New SIPA Research Finds that the CHIPS Act Created Jobs – A Q&A with Eric Verhoogen

Image

Industrial policy has undergone a major revival in the United States, and one key element is the Creating Helpful Incentives to Produce Semiconductors (CHIPS) and Science Act, which committed billions to revitalize the domestic semiconductor industry. A key selling point of the CHIPS Act was its potential to create jobs. Was it successful on that front? In a recent paper, SIPA Professor Eric Verhoogen and his coauthors – Bilge Erten, an associate professor at Northeastern University, and Joseph Stiglitz, a professor at Columbia University – reveal that it created more jobs than experts had initially predicted. 

Verhoogen, who is also codirector of the Center for Development Economics and Policy, focuses on firms and development, exploring what leads firms to increase productivity. His research was recently highlighted at a conference at the Brookings Institution. In this SIPA News Q&A, Verhoogen discusses the motivations behind the project, its political and economic implications, and the future of US industrial policy.

What inspired you to explore the impact of the CHIPS Act on employment? 

Our primary interest is US industrial policy, and what other countries can learn from that. The research was also adjacent to my interest in firms and development. I have previously explored what leads firms to be more innovative and productivity to increase. This was an interesting example of where state intervention can boost innovation and create positive spillover effects. 

What findings surprised you? 

Most notably, the effects of employment were larger than many people, including us, expected. People doubted that the effects on employment would be significant, partly because semiconductors is such a capital-intensive industry. Our findings suggest that the Act created between 15,000 to 16,000 direct jobs in the semiconductor sector and 28,000 to 35,000 indirect jobs in construction and other local industries. 

While jobs were always one of the motivations listed for the CHIPS act, they likely weren't the primary reason. National security concerns and supply-chain resilience were more important, especially concerns about China and the dependence on semiconductors from Taiwan. 

The other important finding was anticipation effects. The employment effects that we see actually took place before the act was signed by Biden. We argue in the paper, and provide evidence, that they started in mid-2021, which is when a precursor bill was introduced in the senate. The act quickly garnered bipartisan support, including from 19 republicans. This made it clear that the bill had significant bipartisan support. Companies were paying close attention and responded quickly to the political developments. 

How does the CHIPS Act differ from previous US approaches to industrial policy?

The CHIPS Act and the Inflation Reduction Act (IRA), passed around the same time, were much more self-consciously industrial policy. Historically, US industrial policy has occurred through defense procurement and tax credits, among other things. While these might be considered industrial policy, they weren’t labeled as such. Innovation and productivity gains were often just positive side-effects of military expenditures. The Biden administration’s policies were much more deliberate and focused, and more willing to take on the “industrial policy” label. 

Is the CHIPS Act evidence of some kind of new economics of national security? 

I personally don’t believe that national security is needed to justify industrial policy. However, the CHIPS Act got bipartisan support partly because both parties were concerned over China, and potentially its relationship with Taiwan. Given the political situation in the United States, it probably was important that there were these bipartisan national security concerns. Another key aspect was COVID, which exposed vulnerabilities in US manufacturers’ supply chains. I wouldn’t describe what happened as a “new economics of national security,” but it clearly was an important motivating factor. 

Do you see any risks in the way the CHIPS Act is being implemented?

A big risk is that people see the CHIPS Act as corporate welfare — that it’s basically shovelling money from government accounts into company accounts, with little effect on what companies actually do. That question was partially what motivated our paper and our results show that it’s not the case. We’re finding big employment effects; companies aren’t simply getting money to do something they would’ve done anyway. By comparing counties with and without semiconductor plants, we were able to isolate the effects of the CHIPS Act and highlight its positive effect on employment. 

It's unclear what the current status of the CHIPS Act implementation is. The act hasn’t been officially repealed, but it hasn’t been moving forward. There haven’t been any new notices of funding opportunities and there’ve been few official announcements to my knowledge. That underlines the great uncertainty about what the US policy regime is, which risks decreasing private investment and possibly undermining the effects of the policy long-term. 

What lessons from the rollout of the CHIPS Act should policymakers keep in mind when designing future industrial policies?

One message is that policymakers need to make sure that money starts flowing in time to be able to take credit during elections. That did not happen with the CHIPS Act. Although we find that there were anticipation effects, the first contracts weren’t finalized until November 2024 when the elections were already happening. The Biden administration couldn’t really take credit for them. I think that was a failure of the policy implementation that should be avoided in the future. 

The other big question is whether governments should give direct grants to companies or tax credits. Direct grants generate less uncertainty about the policy’s total cost, since the grants are pre-determined. In contrast, tax credits depend a lot on how much companies take advantage of them, making it more challenging to predict the cost. Direct grants are arguably more transparent to taxpayers. However, a danger with the direct-grants approach is cronyism — that “insider” companies get favorable treatment. 

Finally, what questions remain unanswered about the long-term impact of the CHIPS Act on the economy?

The data available now are on employment. Eventually, we’d be very interested to explore the effects on productivity, output, and investment of firms in the semiconductor and related sectors. Unfortunately, it will take several years for the microdata on individual firms to become available.