Bryan Caplan  

Orange Rocks and the Minimum Wage

PRINT
Henderson on Oge's CAFE... Nixon's Wage and Price Control...
I've long been deeply suspicious of contrarian research that purports to show that the minimum wage doesn't decrease low-skilled employment.  But Don Boudreaux explains my suspicion far better than I could:
Has any science ever devoted so much time, effort, and cleverness to elaborate attempts to determine whether or not a central and indisputably correct tenet of that science - a tenet used without question to predict outcomes in general - fails to work as an accurate predictor for one very specific, small slice of reality as has been devoted by economics over the past two decades to determine whether or not the law of demand works to accurately predict the effects of minimum wages on the quantity demanded of low-skilled labor?

I'm pretty sure that the answer to my question is 'no.'

I judge from the furious debate over the effects of minimum wages on the quantity demanded of low-skilled labor that were there to exist powerful political and ideological forces that stand to benefit if the general public believes that small orange rocks dropped into swimming pools cause no increases in the water levels of swimming pools, there would be no shortage of physicists who conduct and publish studies allegedly offering evidence that, indeed, the dropping of small orange rocks into swimming pools does not tend to raise the water levels of swimming pools (and, indeed, might even lower pools water levels!).

[...]

And so it is with minimum-wage legislation.  The strong political and ideological interests on the pro-minimum-wage side keep alive the debate over whether or not raising employers' costs of employing low-skilled workers causes employers to further economize on the amounts of low-skilled labor that they hire.  There is no furious empirical debate among scholars over whether or not, say, raising an excise tax on oranges would, ceteris paribus, cause fewer oranges to be bought and sold.  There is no furious empirical debate among scholars over whether or not, say, an increase in the tuition charged to attend college would, ceteris paribus, discourage some people from enrolling in college.  There is no furious empirical debate among scholars over whether or not, say, imposing a poll tax would, ceteris paribus, discourage some people from voting.

Yet because powerful political and ideological interests have a stake in the market for low-skilled workers being immune from the normal operation of the law of demand, a furious debate rages over whether or not employers forced to pay more for labor do or don't further economize on labor.

One picayune criticism of Don's analogy: Comparing the minimum wage to "small orange rocks" leads casual readers to the specious inference that we have trouble detecting the minimum wage's disemployment effect because it, too, is small.  Don's should have simply spoken of "orange rocks" - then explained that we might trouble detecting even sizable disemployment effects because of confounding factors.  Don is naturally well-aware of this point.  He even extends his orange rock analogy by identifying confounding "factors such as rainfall and evaporation, swimmers jumping into and out of pools, and the condition of each of the many pools' drainage and filtering systems."  But I wish he'd driven it home a little more aggressively.

Last point: If I were an immigration skeptic, I'd be sorely tempted to use Don's words to debunk the mainstream view that immigration has minimal effect on native wages:

Has any science ever devoted so much time, effort, and cleverness to elaborate attempts to determine whether or not a central and indisputably correct tenet of that science - a tenet used without question to predict outcomes in general - fails to work as an accurate predictor for one very specific, small slice of reality as has been devoted by economics over the past two decades to determine whether or not the law of supply works to accurately predict the effects of immigration on the wages of native laborers?
The key difference: The Law of Comparative Advantage specifically predicts an ambiguous net effect of immigration on native wages, because specialization and trade raises labor productivity and therefore labor demand.  You can't credibly say the same about the minimum wage. 




COMMENTS (22 to date)
pyroseed13 writes:

"Last point: If I were an immigration skeptic, I'd be sorely tempted to use Don's words to debunk the mainstream view that immigration has minimal effect on native wages."

Exactly, and this is what troubles me about the inconsistency between those who oppose the minimum wage but support more low-skilled immigration. Both the minimum wage and low-skilled immigration imposes significant costs on a small segment of the workforce, but do produce gains for others groups. Yet libertarians, correctly, argue that someone who is not hired or laid off because of a minimum wage hike is much worse off than the person who gets the wage boost and keeps their job. Yet when it comes to immigration, they seem to argue that the gains to everyone else far outweigh the losses to the high school dropout who was unable to find a job, which I don't think is obvious at all.

Similarly, David Card has done his very best to prove the the demand curve for labor doesn't slope downward. Libertarians criticize his minimum wage paper for doing this, but yet praise his work on immigration. Arguing the the "net effect" is ambiguous doesn't do much to reconcile the tension between these views, because most arguments made against immigration are concerned about the distributional effects, not necessarily the net effects.

Ricardo writes:

Come on, Bryan, this one is easy.

1. Start with a swimming pool.

2. Add orange rocks.

3. The water level rises momentarily.

4. However, this brings the water closer to the sun.

5. Some of the displaced water then evaporates, restoring the pool to its original level.

6. Now, remember that the rocks are orange, which is a reflective color.

7. So some of the sunlight that was previously absorbed by the dark floor of the pool is now reflected back into the water.

8. This additional reflected light causes a small rise in the water temperature.

9. The increased temperature causes even more evaporation from the surface.

10. Thus the water level is now actually lower than it was before the rocks were added.

QED.

Disclaimer: any resemblance between this description and models of our climate are purely unintentional.

Hazel Meade writes:

Perhaps we should encourage working class whites to emigrate to Mexico ;)

Mike W writes:

But, is there a net benefit or net loss to workers generally from the decrease in some workers' employment opportunities due to minimum wages and the improvement in some workers' standards of living?

Thaomas writes:

Exactly what is the elasticity of demand that the "law of demand" decrees? What if it is "small" like -0.1? And what if an increases in the EITC seems even more fanciful now than before Nov 9? Might not an increase in the minimum wag be a pretty good second best way of raising the wages of most low income workers?

The "orange rocks" metaphor seems to miss the point about the costs and raising the level of water in a pool and the benefits that might lead one to put orange rocks in it.

MikeP writes:

But, is there a net benefit or net loss to workers generally from the decrease in some workers' employment opportunities due to minimum wages and the improvement in some workers' standards of living?

If minimum wage supporters outright said, yes, indeed, lower skilled workers are disemployed by a minimum wage and are more disemployed by a higher minimum wage, and entry level positions are reduced so opportunities to start low and climb the wage ladder are also reduced, but those costs are worth it in order to make sure that some older and more experienced breadwinners have more wages to support their families, then we could have a conversation around this question.

Do you hear anyone taking a position that sounds like that?

pyroseed13 writes:

@Mike W

To your point, there's nothing about the standard view on the minimum wage that implies that the net costs outweigh the net benefits. Someone could just as easily say "Well, people get more money in their paychecks, and then go out and spend it, which creates more jobs, reducing the disemployment effects of the minimum wage." I think this a ridiculous argument, but it's pretty similar to the one made to deny any negative employment effect of immigration.

Jon Murphy writes:

@pyroseed13:

We want to be careful on straight-up comparisons on the effect of MW the effects of immigration. On top of the comments Bryan Caplan made at the end of the paper, the two are very different changes in the market, which will cause very different results. Minimum wage is a price floor, whereas immigration is a shift in the supply and/or demand curve(s). While there may be some superficial similarities, they are not the same. As such, one could easily, and consistently, dismiss something as "ridiculous" (to use your word) in relation to minimum wage and as reasonable in relation to immigration.

LD Bottorff writes:

The benefits of the minimum-wage flow to politicians who favor increasing it. They get more votes just because they are for it.
The costs are born by employers, investors, and low-skilled workers.
In the case of immigration, the benefits to restricting immigration also flow to the politicians. The costs are born by employers, investors, and immigrants.

Paul Zrimsek writes:

If the nefarious tobacco barons were smart, they would have let the whole cancer issue go and instead sponsored research "proving" that cigarette taxes have no detectable effect on demand for cigarettes, and may even increase it.

Anonymous writes:

@Thaomas

Isn't it relatively well established that the impact on happiness of unemployment is really quite big, much bigger than the impact of even moderate sized wage differences?

Aaron writes:

Minor point, but poll tax is not a tax on voting. It is a tax assessed equally on all people. The reason it is associated with voter discrimination in the US is that some states enacted poll taxes and refused to let non-payers vote. The intention was to disenfranchise poor people, and to increase the effect of paid voters. It also excluded many black voters in the South, but it was primarily economic discrimination and vote fraud, only indirectly racial discrimination.

[url around nick fixed. Please check your links before posting!--Econlib Ed.]

pyroseed13 writes:

@Jon Murphy

I do not agree that the similarities between the minimum wage and immigration are "superficial." These are both policies that disproportionally hurt low skilled workers. Even those accept the Borjas findings on the effects of immigration on high school dropouts argue that we should focus on the benefits that accrue to everyone else...which is an argument that some proponents of the minimum wage also make. I say only "some," because as another commenter here pointed out, many minimum wage proponents deny the existence of a disemployment effect altogether.

ybell writes:

There is one thing about this post (Boudreaux's) that bothers me and that is the sense of anti-evidence. It is one thing to say -- listen, we have strong theoretical priors that minimum wage causes unemployment, so we'd need to see strong evidence to persuade us otherwise. But it is entirely another to say we can a-priori discredit any conflicting evidence just because there is a lot of political motivation to create supporting research.

jon writes:
In the case of immigration, the benefits to restricting immigration also flow to the politicians.
If politicians benefit from immigration restriction, how come so few take that position? Trump obviously made a big deal out of it, but who else in this election favored restricting immigration?
Jon Murphy writes:

@jon:

If politicians benefit from immigration restriction, how come so few take that position? Trump obviously made a big deal out of it, but who else in this election favored restricting immigration?

Hillary Clinton, Bernie Sanders, most of the other GOP candidates. Immigration restrictionism has been a hallmark of US policy since first called for by the Progressives in the early 1900's (and heck, there's been all kinds of restrictionist political parties and legislation since the formation of the nation: the Know-Nothings come readily to mind).

Jon Murphy writes:

@pyroseed13:

My point is that, unlike with minimum wage, you cannot assert based off theory alone that immigration "disproportionally hurt[s] low skilled workers." That is not supported by the theory, unlike minimum wage. That we can say disproportionately hurts low skilled workers.

To make the claim that immigration hurts low-skilled workers, you'd need substantial evidence; a prima facie claim is not sufficient.

robc writes:

@pyroseed13

Both opposition to the minimum wage and being pro immigration are liberty maximizing positions.

I dont think any libertarians (at least this one) should be concerned about anything else.

The fact there there may be utilitarian benefit is just a bonus.

pyroseed13 writes:

@Jon Murphy
I agree, but I did not really argue anything of the sort. The evidence shows that the minimum wage and immigration both disproportionately hurt low skilled workers. All theories must be met with evidence, although all of the usual caveats apply (thing are not always ceteris paribus, etc.)

@robc
I don't think many libertarians would be opposed to the minimum wage or support immigration if they did not believe there were "utilitarian" reasons for doing so. Indeed, Don Boudreaux and Bryan Caplan spend much of their time talking about these policies in terms of benefits and costs.

@ybell
I agree with this, and the reality is that the theory alone cannot really tell you whether any of these policies are necessarily good or bad, just what the likely trade-off is. The magnitudes of the effects matter and can only be assessed empirically. That being said, because the minimum wage is such a strongly politicized issue, I sense there is a tendency for some left-leaning economists to give more preferential treatment to studies which find few negative effects, because that's the outcome they would prefer to be true. Although I suppose you say the same about libertarian economists as well.

robc writes:

@pyroseed13

Most libertarians are deontological libertarians, not utilitarian libertarians, as far as I have been able to count, so you are probably wrong.

But even those of us who care more about the moral side often discuss things in utilitarian terms too.

MikeP writes:

If politicians benefit from immigration restriction, how come so few take that position? Trump obviously made a big deal out of it, but who else in this election favored restricting immigration?

Since precisely zero elected politicians support immigration restricted only for specific individuals with explicit cause individually applied, I have to say every Democrat or Republican in this election favored restricting immigration.

Politicians of all stripes and all positions reserve the right to demagogue against actual or prospective immigrants as the always present nonvoting group that they can impress voters with their action toward.

A writes:

When I read the headline I anticipated a pushback against recent minimum wage studies. But there is no rewarding discussion about spatial controls, endogenous/exogenous framing, and other attribution problems. In fact, this post, and linked article, appear to reject empiricism in favor of elevating certain priors to unexaminable status.

Andrajit Dube et. al. are motivated by blah blah blah. Who cares? They might be the most politically motivated people on earth, but why are they wrong?

Comments for this entry have been closed
Return to top