Tyler Cowen has an insightful post on Marginal Revolution, that provides a lovely insight into the power of economic thinking.
Often, multiple policy questions come down to a single issue. We may not know the answers to any of the questions, but we can at least say that the single issue drives the answer to all of them. So once you decide one issue goes one way, you can't (rationally) believe another issue goes another way, no matter how politically convenient that might be.
The issue here is the "elasticity of labor demand." If wages go down 10%, how many more workers will employers hire? If wages go up 10%, how many fewer will they hire? Already, note, that's one number, and it makes little sense to believe a higher number in one direction than another except for some likely quite transitory adjustment costs.
Tyler
There is a longer history of minimum wage assumptions not really being consistent with other economic views.
Have you ever heard someone argue for wage subsidies and minimum wage hikes? No go! The demand for labor is either elastic or it is not.
Have you ever heard someone argue for minimum wage hikes and inelastic labor demand, yet claim that immigrants do not lower wages? Well, the latter claim about immigration implies elastic labor demand.
Have you ever heard someone argue that “sticky wages” reduce employment in hard times but government-imposed sticky minimum wages do not? Uh-oh. [The link is good too - JC]
It would seem we can now add to that list. Maybe we will see a new view come along:
“Labor demand is elastic when licensing restrictions are imposed, but labor demand is inelastic when minimum wages are imposed.”
This is a bit of economic-ese, so let me translate just a bit. The argument for a minimum wage is that labor demand is inelastic -- employers will hire the same number of workers. They will just absorb the higher wages or pass along the costs to customers. Workers get all the benefit. If labor demand is elastic, employers cut back on the number of employees. Most people lose their jobs and only a lucky (or productive, or willing to tolerate harsher working conditions) few get the higher wage.
The argument for wage subsidies requires the opposite assumption. If we subsidize wages, do employers respond by just paying people less? (Or, of we pay the subsidy to the worker, are the same number now willing to work for lower wages from employers? An often forgotten core result of economics is, it doesn't matter who pays the tax or gets the subsidy.) That is the case if labor demand is inelastic. Or do employers respond by paying the same amount, and expanding the workforce? That is the result if labor demand is elastic. You need elastic labor demand for wage subsidies to work as intended.
Thus, you can't simultaneously be for higher minimum wages and for wage subsidies. That is cognitive dissonance. Or, inconsistency. Or wishful thinking. And very common.
Likewise, just about 99% of macroeconomics is centered on the proposition that wages are "sticky." (1%, consisting of basically me, still has doubts.) If deflation breaks out, wages do not fall. Wages are too high. Employers cut back on workers, and people are unemployed. For this argument to work, (among other things) labor demand must be elastic. And then inducing exactly the same situation by minimum wages must have exactly the same result -- fewer jobs.
The point on immigration is good. The labor demand curve unites a price and a quantity. Thus, if pushing on a price has little effect on quantity, we know that pushing on a quantity has a large effect on price. Or, you have a hard time believing one thing about pushing on a price, and another thing about pushing on a quantity.
If you don't like minimum wage hikes, you can't believe that immigrants are pushing down wages. If you like higher minimum wages, you must believe that immigrants are terrible for labor markets. This proposition should be equally uncomfortable on the left and the right.
(Supply demand graphs are great here. I leave them as an exercise for the reader. If you know how to read them, this was all obvious already.)
Yes, one can complicate the analysis in each case to get a desired result -- add adjustment costs, asymmetries, or whatever. But it is striking that most discussions don't do that -- they don't reconcile that what is good for the goose ought to be good for the gander, and defend why not in a specific case.
Tyler:
Third addendum: Of course there are numerous other ways this analysis could run. What is striking to me is that people don’t seem to undertake it at all.It is interesting in policy debates how people debating each issue seem not to connect with other issues. Clear thinking is hard. "Wait, this comes down to the elasticity of labor demand, and we made the opposite assumption last Tuesday discussing immigration" is not the sort of thought that occurs naturally. It is also a sign that policy-oriented economics is all too often searching for justification of pre-determined answers.
Update: Casey Mulligan complains that Tyler and I both left out the many margins of adjustment that happen in response to minimum wages, such as making hours less convenient, reducing benefits, making work less pleasant, and so forth. Casey's right, but Casey needs to relax! This isn't an extensive minimum wage post. I've blogged about those effects many times before. They're real and important. And the other side may want to complicate things as well. Before we come up for excuses why (say) minimum wages and immigration are different, let us at least acknowledge the simple model in which they are the same, and that to the extent labor demand elasticity bears in the analysis of each, one should use a consistent assumption on that ingredient.
Explained well. Thinking like an economist challenges illogic. A behavioral psychologist would explain cognitive dissonance as thought being inconsistent with behavior or decisions. To wit. A person who smokes (behavior) knows (cognition) smoking can cause cancer. Clearly,logical clarity would preclude smoking.
ReplyDeleteSmoking yields great pleasure. The cost is an elevated risk of cancer. Different folks yield different amounts of pleasure from smoking and have different attitudes toward risk.
DeleteOf course Frank. But if the cost of an elevated risk of cancer or death were immediate, I suspect the rational thing would be not smoking. A personal example. I was a market-maker on the CBOE and the AMEX. In effect I owned a casino. If an at the money option was at 3, my market was 3 bid, 3 1/8 ask. The return is a about 4.16%. The public, as price takers, sold to my bid and bought at my offer. A friend invited me to Las Vegas to gamble. I declined. Just the idea of paying a dollar for a bet that on average is worth 92 cents caused anxiety.
DeleteSmoking and casino: The correct analogy would be that you got pleasure out of gambling! :-)
Delete> Employers cut back on workers, and people are unemployed. For this argument to work, (among other things) labor demand must be inelastic
ReplyDeleteIf the number of workers is lower after deflation, doesn't it mean the labor demand is elastic (as the company is making do with fewer employees)? What am I missing?
I don't quite follow the part about immigration and wages. If labor demand is elastic, then higher minimum wages can potentially lead to lost jobs. If more people are added to the system (through immigration) these additional people also find jobs, but everyone now has lower wages. Why are these two statements incompatible (assuming full employment in the economy)?
ReplyDeleteYou describe a scenario (assuming elastic labor demand) in which a minimum wage and an increase in immigration occur simultaneously.
DeleteThe minimum wage increase will lower employment as you describe. However, it's incorrect to conclude that with increased immigration, "everyone now has lower wages." If lab demand is elastic, employers will simply accommodate new workers without any change in wages.
As John says, the easiest way to grasp this is to think of supply and demand curves. Let the supply curve of labor have a slope of 1 and the demand curve be a horizontal line (perfectly elastic). Increased immigration means a rightward shift of supply curve. No matter how far right (or left) you move the supply curve, it's going to intersect the demand curve at the same price level. Thus, increasing the supply of workers has no effect on wages.
Thanks for the reply, Joel. Makes sense when the demand curve is horizontal.
DeleteBut if the demand curve has a slope of -1 (the standard X looking supply/demand curve), then as we move right price drops (i.e. more immigrants lower wages), but I'm confused on how to interpret a higher price. There's higher supply but lower demand - wouldn't this eventually lead to lower wages and a slide back to the intersection point?
Yes, if the demand curve slopes downward, then both of the following are true:
Delete1) minimum wage increase --> employment decrease
2) immigration increase --> lower wages for competing workers
It's easy to overstate the point, and I think John does so slightly. Because demand curves aren't perfectly elastic or inelastic, this issue is one of degree, not absolutes.
If the demand curve is more flat (elastic), then
1) minimum wage increase --> large employment decrease
2) immigration increase --> small wage decrease for competing workers
while a steep (inelastic) demand curve implies
3) minimum wage increase --> small employment decrease
4) immigration increase --> large wage decrease for competing workers
The point is that lots of people mistakenly think that 1) and 4) are true, whereas 1) implies 4) is false and vice versa. Same for 2) and 3).
As immigrants don't exist in a bubble and increase demand for good and services, wouldn't an increase in immigrants push the labor demand curve to the right regardless of whether the labor demand curve is elastic or inelastic?
ReplyDeleteYes, of course. But the question is what effect that shift will have on the price level. If labor demand is perfectly elastic (horizontal), then the price will remain the same no matter the location of the supply curve. If labor demand is perfectly inelastic (vertical), then a small rightward shift of supply means a much lower price. John's (and Tyler's) point is that some folks talk as if a rightward supply shift has no effect price in some situations and has a significant effect on price in other situations, all while failing to offer an explanation for why the elasticity of labor demand would be different in these different circumstances.
Delete
ReplyDeleteIf you don't like minimum wage hikes, you can't believe that immigrants are pushing down wages. --JC
Am I missing something? I have reservations about higher minimum wages or even any minimum wages, in theory. But I believe that large-scale immigration does push down wages.
I am kind of with Benjamin here.
Delete"Supply demand graphs are great here. I leave them as an exercise for the reader".
Please John, could you include the answers to these exercises?. That would be great after trying so hard.
I agree also. I think John was a little imprecise here. More precise (but less exciting) is "if you believe minimum wages lead primarily to unemployment, you can't believe immigrants have a large downward effect on wages."
DeleteIf supply and demand both have normal elasticities, then the following are both true: 1) minimum wage increases cause deadweight loss, and 2) immigration causes lower wages. It's only when one of those curves is perfectly elastic or inelastic that a statement like "a shift in a supply/demand curve has no effect on price/quantity" is true.
It took civilization a couple of thousand years to figure out the supply and demand diagram, so I reckon that no, a lot of economics is not obvious.
ReplyDeleteBased on my experience as an econ professor, most people do not understand supply and demand even if it has been taught to them. So no, not obvious, at least to the vast majority of people.
DeleteI don't get something:
ReplyDeleteIf job demand and supply are elastic (+1 slope and -1 slope?), then higher minimum wage would cause unemployment (if it is above the crossing point), and immigration would cause higher labor supply at the same price, leading to lower wages.
So I can dislike minimum wage hikes on the ground that they cause unemployment, and immigration on the ground it causes lower wages.
What am I missing here?
Some of these arguments depend on the relative elasticities of supply and demand. Alternately, what if the elasticities are around 1? Then it's not as clear what is the optimal policy?
ReplyDeleteAlso, on immigration, George Borjas argues that the supply of labor from illegal immigrants is inelastic. That doesn't mean that everyone's labor supply is inelastic.
It's true that "policy-oriented" economists are typically unwilling to concede ANY trade-offs associated with their positions once established, which includes refusing to admit potential inconsistencies between various stances. But I'd like to believe those same economists at least undertook a good faith effort -- involving research and deep thinking -- to figure out what their positions should be in the first instance (though perhaps this is just naive on my part).
ReplyDeleteOf course it is disheartening to see so economists take the "admit nothing, deny everything" approach to defending their favored policies. Once upon a time economists EDUCATED the public, whereas now most seem to focus on PERSUADING the public. But hey, that's life in 2019... to be fair and balanced is to be irrelevant. This sad fact applies just as much to policy-oriented economists as it does to everyone else.
I think the problem here, for some commenters, is that immigration increases BOTH the supply of and the demand for (mostly unskilled) labor. The immigrants are consumers as well as producers.
ReplyDeleteIf that weren't true, each June, when a new group of high school and college grads enter the work force, we'd see an increase in unemployment. Especially, as back in the day when the Baby Boomers grew to adulthood, they swamped the numbers of retirees (their grandparents!).
Yes. If labor demand is perfectly inelastic and the same for every firm, but the number of firms is proportional to the population (to use a simple example), couldn't one see both labor supply and demand curves shift to the right, leading to no change in wages?
DeleteGotta distinguish between a shift of a curve and a movement along a curve! :-)
Delete"If you don't like minimum wage hikes, you can't believe that immigrants are pushing down wages." You need to expound on this a bit more.
ReplyDeleteI don't like minimum wage hikes, because I believe they diminish less-skilled employment. OTOH, I also believe that immigrants lower the equilibrium wage in the areas where they compete. Perhaps you could explain better why my two beliefs are logically inconsistent.
Also, the quote is "sauce for the goose..."
A good former student of mine (Rafael Guntin, now doing his PhD in NYU) remembered this column of mine in a Spanish Language newspaper (Feb 2017), with the title "minimum wages and immigration" which makes the same point.
ReplyDeleteI always find this cognitive dissonance baffling.
https://www.elobservador.com.uy/nota/salario-minimo-e-inmigracion-2017226500
A completly different kind of cognitive disonance would be expecting a tax on carbon emisions reducing carbon emision and not expecting a tax on income reducing income.
ReplyDeleteMany people seem to be victims of this cognitive disonance
I found this essay to be remarkably different in tone from Cochrane's other essays. More circumspect, more even handed. It left me feeling hopeful instead of annoyed.
ReplyDeleteCochrane concludes, "It is also a sign that policy-oriented economics is all too often searching for justification of pre-determined answers."
Yes, but ironically, this was often my takeaway from my encounters with Cochrane's previous writings.
How do you arrange with monopsony power? Couldn't you have, without cognitive dissonance, an increase in employment and an increase in minimum wage?
ReplyDeleteCorrect! This is
DeleteSaturday the classic - correct - argument for a minimam wage. It's no longer relevant on account we all have cars, but it'seems logically correct.
I was also thinking on a rather similar form of cognitive dissonance in regards to inequality and diversity. The people who hate inequality want to tax people because they are richer than others to force everybody into the same income level, ideally. At the same time these same people defend diversity in terms of lifestyles saying that people should be free to live as they wish even if they are different from the average (such as gays, trans, differences in customs between ethnic groups, etc).
ReplyDeleteThat is, there is a fundamental inconsistency between the desire to compress differences in economic outcomes by force and the desire to allow for full expression of differences in lifestyle. Specially because very different lifestyles do lead to radically different economic outcomes (compare hippies with workaholics in finance).
For those interested in an economist's perspective on immigration- why an increased supply might be compatible with abolishing the minimum wage- I recommend Bryan Caplan's imminent release of 'Open Borders'.
ReplyDeleteI also highly recommend writing down your policy positions, and logically working through each one of them via the supply and demand of labor, wages, prices, etc.
Already bought it! ... can´t wait
DeleteI wrote a blog post in response, which I'll turn into some notes for my students in my regulation class too. Here's how it starts:
ReplyDeleteA very inspiring post, but it goes to show how useful the graphs are, even for someone as expert as Professor Cochran. What he points out is that you can’t argue both (a) the minimum wage doesn’t destroy jobs, and (b) immigration doesn’t reduce jobs for Americans. Let’s take the supply curve to be upward-sloping, so more people want to work if wages are higher (this is for simplicity, not essential to the argument). Suppose that labor demand by employers is completely inelastic, completely price-insensitive, a vertical demand curve (see Figure 1). Then if we impose a minimum wage, employment is constant (L=100, wage rises from 15 to 20 in the diagram). Suppose, instead, that we start from the old wage and add immigrants. The labor supply curve shifts to the right, with more people willing to work at the old wage than employers want to hire. Employers can reduce their wages and still find enough workers, so in the end, no more people are employed than before, but the wage has dropped and fewer Americans are employed.
http://rasmusen.dreamhosters.com/b/2019/10/elasticities-the-minimum-wage-and-immigration/
Both supply and demand curves are pushed to the right by immigration.
DeleteThe demand curve is not pushed to the right by a minimum wage hike.
True, but the increase in the demand for labor would be indirect, via immigrants increasing the demand for US products, and would be smaller, since not all their wages would go to purely labor-made (no capital or land) products.
DeleteWhat about the labor supply curve? There are reasons it slopes upwards. But the reasons are not the same as it is for goods. Yes, S/D graphs are useful here because we can graph multiple shifts in S/D and see what happens. Wage subsidies are certainly something to graph to see what happens to the Labor Demand Curve. The model may not be perfect, but that's ok. Insert wage floors. Watch what happens.
ReplyDeleteBest,
M
John Cochrane is great, but in this case his (and Tyler's) thinking is too simplistic. A higher minimum wage would have no effect on employment only if demand for labor is perfectly inelastic. If the curve is inelastic but not vertical, employment would fall. Similarly, "if you don't like minimum wage hikes, you can't believe that immigrants are pushing down wages" is only true if labor demand is perfectly elastic. If it's elastic but not 0, a minimum wage would hurt employment, and immigration would drive wages down. In other words, under most reasonable conditions the "inconsistent" positions described above aren't inconsistent in the least. What is dubious is to assume only two possibilities, that elasticities must be at one extreme or the other.
ReplyDeleteIndeed, the Mariel Boatlift "natural experiment" and the various minimum wage "natural experiments" suggest contradictory labor demand elasticity.
ReplyDeleteHowever, isn't there a major difference between the two? While mass immigration increases GDP, raising the minimum wage doesn't.
For sake of argument, let's assume that labor demand is inelastic. An immigration-induced gdp increase will push the demand "vertical line" to the right, and mitigate the negative effect of immigration on wages.
Unskilled immigration in the presence of an enforced minimum wage leads to higher measured unemployment among residents and no change in output.
ReplyDeleteIf unenforced among illegal immigrants, the notional minimum wage leads to a decline in the real wage of the immigrant unskilled and an increase in output.
I have two words.
ReplyDeletemonopsony power
OK,ok I add a comment.
ReplyDeleteWhat if the opposite is true? What if people think naturally and the economists suffers dissonance? If the economist original got it wrong then the agent has limited options, the rules and semantics already established. He has little choice in government adaption but a wage price control, though he knows it a bad choice relative to his other priors. What he needs is something the economist has not yet discovered, like the doctor who used to apply bleeding technique.
So much discussion about what seems confusing about "labor supply" and "demand for labor" when the technique to clear it up is by disaggregating the terms. There are very specific demands for very specific labor, or for people who might be trained for it. The supply is also very specific, so for the general question we have a spectrum of elasticities in "the labor market"(s).
ReplyDeleteOnly people still infected with Marxian fallacies think "workers" are a reified category instead of a fuzzy generalization. Immigration and minimum wages hurt (and help) different people.