Industrial policy doesn't have to succeed right away

Progressive approaches need to fail before they can change.

"Intel D1X Development Fab Hillsboro Oregon" by Intel Free Press, CC BY-SA 2.0.

Ezra Klein has written the definitive piece on the challenges facing progressive industrial policy. In recent years, most progressives have concluded — quite rightly — that they need to offer the American people more than “stasis subsidies” and promises of expanded transfer payments that seem to never make it through the legislature. Progressives need to build — to create green energy and green infrastructure to fight climate change, to create housing so renters can afford to live where the jobs are, and to create high-tech manufacturing industries in order to prevent the U.S. from turning into an industrial backwater. This is a very very good thing! But in order to follow through on those promises, progressives are going to have to abandon the “laundry list” approach that they embraced over the last 50 years, where every government project is sold as a way of paying out a huge assortment of interest groups and micro-constituencies. Klein refers to this as “everything-bagel liberalism”:

The high cost of building [semiconductor fabs] has become a national security liability…But it is very hard to read the guidelines the [Biden] administration just released and see a serious effort to lower costs. The government is adding subsidies with one hand and layering on requirements with the other.

Page 11, for instance, encourages a pre-application that includes an environmental questionnaire “to assess the likely level of review under [NEPA].” Page 20 mandates that applicants prepare “an equity strategy, in concert with their partners, to create equitable work force pathways for economically disadvantaged individuals in their region,” which should include “building new pipelines for workers, including specific efforts to attract economically disadvantaged individuals and promote diversity, equity, inclusion and accessibility.” Page 21 asks for a plan “to include women and other economically disadvantaged individuals in the construction industry,” “strongly encourages” the use of project labor agreements and sets out requirements for “access to child care for facility and construction workers.”

Pages 23 and 24 ask applicants to detail how they will include minority-, veteran- and female-owned businesses, as well as small businesses, in their supply chain and offer seven bullet points detailing how this might be done, including dividing supply chain requirements “into smaller tasks or quantities to expand access” and “establishing delivery schedules for subcontractors that encourage participation by small, minority-owned, veteran-owned and women-owned businesses.” Then there are requirements for “a climate and environment responsibility plan,” as well as community investments in areas like transit, affordable housing and schools.

Many of these are good goals. But are they good goals to include in this project?

When Klein raises these concerns with Commerce Secretary Gina Raimondo, she is astonishingly dismissive, issuing generalities such as “I consider it a fact that a more diverse work force is a more productive work force,” without defending the assumption that splitting the fab construction process among a galaxy of small inefficient contractors is an effective or cost-effective way of creating a diverse workforce. It’s very clear that the Biden administration decided on these requirements for political reasons and reverse-engineered a bunch of ad-hoc economic justifications for them.

Some of those justifications might even turn out to be right, but ultimately they will be overwhelmed by the basic truth that political payout schemes do not tend to lead to productive enterprise. This is not to say that progressive goals, such as a diverse workforce or a robust small business sector, are bad ones. But the methods that progressives are trying to use to accomplish these goals aren’t necessarily the right ones.

For example, take the aforementioned idea of a diverse workforce. Let’s accept for the moment the proposition that it’s productive and socially beneficial to have workers of various backgrounds working together on projects. That says absolutely nothing, however, about whether it’s productive or socially beneficial to have a project worked on by a set of subcontractors with diverse ownership. In other words, maybe it’s a good thing to have a workforce be 50% men and 50% women. But that doesn’t mean that having 50% of your subcontractors owned by men and the other 50% owned by women will give you any of the benefits of workforce diversity! It’s simply porting a concept from one domain — soft quotas for work team composition — over to a completely different domain and assuming it’ll work exactly same way.

But there’s absolutely zero reason to expect it to work exactly the same way! Maybe the male subcontractor owners will occasionally talk to the female subcontractor owners, and maybe this will produce some tiny fraction of the productivity and equality that you get from a diverse work team. But overall, it seems extremely likely that soft quotas for subcontractor ownership will simply force companies to split the work among an inefficiently large number of subcontractors, thus impairing the efficiency of the overall project.

Does that mean that promoting woman-owned businesses is a bad goal? No. But progressives can’t just use a cookie-cutter approach here. And as Ezra suggests, the goal of promoting woman-owned businesses should be unbundled from the goal of building semiconductor fabs. Perhaps the government can provide free marketing and matching services for woman-owned businesses, to help them win contracts. Perhaps the government can subsidize loans to woman-owned businesses. Obviously there will could be legal issues with these alternative ideas, but the point here is to illustrate a principle: It should be possible to promote female business ownership without forcing semiconductor companies to split their subcontracting work into a million tiny pieces.

And this, in turn, is an example of a more general principle that progressives need to understand: Individual projects don’t have to be equitable in order for growth itself to be equitable. Bundling — insisting that every project accomplish every progressive goal down to the most minute level of execution — is a recipe for failure.

My theory is that progressives’ insistence on inserting every goal into every project — what Ezra calls “everything-bagel liberalism” — is a relic of a time when progressives were in the minority and were only occasionally able to get anything done. But progressives are no longer a scrappy rebellion' — they’ve entered an era where they are actually able to do multiple projects over an extended timeline. And that means it’s time to start thinking about efficiency.

This is just how we figure things out

Anyway, Ezra’s post is great and you should read the whole thing. But there’s the danger that skeptics of industrial policy will use these emerging challenges opportunistically, as a way to persuade the general public to dismiss the entire project. For example, the Bloomberg Opinion editors wrote an extremely reasonable editorial about all the challenges that I’ve been talking about, which concludes on an optimistic and highly reasonable note:

The good news is that these are mostly solvable problems. As just a start: Issue “fast track” exemptions for chipmakers under federal environmental law — or, better yet, amend the law to accelerate all such projects and inhibit frivolous lawsuits. Increase visas for skilled workers, prioritize applicants with needed STEM abilities, and exempt foreign graduates with advanced science degrees from the cap on green-card allotments. Slash the costly and counterproductive strings attached to Chips Act funding and accept that the proper venue for enacting the progressive agenda is Congress, not random companies…If lawmakers want to solve them — to start building again — they can’t just push more money out the door. They need to do the hard work that national progress demands.

But Cato Institute Director Scott Lincicome seized on the editorial’s critical sections to denounce the entire idea of industrial policy.

We should expect to see this pattern repeated again and again. Industrial policy will hit challenges and stumbling blocks, and each of these will instantly be trumpeted by both ideological libertarians and opponents of progressive social policy as a reason to scrap the entire idea of using the government to promote investment. I also expect this premature doomerism to prompt some progressives themselves to rally around the existing approach and insist no changes at all need to be made.

Both of these instinctive, tribalist reactions must be resisted. The truth is that industrial policy can work, but it’s difficult and tricky to get right. Whether we succeed or fail, we’re going to make a lot of mistakes along the way. Because we don’t know the exact right mix of policies ex ante, we have to “cross the river by feeling the stones”, as Deng Xiaoping said. (It’s notable that Deng himself made plenty of economic mistakes before settling on the formula that eventually turned China into the economic powerhouse it is today; the Tiananmen Square protests were sparked in part by high inflation resulting from Deng’s policies in the 1980s.)

So all the howling about permitting and onerous subcontracting requirements and “Buy American” provisions that you’re hearing right now — including from Ezra Klein and from Yours Truly — are just part of the process of figuring out how industrial policy works. We’re kicking the tires on our modern state to see what it does well and what it does poorly. Observing the results each step of the way, we’ll correct our course.

So in fact, while it’s important that industrial policy succeeds in a reasonable time frame, it doesn’t have to cruise to an unimpeded total success on the very first try. Just like for a tech startup, it’s ok to fail at first if you pivot nimbly.

And it’s important to remember that criticism of the current approach is key in executing a speedy pivot. China’s government may not be burdened by the same sorts of interest groups and electoral uncertainty that America is, but their dictatorial system makes it very hard for the supreme leader to correct course. Both China’s semiconductor “Big Fund” (a predecessor of the CHIPS Act) and their Made in China 2025 initiative largely failed, but it’s not clear that Xi Jinping or his yes-men learned their lessons from those failures. Biden and the Democrats, however, have a media that’s able to swiftly point out what they’re doing wrong.

Checkism is a blindfold that stops us from correcting course

When we think about industrial policy as an iterative policy involving course corrections, it becomes clear that high-quality feedback is crucial; we need to know when the policy is succeeding and when it’s failing.

And if we fall victim to what I call checkism, we will blind ourselves completely. Checkism is when the apportionment of funds is viewed as a policy victory in and of itself, regardless of how much useful stuff is actually produced by those funds. Many progressives seem to believe that the bottleneck in U.S. state capacity is political will, so that all we need to do is cut big checks and great things will happen. This leads to a lot of progressive triumphalism when funding is actually secured. For example, author Jay Turner recently wrote a Twitter thread in which he trumpeted the announcement of projects as a success of the Inflation Reduction Act:

This prompted former Biden official Jennifer Harris to declare that “I’m a fan of US’ revamped industrial policy because #itworks”.

Now, don’t get me wrong; I’m overjoyed that all of these projects are being announced and that so much money is being committed to the project of building green energy. It’s absolutely worth trumpeting. But it’s just premature to say that industrial policy “works” just because projects have been announced. Turner’s first graph shows 26 projects in planning but only 4 in construction. We don’t know yet how many will be held up by environmental review or other roadblocks or veto points.

In other words, we’re at the beginning of our experiment with industrial policy, not the end. And because checkism declares victory as soon as the numbers go in the spreadsheet, it has the potential to act as a blindfold — short-circuiting the necessary process of failure and course correction that is essential to the actual success of any policy program. The Bloomberg editors are on exactly the right track when they say that “simply writing checks was never going to be enough…If lawmakers want to…start building again…they can’t just push more money out the door.”

So even as proponents of industrial policy criticize the specific ways in which industrial policy is going wrong, we also have to criticize the use of bad metrics that lead us to mismeasure success and failure. And we have to be patient, and learn to live with constant unexpected setbacks and frustrations. But we have to keep pushing, because national security and the fight against climate change simply won’t be solved by having our government sit on its hands and leaving things to the market.

---