Some people argue that education is the answer to some of the big current problems the U.S. economy faces. Want to fix the unemployment problem? That's easy: just provide additional educational opportunities for those having difficulty finding jobs. Want to lessen income inequality? That's easy too: if more people have college degrees, they'll qualify for higher wage work. While these arguments appear to make sense, looking at the data over the past several decades provides the opposite answer: more education would solve neither problem.
Lawrence Mishel of the Economics Policy Institute makes this argument in a new paper. While he agrees that a better-educated workforce would ultimately help U.S. growth, he shows pretty convincingly that these two current economic problems can't be solved with more education.
In fact, it's pretty easy to see that education isn't the answer to solving the unemployment problem if you make one key assumption: the unemployment problem is cyclical, not structural. This means that those unemployed today will eventually be able to find jobs in the respective industries, relying on their current skill sets. If the problem was structural, then specific dead industries segments would be responsible for the high unemployment problem, and those workers would need to develop other skills to find new work.
In his paper, Mishel argues at length that the unemployment problem is cyclical. His argument is relatively straightforward, and pretty compelling. In fact, in some ways it resembles a post here a few weeks earlier, which argued that the unemployment problem is only a little bit structural. It's pretty clear that there simply aren't enough job openings at this time to accommodate all of the unemployed Americans. Once the economy improves enough to spur substantial job creation, work should return in almost all industries.
The one notable exception is construction. Mishel disagrees, however, saying that construction is no different than any other industry. For reasons explained in the post on structural unemployment linked above, however, construction does look different. The incredibly large amount of investment in real estate during the boom years will not return. Eventually construction will be needed, but not to the extent it was during the bubble.
But one structural change does not make a structural unemployment problem. Construction workers laid off since the beginning of the recession only make up about 13% of all unemployed Americans. While that's not insignificant, if every other industry recovered but construction, the unemployment rate would drop to about 6%, which is quite close to the natural rate. And as mentioned, some construction jobs will inevitably return once the housing market supply is worked through and businesses begin expanding again. The industry is, by nature, pretty cyclical.
Since the unemployment problem is mostly cyclical, education wouldn't help. The U.S. is not in a situation where jobs are available, but millions of workers have the wrong skills or not enough education to perform the duties those jobs require. Instead, it's just a matter of economic activity increasing to where firms feel the need to hire back already-qualified workers.
There was some hand-waving in the last argument, because it was fairly straightforward. When it comes to income inequality, we need more statistics. Luckily, Mishel provides plenty. Let's start with a chart showing how income inequality has changed by wage group over a quarter of century:
This point is related to another one economist Tyler Cowen makes in a paper on income inequality published recently. Much of the wage inequality the U.S. has been experiencing has occurred because a small handful of people are able to become very, very rich thanks to modern communications, marketing, and technology. But for everyone else, incomes have not changed much. Education has little to do with this, as a college degree, or even an advanced degree, does not guarantee anyone who obtains one a seat in the top 10% of earners. Obviously a far larger portion or Americans than that obtain college degrees.
Mishel provides another compelling argument tied to productivity-pay gap. Although productivity has grown significantly over the past 15 or so years, median hourly compensation has not -- for either group, college educated or high school educated:
... the pay of college graduates is as disconnected from productivity growth as is the pay of high school graduates. Vastly expanding college enrollment and completion will do nothing to address this problem.
Mishel further shows that the premium in wages that college provides has not been rising as quickly as income inequality if one compares how it has changed each decade since the 1980s. To do so, he compares how the ratio between the 95th and 50th percentile of earners has changed and compares it to the growth in the college premium. So his statistics even exclude the extreme top 5% of earners, where income inequality has changed the most. (See p19-20 of his paper for more on this.)
So you can argue that income inequality is a problem and that education is a good thing. But the two do not appear to be correlated. Most income inequality today is due to a handful of earners at the top earning a lot more than everybody else, and that's not because they are the only ones who are well-educated. Wages have been relatively stagnant for both groups of people who are not in the very highest segments of earners, those who have obtained higher levels of education and those who have not.