Industrial Policy: Shouldn't we actively encourage X, rather than just passively allow Y?
By Scott Sumner
This post will combine two of my interests:
1. Framing effects
2. General equilibrium and the fallacy of composition
Although most Econlog readers can see past framing effects, they have a powerful effect on the general population. My claim is that most people would answer yes to the question in the title of this post. And I also claim they’d continue to answer yes if you reversed X and Y. It simply sounds good.
Do I have any evidence for this claim? Sam Bowman directed me to this tweet:
It’s all about the framing effects.
I’m reminded of this problem when I run across people advocating industrial policy. They tend to advocate policies that sound good, without fully considering the “general equilibrium” effects. The term ‘general equilibrium’ means that everything in the economic model is affected by everything else. For our purpose, just think of a simple downward sloping production possibilities frontier, with “industries of the future” on one axis and “older industries” on the other. Any industrial policy that encourages more of one has the effect of encouraging less of the other.
You might argue that both sectors can be boosted if you are in a recession. Yes, but that’s not industrial policy, that’s demand stimulus. True industrial policy is a set of taxes, subsidies and regulations that push economic activity toward some industries and away from others.
Consider a recent defense of industrial policy in the NYT:
Industrial Policy Isn’t Boring. It’s Big, and It Has United the Left and Right.
Many Republicans and Democrats agree: Governments should intervene to help the industries and technologies of the future.
That seems pretty clear, right? Julius Krein of American Affairs wants to use industrial policy to boost industries of the future, presumably areas such as biotech, AI, autonomous vehicles, fintech, virtual reality, etc.
One problem is figuring out which will be the industries of the future. When I was young, I had no idea that the booming industries of the 2010s would be yoga studios, pedicure establishments, tattoo parlors, massage parlors, SAT coaching, etc. One rarely saw that sort of business during the 1960s.
But here’s the bigger problem. The actual article Krein has written often seems to (implicitly) suggest the exact opposite, that we should be encouraging the industries of the past:
In the wake of the 2008 financial crisis, however, the Reagan-Bush-Clinton neoliberal consensus seems intellectually and politically bankrupt. Party “establishments” appear incapable of proposing, much less advancing, policies sufficient to address major economic challenges. . . .
In addition to improving competitiveness and productivity growth, industrial policy may be the only way to meaningfully increase economic opportunities for struggling regions and populations. More generally, during the last several decades, the benefits of productivity gains have mostly accrued to the largest capital holders. A successful industrial policy would aim to strengthen worker bargaining power while organizing and training a better skilled labor force. Industrial policy also involves, and even depends upon, rebuilding infrastructure. . . .
Just as Republicans and Democrats in the Bush-Clinton era had intense debates over issues like marginal tax rates, all the while advancing an underlying neoliberal agenda, so will future debates play out in an era of renewed industrial policy. Uniting both sides will be a desire to revitalize domestic industry and to use the state to ensure that the economy serves a broader set of stakeholders and public interests rather than only large shareholders.
What does that all mean? If we “revitalize” industry in some “struggling regions”, aren’t we likely to actually promote older industries? The popular story today is that we let the market allocate resources due to a misguided adherence to neoliberal ideology. As a result, we have booming high tech industries in Seattle, California and the Northeast, and a hollowing out of our older industrial sector in the Rustbelt.
I don’t know about you, but to me that looks like a policy that is very much oriented toward the industries of the future. We could adopt an industrial policy of propping up older industrial sectors and struggling regions of the country, but that policy would effectively discourage the industries of the future.
The upcoming reauthorization of the Small Business Administration offers a chance to make changes along these lines. Government agencies could also step in to seed investment funds focused on strategic industries and to incentivize commercial lending to key sectors, policies that have proven successful in other countries (including in Canada, Britain, Israel, South Korea, China, Taiwan and Singapore).
When you compare the US to poorer, less successful economies (which includes all countries on that list except Singapore), what stands out is that the US is completely dominant in the industries of the future. We have most of the world’s top high tech companies, most of the world’s top biotech companies, most of the world’s top research universities, etc. Germany and Japan are better at making cars, but we have Tesla.
If you want an “industrial policy” aimed at bringing America the industries of the future, I’d say we already have it. It’s called free market capitalism. If you want an industrial policy aimed at helping workers in struggling regions, workers displaced by competition from China, that would be the sort of industrial policy that European countries used to engage in during the bad old 1970s. Those policies help to explain why Silicon Valley is in San Jose, not in Oxford, Paris, or Dusseldorf.
Can’t we just help all sectors, industries of the past and the future? Not really. Economics is about making choices. If you help sector A, you discourage sector non-A. Think general equilibrium. Do we want to help or hinder industries like this?
I suppose you could try to thread the needle by promoting both high tech and struggling Rustbelt manufacturing and mining, but then what is being left out? Consider the following industrial policy:
1. We’ll spend lots of money subsidizing high tech industries that employ lots of nerdy young men.
2. We’ll also spend lots of money subsidizing older industries that employ lots of brawny male steelworkers and coal miners.
Yes, you could do both, and it might sound good, but what would that actually mean? Does that policy sound as good if you frame it this way:
“Let’s have an industrial policy that discourages the creation of businesses that traditionally employ women.”
To me, industrial policy is just a bunch of happy talk. It sounds good to encourage the industries of the future. It sounds good to help struggling workers in struggling regions of the country. But until industrial policy advocates make the hard choices, or even show they understand the choices that they are implicitly making, I’ll continue to assume there is no realistic alternative to neoliberalism.
I’m actually heartened by the fact that the policies proposed in the article (SBA? seriously?) are so weak and ineffectual. This suggests that Margaret Thatcher’s, “there is no alternative” is still in effect. Neoliberalism isn’t going away.
In fairness, Krein has written a more nuanced essay that what one might infer from the quotes I cherry picked. So I’ve been a bit unfair to him. He seems to want to boost manufacturing as a share of GDP, and have much of the new manufacturing occur in future oriented industries. But threading the needle that way is much harder than it looks, and real world industrial policies will either end up usefully correcting actual market failures (say with carbon taxes or basic research subsidies), or they’ll get distorted by the political process and end up stumbling over exactly the sort of dilemmas I discuss in this post. Who are we actually trying to help, the nerdy guys or the brawny guys?
And what about women?