Tuesday, December 10, 2013

Conservatives and the minimum wage: still lying after all these years! Plus: wonkeriffic discussion about why the minimum wage does not cause job loss!!

In recent weeks, there has been a lot of talk, and even some action, on one of the most powerful policy tools we have in this country for reducing economic inequality: the minimum wage. See, for example, these columns and blog posts by Arindrajit Dube, Paul Krugman, Jared Bernstein, Brad DeLong, and me, all of them published within a day of each other last week. (Yes, we all got the memo. Clearly, the vast left-wing conspiracy is nothing if not well-organized!). In his speech last week on economic inequality, President Obama expressed strong support for raising the minimum wage.

Even more encouraging are the recent victories for minimum wage increases at the state and local level in places like California, New Jersey, D.C.,  Albuquerque, and Sea-Tac, Washington. Minimum wage campaigns are also active in other cities and states and at the federal level as well. Many of those state and local victories came about through direct initiatives at the ballot box. You see, people really like the minimum wage. It's an issue that consistently polls very well. A Gallup poll released last week found that 76 percent of Americans support an increase in the minimum wage from its current level of $7.25 an hour to $9 an hour. UPDATE: And none other than the Wall Street Journal reports that 63 percent of Americans support a $10 minimum wage.

You know what this means, don't you? Of course you do.

It means the wingnuts will be shrieking and rending their garments about how increasing the minimum wage will throw the country into Great Depression-level rates of unemployment.

And here they come, right on schedule!

Just in the last week, we've seen the following shameless exercises in bogusity: the Heritage Foundation said that raising the minimum wage would "limit the number of jobs available." A top executive at White Castle claimed that such a raise would lead to layoffs. Right-wing economist Douglas Holtz-Eakin argued that increasing the minimum wage "impedes job creation." The Wall Street Journal even charged that there are "mountains" -- did you hear what I said, mountains! -- of evidence that  "minimum wages lead to fewer workers hired." Finally, the Washington Post's ever-delightful Jennifer Rubin -- where do they find these people? -- tweeted that President Obama's inequality speech "suggests he's economically illiterate" and linked to this anti-minimum wage screed by the American Enterprise Institute.

It's time to bring in Kathy Bates, who has something to say about the claim that raising the minimum wage leads to job loss:


Kathy, by the way, is going to be a regular feature of this blog. She will be showing up every time I'm debunking some bullshit claim related to economic inequality. That means she's going to be around quite a lot.

(The above clip is my very favorite moment thus far from this season of American Horror Story: Coven. You do know that American Horror Story is a blast and a hoot and the best, not to mention the gayest, show on TV right now, right?)

Okay, back to the minimum wage.

The fact is, the impact of the minimum wage on employment levels has been studied to death -- it is among the most studied topics in all of economics. For years, it was believed that the minimum wage caused job loss. This is mainly for two reasons. First, simplistic textbook models created a misleading impression of labor market dynamics when a minimum wage is introduced. Secondly, early econometric studies of the impact of the minimum wage were methodologically flawed. (N.B.:  this great paper, published earlier this year by economist John Schmitt of the Center for Economic Policy Research, is an important source for this post. Anyone interested in this topic should read it).

But beginning with the series of groundbreaking studies published in Myth and Measurement, the 1997 book about the minimum wage written by economists David Card and Alan Krueger, the economics profession slowly began to rethink the impact of the policy. Card and Krueger's book looked at the impact on employment of state minimum wage increases, using states that lacked such laws as controls. To the consternation of much of the economics profession (Card spoke about some of the nasty attacks he received from fellow economists here), they found that minimum wage laws did not have large employment effects.

Meta-studies, or "studies of studies," confirmed Card and Krueger's results. See this one, for example, and this one.

By contrast, in several studies, economists David Neumark and William Wascher have continued to find that the minimum wage is associated with a disemployment effect. But economist Arindrajit Dube has persuasively argued that their research is flawed because "it does not properly account for differences between high- and low-minimum-wage states. Essentially, they make the unrealistic assumption that low-wage employment trajectories are similar in states as diverse as Texas and Massachusetts." When the differences in these trajectories are controlled for -- as Dube and his colleagues do in this major research study -- employment effects are small.

Most other econometrically rigorous studies of the minimum wage on employment show a similar effect: employment effects are small or nonexistent. Why is this? Why don't the real-life effects of the minimum wage work out the way they do in the textbooks?

The textbook model, of course, is just that: a model. Now, there's nothing wrong with models. In fact, in economics, models are necessary, and good ones can be analytically powerful. A model is a stylized representation of reality. That "stylized" part is necessary. A model has to be a simplified version of reality to be useful, so it is unrealistic by default. Make it too unwieldy -- I'm reminded of that wild Borges story about the map that's the same size as the kingdom -- and it won't be workable or yield predictions easily. But a model that distorts reality too much by leaving out the wrong variables is useless. It will make bad predictions that tempt policymakers to apply the wrong policy prescriptions.

And that's just what happened with the standard labor market model as applied to the minimum wage. That model, which predicts that the introduction of a minimum wage would cause major job loss, turned out to be way off base. It doesn't fit the empirical data. So what, exactly, is going on here?

In the standard labor market model, known as the perfect competition model, the market as a whole -- that is, the supply of labor (all workers seeking a job) and the demand for labor (all jobs being offered by all firms) -- determines the wage. The market-clearing wage occurs at that point where labor supply equals labor demand.

In the perfect competition model, no single firm has the power to determine the wage; it simply accepts the wage that the market as a whole has determined, and that is what it offers to its workers. In this model, workers are extremely wage-sensitive, so much so that if any single firm cuts wages by even one cent, all the workers at that firm will immediately quit and find employment elsewhere.

Now that is a very strong assumption -- indeed, I would call it nothing less than a heroic assumption. A growing number of economists agree with me, and have been arguing that there is an alternative model of the labor market that is a much more plausible fit for the data and for common-sense observations about how the real world works. This model is known as the monopsony model. The literal meaning of monopsony is "one buyer" (just as the literal meaning of monopoly is "one seller").

But the monopsony model for labor markets does not necessarily imply "one employer." What the theory does assume is that the employer has what is known as "market power," and therefore is not a "wage-taker" (i.e., does not have to offer the market wage). In this model, it's hypothesized that it's the employer, not the market, which sets the wage. Therefore, in the monopsony case, the employer will offer below-market wages. Moreover, it's assumed that the source of the firm's market power is the force that binds an employee to an employer, so that if wages were cut, at least some of the employees would stay.

What are the forces -- or frictions -- that prevent workers from automatically quitting if wages were cut? Frictions include the search costs of getting a new job, incomplete information about what other jobs might be available, mobility costs (the costs of traveling to a new job), personal preferences (such as strongly preferring a certain type of work, or one's co-workers, or a benefit offered by the job), and the firm-specific training and skills that a worker may have.

If you ask me, the monopsony model clearly appears to be a better fit for the way the world works than perfect competition is. The assumptions of the perfect competition model -- complete information, zero mobility costs, employees at a firm being so wage-sensitive that they'd all quit even if their employer cut their wages by only one cent -- seem unreasonable. On the other hand, the assumptions of the monopsony model -- that the employer sets wages, and that there are important frictions in the labor market -- seem highly plausible, at least in the case of most employment situations.

Finally, here's what I especially like about the monopsony model, and I why I find it far more persuasive that the competitive model: it models power -- in this case, the power that employers have over employees. The Achilles heel of mainstream economics is that the concept of power is almost never incorporated into its theories and models. This is a huge intellectual shortcoming. In the standard labor market model, for example, the employer/employee relationship is seen as symmetrical, with neither being more powerful than the other. Each has the equivalent power to terminate the relationship if a better offer comes along. Which is just so wrong, on so many levels.

Monopsony is a useful concept, in part because it explains some interesting, and sometimes puzzling, features of labor markets. For example, in his book on monopsony, economist Alan Manning argues that monopsony helps explain features of the labor market that the competition model cannot explain, such as the gender gap in wages, evidence that employer characteristics are correlated with wages, and evidence that the presence of unions raises non-union wages.

One of the key features of the monopsony labor markets that they are inefficient -- that is, when you work out the details, you'll find that both wages and employment are always below ideal levels. This suggests that government intervention via a minimum wage could improve both efficiency and equity. By contrast, in the competitive model, the assumption is that the market always functions ideally and that government intervention would impair its efficiency.

So if we accept that the labor market acts more like a monopsony than it does like a perfectly competitive market, what impact does a minimum wage increase have in such a market? For one thing, employment levels may be increased if, as is in the case in the traditional monopsony model, the employer had been unable to fill vacancies because it was paying employees less than their marginal product. For another, it's likely that the rate of labor turnover will be reduced, resulting in cost savings for the employer. This appears to be the most important channel of adjustment, in the minimum wage scenario.

The reason I'm blogging about the minimum wage on this site should be obvious. The minimum wage is one of the most important policy tools we have for reducing economic inequality. These two economic studies find that the minimum wage reduces wage inequality, particularly for women. Here's what economist Arindrajit Dube had to say about the subject in his excellent recent New York Times piece
While there is some disagreement about exact magnitudes, the evidence suggests that around half of the increase in inequality in the bottom half of the wage distribution since 1979 was a result of falling real minimum wages. And unlike inequality that stems from factors like technological change, this growth in inequality was clearly avoidable. All we had to do to prevent it was index the minimum wage to the cost of living.
There are other reasons to support raising the minimum wage besides its inequality-reducing effects. A significant increase would reduce poverty, stimulate the economy, disproportionately benefit women and people of color, and put the U.S. minimum wage  in the same range as that of its international peers, instead of shamefully near the bottom of the pack.

At its most basic level, however, raising the minimum wage is the decent thing to do. The people who flip your burgers, ring up your groceries, scrub your toilets, and care for your children do work that is physically taxing and often emotionally demanding as well. Often, they lack benefits like health care and paid sick leave, as well as any real way to advance on the job to a position that might allow them to enjoy middle-class security. In contrast to many in high-paid professions, they do the kind of work that is vital to keep society functioning, and yet they receive precious little in the way of respect or gratitude in return. It seems to me that compensating them decently for a hard day's work is the very least we can do for them -- and for ourselves.

22 comments:

  1. "The Achilles heel of mainstream economics is that the concept of power is almost never incorporated into its theories and models. This is a huge intellectual shortcoming."

    Fantastic observation. (I can almost see M. Foucault's shadow.)

    Enjoying the blog - please keep it up.

    ReplyDelete
  2. Thanks for focusing on monopsony power when discussing the minimum wage! This is spot on, explaining why so many workers work for sub-living wages (power), and why employment would go up rather than down with a minimum wage hike.

    I believe it also relates to the issue you mention in your Political Animal post today about today's pressure to work extra long hours -- that's another expression of employer monopsony power, the ability to demand pretty extreme work hours.

    ReplyDelete