Employment status changes put millions at risk of losing SNAP benefits for years

As the Senate moves toward conference with the House of Representatives on the 2018 Farm Bill, the central policy issue the two chambers must resolve is whether to impose even more work requirements on those receiving Supplemental Nutrition Assistance Program (SNAP; formerly the Food Stamp Program) benefits. New research shows that many beneficiaries would have difficulty meeting the requirements and would lose their benefits.

The House plan would impose mandatory work requirements on SNAP participants ages 18 through 59 who are neither disabled nor raising a child under age 6. This would entail proving every month that they are working or participating in a work program at least 20 hours per week or that they qualify for an exemption. Those who cannot comply would lose access to benefits for a period between 12 and 36 months.

These new requirements would pose a challenge for millions of workers, especially those with low levels of education. Earlier this week, the Center on Budget and Policy Priorities released a new report by Kristin Butcher and Diane Schanzenbach that describes the characteristics and employment outcomes for adults who do not have children under the age of 6 at home and who are likely to participate in SNAP or Medicaid. They find that these individuals typically work in jobs with low and stagnant wages. Their jobs are also more volatile, with higher rates of unemployment, less employment stability, lower tenure levels, and more displacement.

To understand the likely challenges in meeting proposed work requirements, we dig deeper into monthly employment stability, finding considerable churn in the labor market across the 20 hours per week threshold, making it tenuous for people subject to work requirements to retain their SNAP benefits.

Over 16 months, between one in five and one in three adults 18 to 59 without young children at home switches between working more than 20 hours per week and a different employment status: less than 20 hours per week, seeking employment, or being out of the labor force.

Employment status transitions are frequent over consecutive months and years

Using data from the Current Population Survey (CPS) from June 2016 to May 2018, we describe the extent to which adults between ages 18 and 59 without young children at home transition in or out of the labor force, as well as between employment statuses.

[1] Because the CPS surveys the same households for four consecutive months in two consecutive years, we are able to identify the type and frequency of employment transitions over both four and 16 months.[2] This sample is a close approximation of those whom the work requirement proposal would target, but data limitations prevent us from limiting the analysis to non-disabled adults only.

We categorize each individual in each month into one of four categories: (1) employed and worked more than 20 hours a week, (2) employed and worked less than 20 hours a week, (3) unemployed and seeking employment, or (4) not in the labor force. One is considered to have a stable employment status if she does not change categories in each of four consecutive months or in eight of 16 months, and is considered to have made an employment status transition if she switched between any of these categories at least once. There is no employment status transition when a worker changes jobs but works more than 20 hours a week at each job.

Our focal and conservative estimate makes these categorizations using actual hours worked, but allows reports of usual hours worked to replace actual hours worked if the respondent reported being employed but not at work in a given week. This empirical choice reduces the number of month-to-month transitions because respondents are able to smooth week-to-week variability in hours worked if they are not working that week. However, the work requirements would generally not allow a worker to retain their SNAP benefits if their actual hours fell below the threshold even if they “usually” worked more than 20 hours a week per month.[3] For this reason, we also consider a second estimate that differs by how we count employed workers not at work; for the second estimate, displayed in the appendix, a worker is coded as working zero hours if she was employed but not working that week.

Figure 1 describes the distribution of stable and changed employment statuses in the last two years (June 2016 to May 2018) by the more conservative estimate, in which actual hours are used except for employed workers not at work that week.

By the conservative categorization, one in five adults made at least one transition in four months and three in 10 did so over 16 months. The share of switches between working more than 20 hours per week and another employment status were 69 percent over four months and 63 percent over 16 months. Because the 16-month estimates only observe transitions that occur within or between the two four-month periods, these are low estimates of the total number of transitions that occurred over the entire time period.

Over 16 months, at least 19 percent of adults between the ages of 18 and 59 without young children at home switched between working more than 20 hours per week a month and less than 20 hours per week, seeking employment, or being out of the labor market. Over four consecutive months, 14 percent of working-age adults without young children transitioned between being employed more than 20 hours per week and another employment status.

In other words, out of four months in 2016-2018, 14 percent of adults in this group spent at least one month employed more than 20 hours per week, and at least one month employed fewer than 20 hours per week, unemployed, or out of the labor force. Over 16 months, that number grew to 19 percent. Using only actual hours worked, the alternative measurement, 21 percent of adults 18 to 59 without young children at home spent at least one month employed more than 20 hours per week, and at least one month not out of four months in the past two years. Over 16 months, that number grew to 34 percent.

Approximately 62 percent were employed more than 20 hours a week in each of four consecutive months, falling to 55 percent over 16 months. Very few people were consistently employed at less than 20 hours per week over a four-month period (1.1 percent) or 16-month period (0.4 percent). While in any given month, the monthly share of unemployed varies between 3 and 8 percent, fewer than 1 percent were unemployed but seeking employment consistently (0.7 percent over four months; 0.1 percent over 16 months.)

Approximately 14 percent of adults in our sample were consistently out of the labor force over 16 months. Women were more likely to be consistently out of the labor force than men (16.7 percent versus 10.5 percent) and those with a high school degree or less were more likely to be consistently out of the labor force than those with more education (20.8 percent versus 9.7 percent).

Frequency of employment status transitions similar during recession and expansion

How many transitions between different employment statuses were made between employment statuses in four-month or 16-month period among adults without children under 6? Over a four-month period, one could make up to three transitions; over a 16-month period, one could make up to seven transitions measured in the data.

Figure 2 shows the number of transitions over the past two years (June 2016-May 2018) as well as for the start of the Great Recession (June 2008-May 2010) by the more conservative sample.[4]

We cannot observe if there were transitions between statuses in the eight months between four-month periods. As an illustrative example, image an adult was in the sample for June-September 2016 and June-September 2017. If that person made a job status transition from October 2016 to May 2017 and switched back to the original status by June, we would not observe that transition.

While there was slightly more stability in the 2016-18 period than in the 2008-10 period and a slightly larger share of multiple transitions in 2008-10, the composition of stability and transitions is remarkably similar. Even in the stronger labor market with substantially lower unemployment in 2016-18, employment status churn was common. The frequency of transitions and multiple transitions matters because administrative expenses are much higher among those who are switching between employment statuses.

Conclusion

Over 16 months, between one in five and one in three adults 18 to 59 without young children at home switches between working more than 20 hours per week and a different employment status: less than 20 hours per week, seeking employment, or being out of the labor force.

Though these estimates are substantial, they do not represent all who might lose benefits or fail to obtain benefits for which they are eligible due to the work requirements. As others have written, the administrative burden of employment or employment assistance verification would be substantial, expensive, and constitute a barrier to entry for potential beneficiaries. These factors could lead eligible workers to not seek benefits or to lose benefits through administrative failures.

Furthermore, the volatility of employment statuses and potential for oversight error is met with a persistent penalty. That is to say, dipping below 20 hours per week in one month could result in a loss of benefits between one and three years. A server whose employer cuts her weekly work to 19 hours for one month would face the drastic penalty of losing SNAP benefits for up to three years.

Those who benefit from SNAP typically receive low wages, experience higher rates of unemployment, and have less employment stability. Month-to-month employment and hours worked are highly variable among those potentially affected by a SNAP work requirement. Using a conservative measure of monthly employment status shows that many people—one in five over 16 months—could be exposed to sanction under the House work requirement proposal that SNAP benefit receipt be contingent on monthly participation in employment or employment assistance of at least 20 hours per week. By actual hours worked, this number grows to one in three.

[1] Data come from IPUMS CPS monthly surveys from June 2008-May 2010 and June 2016- May 2018. The sample is limited to US citizens, non-active military, aged 18-59, with no children under age 6 in their household. It also limited to those who could be successfully matched across 4 contiguous months for the 4-month group and those who could be successfully matched across 4 contiguous months and then those same 4 months 8 months later for the 16-month time period. Hours worked per week are determined by the actual hours worked in the reference week (the week prior to participation in the CPS) across all jobs. A person transitioned between statuses if their employment and labor force status changed at any point over the 4- or 16-month period, and they are considered “stable” if there was no change during the 4- or 16-month time period. Number of transitions is calculated as the number of months that the person’s labor force and employment status did not match their status in the next month.

[2] We do not observe changes in employment status during the 8 months between survey waves.

[3] Participants may be able to retain benefits if they can prove that they were not working 20 hours per week in a month for an allowable reason.

[4] We cannot observe if there were transitions between statuses in the eight months between four-month periods. As an illustrative example, image an adult was in the sample for June-September 2016 and June-September 2017. If that person made a job status transition from October 2016 to May 2017 and switched back to the original status by June, we would not observe that transition.

Appendix Figures