Lucie Kalousova, Patrick Wightman, Sheldon Danziger and Italo Gutierrez, National Poverty Center, Gerald R. Ford School of Public Policy, University of Michigan
For many Americans, the most palpable manifestation of the Great Recession (December 2007 – June 2009) was the steep increase in unemployment. The national unemployment rate reached 10 percent in fall 2009 and was even higher in local labor markets in many parts of the country. In metropolitan Detroit, an area whose economy plunged deeper than most, the unemployment rate exceeded 25 percent. Given the extent of economic uncertainty, even workers who would ordinarily have been securely employed lost jobs, and many who managed to keep their positions feared layoffs.
This brief examines the job search behavior and success in finding new jobs for metro Detroit residents, both those who were unemployed and those who were working but searched for jobs. We focus on the predictors of a successful search during the slow economic recovery. Our analysis reveals that successful job seekers sent more job applications on average than did those who searched but did not find a new job, regardless of whether they were employed or unemployed. Also, the number of applications sent by unsuccessful job seekers was approximately the same for both the employed and unemployed. Importantly, these results highlight that the chances of finding new employment decrease precipitously with the duration of an unemployment spell.
Michigan Recession and Recovery Study (MRRS)
Since late 2009, the MRRS has followed a stratified random sample of English-speaking adults who lived in Southeastern Michigan (Macomb, Oakland, and Wayne counties) and were ages 19 to 64 in late 2009/early 2010, at the first interview. The MRRS oversampled African Americans and includes mainly African American and non-Hispanic white respondents, reflecting the residential composition of the area (2). To date, respondents have been interviewed twice. We use data from the 847 respondents who participated in both waves of in-person survey interviews (3).
The MRRS survey instrument is unique in its depth and breadth, covering many domains, including employment and labor market experiences, housing instability, material hardships, income, assets, financial problems, credit and debt, health and mental health, demographic characteristics, and use of public programs and private charities. More information about the study and related papers and policy briefs can be found at: http://www.npc.umich.edu/research/recessionsurvey/index.php.
During the first survey interview, all respondents were asked about their labor market status, i.e. whether they were currently employed or not working. Regardless of their answer, all respondents were then also asked whether they were searching for a job in the last 30 days. If they answered in the affirmative, a follow-up question established the intensity with which they were searching by inquiring about the number of job applications they had submitted during their most recent search. We limit our analysis to the 241 respondents, 123 employed and 118 unemployed, who were searching for jobs at Wave 1. We exclude the respondents who self-identified as out of the labor force (i.e. students, retirees, homemakers, discouraged workers) and employees who were not searching. Detailed information on the construction of all measures can be found in the Appendix to this brief.
Who is looking for work?
In Table 1 we present the sociodemographic characteristics of respondents who were looking for jobs at the first wave of data collection. We find significant differences between the employed and unemployed searchers in educational attainment, poverty prevalence and age. There were fewer respondents with a bachelor’s degree or more among the unemployed searchers compared to the employed searchers (15 vs. 32%). Moreover, a greater proportion of the unemployed were classified as below the poverty line, relative to those who were employed (30 vs. 13%). Finally, unemployed job seekers were less likely to be in prime working age—more than half, 52% of them, were between the ages of 19 and 34, in contrast to 42% of the employed who were searching. Additionally, 28% of the unemployed were older than 50 years, compared to only 12% of the employed job seekers.
Table 1: Weighed Characteristics of Michigan Recession and Recovery Study Respondents Searching for New Employment, N = 241
Search Intensity and Success
Because the MRRS uniquely asked both the employed and unemployed about the details of their job search, we can contrast their search behaviors. We find that unemployed searchers were on average seeking new employment much more aggressively than employed searchers, with the former reporting a mean of 18.9 applications in contrast with 8.5 reported by the employed. The unemployed searchers appear to be willing to accept lower wages compared to employed searchers, although this difference is not statistically significant. When the survey re-interviewed all respondents at the second wave of data collection, those who were formerly unemployed were more likely to have transitioned into a new job compared to those who were searching for a job while employed—54 vs. 15 percent.
Table 2: MRRS Search Behavior and Success of Employed and Unemployed Searchers
Using a multivariate regression framework, we evaluated the number of applications sent, on average, by employed and unemployment searchers, stratifying by job search success, when other attributes are held constant. These attributes include race (Black or not), gender (female or not), education (bachelor’s degree or more vs. less), age, and health (poor or fair health or not). Based on the resulting estimates, we generate the predicted number of applications submitted.
Figure 1 presents the predicted number of applications sent by a white male with less than a bachelor’s degree, who is between the ages of 35 and 50, and whose health is good, very good or excellent. We find that there was no significant difference in the number of applications sent by unsuccessful job seekers by employment status—8.1 for the employed and 7.7 for the unemployed. Those who found new employment, both among the employed and unemployed, searched with significantly greater intensity. The formerly unemployed sent the largest number of applications overall, approximately 19.5. The employed job seekers who changed jobs between the first two surveys reported sending about 18.7 applications. There was no statistically significant difference in applications sent by the two groups of successful searchers.
Figure 1: Estimated Number of Applications Sent; Stratified by Employment Status at First Interview and Success of Job Search
We also examined how likely unemployed job seekers were to find a new job during the study period, depending on the number of months they had been unemployed at the first survey wave. We display the re-employment probabilities in Figure 2, categorized by the number of months a respondent reported being unemployed at the time of their first interview. Like in the analyses shown above, we hold other attributes of the job seeker constant. The predicted probabilities below were estimated for a white male with less than a bachelor’s degree, who is between the ages of 35 and 50, and whose health is good, very good or excellent.
Figure 2: Predicted Probability of Re-employment by Number of Months Unemployed at First Wave of Data Collection
We find the greatest probability (over 70 percent) of being re-employed for those who experienced between six and twelve months of unemployment at the time of their first interview. The probability of re-employment did not increase with longer unemployment spells. In fact, those who were unemployed for longer periods were substantially less likely to become re-employed by the second wave of the study. Respondents who were unemployed between one and two years had an approximately 40 percent probability of finding a new job, while those who had searched for a job for more than two years had a probability of only 10 percent of becoming re-employed by the second wave of the study.
The current economic recovery has been dubbed as ‘jobless’ for the slow rate at which the American labor market reemployed the workers who lost their jobs during the Great Recession. These workers face competition not only from other unemployed job seekers, but also from the employed who are applying for better jobs. This brief has shown that the employed job seekers in metropolitan Detroit are generally more advantaged compared to the unemployed job seekers, as they are more likely to be college graduates and less likely to be poor. Moreover, the disadvantage of unemployed seekers likely grows with the duration of their unemployment spell; the long-term unemployed were the group least likely to secure a new job during the study period. As manifest by the lower average number of job applications the employed seekers report, they are likely more selective in deciding what opportunities to apply for, and only 15% of them changed jobs between their interviews. Nevertheless, employed job seekers who found new jobs sent job applications with a level of intensity similar to that of unemployed searchers. These findings highlight the unevenness of the competition for job openings. With an unprecedented number of Americans classified as long-term unemployed (5), we may expect the disparity between the employed searchers and unemployed searchers to widen.
(1) The data collection for the Michigan Recession and Recovery Study (MRRS) was supported in part by funds provided by the Office of the Assistant Secretary for Planning and Evaluation at the U.S. Department of Health and Human Services, the Ford Foundation, the John D. and Catherine T. MacArthur Foundation and the Office of the Vice President for Research at the University of Michigan. This brief is made possible by a grant from the Rockefeller Foundation. Tedi Engler and Danielle Battle provided valuable assistance preparing the MRRS data.
(2) Survey weights are used in all analyses reported here to make our results representative of the population 19 to 64 in the study area.
(3) A total of 914 respondents were interviewed at wave one, with a survey response rate of 82.8%; 847 of these respondents were re-interviewed in spring/summer 2011, for a wave two response rate of 93.9%.
(4) “P for diff” refers to the probability that we would observe the reported values if there was no difference in the population.
About the Authors
Lucie Kalousova is a Ph.D. candidate in the joint program in Sociology and Health Policy at the University of Michigan. firstname.lastname@example.org.
Patrick Wightman is a Senior Economic Analyst with the Oregon Health Authority. email@example.com
Sheldon Danziger is President of the Russell Sage Foundation. When this research was in progress he was the H.J. Meyer Distinguished University Professor of Public Policy and the Director of the National Poverty Center at the Gerald R. Ford School of Public Policy at the University of Michigan. firstname.lastname@example.org
Italo Gutierrez is an economist with the Rand Corporation. email@example.com
Appendix: Measures Used in This Brief