As I have noted before, the concept of industrial policy is making a resurgence. But is beginning to sound like the old song, “You say tomato, I say tomahto” (or the modern version “you say GIF, I say gif”). In this case, you say industrial policy, someone else says innovation policy – and they really mean technology policy.
Here are a few examples. Caleb Watney in “Untangling innovation from industrial policy” argues that we shouldn’t use the term anymore — it is too encompassing and therefore vague. Adam Thierer in “On Defining ‘Industrial Policy‘” takes the narrower approach, focusing on “developing or retrenching selected industries to achieve national economic goals” (a definition he borrows from historian Ellis Hawley as published in the 1986 AEI book The Politics of Industrial Policy). Dylan Gerstel and Matthew P. Goodman (From Industrial Policy to Innovation Strategy: Lessons from Japan, Europe, and the United States) seem to wrap industrial policy in the rubric of innovation policy, by which they really mean technology policy.
These discussions recount the debate over industrial policy in the 1980s. As someone who was heavily involved in those debates, the current discussions have a feeling of déjà vu. Back then, I tried to make sense of the various conceptual frameworks and approaches to industrial policy (see “A Reader’s Guide to the Industrial Policy Debate“). Many of today’s arguments fall into those same rubrics.
In some cases, the debate is on aid to a specific industry, (e.g. save the semiconductor industry, save the auto industry, save the …) – what I have called the problem-solving approach. The danger in this approach is that it fails to consider the system-wide effect of these actions on the economy as a whole. In other cases, the debate focuses on a specific policy (e.g. trade protection, Buy America, R&D tax credits) in an instrument-specific approach. The danger here is in the law-of-the-hammer: give a child a hammer and everything looks like a nail. In other words, the policy instrument is used not necessarily because it is appropriate but because it is available.
Another version, industrial policy is essentially manufacturing policy (what I labeled back in the 1980s as the “reindustrialization” approach). And while I am a strong manufacturing-matters advocate, I also strongly believe that industrial policy goes beyond manufacturing.
Two other versions of industrial policy routinely appear in the debate. I had labeled these as the “industrial triage” and the “reallocation” approaches. In industrial triage, some industries are doing fine while others are doomed to decline so attention should be paid to those where government help would be the most effective. The reallocation approach is similar in that it seeks to move resources from “sunset” industries to “sunrise” industries. While not using the same terms as before, part of the current debate over “industries of the future” is built upon this argument. Of course, both of these are subject to the classic criticism of “picking winner and losers.” Advocates for these approaches often respond with the argument that picking winners is exactly what we should be doing.
Only occasionally raised in the debate is the most comprehensive version of what I labelled “structural industrial policy” that focuses on the entire production system and multiple economic objectives. The earlier example I gave of this approach is the Japanese developmental state. It should be noted however that such an approach is difficult to establish and maintain due to a number of issues including political capture.
I would also note that the politics of industrial policy has changed very little over the past few decades. With some notable exceptions (such as Senator Mario Rubio), liberals argue for a more activist industrial policy while conservative argue for less. As a result, we tend to have a fear driven policy that appeals to both liberals and conservatives. Today, we have a fear-of-China industrial policy. In 1980s it was a fear-of-Japan industrial policy. In 1950s and 60s had a fear-of-Russia industrial policy. One could even argue that Hamilton’s Report on Manufacturing and Henry Clay’s American System were in part fear-of-Britain industrial policies.
I understand the political value of the fear-of approach and the reluctance to advocate for a systemic structural approach. Absent a national security rationale, such a more comprehensive approach brings attacks of “centralized planning.” But without the more systemic structural view, industrial policy devolves into piecemeal reactive actions. Such an outcome simply reinforces the critic that industrial policy is ineffective and captured by special interests. To avoid that outcome, we need to raise the debate to a higher conceptual level.