You’re reading Economic Forces, a free weekly newsletter on economics, especially price theory, without the politics. You can support our newsletter by signing up here:

Subscribe now

I was just a spry young lad during the Great Recession, focused mainly on playing college football. (Please, no brain damage jokes. Too soon.) Before 2008, I had no clue about economics, but the Great Recession opened my eyes to the power of economic reasoning. Like those generals fighting the last war, I often go back to that defining moment for how to think through economic policies. Most of the Great Recession retrospectives focus on monetary policy and what the Fed did, primarily related to the financial crisis.

Another policy that was hotly debated at the time but mostly forgotten now was the decision to bail out the auto manufacturers: GM and Chrysler. At the time, I remember there being a big fuss. Lots of people asked why should the government help these companies, in particular, their union employees? Companies like Toyota were relatively unaffected because they could keep costs low without high union costs. I don’t know if this was the proper interpretation of the events, but it was the popular narrative.

Such a bailout has two positive impacts, according to proponents. First, it props up companies, just as the financial bailouts did, and prevents a broader collapse of the overall economy. Second, and this is unique to the auto bailouts, it was explicitly protection for the autoworkers and their jobs at the automotive plants.

This newsletter will argue that protecting jobs (at an automotive plant or elsewhere) is perverse. We should try to project the worker instead. I am making a normative argument—yes, grounded in economics—about how we should think about these policies. To be clear, it is not strictly the type of economic argument I usually write on.

Let’s consider the distinction between subsidies for jobs and workers. The most pervasive form of subsidies for an explicit job is through employment protection. Employment protection is a benefit given to workers, but only if they remain in their current job. For example, a law that prevents companies from firing workers will make it harder to fire workers, which benefits the current workers as long as they stay in their current jobs. However, for new workers entering the labor market or for current workers finding new jobs, it also makes hiring them more costly, which hurts workers. The stated purpose for such a law is that it’s terrible when people lose their jobs, and therefore policy should stop bad things.

Subsiding the job, in whatever form, has adverse effects. Raising the costs of hiring workers throws sand in the gears of the economic process by which workers move from unproductive firms to productive firms. For economic growth, this reallocation needs to occur within and across industries. A policy of propping up coal mines keeps workers from learning new skills and moving to another industry. 

This is the exact opposite of what a dynamic economy needs. Workers in an economy need to have the opportunities to relocate, to move from low value-added jobs to high value-added jobs. Reallocation is a major source of increases in productivity that economies need to harness. With the auto bailout, workers received the benefit only if they remained working with GM. In effect, the government indirectly paid workers to continue working at unproductive jobs.

Europe as a whole, but particularly an economy like Spain, which I know a bit about, is the classic example of subsidizing the job through employment protection. But the dampening effects of employment protection on economic productivity and growth are all over.

If the goal is to protect people who lose their jobs from the terrible consequences of losing a job, economic reasoning encourages us to look at whether there is a way to get the benefits with lower costs. I do not have a magic policy. If there were one, someone before me would have found it and implemented it. But I think there is a basic framing that we should apply to such policies.

Subsidize the worker, not the job.

In the auto bailout, it would have been more efficient if we could find a way to help those same workers, subsidize them directly, without locking them in an inefficient job from a societal perspective. For example, Congress could pay them all of their lost wages due to GM’s failure. Even unemployment benefits are better. But definitely do not keep them in a wasteful job! Even if you think the China shock is a big deal, help the workers. Don’t use tariffs to keep workers in wasteful jobs.

This is the same argument that one sees in favor of school vouchers. Instead of giving money to the schools, give money to the students/parents. This has two main benefits. First, the money goes directly to the people you say you are trying to help. Second, you induce competition for quality schools.

For workers, by subsidizing the workers directly, the money goes to the thing you claim to care about (the well-being of the workers who are hurt by lost jobs). At the same time, you induce competition among employers to hire these workers. Contrast this with a robust employment protection system, where you dampen competition for the workers.

The Josh Hendrickson in me must ask, why don’t we see so many subsidies for jobs instead of workers? We tend to see policies that prop up industries and jobs, not the people. As always, there are two types of answers to why policies exist.

The first response (the true Hendricksonian response) is that the system of subsidizing jobs is efficient! There is a transaction cost that I am ignoring in my analysis above that raises the cost of subsidizing the worker. If I took that into account, my supposed efficiency gain would disappear. One possibility could be that it is harder to identify the relevant workers than it is to identify the relevant markets you want to subsidize. If we tried to subsidize every farmer, everyone with a garden would try to fake being a farmer. Instead, we subsidize corn in a world of second-bests. I’m not saying that is the most reasonable transaction cost, but it is one.

The second reason we may see the policy is rent-seeking. We got auto bailouts because the unions have political power, not for efficiency. We have corn subsidies because the corn farmers use the government to curry favor and get more money for themselves at the expense of the consumers. Because farmers are a concentrated interest group and the costs to consumers are dispersed, they have an advantage in the political sphere and use it. We have subsidies for schools instead of students because policymakers want to subsidize the teachers and administrators, not the students.

However, a rent-seeking theory needs to explain why one group gets the rents and also the form of the rents. One could imagine paying the current farmers directly instead of subsidizing the industry, even after they leave farming. You literally buy off the current farmers.

A better theory of rent-seeking gives us a reason why the subsidy looks the way it does. For example, in Chicago Price Theory, the authors argue that we see an ethanol subsidy because it is a way to benefit the farmers twice while only paying one. You raise the value of corn which increases the value of land for corn and the land for soybeans. This is more politically popular since it has a lower price tag.

How is the job subsidy cheaper for politicians than the worker subsidy? Let me know below.