Doruk Cengiz, Arindrajit Dube, Attila Lindner, and Ben Zipperer write,
Our method infers the disemployment effect of the minimum wage by tracking the changes in the number of jobs throughout the wage distribution following a minimum wage increase. The changes at the bottom of the wage distribution—in particular the missing jobs below the minimum, and the excess jobs at or just above the minimum—reflect the effect of the minimum wage on low-wage workers.
The idea is that if the minimum wage increase has a big effect, then you should see a noticeable drop in the rate of growth of jobs that are just below the minimum wage. They do not find such a drop. They conclude that an increase in the minimum wage has minimal adverse employment effects.
Could you provide a better explanation of this paper? On its surface, the statement “The idea is that if the minimum wage increase has a big effect, then you should see a noticeable drop in the rate of growth of jobs that are just below the minimum wage” makes no sense. By definition, there are zero jobs below the minimum wage, right?
I think it’s saying that if a job had previously earned X$ per hour, and the minimum wage was increased above X$ per hour, then you might expect the number of employment opportunities in that job to decline, or at least grow more slowly. It seems to be saying that the growth rate of such jobs is unchanged, indicating that the increase in the minimum wage doesn’t affect employment much*.
*Within the range of increases studied. If you increased the minimum wage to a billion dollars an hour, things would probably be different.
Which Workers Are Exempt From the Federal Minimum Wage?
Even if your business is covered, federal law does not require you to pay the following workers the federal minimum wage:
independent contractors (only employees are entitled to the minimum wage)
outside salespeople (a salesperson who works a route, for example)
workers on small farms
switchboard operators employed by phone companies with no more than 750 stations
employees of seasonal amusement or recreational businesses
employees of local newspapers having a circulation of less than 4,000
newspaper deliverers, and
apprentices, students, and learners, as defined by federal law.
Also special rules on employees who receive tips.
The journal article is Greek to me but you characterization makes perfect sense. If that is a precise statement of their hypothesis then they’ve started out with a deeply flawed premise by measuring the loss of growth of jobs that paid just below the new minimum wage before the new law. They need to be looking at the loss of job growth in jobs that had paid the old minimum wage or slightly above it. That seems so obvious I’m astounded they didn’t do that.
I’m not an economist, but I’ll be impertinent enough to ask this: Assuming the correctness of their premise that the evidence shows that there was no post-min-wage-hike drop in the growth rate of jobs that formerly paid below the new minimum wage, before you could conclude that the hike had no effect, wouldn’t you have to investigate in what sectors and locations the new jobs were appearing, and in what sectors and locations jobs were disappearing (if any), and test the hypothesis that, overall, the negative effect of the hike was overridden by other factors?
“We show that the minimum wage is likely to have a negative effect on employment in tradable sector, and manufacturing in particular—with an employment elasticity with respect to own-wage of around -1.4—although the estimates are imprecise. At the same time, the effect of the minimum wage is close to zero in the non-tradable, restaurant, retail and other sectors—which together comprise the vast majority of minimum wage workers in the U.S. This evidence suggests that the industry composition of the local economy is likely to play an important role in determining the disemployment effect of the minimum wage (Harasztosi and Lindner 2016).”
To add to the impertinence…”before you could conclude that the hike had no effect…” add
calibrate the “strength” of the claim to the real magnitude of the increase being proposed — as I suspect that not all the historical increases challenged the demand curve in exactly the same way.
These studies annoy me… the notion that minor changes to minimum wage produce minor changes to employment rates is compatible with the notion that the existence of minimum wage has had a major effect the number of low skill jobs, as firms have had decades to adapt. I think that objection is much more serious – how many low-skill stepping stone jobs never appear the first place? How could this question even be approached?
Thank you!
My understanding is that this issue is discussed under the label “The Truncation Problem.” Bastiat may have written on it 100 years ago.
“Greek letter economics” is mostly beyond me. My guess is that many of the largest effects of minimum wage policies are felt by unskilled youths, and that many of those effects are “dynamic” rather than static. Many youths don’t get their first job as early, while meanwhile they are kept in school longer, and the opportunity cost of school is less because honest work in the formal sector is harder to find.
This involves dynamic effects and sociological impacts, and it can’t easily be measured in a “comparative statics” approach.
Walter Williams makes the argument better than I can, and he has more street cred and formal training in economics. In the end it’s an empirical question, and much of the question is “how big? How much? How important?”
He makes his argument repeatedly. One example is here:
http://jewishworldreview.com/cols/williams011316.php3
The minimum wage can be set where it has no effect. The law may or may not matter, it depends on where the wage quants were set in the first place. There is such a thing as the defacto minimum wage, the quants at that level are quite coarse.
It sounds like they’re counting only the number of jobs — what about hours? Haven’t other studies showed one of the strongest effects is on hours worked?
As someone who has employed a lot of people, my response to a higher minimum wage would be not to contract employees, but to use the higher demand for the job (due to a higher required compensation) to increase the number in my temporary pool.
I would then use the expanded pool to selectively offer the best hours and shifts to the best employees. Thus the employer gains from the expanded pool and superior talent chosen in that shift, partially offsetting the higher wage. The better employees gain as well, getting MORE and better hours (the shift desirability is itself a benefit). The losers are the below average workers who get fewer hours, crappy shifts and get demoted down the shift totem pole.
The net result is probably slightly fewer hours worked, with possibly higher net employment. The fact that most economists don’t take this feedback into consideration just points to the poverty of empirical understanding within economics. Don’t confuse us with messy details that screw up our pretty math models!
I’d also expect restaurants to become more likely to send workers home when business is slow with a higher minimum, which would also reduce average hours worked per week.
There are a number of “hipster food” places around me that are pretty expensive (not fine dining, but too damn much for a fancy sandwich). They appear to employ white college students (or recent grads) and pay well above minimum wage (sometimes due to tips). At $15 (maybe it would need to be more in NYC) they seem to get relatively attractive and attentive workers rather then the kind of minimum wage dregs you observe in fast food. The flip side is I don’t see anyone on a normal wage affording their food as anything other then a treat.
My own mental model of what higher minimum wages would mean would be fewer workers at a higher productivity level and the service sector moving more towards higher end goods over time. I suppose the lower end of the market would move towards more home production (eat out less), which wouldn’t show up in GDP but work would be getting done.
I’m no economist, but I always heard that the demand curve sloped downward. I guess not.
It’d be nice if more studies tried to actually identify who absorbed the cost of minimum wages increases. In the case of employers who sell inelastic goods and services, the increase may simply be passed on to consumer. Since many such goods are necessities (foot, etc., the things sold by retail stores) wouldn’t a price hike in inelastic goods basically amount to a regressive tax?
Also, there’s the issue of effect size. If the increase raises the average wage for people in a given sector by a small enough amount, aggregate statistics likely won’t observe either a negative or positive effect. Many of minimum wage studies really just prove the trivial claim that if a policy change is small enough in magnitude, it doesn’t have any observable effect, which isn’t useful information.
All McDonalds are now getting touch screen order kiosks. There seem to be two reasons:
1. Fewer people need to be employed as order takers.
2. People tend to order more from the touch screen, and some people like using the kiosk better.
If you raise the minimum wage from $5 to $10 not only do those making $5 see their wage go $10 but everyone making over $5 but less than$10 also see their wage go to $10. I presume these are the employees this paper is discussing.
As a first approximation, after the wage hike the number of people making over $5 but less than $10 should go to zero, right? You should get a large increase in the number of pe0ple making $10 — more than the number of people that were previously making $5.
If this is the case the entire premise of the paper does not make any sense.
What am I missing ?
Presumably the people making over $5 but less than $10 are now making the new $10 minimum wage. This has to be the dominate reason that the US data always shows an increase in the population making the minimum wage increase after it is raised.
But I never see either the advocates or opponents of the minimum wage discussing this.
That second sentence should be:This has to be the dominate reason that the US data always shows an increase in the population making the minimum wage after it is raised.
The final (( increase )) is edited out.