TEEN UNEMPLOYMENT AND THE MINIMUM WAGE
Recently the Wall Street Journal editorial page and numerous conservative/libertarian bloggers have shown charts comparing the last few years spike in the teen age ( 16-19 year olds) unemployment rate with the rise in the minimum wage in an attempt to blame rising teen unemployment in 2008 and 2009 to the minimum wage hike. But more detailed data from the BLS shows that teen employment at the minimum wage rose 46.1% in 2008 and 50.1% in 2009. This detailed data implies that the rise in the teen unemployment rate from around 15% in 2007 to over 25% in late 2009 was entirely due to the great recession, not the rising minimum wage. This data is so compelling that I think at a minimum these conservative/libertarian at least owe us an explanation as to why it does not disprove their argument.
On 9 February–Minimum wage disinformation –I published a chart that showed since 1960 there have been 18 increases in the minimum wage. Nine of those increases were accompanied by a falling teen unemployment rate and nine were accompanied by a rising teen unemployment rate. The difference was that the nine times teen unemployment rose were also associated with a recession while the other nine occurred in an economic expansion. This chart and regression analysis clearly implied that the bulk of the increases in teen unemployment since 1960 was due to recessions, not rising minimum wage.
Now the BLS has published more detailed data on teen employment at and above the minimum wage in recent years. This data contains several surprises that tend to strongly contradict the standard theory of teen employment and the minimum wage taught in most economic textbooks.
First it shows that during the years 2002 to 2007 when the nominal minimum wage was unchanged, teen employment at the minimum wage was less than 10% of teen employment. Over 90% of teens were paid more than the minimum wage.
Second, over this period of a flat nominal minimum wage the real minimum wage was falling and each drop in the real minimum wage was associated with a drop in the number of teens employed at the minimum wage. Teens clearly though that school work, public service — so important in college applications — sports or leisure activity like video games such as 먹튀검증 was a more valuable use of their time than working at a minimum wage job. Consequently, firms had to offer a higher wage than the minimum wage to induce teens to work for them. This is why the share of teens paid more than the minimum wage rose from 2002 to 2007. This should not be any great surprise to an economist. It is standard economics 101 that if you want more of something you offer a higher price. I have never really understood why most economics instructors spend much of an economics course teaching that if you want more of something — raise prices. Than they turn around and teach that you will get more teen minimum wage employment if you offer a lower price. YEAH RIGHT — and they wonder why the students have trouble getting it.
Third, the data shows that from 2007 to 2009 while the minimum wage rose from $5.15 to $7.25 — a 50% increase — teen employment at the minimum wage rose from 373,000 to 818,000 — a 119 % increase. Moreover, teen employment at more than the minimum wage fell from 5,061,000 to 3,579,000, almost a 30% drop. The drop in teen employment at more than the minimum wage accounted for 130% of the 1,037,000 fall in teen employment from 2007 to 2009. Again, this should not be a big surprise to economist. A large increase in the real minimum wage generated a large increase in teen employment at the minimum wage. It is standard economics 101. This data also implies that essentially all of the drop in teen employment from 2007 to 2009 was due to the Great Recession, not higher minimum wages. I suspect that if we had this detailed data for other periods of a rising minimum wage it would show a similar pattern.
In my blogging about the minimum wage I have never taken the position that raising the minimum wage does not lead to weaker employment. All I have done is point out that the available data completely contradicts the theory that the minimum wage causes weaker employment. Moreover, the more data that becomes avilable, the more it condraticts standard textbook theory.
I have noticed bloggers such as Mark Perry at Carpe Diem, Don Boudreaux at Cafe Hayek, Alex Tabarrok at Marginal Revolution, Megan McArdle at the Atlantic, Ironman at Political Calculations, or Greg Mankiw have advanced the thesis that the recent rise in teen unemployment is due to higher minimum wage, but they need to explain why a doubling of teen minimum wage employment does not contradict their analysis.
This data can be found at
Characteristics of minimum wage workers
Because the BLS reports each years data in a single file, if you want to look at data for multiple years it is easier to go to the top left of the BLS home page and search for “characteristics of minimum wage”.
The search will show a Google search of each of the different years files and moving back and forth from Google to BLS is the best way to switch from the file for one year to another. Also, since the data for 2009 has just been reported it has not had time to collect many hits, so it will be on page 2 or 3 or the Google search.
I think their view is if we eliminate the minimum wage and unemployment then unemployment will cease to exist; it will have been defined out of existence. Anyone that wouldn’t work would do so by choice, macroeconomy be damned.
This displays the (rhetorical) predilections of conservative and libertarian economists and bloggers: they constantly cherry-pick particular “textbook” (read: grade-school) economic theories even when their simplistic constructions are contradicted by the real-world evidence.
Viz, Mankiw:
“Policymakers who want to promote growth would not go far wrong ignoring most of the vast literature reporting growth regressions. Basic theory, shrewd observation, and common sense are surely more reliable guides for policy.”
http://www.asymptosis.com/mankiw-post-friedman-ergo-propter-friedman.html
If a teen is already earning above a new minimum wage then he or she would be earning more that both the old or new minimum wage so the change would have no impact on their cost to their employer.
The econometrics is a little tricky here since the only teens you expect to be impacted would be those at the old minimum wage (or below the new proposed higher minimum wage). For these employees either they are not worth anything more to an employer or more likely the employer is earning some rent on them by keeping their pay less than what its actually worth to the employer. Other things that can confound the analysis is the overall strength of the labor market. If its very strong then the market might take the wage increase in stride and maybe pass it on to customers. However, If the economy is weak businesses may not be able to pass along costs and might decide to keep a smaller staff and it would be hard to untangle what percent was do to the drop off in business and what was due to the higher wages.
Economics 101 tells us that putting a wage floor (minimum wage) creates an excess supply of labor if the market wage without the wage floor is below the floor. This means that employment arrangement that would have been done without the floor just don’t get done and the person remains unemployed.
I don’t see way to easily use the BLS data since it makes a difference if the teen is either at minimum wage or above it.
Cantab– if your analysis of a wage floor is correct why did teen minimum wage employment fall from 2002 to 2007 when the real minimum wage fell. In real terms the floor was falling so teen minimum wage employment should have risen. But it fell — just the opposite of what your analysis implies.
Cantab– you are right that part of the growth in minimum wage employment is due to wage compression at the bottom end of the scale. By wage compression I mean when the wage is increased from $5 to $6 everyone that was making between $5 to $6 now making $6. So someone who had been making $5.50 and counted as being employed at a wage above the minimum wage is now employed at the minimum wage.
But in this case when minimum wage employment doubled and more than minimum wage employment fell more than total teen employment fell I think it is safe to conclude that wage compression only accounted for a very small portion of the change. Especially since minimum wage employment was well under 10% of total teen employment before the wage hike. That was a big surprise to me. I was under the impression that a much larger share of teen employment was at the minimum wage. I know it is standard practice in many firms that hire teens to give them a raise after a few months.
But my bottom line, and what I am really protesting is those who try to blame the entire rise in teen unemployment to the minimum wage and none of it to the Great Recession.
spencer,
There are several different things going on here. First, the employment effect of the minimum wage in general is a slightly different issue than the effect of the minimum wage on overall employment. Clearly teens have a more flexible (i.e., elastic) reservation wage than do adults. And there’s every reason to believe that a teen’s reservation wage is strongly dependent upon the strength of the overall economy. A weak economy not only depresses teen employment cyclically, but there is also some evidence that it affects the number of teens entering the labor force. The opportunity cost of a college education is a lot higher during boom times, so we would expect more teens to delay college and enter the workforce.
I don’t think anyone disputes that the cyclical nature of the economy strongly affects teen employment. That’s always struck me as a good reason to look at teen unemployment rates as a kind of leading indicator of general employment trends.
As Cantab said, the econometrics here can be tricky. Since we’re dealing with a censored dependent variable, so you would have to use something like a Tobit or Heckit model.
The Iron Law of economics still applies, no matter how the followers of Holy Saint Mises try to pretend it doesn’t. People do not voluntarily starve to death. They just don’t. If wages are not sufficient to prevent starvation, they do something else, like sell crack outside of elementary schools or something. In short, yeah, if everybody gave their labor away for 1 cent per hour there would be no unemployment, but the Iron Law says it’s never going to happen — once you get below the number needed for basic subsistence, people are going to find some other way to survive, generally a way that is a) illegal, and b) violent, because people simply do not voluntarily starve to death just because the followers of Holy Saint Mises believe they should. Just ask Marie Antoinette about that one. Oh wait…
Regarding teens and unemployment, most businesses follow a “last in, first out” policy. I.e., they fire their most recent hires first when the economy goes into a downturn. By definition teens are going to be recent hires, and thus the most likely to be fired when the economy goes into recession. If you look at the recession of 2001-2002 teen unemployment soared there too, but there was no minimum wage hike. I bet if you look at the percentage decline in GDP vs. percentage rise in teen unemployment, you’ll find a similar ratio now that you found then. Anybody up to digging up those numbers?
The real reason for the minimum wage is to raise productivity which drives economic growth. We wouldn’t have had an industrial revolution if wages in England hadn’t been so high. Why innovate when you can throw cheap labor at a problem? The U.S. always had a high minimum wage since land was cheap. You had to offer wages higher than basic farming and then some. Low wage countries are poor, and the low wages are as much or more the cause than the effect.
That’s because those countries do not have social welfare provision systems such as seeting a guaranteed minimum income.
Alex
Custom Logo Design
On the wages issue during England’s industrial revolution–the evidence indicates that wages were highest in the regions where industrialization was most heavily concentrated.
Hmmm!
Bad Tux — Since 1950 the teen unemployment rate has averaged 3.8 times the total unemployment rate. This cycle the teen unemployment rate has been almost exactly what this average said it should be.
“I don’t think anyone disputes that the cyclical nature of the economy strongly affects teen employment.” — slugger
“I have noticed bloggers such as Mark Perry at Carpe Diem, Don Boudreaux at Cafe Hayek, Alex Tabarrok at Marginal Revolution, Megan McArdle at the Atlantic, Ironman at Political Calculations, or Greg Mankiw have advanced the thesis that the recent rise in teen unemployment is due to higher minimum wage…” — spencer
spencer identifies people by name who make the argument that slugger says nobody makes. So we have two testable, competing hypotheses. (My personal impression is that slugger’s use of “nobody” makes him wrong. Never. Always. Nobody. Everybody. Those terms make statements wrong. Always.)
Now, the reality of slippery rhetoric is that one tries to say what one wants people to believe while allowing one’s self a way out if challenged by a spencer. The point that I believe spencer is making is that the data suggest – more all the time – that the impact of minimum wage hikes are quite small. Those chatterers who focus strongly on the supposed job-killing influence of a minimum wage don’t need to come right out and say the floor is a big deal or a small one, more important that cyclical factors or less important. We can reasonably count up their statements about each and see if they are in reasonable proportion.
kharris,
I don’t closely follow all of the economists that spencer cited, and I surely don’t follow Megan McArdle because she is not an economist, but I do follow Mankiw and I think you’re distorting the argument. First, all economists believe in two curves…a demand curve and a supply curve. When economists talk about the effect of the minimum wage, they generally have in mind the effect on the labor supply curve and assume no change in the labor demand curve. A recession clearly affects the labor demand curve, and I’m pretty sure that Mankiw is aware that wages are set by both curves, so given those facts I think it’s pretty safe to conclude that Mankiw also thinks recessions play a big role in teen unemployment. The net effect on employment depends on both curves. In a recession the labor demand curve shifts to the left, but it may also flatten out so that a small increase in wages has a larger effect on labor hours demanded. The minimum wage shifts the labor supply curve to the left, but in this case the new equilibrium point for labor hours demanded is determined by the wage rather than the other way around. So in a sense the slope of the labor supply curve is fairly irrelevant in the neighborhood of the minimum wage. But the slope of the labor demand curve is VERY relelvant. And I’m pretty sure Mankiw knows this, as I’m sure you do as well. As long as we have elastic labor demand curves during bad times and inelastic labor demand curves during good times, then it’s fair to say that the business cycle will be the predominant factor in minimum wage employment.
Now here’s what Mankiw actually says about teen employment (this is from his textbook):
The minimum wage has its greatest impact on the market for teenage labor… Many economists have studied how minimum-wage laws affect the teenage labor market. These researchers compare the changes in the minimum wage over time with the changes in teenage employment. Although there is some debate about how much the minimum wage affects employment, the typical study finds that a 10 percent increase in the minimum wage depresses teenage employment between 1 and 3 percent. In interpreting this estimate, note that a 10 percent increase in the minimum wage does not raise the average wage of teenagers by 10 percent. A change in the law does not directly affect those teenagers who are already paid well above the minimum, and enforcement of minimum-wage laws is not perfect. Thus, the estimated drop in employment of 1 to 3 percent is significant.
Notice that he does not say that the minimum wage is the main determinant of teen unemployment, he says that the minimum wage has its greatest impact on teen employment. Those two statements sound similar, but they have very different meanings. In other words, Mankiw is not saying that the minimum wage has THE greatest impact on teen employment, he is saying that the minimum wage has ITS greatest impact on teen employment. Understanding the difference between those two statements is critical to understanding what Mankiw is actually saying.
2slugs:
As Laurent Guerby and I have pointed out repeatedly, there has been a surplus of labor since Oct 2001. Minimum wage has little to do with such an issue regardless of how many ways one chooses to argue it.
I am agreeing with you overall; but, I am doing so for different reasons.
You’re data proves nothing of the sort.
All it proves is that before the minimum wages were increased in recent years a great many teens were already earning above what were then the minimum wage.
The increases in the minimum wage did greatly increase the numbers at the minimum wage as the government price floor was only catching up with market prices.
Yes the recession has caused a temporary increase in teen unemployment as the demand curve has shifted dramatically. However those temporary increases are now becoming more permanent becasue the new clearing price is below the now raised price floor.
run75441,
Just to be clear, I support the minimum wage. And I supported increasing it. My argument with kharris is simply that I think Mankiw gives more credence to cyclical effects than kharris thinks he does. Let’s take Mankiw at his word when he says in his textbook that 10% increase in the minimum wage reduces employment by 1% to 3%. And let’s take the high end of the estimate. Since the minimum wage increase by 50%, the we would have expected teen employment to drop by 15%, or 373,000 x 10% = 55,950. Call it 56K. Given what actually happened to minimum wage employment and teen employment overall, that number is background noise. The cyclical effect clearly dominated the employment picture for teens. And if I can do arithmetic, then I’m pretty sure Mankiw can do it as well. So from that I conclude that Mankiw would not disagree that the recession was the dominant factor in teen unemployment.
I agree with Andrew; the minimum wage didn’t affect unemployment because the market wage was already above it.
As a thought experiment, imagine if the minimum wage was $100 / hour for college students. I suspect that would have an effect on unemployment.
As for the data, correlation does not imply causation. It’s nice to think about, but when the data contradicts the concept of supply & demand for labor, it’s probably for another reason, not because supply & demand are wrong.
I noticed that all official data regarding unemployment and especially teen unemployment are not exact, due to the fact that many of them are hired without documents.
good morning Cynthia:
Could you be more specific on “official data.”
Higher minimum wage = more equal distribution of wealth = more consumers with enough money to buy (2 people with $500,000 will spend more than 1 person with a million) =
more jobs = increase in demand for jobs = More increase in wages
Where I live, Orange County, CA, 60% of the homeless population (counting those who move between motels, crashing at a friends house, and sleeping on the streets) is working over 40 hours a week! They don’t earn enough to save a deposit for an apartment. A tiny one bedroom here costs $1250 a month (if you sign a 9 month lease, otherwise it’s $100more); that does not include utilities (they charge for sewer and trash service!) and your tiny garage space ($60 per month extra). Our county have the nation’s 2nd highest percapita population of people without a permanent address (homeless+motel living): 35,000.
Below me, in their tiny 1 bedroom are 4 Mexican adults and their 3 children, all crowded together with only 320sqft.
Even if a minimum wage did cause unemployment, there MUST be some solution to the horrible poverty in this country. You can’t just say that a minimum wage would just make things even worse. I don’t buy the argument that “a family living on minimum wage is better than a family that’s homeless”
Zack:
It is called “Job Creation” something this country has been lacking for a decade now. When manufacturing started to be moved overseas, it removed the one area a man or woman with a high school education could work and with a good work ethic win a ticket to the middle class. Now we have people with coleege educations not able to get jobs. The most recent MBA class of University of Toronto Rotman (the alleged best in Canada) has a job placement of 34% and has a history of 90%. It is bad on this continent.
PARTIN N HAVIN FUN WIT MY FRIENDS.
Teen unemployement and minimum wages depends on the economic conditions of the country. When the economic conditions of the country improves automatically unemployment decreases. But it is very important for the government to decrease the unemployment in the country because due to unemployment the crime rate increases and and when teens are not able to afford or purchase some expensive things or any thing thay like, they comit crime. The good way is to participate in penny auctions if you dont have too much monay and you want to get an expensive thing you like
I’d like to see an inflation-adjusted minimum wage chart. Nominal data is meaningless over such a long time-frame.
Garret Peterson
Here are a few charts and an article that should show it: http://www.newyorker.com/news/john-cassidy/the-minimum-wage-debate-in-charts