Monday, February 18, 2013

EITC and Minimum Wage

I wrote a post in January arguing for an increase in the Earned Income Tax Credit, an income transfer which goes largely to poor families, as an alternative to an increase in the federal minimum wage. Some  left-leaning economic commentators responded that I had imposed a false choice between the EITC and the minimum wage. The two, they said, were complements which should be increased together.

I was a month early to the debate.

President Obama, in his State of the Union address, called for an increase in the minimum wage. And the case that the EITC and minimum wage complement each other has become louder. It's appeared, most recently, in pieces by Mike Konczal of the Roosevelt Institute and Paul Krugman in the New York Times. Krugman writes:
[I]t’s important to understand how the minimum wage interacts with other policies aimed at helping lower-paid workers, in particular the earned-income tax credit, which helps low-income families who help themselves...But it has a well-known defect: Some of its benefits end up flowing not to workers but to employers, in the form of lower wages. And guess what? An increase in the minimum wage helps correct this defect.
The EITC probably does depress pre-transfer wages. The argument isn't new. I can track it back to Bluestone and Ghilarducci (1992), at least. But before leaping to the argument that a minimum wage would shift more of the subsidy to workers, it's worth asking why it does so.

The labor supply must be somewhat elastic to income if employers capture some share of the gains from the EITC. (If labor supply were perfectly inelastic, then the EITC gains would go entirely to workers.) This fact, however, implies something else -- namely, that there should be some effect on unemployment from the minimum wage.

To the extent that the minimum wage does not generate unemployment, in other words, the EITC will not generate an increase in the "surplus" which goes to employers. Both effects come from elasticity.

Second, I think we should be wary before associating an economic effect (that employers should receive some of the transfer) with a normative statement (that employers don't deserve it). I think there are a pair of underlying assumptions here: (1) that transfers to employers are regressive, and (2) that the minimum wage is the right policy to reduce such transfers.

On the former, it will always be the case that some "employer surplus" goes to small business owners who are EITC-eligible themselves, or receive some other form of subsidy from government. So I don't find it a particularly convincing argument that we ought to write off that portion of the surplus as entirely wasted.

And, on the latter, I certainly do not find it convincing that the best policy solution, if we want to target the EITC's gains as narrowly as possible to the poor, is the minimum wage. Why can't we just increase the progressivity of the corporate income tax? We could eliminate the employer gains which go to wealthier employers that way, too. (In a formal parliamentary-debate setting, these benefits are "non-unique" and do not count to either side.) And, while we're talking about targeting the gains: The average income of an EITC beneficiary family is substantially lower than that of a family with at least one minimum-wage earner.

Here's another thing to think about. The EITC doesn't benefit all workers equally -- in particular, it tends to go to families or singles with children. The transfers are much, much smaller for the childless. The argument that the two policies are complementary, Neumark and Wascher find, may not be as strong when you consider that EITC recipients are a subset of the poor. EITC non-recipients may get hurt by both.

In general, I worry about overstating the strength of consensus of the research for either side in the conversation about the interaction of the minimum wage and EITC. And I remain open, as always, to persuasive argument. My evolving view at this time is that (1) increasing the minimum wage provides no unique benefits as a complement to the EITC which could not be matched by other tax policy, and (2) that the minimum wage may be uniquely disadvantaged by the vice of being a price control rather than a tax/subsidy.


  1. This is a very good post and right where the conversation about the minimum wage needs to be.

    There is an ocean to discuss in this vein, but I will just add the following small bit: the minimum wage might be passed along to consumers in the form of higher prices. There is nothing stopping an employer from responding to higher labor costs by raising prices. To the extent that labor demand is inflexible (i.e. to the extent that the minimum wage does not create unemployment) the employers' incentive to raise prices is stronger. Likewise, to the extent that employers cannot pass along the cost to consumers, they have an incentive to cut hours etc.

    So an an increase to a $9 minimum wage is only nominal, and the "real" increase may be something different, say $8.60 or $8.90. I'd like to see more discussion of the real vs. nominal effects.

    1. True. I think the econometrics would get really complicated in trying to discern the inflation effect from a minimum wage hike. Obviously, it wouldn't fully cancel out the nominal increase -- the real minimum wage has fluctuated -- but explaining why becomes an interesting question for micro theory.

  2. To the extent that the minimum wage does not generate unemployment, in other words, the EITC will not generate an increase in the "surplus" which goes to employers. Both effects come from elasticity.

    Remember, these are 2 different is the worker's income elasticity of labor supply, and the other is the firm's price elasticity of labor demand...

    Remember that a higher EITC does not increase the firm's costs in a way that is proportional to wages, while a minimum wage does. That will lead to two very different incentive effects.

    The argument about EITC lowering wages is that workers who get higher income from EITC are willing to work for a lower wage. Theoretically, this elasticity could be very high, while firms' price elasticity of labor demand could be very low (i.e. firms don't care much about wage changes for low-wage workers, but low-wage workers care a lot about their total income).


  3. Some of its benefits end up flowing not to workers but to employers, in the form of lower wages. And guess what? An increase in the minimum wage helps correct this defect.
    Not sure you should not have been more critical of Krugman in the above quote. Proponents of the minimum wage, including Krugman, generally argue that it does not reduce employment too much because employers are able to pass the cost on to consumer as all employers are subject to the same cost “bump”. Why then does he label employers as beneficiaries of EITC? Surely, consumers also capture most of the “benefit” of EITC that flows through the lower wage channel as well.

  4. I could see this evidence as supporting the current $7.25 federal minimum wage, but I fail to see how it justifies an increase all the way to $9.00. A figure which is even higher than the already absurd Californian minimum wage. Krugman is so quick to dismiss the negative impact of high real wages caused by price controls on employment, but it just as quick to support the idea of high real wages causing mass unemployment when caused by a nominal shock. How New Keynesian economists can vilify deflationary policies that cause higher real wages and prices, but support policies that raise wages and prices directly, causing the exact same macroeconomic situation just with less flexibility baked into the system, really boggles the mind.