Book Review: The Redistribution Recession

The Redistribution Recession: How Labor Market Distortions Contracted the Economy by Casey Mulligan, an economist at the University of Chicago, looks at the extent to which Americans will withdraw from the work force if you pay them to stay home and not work. This is kind of a companion piece to my review of Peter Wallison’s book on what caused the Collapse of 2008. Wallison asks “Why did things fall apart?” and Mulligan asks “Why can’t Americans put their economy back together [in less than 7 years]?”

The book is written for professional economists but readable for anyone with an undergraduate Econ 101 background.

If you live in an expensive city and are acquainted with families collecting welfare the book confirms what you might have noticed, i.e., that it would be irrational for the adults in the family to enter the workforce. Here in Cambridge, Massachusetts, for example, the welfare families that I have spent the most time with occupy apartments with a market rent of about $4,500 per month ($54,000 per year in post-tax income) in a building with a swimming pool, two gyms, and a variety of other luxurious facilities. Their health care is free through some combination of Medicaid, Obamacare, and a city-run health system. Their food is mostly free through food stamps. They can get cash from TANF and some similar programs. They would need to earn at least $160,000 per year pre-tax to obtain the same standard of living at market prices. However, even if someone were to offer the adults in the family a $160,000 per year job it would not be rational for them to accept it. If they were ever to lose that job it would take many years of paperwork, bureaucracy, and waiting lists to get back to their current welfare lifestyle.

Able-bodied American adults who live off alimony and child support, as chronicled in Real World Divorce, face a similar situation and, according to the attorneys we interviewed, respond similarly. If an adult tapping into a former spouse or sexual partner gets a high-paying job there is a risk of the defendant going to court and getting the payments reduced due to a lack of need. If the job were lost for some reason it is possible that the original cash flow might never be reestablished.

Mulligan makes some attempt to model these high switching costs and long switching delays with (a) a “model of take-up” of government benefits (Appendix 3.2), and (b) program participation weights, which make a dollar of assistance worth less to the recipient than a dollar of cash picked off the sidewalk. But the core of Mulligan’s analysis is that income-based government handouts function as a high marginal tax rate in that if the low-income person earns a bit more he or she qualifies for fewer government benefits and therefore has only slightly more spending power than before. Mulligan looks at the 2009 expansion of means-tested welfare in at least the following areas:

• unemployment benefits
• food stamps
• mortgage forgiveness
• non-mortgage debt forgiveness
• health insurance subsidies (Obamacare; not yet implemented during the period of Mulligan’s analysis)

Underlying Mulligan’s work is an assumption that policymakers (and family court judges) reject:

A basic economic principle in this chapter, and much economic analysis of the labor market, is that the pecuniary reward to working is an important determinant of how much people work. When social programs increase what they pay to someone who does not work, they diminish this difference, and the result is that some people work less.

Nobody who believed in this principle would offer to pay people to sit home and play Xbox for 99 weeks (January 2011 posting and January 2013 follow-up citing economic analysis). Yet that is apparently what the American people wanted their elected government to do. Thus it may be the case that Mulligan’s book will have limited impact because more than half of Americans disagree with the assumption that payments to non-workers encourage non-working.

Mulligan makes some effort to justify this assumption:

Before the recession began, well over one hundred million Americans were not working. To be sure, some of them could find no reward in the labor market and would be stuck without gainful employment no matter how lean the safety net got. But many others were not working by choice. We all know skilled stay-at-home mothers or fathers who could readily find a job but believe that the pay from that job would not justify the personal sacrifices required. They are examples of people who deliberately do not maximize their income. Others are people who turn down an out-of-town promotion in order to avoid relocating their family, and workers who eschew higher-paying but less safe occupations. Earning income requires sacrifices, and people evaluate whether the net income earned is enough to justify the sacrifices. When the food stamp or unemployment programs pay more, the sacrifices that jobs require do not disappear. The commuting hassle is still there, the possibility for injury on the job is still there, and jobs still take time away from family, hobbies, sleep, etc. But the reward to working declines, because some of the money earned on the job is now available even when not working.

And he cites research that tries to put numbers on the correlation between increased welfare and reduced working hours:

Decades of empirical economic research show that the reward to working, as determined by the safety net and other factors, affects how many people work and how many hours they work. To name a small fraction of the many studies: Hoynes and Schanzenbach (2012) show how potential participants stopped working or reduced their work hours when the food stamp program was introduced. Studies of unemployment insurance (see Appendix 4.2) find that program rules have a statistically significant effect on how many people are employed, and how long unemployment lasts. Yelowitz’s research (2000) shows how a number of single mothers found employment exactly when, and where, state-level Medicaid reforms increased their reward from working. Gruber and Wise (1999) and collaborators show how the safety net for the elderly results in less employment among elderly people. Autor and Duggan (2006) and the Congressional Budget Office (2010) explain how the number of disabled people who switch from work to employment-tested disability subsidies depends on the amount of the subsidy relative to the earnings from work. Murphy and Topel (1997) show how poor wage growth among less-skilled men helps explain their declining employment rates during the 1970s and 1980s. Jacob and Ludwig (2012) show that means-tested housing assistance reduces labor force participation and earnings among able-bodied working-age adults.

But most of the book assumes that readers are convinced by Econ 101 that economic incentives will change peoples’ behavior and proceeds with analysis based on that assumption.

If you’re wondering why you can’t get anyone to come over and fix your roof when supposedly millions of Americans are unemployed, Mulligan explains:

Traditional labor and macroeconomic theory predicts that marginal labor income tax rates and binding minimum wages distort the labor market and thereby reduce aggregate labor usage, reduce aggregate consumer spending and investment, and, in the short term, increase wages, labor productivity, and the usage of factors that can take the place of labor hours. As a result of greater labor productivity, part of the population—those (if any) not subject to the marginal tax rates or minimum wages—actually works more, even while aggregate work hours are less.

In other words we’re trending toward a society where about half of the working-age adults will kill themselves with 60-80-hour weeks while the other half will relax on the sofa.

Just how much has the welfare state expanded?

With the exception of Medicaid, subsidies flowing to the unemployed and to financially distressed households in the forms of consumer loan forgiveness and government transfers almost tripled after 2007. A minority of that increase is due to an increase in the number of people who would have been eligible for subsidies under prerecession rules, and a majority is the result of more than a dozen changes in benefit rules made possible by several new federal and state government statutes.

Mulligan says that “businesses perceive labor to be more expensive than it was before the recession began” and that’s why they aren’t hiring a lot of workers. Shouldn’t we then see a huge spike in capital spending as companies buy robots to replace the former human employees? Mulligan gives us the academic version of “no”:

Assuming, as economists usually do in aggregate analysis, that capital enhances the productivity of labor, and labor enhances the productivity of capital, then the efficient reaction to less labor is to have less capital. Investment is the rate of change of the capital stock, so even small reductions in the capital stock may be achieved by large investment reductions for a short period of time. For this reason, investment is expected to decline by a much greater percentage than consumption in the short term, and by the same percentage in the long term. In this view, the investment decline is entirely a reaction to the labor market, and not a cause of the low rates of labor usage.

Welfare is not equally available to everyone, however, and Mulligan digs into the demographics. Single mothers are in the best position to get cash without working. Custody of their child can yield payments from the father (over $100,000 per year tax-free if they use the information in Real World Divorce thoughtfully). If they didn’t choose a high-income father for their child(ren) then “Single mothers are much more likely than [married women without children] to be eligible for SNAP, Medicaid, and other means-tested government programs when not working because (1) they do not have a spouse present whose income by itself would likely put household income above FPG and (2) the latter group has no children (children are the target of a number of means-tested programs).” Mulligan finds that American women behaved in accordance with Econ 101. The married women who could not get welfare worked similar hours to their 2005-2007 hours. Single mothers, however, even those with at least some college education, dramatically reduced their working hours as welfare became more lucrative.

What about old people? Mulligan notes that they actually increased their working hours in response to the recession due to a lack of new welfare programs available to the elderly: “marginal tax rates [including the tax-like effect of reduced welfare eligibility] for the nonelderly increased sharply, while marginal tax rates for the elderly hardly changed.”

What about the minimum wage discouraging employment from the employer side? Mulligan looks at this carefully: “My 2011 paper (Mulligan 2011c) estimated a monthly time series model of national part-time and full-time employment per capita for each of twelve demographic groups distinguished according to race, gender, and age, relative to prime-aged white males, whose employment rates were assumed to be unaffected by the July 2009 minimum wage hike. I used the model to estimate the amount and composition of employment losses due to the hike for the average month between August 2009 and December 2010, and found that lower-skill groups had the greater employment losses. The net nationwide employment loss estimate was 829,000, which includes employment gains among more skilled people. … Thus, the minimum wage hikes since July 2007 might explain about roughly one-third to one-half of the employment decline among persons aged sixteen and over who were neither elderly nor household head or spouse.” [Personal experience: I have taken on a recent high school grad (friend’s son) as an intern in my web development business. He was an above-average student in a Harvard University intro CS class and also completed AP Computer Science in high school plus an additional programming class. His current productivity is about 1/100th of a $25/hour Ukrainian or Filipino contract programmer so any wage+benefit package above 25 cents per hour would be above the market-clearing price. Yet he can never reach the productivity of the foreign contractors unless he can get substantial work experience.]

Mulligan tells us that employers don’t want to pay more for workers with the same skill level: “To the extent that wages and other employment costs increased since 2007 without a commensurate increase in total factor productivity, it is no puzzle that employers do not want to return to the number of employees they had before the recession began. Wages and employment costs have been greater since 2007 in part because of the federal minimum wage hikes, but more significantly because the safety net expansions have given employees, and prospective employees, less reason to accept reductions in their compensation.”

One of the biggest surprises in the book is the importance of federal mortgage loan forgiveness: “This chapter shows how all of these outcomes, and more, may be a direct result of stark incentives created by the FDIC and HAMP programs (hereafter jointly referenced as FH) and their practice of targeting the ratio of housing expenses to borrower income. The FH programs offer modifications on the basis of borrower income reported to the United States Internal Revenue Service. The first section of the chapter shows how the programs resemble government safety net programs, except that the marginal income tax rates from mortgage modification far exceed 100 percent in some instances.” It turned out that a person’s mortgage payments would be reduced by $1.31 for every $1 fall in income. In other words a lower-income American with a mortgage would have more spending power by working fewer hours and/or quitting altogether. Mulligan also notes that as the program was unfavorable to lenders they had a huge incentive to “promote borrower confusion and uncertainty about the disposition of their modification application.” (i.e., the bureaucratic run-around might not have been accidental!)

How about the theory that spreading cash around to low-income Americans will grow the economy because low-income people spend all of their cash instead of saving any? Mulligan points out that this may not be true in practice: “… some of the most labor-intensive industries are hotels, coal mining, and restaurants, whereas farm commodities and cell phone services are some of the least labor-intensive. If the safety net redistributes resources from people who spend a relatively large fraction of their resources on restaurants and hotels toward people who spend a large fraction on groceries and cell phone services, the redistribution may well reduce national labor demand rather than increase it. … The fact that unemployed people tend to consume their benefits when they receive them is an important indicator of the insurance benefits of the UI program, because it tells us that the benefits are especially important for maintaining their living standards (Gruber 1997). But this fact is irrelevant for understanding aggregate labor demand, unless it happens to be that the consumption items purchased by the unemployed are more labor-intensive in their production than are other goods and services in the economy. … it is possible that the poor use their safety net benefits to purchase consumer items that are of less-than-average labor intensity, and the rest of the country reduces spending on labor-intensive items, so that the net result of redistribution is to reduce aggregate labor demand.”

Mulligan considers a theory popular among pro-handout politicians and their non-economist supporters: “safety net expansions do not prevent people from finding jobs during a recession because there are no jobs to be found.” He points out that “This exposition is literally untrue because millions of jobs were found every month during the recession (de Wolf and Klemmer 2010), but I take it to mean that good jobs are difficult to find during recessions. But a lack of good jobs might mean that the safety net is even more potent at keeping people from working during recessions because safety net benefits look more generous when compared to the pay from a bad job than when compared to the pay from a good job.” He then tests the theory carefully by looking at the seasonal teen labor market. These young workers become available during Christmas and summers both in recessions and expansions. Mulligan looked at thirteen summers and fourteen Christmases during recessions since 1948 and found that teens were able to get jobs at roughly the same rates in both recessions and expansions. As childless teens are ineligible for welfare this tends to confirm Mulligan’s theory that people who don’t have an alternative way to get cash will resort to finding a job.

With 363 pages of analysis and mathematical models, Mulligan shows that essentially all of what we have observed since 2009 can be explained by the following:

• many able-bodied Americans will not work if they have a reasonably attractive alternative
• the federal government, starting in 2009, made not working much more financially rewarding for tens of millions of working-age adults

I have a feeling that the high switching costs and long switching delays to getting back onto welfare are not completely accounted for in Mulligan’s models. Housing, health care, food stamps, and cash don’t automatically start flowing as soon as a job is lost and people hate uncertainty about where their housing, health care, food, and cash are coming from (look how many people are willing to take boring jobs with mediocre pay if they believe that there is very high job security). If the cost and delay of getting back onto welfare were weighted more heavily, Mulligan’s models might show an even larger drop in employment with increased welfare benefits. As it is, however, his model matches pretty closely what happened in terms of employment and capital spending.

This is an important book and deserves to be read by anyone interested in the U.S. economy. There is a lot of interesting information still accessible even to those whose econ/math background isn’t adequate for fully appreciating Mulligan’s mathematical model. And given the political debate around raising the minimum wage to $15 per hour (uncharted territory for the U.S. economy where the historic minimum wage peak was $8.54 in 2014 dollars, reached in 1968 (Pew Research)), the book is highly relevant for policymakers going forward.

More: Read the book.

Related:

• “The Labor Market Effects of the VA’s Disability Compensation Program” by Mark Duggan, econ professor at Stanford, November 2014: 18 percent of people who can get disability benefits will drop out of the labor force.