Recently, a number of politicians including Bernie Sanders, Kirsten Gillibrand, and Cory Booker expressed vocal support for a “federal jobs guarantee.” This initiative would operate through a bunch of state firms building things like the Works Progress Association (WPA), and they have a good case for it. American infrastructure is slowly crumbling, and no one is rebuilding it. Perhaps Cory Booker’s WPA could. No one could be stuck or unable to receive even a low-wage job, because the government would provide help. Second, when unemployment is low, low wages rise, reducing inequality.

There are indeed good reasons for a federal jobs guarantee. Still, we can do better. As a country, we can buy a lot of parks and bridges, and do so often during recessions, when government borrowing is cheaper and unemployment is dire. There are three ways that a federal jobs guarantee would differ from buying infrastructure through auctioning off the rights to build public works to private firms.

First, a WPA would produce less effectively than private construction could produce with the same amount of money. State-owned firms have lower total factor productivity than private firms.

Second, buying construction work during recessions stabilizes demand for construction, an industry especially deep in the thralls of business cycle volatility. Here’s the thing: if you steady the demand for construction, you can make all construction more efficient. Let me tell you a story. Something has happened in construction over the last 50 years. Given an hour of labor, construction firms are able to produce less value than they could in, say, the late 1970s. In fact, they can only produce half as much value.

Now, all three libertarians at Wes, when reading this, will quickly blame burdensome regulation, and they’re right! Regulation accounts for a whopping… one-eighth. By best estimates, one-eighth of the decline in productivity is due to new regulations. So what accounts for the rest? Well, perhaps the most plausible is that construction has become much less capital intensive since the ’70s. Essentially, construction firms used to have a lot more equipment per-worker than they do now. Why? Well, 1939 until 1975 was a period of very few recessions, rapid population growth, and quickly growing output in America. This meant that there was constant demand for construction, which made it efficient for construction companies to invest heavily in equipment since they knew there would be demand for its use. Since then, growth (both in wealth and population) has slowed, and construction has been much more sporadic. This means that it is no longer as useful to invest in equipment because there is a good chance that it will rust during a long recession and/or lull in demand for new infrastructure.

However, this is exactly the kind of issue that governments can efficiently fix. When private demand is low, supplement it with public demand. This is called an “Automatic Stabilizer.” The idea is that during recessions, when private demand is low and interest rates are low, the government can both ameliorate the recession and create public goods cheaply (by borrowing cheaply) through deficit spending.

This is most efficient when the government funds industries like construction. Imagine if the government, facing low private demand, supplemented it with public demand, and decided that they would buy, say, fine champagne.

Surely, this would help France’s economy, although it wouldn’t do much here. In contrast, construction is largely US labor, and the inputs (wood, cement, customized manufacturing products) are expensive to ship or require skilled labor, which is largely domestically sourced. The government’s best way to aggregate demand during recessions is to spend their resources on construction, or perhaps education and public art because the money will go to Americans instead of abroad.

Lastly, it’s much better for the federal government to guarantee certain spending during recessions than to simply spend during them, because the guarantee allows firms to invest, knowing that their capital will not go to waste.

So, by having the federal government guarantee a certain amount of public spending when unemployment rises above a certain percentage, we can both ameliorate recessions and incentivize private construction firms to invest, making construction cheaper across the board.

Through a WPA system, you don’t get private firms to invest. Plus, since public firms are always less effective than private firms at doing the same job, you get less infrastructure. This WPA would be particularly inefficient, because you couldn’t fire anyone, or the idea is kind of ruined.

But, granted, I did not consider the third way that a Works Progress Administration would differ from auctioning off more construction projects. There are some persons who are extremely unlikely to be able to find employment, even in good times. This gap is labeled by economists as “structurally unemployed.” This group is largely made of people with disabilities, some non-English speakers, single parents busy/tired from raising children, and ex-convicts. Even if the economy were at approximately full-employment all the time, these people would still be unlikely to find jobs, while in Cory Booker’s plan, they would be guaranteed them.

But is this what we want? Do we want them to be stuck in make-work jobs paying $24 thousand a year? Or is that a band-aid withering the political will to find a better solution? Consider that such public employment does not make workers more desirable employees, and doesn’t give them any future opportunities.


Thomas Hanes is a member of the Class of 2020 and can be reached at

  • Man with Axe

    In those make-work jobs, what do you do with an employee who is really bad, i.e., he fights with other workers, he comes to work inebriated if at all, he does little to no work, etc.? Do you fire him or do you find a different job for him so that he becomes someone else’s problem?

    I’m not sure I believe the 1/8 statistic on the effect of burdensome regulation. The Empire State Building was built in 13 months. The Pentagon took about 16 months. Today, you couldn’t get an environmental impact statement approved in that amount of time.

    I wonder how much of the inefficiency comes from Davis-Bacon and similar legislation designed to protect union interests?