The Agenda

The Minimum-Wage Debate and the Case of the Missing Dishwasher

The idea of a statutory minimum wage is not new. It has been around for over a century. Yet it has nevertheless become the central domestic policy issue of our time. When you consider the domestic policy debate of the last few years, it’s not actually that surprising that we’re debating a big increase in the federal minimum wage.

The first two years of the Obama presidency saw dramatic, and in some cases very expensive, policy change and alarmingly high deficits, driven largely by collapsing tax revenues. This transformative moment, which contemporary progressives don’t fully appreciate, sparked a political backlash, which limited President Obama’s room for maneuver. Even after the president’s reelection, the White House finds itself tightly constrained, in political and fiscal terms. So while President Obama might want to fight for comprehensive climate change legislation to secure his legacy, or granting unauthorized immigrants a path to citizenship or a new transfer program, he has found those roads blocked by determined Republican opposition, politically vulnerable members of his own legislative coalition, and the indifference of the broader public to many of the causes he cares about most, he finds himself in a position in which raising the federal minimum wage is his most attractive option.

Among other things, raising the federal minimum wage generates no explicit cost to taxpayers (though raising wages for federal contractors does have an explicit cost) and it is a “wedge issue,” i.e., it is an issue that divides the opposing (conservative) coalition. The White House would be foolish not to press its advantage on this issue. To most U.S. voters, raising the minimum wage seems like a commonsense, cost-free way to better the lives of American workers. Calls for raising the minimum wage are an excellent way to change the subject from, say, the problematic implementation of Obamacare, which has already caused considerable dislocation in the individual insurance market, and which threatens to cause similarly severe dislocation in the small group health insurance market in the coming year.

Moreover, the past two decades have seen new empirical work which suggests that modest increases in the minimum wage might prove less economically damaging than had previously been understood. These findings are contested, and not always very well understood (least of all by me). There are many different factors at play. Let’s accept for the moment that increasing the minimum wage by X amount does not reduce total employment levels. Does that mean we’re in the clear? Not necessarily. (Thanks very much to a friend of The Agenda who took the time to shed light on a subject I find pretty bewildering.)

One can imagine a scenario in which increasing the minimum wage doesn’t reduce the total number of full-time equivalent employees, but it changes the composition of the workforce, as less productive workers find themselves pushed out of the workforce and more productive workers are drawn into it. (In a related vein, we might see a shift of less productive workers into occupations that are subject to lower statutory minimums, like tipped restaurant servers.) Recently, Kirk Johnson of the New York Times reported on low-wage workers who travel long distances from states with low (or no) state-level statutory minimums to states with high (or higher) state-level statutory minimums. Though Johnson doesn’t give us a good sense of the number of workers who travel across jurisdictions to take advantage of higher minimum wages, the existence of such workers adds a wrinkle to the scholarly debate over the impact of minimum wage laws. To isolate the effects of minimum wage laws, scholars have tried to exploit differences in employment outcomes across contiguous counties. The idea, as I understand it, is that if minimum wages actually matter, we should see a significant discontinuity between County A in a state that chooses not to raise its state-level statutory minimum and neighboring County B in a state that does choose to do so. But if ambitious, capable workers from County A start streaming across the border to County B to earn higher wages, the picture starts to get blurrier. It could be that the reason we don’t see a significant discontinuity is because the composition of the workforce in both counties is adjusting to the new wage differential. I doubt that this is a huge factor, particularly over short periods of time. Yet it seems premature to rule it out as a contributing factor.

Similarly, one can imagine that while the number of workers remains the same, the intensity of minimum-wage jobs increases. Consider the case of Amazon UK’s Rugely fulfillment center, where workers drawn from the ranks of the long-term unemployed struggled to keep up with the demands of packing and shipping customer orders. Workers who needed more time and training to achieve a given level of productivity were quickly weeded out of Rugely, as Amazon UK found it too expensive to retain them. Essentially, Amazon UK’s hiring and firing practices reflected Michael R. Strain’s observation about the risk employers take on when they hire a worker at a given wage. “Because of the federal minimum wage,” Strain writes, “the company knows that it has to take at least a $7.25-an-hour chance on a worker.” When we reduce the minimum wage, we in effect “significantly mitigate employers’ risk from hiring a long-term unemployed worker.”

Johnson’s (legitimately awesome) article includes a useful illustration of how an increase in the minimum wage might impact the intensity of work effort. One of Johnson’s interviewees, 20-year-old Carly Lynch of Idaho (where the $7.25 federal minimum wage is the law of the land), travels 20 miles a day to Ontario, Oregon (which has a state-level minimum wage of $9.10), where she works at a bar and restaurant. Rather impressively, Lynch doesn’t just work long hours at a bar and restaurant. She also competes in rodeos as a barrel racer, a difficult, resource-intensive undertaking, and so the promise of a higher wage is particularly attractive to her. But Lynch earns her higher wage. One of her employer, restaurant owner Angena Grove, explained to Lynch that “she would have to work harder than before for that money”; servers in the Oregon restaurant cover five tables rather than the three Lynch covered in her Idaho restaurant, and “higher labor costs meant getting rid of the dishwasher.” That is, Grove and her co-owner, her husband Shawn, could no longer outsource the labor-intensive, less-skilled task of dishwashing to a low-wage employee. They had to rely on their servers to pick up the slack.

This particular substitution is of particular interest to those of us who follow the immigration debate, as one of the classic examples of how less-skilled natives and less-skilled immigrants complement each other in the workplace is that less-skilled natives tend to take jobs that require a high degree of English language proficiency, like being a restaurant server, while less-skilled immigrants tend to take jobs that do not, like the job of dishwasher, or other tasks that don’t require much interaction with customers. (On a tangential note, I get the sense that Lynch is what we’d call a “go-getter,” and the real dilemma she faces is that for a variety of reasons, rural Idaho is not a great place for someone with her drive and talent to find remunerative work, and to build the human and social capital she needs to increase her productivity; on the other hand, it’s probably a good place to live if you want to maintain horses and pursue a serious career in rodeo.)

This intensity point has lots of interesting implications. Last month, the economists Mark Bils, Yongsung Chang, and Sun-Bin Kim released a working paper (gated) on “How Sticky Wages in Existing Jobs Can Increase Hiring.” The paper does not explicitly address minimum wage laws. Instead, it presents a matching model of employment in which wages are flexible for new hires but inflexible for workers who’ve already been hired. The basic idea is that if wages are sticky, employers will demand more work effort from their workers during recessions; this in turn will lower the marginal product of labor for new hires, thus making hiring less attractive. This model doesn’t necessarily reflect how the U.S. economy works in practice. It does, however, seem consistent with recent “jobless recoveries,” in which employers have been reluctant to take a risk on new employees, yet they have devoted considerable effort to raising the productivity level of their existing workforce. (The Bils et al. model also brings to mind Michael Schrage’s “hireless economy” thesis.)

So why might the Bils et al. model be relevant to the minimum wage debate? It dovetails with the work of Jonathan Meer and Jeremy West of Texas A&M University, who have found that the impact of wage floors is more apparent in job growth than in employment levels. Specifically, Meer and West find that higher wage floors tend to reduce net job creation, particularly for younger workers and in industries with a higher proportion of low-wage workers.

The Congressional Budget Office has released a new analysis of the potential effects of a minimum-wage increase. Critics of a minimum-wage increase are highlighting the CBO’s projection that a minimum-wage increase would reduce total employment and proponents are highlighting its finding that it would increase household incomes at the bottom of the distribution. Josh Barro of Business Insider argues that if the CBO is right and the impact of a minimum-wage increase on family incomes for the poor is substantial and the employment impact is relatively modest, we ought to consider raising the minimum wage by even more than in the latest White House proposal. Josh is careful to leave open the possibility that the CBO analysis is incorrect in its sanguine assessment. I was impressed by the breadth of the sources the CBO analysts drew on to make their assessment. My concern is that the CBO analysis captures the effects of a minimum-wage increase up to 2016, and that if a minimum-wage increase really does dampen net job growth, its negative effects will become more pronounced over time. Indeed, the CBO allows for this possibility. Among other things, the CBO notes that:

Employment reductions after a minimum-wage increase are probably larger over a longer term, in part because those reductions may be less attributable to the elimination of existing low-wage jobs than to slower growth in the number of low-wage jobs, which is difficult to detect in short-term studies.


Raising the minimum wage from $7.25 to $10.10 represents a 39 percent increase, which would be larger than most of the increases that have been studied, and CBO expects that employment would be more responsive to a larger increase. Many employers incur adjustment costs when they reduce staffing (especially if that requires restructuring their operations), which may deter them from laying off low-wage workers in response to a small increase in the minimum wage. But the savings from not having those employees are more likely to exceed the adjustment costs when the minimum-wage increase is large.

Just how large is it? Compare it to other bumps in the minimum wage:


Although the percentage increase in the federal minimum wage from 2007 to 2009 was similar to the one projected under the $10.10 option, the fraction of the workforce affected under that option would be about five times as large (see Table A-1). When a greater proportion of a firm’s work hours are affected by the minimum wage, the adjustment cost per worker of reducing staffing (again, especially if the firm is restructuring its operations) is probably smaller, making the firm more likely to reduce employment.

These cautionary notes are all in the Appendix of the widely-discussed CBO analysis. They’re not likely to get as much attention as the headline numbers. That is unfortunate. If we really do see bigger employment impacts over time, and if these employment impacts disproportionately impact foreign-born non-citizens who are ineligible for transfer payments, the poverty impact beyond 2016 could prove substantial. We can’t expect the CBO to think through these fine-grained questions, which is why we need to think for ourselves.

I eagerly await more work on the central issue that Meer and West have focused on in their research, namely the impact of wage floors on net job creation.


The Latest