Industrial Policy and Economic Security

Takeaways from two days of illuminating discussion

On Tuesday and Wednesday, Larry Summers and I convened a conference at the Harvard Kennedy School on Industrial Policy and Economic Security. At a time when Beijing and Washington are weaponizing many aspects of their economic relationship—and as the U.S. government levies vast tariffs and takes equity stakes in leading firms—we saw a need for more conceptual clarity around the concept of economic security.

Here’s a YouTube link to the conference.

Some key questions included:

1. What to do about ailing industrial champions, looking at Motorola in the past (relevant for 5G) and Intel and Boeing today—de facto national champions facing severe challenges.

2. The implications of China’s auto export surge. Should we allow Chinese cars to surge into U.S. and allied markets? If we ban them—as we de facto have in the U.S.—is there enough competition to force legacy auto firms to offer better products at lower prices?

3. What metrics define the AI race, who is winning, and what role, if any, does government policy have to impact it?

At the conference, I had the chance to interview Brian Deese, former director of the National Economic Council, on lessons he’d learned from his time in office. Larry concluded the conference with a fascinating conversation with Jake Sullivan, on why “mutually assured economic destruction” doesn’t provide perfect deterrence and whether nuclear-age theories of “escalation management” and “escalation dominance” apply in the sphere of weaponized interdependence.

(This same week, policymakers in Brussels and in Washington were asking themselves whether they could “Escalate to Negotiate,” as one smart ECFR paper put it, in response to Beijing’s moves on Nexperia and rare earths.)

The study of international economics clearly needs to take geopolitical competition more seriously. But scholars and practitioners of economic statecraft also need more focus on what economists call “willingness to pay.” How much should inefficiency should we tolerate for the sake of geopolitical resilience? The answer isn’t infinite—costs incurred today make us poorer and therefore less capable of mustering power in the future.

I was reminded of a talk Keynes gave in 1933—at the depths of the depression, after most of the world had left the gold standard, amid experiments in autarky ranging from the Smoot-Hawley tariff in the U.S. to Stalin’s shock industrialization—titled “National Self-Sufficiency.”

Why, Keynes asked, had the world come to reject the free trade ideas that had constituted mainstream economic thought and UK economic policy in the 19^th century? The ideas of Ricardo and Cobden had not been disproven, but they had receded in importance relative to other considerations. Keynes saw several that stood out.

First, economists still believed that free trade produced wealth, but they no longer believed it guaranteed peace:

It does not to-day seem obvious that a great concentration of national effort on the capture of foreign trade, that the penetration of a country’s economic structure by the resources and the influence of foreign capitalists, that a close dependence of our own economic life on the fluctuating economic policies of foreign countries are safeguards and assurances of international peace.

Second, free trade and investment flows were easier for the British to embrace when the British dominated them. The rise of U.S., German, Japanese, and other trade and investment flows might contribute to the division of labor, but the political impact was less obviously positive for Britain.

Keynes wrote:

The investment of British savings in rails and rolling stock to be installed by British engineers to carry British emigrants to new fields and pastures, the fruits of which they would return in due proportion to those whose frugality had made these things possible, was not economic internationalism remotely resembling in its essence the part ownership of the A. E. G. of Germany by a speculator in Chicago, or of the municipal improvements of Rio de Janeiro by an English spinster.’

Third, the measuring of outcomes on financial metrics alone was seen to have sidelined other equally important social or political considerations.

The nineteenth century carried to extravagant lengths the criterion of what one can call for short “the financial results,” as a test of the advisability of any course of action sponsored by private or by collective action. The whole conduct of life was made into a sort of parody of an accountant’s nightmare.

Keynes was careful not to endorse any of the more radical efforts to race away from international trade and capital flows. He criticized Soviet shock industrialization and mocked Mussolini. He urged caution, retaining an economist’s appreciation for the “pecuniary test,” even as he criticized excessive devotion to it:

But once we allow ourselves to be disobedient to the test of an accountant’s profit, we have begun to change our civilisation. And we need to do so very warily, cautiously and self-consciously. For there is a wide field of human activity where we shall be wise to retain the usual pecuniary test.

There are many echoes in today’s calls for self-sufficiency, strategic autonomy, de-risking, or decoupling. So it’s worth remembering, as we debate the same themes that Keynes wrote about in 1933, what he got wrong—in particular, the prediction that the cost of autarky would be low.

I am not persuaded that the economic advantages of the international division of labour to-day are at all comparable with what they were…Experience accumulates to prove that most modern mass-production processes can be performed in most countries and climates with almost equal efficiency.

He based this conclusion, reasonably but wrongly, on the thesis a shift to a services economy would make trade less relevant:

Moreover, with greater wealth, both primary and manufactured products play a smaller relative part in the national economy compared with houses, personal services and local amenities which are not equally available for international exchange; with the result that a moderate increase in the real cost of the former consequent on greater national self-sufficiency may cease to be of serious consequence.

The computer I’m typing this on is strong evidence to the contrary. It may be true that the consumer basket of 1933 could be provided relatively capably by each of the world’s 10 largest economies. The same was perhaps true in 1963. Today’s tech, aerospace, and auto supply chains are different. There’s no way to produce comparable complexity and comparable efficiency within a single country—and certainly not within any country beside the U.S. (Some European countries tried to build national computer industries in the 1960s, 70s, and 80s—it didn’t work.)

So how much inefficiency should we swallow in exchange for the security of self-sufficiency? It’s a tricky calculation, because security gains are probabilistic (how likely is an adversary to test our reliance on products from abroad?) and because the long-term economic costs are highly speculative. I’m a believer that, like in 1933, there was a strong case for rebalancing economic and security concerns in favor the latter. In 1933, the Western powers were far delayed in their rearmament and far too willing to transfer technology and industrial capacity to obviously adversarial powers. But that concern must be balanced against the downsides. Being the world’s richest and most productive economy is a key input to our security, too.