Unemployment and Bankruptcy in the Middle Class
Teresa A. Sullivan, Elizabeth Warren, and Jay Lawrence Westbrook—
Just how significant job problems are for bankruptcy filers can be determined in at least two ways. One is to ask people in bankruptcy systematically about their work histories. The second is to ask them why they filed for bankruptcy and to count how often someone gave a job-related reason. We used both approaches with a surprising result: when we totaled the systematic answers and analyzed the free responses, more than two out of every three bankruptcy filers (68 percent) reported a job problem.
Unemployment has long been the most direct measure of job difficulty, and we sought to identify it by asking several different questions. We asked each person in our sample, including both husbands and wives in the case of joint filings, whether they were currently seeking work. The question we used is approximately the same as that used by the Bureau of Labor Statistics to measure unemployment: ‘‘If not currently holding a job, did this person actively SEEK work during the past four weeks?’’ The precoded possible answers were ‘‘Yes, sought work,’’ ‘‘No,’’ and ‘‘Has a job now.’’ By using the government’s question for measuring unemployment, we sought to maximize the comparability of our data with data for the overall U.S. population.
We also asked each person if he or she had experienced an interruption of work-related income for at least two weeks during the preceding two years. The possible answers were ‘‘Yes,’’ ‘‘No,’’ and ‘‘Not employed during this time.’’ This question is not asked by government interviewers, and it is open to somewhat more interpretation. A worker might have left work voluntarily for a number of reasons. Alternatively, the worker might have experienced short-term layoffs or job searches. Regardless of the reason for the work interruption, the answer to this question is a useful indicator of financial stress because of the importance of work-related income to the finances of nearly all families. Finally, in our open-ended question about the reasons for bankruptcy, our respondents could write any reasons they wished to explain their bankruptcy filings. We coded up to five reasons from each response in the order in which the respondent mentioned them. Bankruptcy filers could list multiple work-related problems, such as a cutback in hours followed months later by a layoff. Through this combination of open and closed questioning, we developed a more comprehensive view of the job problems of the debtors.
Unemployment is the most severe of the job problems that show up in bankruptcy. A total of 490 respondents in our sample reported that they did not have a job and were actively seeking work, which is the government’s definition of unemployment. This number included 332 primary petitioners, either people filing alone or one spouse of a couple, and 158 joint petitioners, nearly all of whom were wives. Asa proportion of all the valid responses, these numbers yield 17.9 percent unemployment, more than two and a half times the prevailing national rate in 1991 of 6.7 percent. But the government’s statistic excludes from its denominator any adult over the age of sixteen who is neither working nor actively looking for work. Such people are not considered to be in the labor force. If we eliminate the people in our sample who are not looking for work and who do not have a job—making our statistic directly comparable to the government’s—the unemployment rate among bankrupts rises to 21.4 percent, more than three times the national rate.
In our first study of consumer debtors who filed for bankruptcy in 1981, we did not survey the debtors directly. When we examined their court files to try to determine unemployment rates, all we had were the debtors’ responses to a question about their occupations at the time of filing.Even among those who are unemployed, some will list the occupation they held when they were employed rather than identifying themselves as unemployed. Nonetheless, from the answers to that question, which almost certainly undercounted the number of unemployed debtors, we found that between 7 percent and 17 percent of them were unemployed at filing. Interpreting ambiguous responses to a question not directed specifically at unemployment, we estimated that perhaps 14 percent of the debtors were unemployed, compared with a prevailing national rate of 7.6 percent.In Judge Sellers’s 1997 study in Ohio, the cases indicated an unemployment rate of about 20.5 percent, a rate about four times the rate of unemployment in Ohio in that year. These data show a remarkably consistent pattern of the unemployment rate in bankruptcy as a multiple of the unemployment rate in the general population. This finding persists over a period of sixteen years and in good times as well as bad. Without claiming an undue precision for these numbers, we can nevertheless assert that the unemployed are substantially overrepresented among bankrupt debtors.
For most workers whose work contracts exclude golden parachutes and hefty severance payments, unemployment is financially devastating. By itself, unemployment is a major contributor to the woes of the people in bankruptcy. The double-digit unemployment rates of the magnitude found in our bankruptcy sample are usually reported only for groups of workers believed to be extremely marginal to the labor force—teen-aged males (19.8 percent), black single mothers (13.9 percent), and high school dropouts (11 percent).
From The Fragile Middle Class by Teresa A. Sullivan, Elizabeth Warren, and Jay Lawrence Westbrook. Published by Yale University Press in 2020. Reproduced with Permission.
Teresa A. Sullivan is President Emerita and University Professor of the University of Virginia. Formerly Leo Gottlieb Professor of Law at Harvard Law School, Elizabeth Warrren is now the senior United States Senator from Massachusetts. Jay Lawrence Westbrook is Benno C. Schmidt Chair of Business Law at the University of Texas School of Law.