Does the government subsidize low-wage employers?

Three months ago a California think tank published a tabulation of the federal and state dollars that boost the incomes of Americans who work but earn so little they qualify for government aid. The authors of the report as well as many readers seemed shocked by cost of the aid. Berkeley’s Center for Labor Research and Education estimated that the national and state governments paid out $153 billion in 2013 to finance health benefits, food stamps, and cash assistance to people in families containing a breadwinner who works at least part time and at least half the year.

A remarkable feature of the reaction to the report is that many readers interpreted the government aid dollars to represent a subsidy to low-wage employers (for example, here, here, and here). According to this view, government assistance to low-income families constitutes a handout to Walmart, McDonalds, and other low-wage employers. The assistance allows these companies to pay their workers lower wages than would be possible in the absence of the government aid.

For the majority of programs analyzed by the Berkeley researchers, this interpretation of government assistance payments is flatly wrong. Instead of subsidizing low-wage employers, most assistance programs reduce the availability of low-skill adults who are willing to work for low pay and lousy benefits. By shrinking the pool of workers willing to take the worst jobs, the programs tend to push up rather than push down wages at the bottom of the pay scale. Low-wage employers do not receive an indirect subsidy from the programs. Many must pay somewhat higher wages or recruit more intensively to fill their job vacancies.

There are two important exceptions to this generalization: the Earned Income Tax Credit (EITC) and child care subsidies targeted on working parents who earn low incomes. Because benefits under these programs are only payable to low-income families containing a parent who is gainfully employed, this kind of government subsidy encourages adults in eligible families to enter or remain in the job market rather than to drop out of it. By boosting the supply of potential low-wage workers, the two programs can put downward pressure on pay, indirectly benefiting employers who depend on less-skilled workers. Even in these cases, however, the main effect of the aid is to lift the net incomes of breadwinners earning low pay.

The authors of the Berkeley study highlighted the cost of four main programs: Medicaid, Temporary Assistance for Needy Families (TANF), the EITC, and the Supplemental Nutrition Assistance Program (SNAP or food stamps).  In all four programs a sizeable slice of benefit payouts goes to families containing a poorly paid wage earner. In three of the programs, however, a sizeable or even bigger slice of payouts goes to families without a worker. For many income-tested aid programs, including both TANF and SNAP, monthly benefits are typically more generous when family earnings are low or zero rather than high. As family earned incomes increase, benefits under the programs are cut back or eliminated.

Most careful analysis of the impact of this kind of means-tested program concludes the programs discourage work. The availability of health insurance, food coupons, and cash assistance when potential breadwinners do not work means that paid employment is less necessary. The fact that government benefits are reduced when the breadwinner’s earnings rise means that work is financially less rewarding. Both these effects tend to reduce, at least modestly, the amount of paid work that eligible breadwinners are willing to do. I do not argue the impact is large or that it affects most adults who are potentially eligible to collect means-tested benefits. On balance, however, benefit programs offering more generous payments to people with zero earnings than to people with comfortable incomes tend to reduce the supply of workers who are willing to accept very low pay.

Three decades ago an overwhelming share of government aid to working-age families followed this pattern. Benefits were substantially higher for working-age adults who did not work at all compared with working-age adults who had modest earnings. For example, adults who had low earned incomes found it hard to obtain government-subsidized health insurance, and this was the case even if the adults headed families containing children.  In order to qualify for Medicaid, working-age adults and their children might have to qualify for cash public assistance. In many states, this required families to have no or very low earned incomes. 

Eligibility limits for Medicaid and a new program, the Children’s Health Insurance Program (CHIP), have since been raised, making it easier for children and parents to qualify for government-subsidized insurance. This does not represent a subsidy to low-wage employers. It is a humane form of aid that ensures low-income adults and their children have access to affordable health care. Such assistance in combination with food stamps also permits low-income adults to be a bit choosier about accepting or declining low-pay or very inconvenient jobs.

The EITC and government child care subsidies offer direct inducements for low-wage workers to enter the job market, even if they must accept positions that do not offer adequate pay, decent fringe benefits, or convenient hours of work. The wage supplement offered by the EITC and the price discount implicit in a daycare subsidy make it feasible for some parents to work who would otherwise be better off remaining at home. Voters and policymakers understood this logic when the EITC and daycare subsidies were liberalized in the late 1980s and 1990s. The goal of the programs was to improve the standard of living of low-income families while simultaneously encouraging parents in these families to increase the share of family income derived from gainful employment. These incentives worked, especially when we were at or near full employment. The percentage of single mothers who held jobs increased.

The success of these two programs in boosting employment rates also succeeded in boosting these families’ net incomes and self-support, the primary goals of the policy shift. An indirect consequence of this success is that some parents who would have been out of the job market before the late 1980s are now working or looking for work. Their availability for work helps hold down wage costs for low-wage employers. If voters think this indirect effect of the EITC and child care subsidies is undesirable, they should push lawmakers to boost the minimum wage or improve the fringe benefits available to poorly compensated workers.

The public programs singled out in the Berkeley report are aimed at improving the health care access, nutrition, and net incomes of the nation’s most disadvantaged families. Three of the programs provide more assistance to nonworking Americans and their dependents than to families containing a working breadwinner. It is hard to see how this kind of program can be said to provide a subsidy to low-wage employers. Two programs that restrict benefit payments to working breadwinners—EITC and daycare subsidies—surely succeed in improving the net incomes of working but struggling parents. Indirectly the two programs increase the supply of low-wage workers, putting some downward pressure on wages. Labeling these programs as “subsidies” to low-wage employers has some merit, but it fundamentally misrepresents the distribution of benefits conferred by the programs. The main effect of the subsidies is to lift the net incomes of the working families that receive them.