Minimum Logic, Part 2

Yesterday’s post touched on several related points, and I’m afraid the most important one got buried near the end, so I want to repeat it:

1) In the presence of an effective minimum wage, all benefits of the earned income tax credit are transferred to employers. This is, as they say, a matter of Economics 101. (Edited to add: As Bennett Haselton points out in comments, I should have said “dissipated”, not “transferred to employers”. The point remains that the benefits don’t go to the workers, which, for this discussion, is what matters.)

2) Paul Krugman argues that we should have an effective minimum wage in order to prevent some of the benefits of the earned income tax credit from being transferred to employers.

In this context, it should be remembered that Krugman ordinarily reserves his deepest scorn for those who, according to Krugman, willfully ignore the lessons of Economics 101.

Let’s review the argument for 1), with reference to the graph below. In the presence of, say, a $5-an-hour minimum wage, employers will hire 1000 workers. Because more than 1000 people want to work, employers can extract extra concessions in the form of reduced on-the-job-training, shorter breaks, and harsher working conditions. They can get away with exactly $1-an-hour’s worth of this, because even at an effective wage of $4, there are still 1000 people willing to work.

Edited to add: I am assuming that these concessions are of relatively little value to employers (otherwise they wouldn’t have waited for the EITC to demand them!), so that the quantity of labor demanded does not change.

Now let’s add a $3-an-hour earned income tax credit, which shifts the labor supply curve to the dashed position. Ordinarily, this would lead to a lower equilibrium wage, transferring some of the benefits of the EITC to employers. But in the presence of the $5 minimum, wages can’t drop, and employment remains fixed at 1000, though now even more people want to work, allowing employers to impose even harsher conditions until the effective wage drops to $1 an hour (the wage at which there are still 1000 people willing to work). This process transfers all the benefits of the EITC away from the workers.

Even if employers are unable to extract all of these benefits (say because working conditions are already so bad they can’t get worse), workers will find some way to compete for the limited number of jobs until the effective wage (that is, the wage minus the cost of that competition) is driven down to $1. This, together with the $3-an-hour EITC, puts the workers right back where they started.

Yesterday I made the point that once you’ve got a binding minimum wage, the level of that wage is irrelevant to the distribution of EITC benefits, so that Krugman’s argument for raising the minimum wage falls flat. What I wish I’d emphasized more strongly is that in any event, the effect of that minimum wage is exactly the opposite of what Krugman claims. Without a minimum wage, workers keep some positive share of their EITC benefits. With it, they keep none.

Share/Save

19 Responses to “Minimum Logic, Part 2”


  1. 1 1 acarraro

    I don’t really completely follow the logic.

    If employers can circumvent the minimum wage by cutting benefits and so on, won’t the minimum wage simply have no effect? Why would employer only ask for 1000 workers? Surely they don’t really care about the official wage but only the true wage. The wage composition will simply change and there would be no constraint.

    I don’t understand why employers wouldn’t increase employment if there is a way for workers to always decrease the effective wage…

    Krugman logic is based on the assumption that employees will not or cannot accept worst conditions after a bump in EITC. I think it’s a possible assumption, which needs to be verified, but it’s as valid an yours in absence of data to the contrary…

  2. 2 2 Harold

    “workers will find some way to compete for the limited number of jobs until the effective wage (that is, the wage minus the cost of that competition) is driven down to $1″

    Nobody.really comented yesterday that employers are limited in just this way. I am thinking that you mean the workers will compete by working harder to get or keep the jobs. However, given the transaction costs of turnover for the employer and job seeking for the worker, would these costs soon overcome the additional benefit to each of excessoive competition? That is, there may be no way for the workers to compete until all the benefits are transfered to the employer.

  3. 3 3 J-man

    Yeah, I think the demand curve, which represents the employers willingness to pay for labour, should incorporate all the costs associated with employing staff. To put it another way, the total price an employer pays for staff is more than the price staff receive for working. This is a dead weight loss caused by taxes, insurance and regulation.

    If we assume equilibrium in these models, we must have rationality. If we have rationality then employers must have already reduced excess non-wage costs, or profit maximization.

    However, if you assume rationality, that merely removes the first movement of the price to $4. I would suggest that the employer is unable to gain any of the eitc, unless he was to auction off employment positions. Let’s assume this would be considered illegal. Then the increased competition would see some proportion of employers successfully auctioning their staff positions to the highest bidders, thus receiving some of the value of the eitc. 1000 people would still be employed and the balance of the value of the eitc would become a dead weight loss as a result of trying to avoid detection in the employment black market.

  4. 4 4 John Hall

    Well, workers for whom the minimum wage is binding, no?

  5. 5 5 Harry Chernoff

    Steve:

    If the idea is to take the advice of Nobel Prize winners, we should at least pick the appropriate ones. In the case of the minimum wage and the EITC, it’s not Paul Krugman, it’s Ned Phelps. I believe it’s accurate to say that his recommendation would be a wage subsidy rather than the minimum wage or the EITC, though he’d certainly take the EITC over the minimum wage. Why don’t you forward your comments to him and see what he says?

  6. 6 6 Rob Rawlings

    I’m not sure I agree with Steve’s analysis:

    If employers and workers can always negotiate a real wages package of money wages +or- benefits then even with a minimum wage the labor market will always be in equilibrium. (Imagine Steve’s chart but with “real wage” rather than wages.

    Then if an EITC is introduced the supply of labor increases and we move to a new equilibrium.

    Assuming a normally shaped supply of labor curve the EITC benefits will always be shared between employer and worker.

    Its only if the supply of labor curve is vertical that we get Steve’s results.

    Did I miss something?

  7. 7 7 Dirk

    Can somebody explain to me when we shift the curves because of policy (like the dashed line created when we talk about the EITC increase)?

    Why don’t we get a similar dashed line shifting the demand curve left because of the minimum wage?

  8. 8 8 Otis

    @Dirk The simple answer is that price is being measured on the vertical axis. Income is a non-price determinant and hence the shift when it changes.

  9. 9 9 Bennett Haselton

    I think there is a logical error in the first graph, along with this argument:

    “In the presence of, say, a $5-an-hour minimum wage, employers will hire 1000 workers. Because more than 1000 people want to work, employers can extract extra concessions in the form of reduced on-the-job-training, shorter breaks, and harsher working conditions. They can get away with exactly $1-an-hour’s worth of this, because even at an effective wage of $4, there are still 1000 people willing to work.”

    But consider *why* an employer would extract these extra concessions — they’re not doing it to be mean, they’re doing it to save themselves money. Reduced training, shorter breaks, and harsher working conditions all mean that it effectively costs the employer less per hour to employ the workers.

    That means the wage that they’re paying effectively drops from $5 to something less than $5, which moves us rightward on the demand curve and means that they are employing more than 1000 workers.

  10. 10 10 Alan Gunn

    Complicating facts here are that not all low-income workers get the full EITC (you have to have children for that), and most of the people earning the minimum wage aren’t poor (teenagers, etc.) and so get no EITC. How these things affect the analysis is well beyond my ability to figure out, but I’m pretty sure it makes the EITC a vastly better anti-poverty program than the minimum wage. I think there was fairly widespread agreement, across political lines, about this, maybe 30 years ago (when the New York Times still ran editorials criticizing the minimum wage). Today, it’s more like a religious-committment thing.

  11. 11 11 andy

    “Paul Krugman argues that we should have an effective minimum wage in order to prevent some of the benefits of the earned income tax credit from being transferred to employers. ”

    It often seems to me that some people care more about the rich not getting some transfers than about helping the poor.

    It seems to me that combining EITC with minimum wage just creates a situation with even more involuntary unemployment where it might as well happen that people that need the job most fall on the ‘unemployed’ side – because people that didn’t need the job that much were priced out of the market.

  12. 12 12 Steve Landsburg

    Bennett: You’re absolutely right; I was sloppy about this. My assumption was that this stuff is costly to workers but of quite small value to employers, so that the effect you’re talking about is small. Therefore the benefits are mostly dissipated, not transferred to employers. I should have said this.

  13. 13 13 Rob Rawlings

    @Steve re 12.

    I see you concede Bennett’s point that a reduction in worker benefits “moves us rightward on the demand curve and means that they are employing more than 1000 workers’.

    Given this , an increase in EITC will cause the supply of labor to shift , and will then cause further adjustment in the wage/benefits package which will again cause movements along the demand curve.

    The result of these shifts in the supply curve and along the demand curve surely mean it is not true to say that “Without a minimum wage, workers keep some positive share of their EITC benefits. With it, they keep none.”.

  14. 14 14 Steve Landsburg

    Rob Rawlings: With the minimum wage, workers must engage in some sort of non-price competition for the limited number of available jobs. If that competition takes place in a way that benefits employers and therefore increases the number they will hire, then you are right and I was wrong. If that competition takes place in ways that do not significantly benefit employers, then it does in fact dissipate all the benefits, so that workers keep none.

  15. 15 15 Rob Rawlings

    Steve: Thanks for replying. Yes, I agree with that.

  16. 16 16 Bennett Haselton

    I saw your edits, but I can’t make sense of the following justification:

    “I am assuming that these concessions are of relatively little value to employers (otherwise they wouldn’t have waited for the EITC to demand them!)”

    The scenario you’re describing takes place in your narrative before the EITC is added, so these employers didn’t “wait for the EITC” to demand these concessions.

    I thought at first perhaps you meant to write “minimum wage” instead:

    “I am assuming that these concessions are of relatively little value to employers (otherwise they wouldn’t have waited for the minimum wage to demand them!)”

    But that doesn’t make sense because in the absence of the minimum wage, employers would just save money by lowering wages instead. If a new minimum wage is then suddenly enacted, that’s when employers would have to start saving money by the more inefficient means of extracting concessions, so they *do* in fact “wait for the minimum wage to demand them.”

  17. 17 17 nivedita

    This seems to be one of the sources of the argument that the policies are complementary: Lee and Saez

    Although one of the points of the paper seems to be that it can be good to reduce the minimum wage (while still keeping one) at the same time as increasing the EITC.

    My understanding is that there is lots of empirical literature showing that the low-wage labor market is not an “Econ 101″ market. So I’m not sure how relevant your argument in this post is. With the same assumptions, a minimum wage can NEVER change workers’ effective wage, but don’t we know that this is empirically false?

  18. 18 18 nivedita

    In fact, your dissipation argument, if true, shows that the minimum wage is destructive even to the workers who have a job — there’d be both more employment and the workers would have a higher effective wage in the absence of the minimum wage.

  19. 19 19 Bob Murphy

    Hey Steve,

    I hope you’re still checking this one. Does your argument rest on the assumption that after a minimum wage is imposed, a worker is indifferent to whether he is employed or unemployed?

    I can understand the case for such a view, but I don’t think that’s Econ 101. I’m quite certain Krugman doesn’t believe it, for example.

    Or am I totally misunderstanding how your argument works?

  1. 1 Why Is There So Much Economic Illiteracy? Paul Krugman Is One Reason
Comments are currently closed.