are roughly 19 to 49 years of age, with correspondingly higher age intervals for those with more schooling We define labor market experience as the minimum of (a) age minus years of education minus seven and Since the denominator for Ui, Ni, and Oi is always 52, the corresponding jobless rate is is simply the sample average of weeks in the state divided by 52 10 Table Changes in Nonemployment by Reason Changes 1967-69 to 1999-2000 Reason For Non-Employment Statistic No Work ILL Other Total OLF 0.6% 1.7% 1.7% 4.0% Unemployment 0.6% 0.0% 0.0% 0.6% Total 1.2% 1.7% 1.8% 4.7% Changes 1988-89 to 1999-2000 Reason For Non-Employment Statistic No Work ILL Other Total OLF -0.4% 0.8% 1.0% 1.5% Unemployment -1.5% 0.0% 0.0% -1.5% Total -1.9% 0.8% 1.1% 0.0% 42 Figure Comparison of Recessions: Contribution of Ill and Disabled 0.16 Share of Total Change 0.12 Nonemployment of Ill & Disabled 0.08 0.04 69-71 73-75 -0.04 43 79-82 89-92 44 Table Unemployment and Non-employment by Weeks Category Non-Employed Weeks to to 12 13 to 25 n/p 67-69 71-72 72-73 75-76 78-79 82-83 88-89 91-92 99-00 6.3% 9.1% 8.8% 12.4% 10.2% 15.2% 11.0% 13.8% 11.0% 17.2% 12.9% 12.4% 9.9% 13.1% 8.1% 10.3% 8.8% 8.0% 18.9% 18.8% 17.2% 16.6% 17.7% 13.9% 13.2% 12.7% 9.7% Unemployed Weeks to 13 14 to 26 27 to 39 u/p 67-69 71-72 72-73 75-76 78-79 82-83 88-89 91-92 99-00 2.4% 1.5% 1.6% 1.1% 1.5% 0.8% 1.0% 0.8% 0.7% 2.2% 4.3% 3.8% 6.8% 4.3% 9.0% 4.3% 6.3% 2.8% 30.3% 20.4% 20.6% 17.8% 27.0% 13.7% 22.6% 16.9% 20.4% 42.3% 40.6% 39.9% 31.2% 34.2% 28.9% 33.5% 31.7% 33.4% 45 14.9% 21.5% 21.0% 22.8% 19.0% 21.7% 18.5% 20.6% 18.2% 26 to 38 18.0% 21.2% 21.7% 22.5% 20.2% 20.6% 19.0% 19.1% 13.9% 40 to 49 8.9% 10.9% 10.4% 14.5% 11.8% 17.9% 14.8% 16.4% 14.6% 39 to 51 14.8% 15.9% 15.3% 17.2% 14.4% 18.3% 15.2% 14.9% 12.2% 50-52 3.6% 6.6% 8.2% 13.7% 8.0% 17.7% 10.5% 14.3% 13.4% 52 28.8% 29.7% 31.9% 32.7% 33.2% 38.3% 41.3% 43.8% 55.5% Figure 11 Entry Rates and Durations for Unemployment 1967-2000 2.1% 5.5 Duration (Months) 1.9% 4.5 1.6% 3.5 1.4% 1.1% 2.5 0.9% 0.6% 1.5 67 72 77 82 87 Year 46 92 97 Average Durat ion (Mont hs) Ent ry Rat es (% Per mont h) Entry Rate (Per Month) Table Entry Rates and Duration for Unemployment and NonEmployment 1967-2000 Years 1967-69 1971-72 1972-73 1975-76 1978-79 1982-83 1988-89 1991-92 1999-00 Period Peak Trough Peak Trough Peak Trough Peak Trough Peak Unemployment Entry Duration Non-Employment Entry Duration (per month) (months) (per month) (months) 1.1% 1.5% 1.4% 1.8% 1.5% 1.9% 1.3% 1.5% 0.7% 2.1 3.2 2.9 4.1 3.1 5.1 3.5 4.5 4.2 1.0% 1.2% 1.1% 1.5% 1.4% 1.5% 1.1% 1.3% 0.8% 6.7 8.6 8.5 9.6 8.4 12.0 11.0 12.4 15.1 47 Figure 13 Unemployment Rates by Wage Percentile Groups 1967-2000 20.0% to 10 Unemplo yment (% o f Weeks) 11 to 20 21 to 40 16.0% 41 to 60 61 to 100 12.0% 8.0% 4.0% 0.0% 67 72 77 82 87 92 97 Year Figure 14 OLF Rates by Wage Percentile Groups 1967-2000 25.0% to 10 11 to 20 20.0% 21 to 40 OLF Rate (% of Weeks) 41 to 60 61 to 100 15.0% 10.0% 5.0% 0.0% 67 72 77 82 87 Year 48 92 97 Figure 15 Non-Employment by Wage Percentile Groups 1967-2000 40.0% to 10 Non-Employment (% of Weeks) 35.0% 21 to 40 41 to 60 30.0% 61 to 100 11 to 20 25.0% 20.0% 15.0% 10.0% 5.0% 0.0% 67 72 77 82 87 Year 49 92 97 Table Changes in Employment & Unemployment by Percentile Unemployment Rate Changes by Percentile Group 1967-2000 Start End to 10 Percentile Group 11 to 20 21 to 40 41 to 60 61 to 100 1967-69 to 1988-89 6.4% 4.9% 2.6% 1.5% 0.3% 1988-89 to 1999-00 -4.6% -3.1% -1.6% -1.0% -0.5% 1967-69 to 1999-00 1.8% 1.8% 1.1% 0.5% -0.1% OLF Rate Changes by Percentile Group 1967-2000 Start End to 10 11 to 20 21 to 40 41 to 60 61 to 100 1967-69 to 1988-89 10.9% 7.2% 2.9% 1.3% 0.0% 1988-89 to 1999-00 0.8% 2.4% 2.6% 1.3% 0.9% 1967-69 to 1999-00 11.7% 9.5% 5.5% 2.6% 0.9% Non-Employment Rate Changes by Percentile Group 1967-2000 Start End to 10 11 to 20 21 to 40 41 to 60 61 to 100 1967-69 to 1988-89 17.3% 12.1% 5.6% 2.8% 0.3% 1988-89 to 1999-00 -3.8% -0.8% 1.0% 0.3% 0.4% 1967-69 to 1999-00 13.5% 11.3% 6.6% 3.0% 0.8% Full Year Non-Employment Changes by Percentile Group 1967-2000 Start End to 10 11 to 20 21 to 40 41 to 60 61 to 100 1967-69 to 1988-89 10.2% 6.3% 2.9% 1.6% 0.5% 1988-89 to 1999-00 1.7% 2.7% 2.4% 1.3% 0.9% 1967-69 to 1999-00 12.0% 8.9% 5.4% 2.9% 1.4% 50 Figure 16 Cyclical and Secular Changes in Non-Employment 20.0% Recessions (Average) Change in % o f Weeks 15.0% 1967-69 to 1988-89 1988-89 to 1999-00 10.0% 5.0% 0.0% to 10 11 to 20 21 to 40 -5.0% Wage Percentile Group 51 41 to 60 61 to 100 Table Earnings, Household Income, and Presence of Working Spouse (In $1996) Year 1967-69 1972-73 1978-79 1988-89 1999-00 Percentile to 10 Male Earnings $11,134 $11,354 $11,479 Working Spouse Household Income 38% 36% 36% 31% 29% $3,050 $3,392 $3,766 $3,975 $4,770 $22,965 $26,366 $28,140 $25,952 $29,194 Male Earnings $23,746 $25,213 $23,579 $21,522 $21,507 Spouse Earnings 21 to 40 61 to 100 $9,584 Working Spouse 48.6% 48.0% 47.9% 46.4% 40.4% Spouse Earnings $5,132 $5,773 $6,133 $7,410 $8,091 Household Income $35,394 $39,137 $40,770 $40,773 $43,431 Male Earnings $44,213 $49,375 $47,147 $51,449 $59,843 41.4% 44.5% 50.4% 55.3% 53.6% $4,786 $5,918 $7,219 $11,537 $14,993 $54,456 $61,722 $61,861 $71,937 $88,507 Working Spouse Spouse Earnings Household Income to 100 $8,625 Male Earnings Working Spouse Spouse Earnings Household Income $31,230 $34,121 $32,686 $33,193 $36,789 44.4% 45.0% 48.0% 48.9% 45.3% $4,713 $5,548 $6,388 $8,895 $10,854 $42,263 $47,538 $48,560 $52,821 $61,541 52 Figure 17 Male Unemployment by Presence of Working Wife: 1967-2000 (o) no working wife (+) working wife 0.14 0.12 share of population 0.10 0.08 0.06 0.04 0.02 0.00 70 75 80 85 Year 90 Nonemployed by Working Wife: 1967-2000 53 95 100 Figure 18 Male Non-participation by Presence of Working Wife 1967-2000 (o) no working wife (+ ) working wife 0.08 0.07 share of population 0.06 0.05 0.04 0.03 0.02 0.01 0.00 70 75 80 85 Year 90 95 Out of the Labor Force by Working Wife: 1967-2000 54 100 Figure 19 Male Full Year Non-employment by Presence of Working Wife 1967-2000 (o) no working wife (+) working wife 05 share of population 04 03 02 01 70 75 80 85 Year 90 95 Full-Year Non-Employment by Working Wife: 1967-2000 55 100 Table Wages and Non-employment Change 1972-73 to 1988-89 Estimated Partial Elasticity to 10 11 to 20 21 to 40 41 to 60 61 100 Log Wage Actual Change Predicted Change Change 0.287 10.2% 7.0% -24.8% 0.217 5.7% 4.7% -22.3% 0.170 2.6% 2.7% -16.5% 0.126 0.8% 1.1% -9.2% 0.048 -0.1% 0.1% -0.2% to 100 2.2% 2.0% -9.9% Change 1988-89 to 1999-00 Estimated Partial Elasticity to 10 11 to 20 21 to 40 41 to 60 61 to 100 to 100 Log Wage Actual Change Predicted Change Change 0.287 -3.8% -0.7% 2.3% 0.217 -0.8% 0.1% -0.4% 0.170 1.0% 0.3% -1.5% 0.126 0.3% 0.2% -1.2% 0.048 0.4% 0.1% 7.0% 0.0% 0.1% 2.4% 56 ... (t ) 22  contemporaneous rate of exit from unemployment, and [1  u (t )]u (t ) is the expected number of spells of unemployment per year at the current entry rate, equation (2) has a natural... calculations for unemployment, where the number of spells is recorded In that case the bias varied little over time, lending some confidence that this method should not be too far off 26 IV: Unemployment,. .. Shimer (1998) finds that aging of the labor force is important in explaining the decline in the unemployment, particularly compared to the late 1970s Katz and Krueger (1999) investigate to what