Jared Bernstein, a former chief economist to Vice President Joe Biden, is a senior fellow at the Center on Budget and Policy Priorities and author of 'The Reconnection Agenda: Reuniting Growth and Prosperity'.
Source: See data table at end of post.
Source: See data table at end of post.

One of the most destructive ideas in poverty policy is what supporters, such as House Speaker Paul D. Ryan and Republican presidential candidate Jeb Bush, call “opportunity grants” and what the rest of us call block grants.

The idea is to take a set of programs, such as the Supplemental Nutrition Assistance Program (SNAP), housing vouchers, child care, and more, and turn them into a consolidated block grant, which means providing states with a fixed amount of funding to run the programs. When Bush claims that he’ll end food stamps, this is what he’s talking about. Because welfare reform turned cash assistance to poor families into a block grant in the mid-1990s, we have a reference point whereby to judge the effects.

The main reason this idea is so destructive is that it undermines the essence of the safety net, or its countercyclical function. The figure above makes the case (as the figure’s a bit gnarly, I pasted in the data below). It shows that when the last downturn hit, SNAP caseloads quickly responded to the loss of income among low-income households, while Temporary Assistance for Needy Families (TANF) hardly responded at all. The opportunity grant threatens to turn SNAP into TANF, killing the former’s countercyclical aspect in the same way block grants killed it for TANF.

We can learn a lot more about this bad idea from studying how SNAP worked in the last recession and thereafter.

Its countercyclical response in the figure is undeniable. Given that, some critics try to move the goalposts by granting that SNAP is responsive at the start of a downturn but arguing it’s less so later in the expansion, implying that it’s taking too long for caseloads to fall as the economy has improved. There’s no question that SNAP caseloads, which are now slowly coming down, remained elevated as the unemployment rate fell. But for a number of reasons, that proves little.

The decline in the unemployment rate has overstated the improvement in the job market. When people give up looking for work and leave the labor force, the unemployment rate declines. That partially explains the fall in the jobless rate and is one reason it correlates less with the SNAP rolls. What’s needed is a measure that records when people leave the job market, one like the employment rate, which measures the share of the working-age population with a job. As you see in the above figure, the employment rate fell sharply in the downturn as SNAP rolls spiked up. Now, as the employment rate begins slowly climbing, we see SNAP slowly starting to come down.

Check out some of these striking correlations (a statistic that runs between -1 and +1, with 0 meaning uncorrelated). Over the full period for which I have data, here’s what you get:

snapcorr2


As you’d expect, higher unemployment is correlated with higher SNAP participation and higher employment rates are (negatively) correlated with lower SNAP participation. The employment rate is somewhat more closely correlated with SNAP participation over the long run, but since 2007, the difference is much larger, as unemployment is hardly correlated at all while employment rates, which are recovering much more slowly, have grown more correlated with SNAP.

There’s a known lag between the recovery and the decline of SNAP rolls. The non-partisan Congressional Budget Office has looked at this question and concluded that as the economy continues to improve, SNAP will continue to fall. But it notes the asymmetry you see in the figure above: “SNAP participation generally declines when the economy improves, though typically with a substantial lag. CBO projects that participation will follow that pattern again in the coming years.”

Why is that? Surely, economic inequality is a critical factor here. The last few recoveries have started out “wageless” or “jobless.” In this expansion, the finance sector recovered way before the rest, and even six and a half years into the recovery, we’re still not at full employment. Poverty rates have been pretty stable, in large part because of the very countercyclical function “opportunity grants” are threatening. But as I’ve documented in great detail, until we get to and stay at full employment, the benefits of growth will not adequately lift the poor.

In fact, rather than revealing undue generosity, what the SNAP rolls’ asymmetry reveals is how quickly poor people get whacked by recessions and how slowly they recover.

SNAP expert Dottie Rosenbaum provides further explanations as to why the SNAP rolls rose as much as they did and took so long to start falling, including higher participation rates among eligible people who weren’t getting nutritional support when they should have been. Rosenbaum suggests that this increase reflects “federal and state efforts begun before the recession to reach more of the eligible households — primarily working families and seniors — by simplifying SNAP policies and procedures.” That, by the way, is a good thing.

When hit with such evidence about the importance of SNAP’s countercyclical function, Ryan (R) argues that policymakers could come up with a formula to adjust the block grants to go up in recessions. But besides being administratively challenging — suppose, if he’s talking about a discretionary formula, Congress doesn’t want to go there — this makes zero sense: “Here’s a neat idea! Let’s break SNAP and then fix it again so it works the way it does now.”

So back off Jeb! et al. SNAP works. In fact, it doesn’t just work for a day. It works for a lifetime, as research tracking children who received nutritional support when they were kids finds a spate of positive outcomes in adulthood (see Figure 1 here).

I sincerely welcome input from all sides of the aisle as to how we can amp up our policies’ anti-poverty effectiveness. This block granting idea, however, pushes hard the other way.

data_fig2