Introducing Factory Settings

A new series about how CHIPS succeeded

In the last few decades, only a handful of government interventions have been near-universally regarded as successes. Operation Warp Speed, which accelerated the development and distribution of COVID vaccines, is one of them. So is CHIPS, the Creating Helpful Incentives to Produce Semiconductors Act. CHIPS successfully spurred a massive investment boom in semiconductors on American soil.

In Factory Settings, a new series from IFP, the inaugural CHIPS leadership team will explain how they pulled that off.

How do we know CHIPS succeeded?

In the years between 2007 and 2020, annual domestic construction spending in computers, electronics, and electrical manufacturing averaged $3.3 billion. After the program was authorized in 2020, and in the nearly five years since, it’s averaged $71.8 billion per year. That’s nearly twenty times higher.

In the coming years, the federal government will try to increase our resilience in domains like energy, critical minerals, and advanced pharmaceuticals. By default, those interventions are likely to fail. If they do, they’ll deepen Americans’ sense that the government can’t do its job.

The CHIPS office succeeded in dramatically increasing investment in a critical domain, and it executed that goal quickly and effectively. But to succeed, the team had to optimize an operating model that too often fails.

The lessons the CHIPS team learned are applicable whether you’re a socialist or a libertarian, an industrial policy fan or skeptic. If you care about making government work better, the story of CHIPS has lessons for you.

With that, here are Mike Schmidt, Todd Fisher, and Sara Meyers, the inaugural leadership of the CHIPS program, with Factory Settings.

- Arnab Datta, Director of Infrastructure Policy at IFP

---

In 2022, Congress gave the Commerce Department $39 billion in incentives to bring the production of semiconductors back to America. The three of us were pulled into an awesome task: build a $39 billion start-up in government that would negotiate, deal-by-deal, with the world’s most sophisticated and geopolitically significant companies to restore US manufacturing.

Over the course of more than two years, we went all in. With no team, no program, and no vision for what success would look like at the outset, we worked with an extraordinary group of public servants to build from the ground up. And, against what often felt like long odds, our team largely succeeded. We implemented the program (awarding $34 billion across 20 deals), catalyzed massive investment (with all five leading edge manufacturers building on our shores), and left behind a policy apparatus that has endured a change in administrations.

That’s a hopeful story — one that demonstrates that we still have the ability to do big bipartisan things as a country, and one that shows a path toward rebuilding industrial capacity in the face of alarming strategic vulnerabilities. However, we are launching this project not out of hope, but out of concern for the country’s capacity to replicate this kind of success going forward.

The challenge, we learned, is that good execution is much, much harder than it needs to be. Our experience suggests that the government sets itself up to fail in countless ways — whether through burdensome cross-cutting statutory requirements, process-intensive regulations, or oversight-conscious procedural norms. As we battled through maddening legal and procedural thickets, we thought, “there has to be a better way.” With no existing playbook for modern industrial policy, we forged one for ourselves.

But before we dig deeper into the motivations behind this project, a bit more about the three of us.

Our team

From left to right: Sara Meyers, Todd Fisher, and Mike Schmidt.

Mike: Before Secretary Gina Raimondo recruited me as the Director of the CHIPS Program Office, I’d spent most of my career in government. Early on, I was a policy generalist, with roles spanning financial regulation, consumer protection, infrastructure, economic development, housing, and tax policy. From there, I’d focused on execution. I ran New York State’s revenue agency for more than two years; after that, I led implementation of the Child Tax Credit program at the Treasury Department. CHIPS presented an opportunity to do implementation in its rawest form, a blank slate that would require me to flex from strategy to policy to execution. I was motivated by the chance to show that government can still do big things.

On my first day at Commerce, I met Todd Fisher, who would become the program’s Chief Investment Officer.

---

Todd: I’d left a long career in the private sector, including 25 years at KKR as a private equity investor and C-suite executive, to pursue a second career with the goal of having a more direct positive impact on the world. I spent a few years helping scale a large workforce not-for-profit. When Gina Raimondo became Commerce Secretary, I sent her a cold email, explaining how I’d always wanted to work in the public sector and have a greater impact on the US and the world. She brought me in to run some economic development programs. When CHIPS passed, Raimondo called me again, and I jumped at the chance to use skills from my long private sector career on a national security priority.

Mike then recruited Sara Meyers to dual-wield the titles “Chief of Staff” and “Chief Operating Officer.”

---

Sara: I’d grown up in government, starting as a Presidential Management Fellow in 2009 and rising to be the Deputy Assistant Secretary for Operations at the Federal Housing Administration. Throughout that experience, I’d worked on a series of start-ups in government, including high-profile projects like the American Recovery and Reinvestment Act and the Hurricane Sandy Rebuilding Task Force. I then served as the #2 executive in DC’s Office of the State Superintendent for Education, an agency with ~2,000 employees. I’d always been drawn to work that demanded cutting through bureaucracy, bringing clarity to chaos, and building new systems, so I immediately saw the upside of joining the CHIPS effort. I also always sought out roles where I could learn from and laugh with colleagues. Mike sealed the deal by offering that I check his references.

---

The three of us were surrounded by great people, many of whose voices will be included in this project. The CHIPS Program Office, as our team was called, would grow to 180 people. It was as dedicated and talented a group as you will find in government or the private sector. And we had fantastic colleagues, from Secretary Raimondo and her senior team to policymakers and lawyers across the Commerce Department.

Why we’re launching this project

It’s hard to do pretty much anything in government.

The stakes of our work at CHIPS were self-evidently high. US semiconductor manufacturing had declined from about 40% of global production to just 10% over four decades, leaving us with a glaring vulnerability. Production of leading-edge logic chips — the critical technology behind AI — had dropped to zero, with nearly all of them now made in Taiwan. At the same time, China’s posture toward Taiwan had become increasingly aggressive, putting the entire modern economy at risk. The national security stakes meant that our project was backed by the full might of the United States government: a bipartisan bill championed by the President and by his Secretary of Commerce.

But though the work was urgent and important, and the resources at our disposal were enormous, we encountered maddening procedural hurdles to successful execution. In the years after joining, we found ourselves (among other examples):

• Trying to stop endangered bats from derailing one of the largest industrial projects in the history of the country.

• Telling companies that if they accepted CHIPS incentives, federal wage regulations would require them to pay hundreds of millions in backpay to tens of thousands of construction workers who no longer worked for them.

• Sitting in a conference room in Taiwan while earnest government lawyers explained that the “Fly America Act” would require anyone taking a flight supported by the program to use only US flag air carriers.

• Fighting the Office of Management and the Budget for yet another “emergency exemption” to the Paperwork Reduction Act to allow us to ask industry basic questions.

To be clear, our frustration isn’t that implementation of the program was administratively complex. To the contrary, administrative complexity is part of why we loved our jobs! Everyone wants to be a policy person with a lofty title — fewer people actually want to execute. But execution matters just as much as policy. When good policy meets poor execution, it is the reputation of the execution that sticks.

Our challenge was compounded by the fact that there was no playbook for industrial policy at the scale of CHIPS in recent history. We confronted big questions with a blank slate: What does it mean for industrial policy to advance economic and national security? What is the role of production in allied countries? What market failures are you trying to solve? How do you size supply-side incentives? How do you structure deals to advance strategic objectives while protecting taxpayer dollars? Should taxpayers get any financial upside? How do you drive demand for US production?

This brings us to another motivation behind Factory Settings: getting industrial policy right. As the US-China relationship evolves into a battle of chokepoints, and as AI assumes center stage in 21st-century geopolitics, industrial policy is becoming an increasingly important tool of economic statecraft. After leading the largest industrial intervention in generations, we believe strongly in the critical importance of effective industrial policy for US national security. But we also believe that industrial policy is like cholesterol: there are good kinds and bad kinds. Done poorly, industrial policy can do more harm than good.

Accordingly, you can expect three things from this series:

1. Unpacking the key decisions we made while implementing the program — from the operational to the tactical to the strategic. Some of those choices worked beautifully, others were imperfect, and a few were downright messy. Our hope is that these deep dives will be useful for future program implementers who can learn from both our successes and our missteps. And for a broader audience, we think these stories will reveal a lot about the realities of operating inside government.

2. Broadly applicable lessons about enhancing the state’s capacity to govern effectively. We hope to contribute to a broader conversation about how procedural rules and norms constrain the government’s ability to deliver. In the short term, there are specific legislative opportunities to make government work better: reauthorizations of the Defense Production Act, the Development Finance Corporation, the Export-Import Bank, and the government’s Surface Transportation programs, to name a few. And at the same time, major new industrial financing programs — like DOE’s Energy Dominance Financing Program (~$200 billion in lending authority), DOD’s Capital Assistance Program (~$100 billion), and DPA Title III appropriations — create opportunities to apply these lessons.

3. A policy conversation about what effective industrial policy looks like. We worry that the enthusiasm for industrial policy today risks outpacing its analytical underpinnings. The solution is not to dampen the enthusiasm but to strengthen the analysis. Industrial policy needs to be thought of as an emerging field of statecraft, one that carefully calibrates economic tools to national security objectives and weighs the benefits of interventions against their fiscal costs and policy drawbacks. We hope that studying our experience can help shape future industrial policy efforts for the better.

Core themes

While the individual pieces in this series will usually center on concrete decision points or challenges, taken together they’ll revolve around five core themes. These themes shed light on both why we were successful, and where we ran into trouble.

1. Setting and prioritizing clear goals.

The CHIPS Act gave us an unambiguous statutory mandate: revitalize semiconductor manufacturing in the United States. To translate that mandate into specific objectives, we put a lot of effort early on into developing a more specific Vision for Success, which laid out concrete goals across industry segments — leading-edge logic, leading-edge memory, current-generation and mature-node chips, advanced packaging, and the upstream supply chain.

Publishing the Vision was not a risk-free choice. It consumed significant resources early on and gave us a scorecard that we could either succeed or fail against. But in practice, it became indispensable. It shaped companies’ funding applications, publicly explained the tradeoffs we were making (which helped us manage negotiations with industry), and served as the backbone of our internal decision-making.

Of course, setting goals only matters if you prioritize them, and prioritization requires tradeoffs. While the CHIPS Act contained some specific legislative provisions guiding our programmatic design choices — on how we supported small businesses, for example — it ultimately gave Commerce huge discretion to advance the program’s mandate. That discretion was fundamental to our successes. We struggled most when others controlled the program’s fate: under the National Environmental Policy Act, for instance, district court judges could make or break projects; under the Davis-Bacon Act, the Department of Labor held regulatory and enforcement authority. Those frameworks serve policy objectives, but onshoring semiconductor manufacturing isn’t one of them.

And yes, we know that some may raise eyebrows here. The CHIPS program was the original target of the “everything bagel” critique, which argued we piled on unrelated policy priorities in a way that distracted from our core mission. We’ll dig into what that critique missed, what it got right, and what it means for policy design going forward.

2. Hiring and retaining talent in government.

For us, the best part of working on CHIPS was the team, which had around 180 people at its peak, before DOGE-related cuts under the early Trump Administration. The program pulled in talent from across the semiconductor industry, finance, consulting, defense and intelligence, workforce experts, environmental permitting, government operations, and external affairs. We recruited hard, hired fast, built a mission-driven culture, and kept top talent far longer than we expected.

Federal hiring is byzantine, and we had structural advantages: significant admin funding, special statutory hiring authorities, and additional excepted service and direct hire flexibilities from OPM. We also poured our energy into building the team, with the Secretary herself leading on key hires and senior Commerce staff pushing the effort forward daily. And as the team grew, culture-building was deliberate, helping us keep people through inevitable stretches of fatigue. Even then, we were still tripped up at times by the system.

3. Navigating risk.

When government money is on the table, the natural instinct is to obsess over risks of commission — like funding a deal that goes bad. This is often shorthanded as “Solyndra risk,” after the misbegotten Department of Energy loan in President Obama’s first term. But while we learned lessons about underwriting rigor from Solyndra, we felt strongly that we couldn’t be paralyzed by fear of mistakes. Instead, we were just as focused on the risk of omission.

Beyond investment risk, there was the constant calibration of process risk, which usually manifested in the form of either oversight risk (would an inspector general or congressional committee call us out?) or litigation risk (would we get sued and lose under the APA, NEPA, or another statute?). We took those risks seriously on a range of issues, including the iterative nature of our investment process, how we engaged chip customers, governance structures within the office, and the terms of our final deals. But again, we were not paralyzed by those risks. We tried to move at the commercial pace of the industry while ensuring that our decisions would hold up under scrutiny. Our constant refrain was that “our biggest risk is we don’t get this done.”

These lessons around risk calibration are especially relevant across the whole government. In our view, the core underlying culture of government is far too risk-averse. This needs to be rebalanced across government. That will mean both changing policy, and changing culture.

4. Matching the tools of industrial policy to its objectives.

We believe that every industrial policy tool — grants, tax credits, loans, equity, tariffs, demand incentives — comes with downsides, from financial cost to market distortion. Effective policy means tailoring the tool to the problem, maximizing strategic benefits while minimizing drawbacks. We’ll talk about where CHIPS succeeded and where it fell short. And we’ll analyze how novel industrial policy tools are evolving under the Trump Administration: revenue-sharing tied to export licenses (Nvidia/AMD), equity in exchange for grants (Intel), golden shares tied to CFIUS approvals (Nippon Steel), or equity and demand guarantees for rare earths (MP Materials). Each of these approaches implicates questions we faced in CHIPS around the role of profit motives, the use of coercive tactics, and more. These are complex, live issues. We’re going to use Factory Settings to tease them out.

5. Operational execution.

Finally, the foundation of everything we did: a focus on operations. Government often prizes policy expertise (and rightly so), but it rarely elevates operations to the same level. We tried to change that. We rarely nailed it on the first try — we had an internal “guinea pig” award for the unlucky souls who went through new processes first — but we learned, iterated, and improved quickly. In this series, we’ll document the nitty-gritty in detail, so future programs can build on what worked and avoid what didn’t.

“Nothing about this feels repeatable”

About a year into our joint tenure at the CHIPS office, Todd was having dinner at Mike’s house after work. We were exhausted and demoralized, fatigued by tough commercial negotiations and seemingly endless procedural thickets. Todd remarked that he had never worked so hard in his life — a strong statement, given his 25 years as a partner and executive at one of the world’s hardest-charging private equity firms.

After a couple of cocktails, Mike asked: “Do you think we are going to get this done?” Todd replied: “I do. But nothing about this feels repeatable.” At the end of the day, we did get it done, and we are very proud of that.

But successful execution in government should be the norm, not the exception. We hope we can help policymakers learn from our experience — what worked and what didn’t, the advantages we had and the obstacles we faced — so that future implementation of industrial investment programs, and government programs more broadly, can be easier, more effective, and better informed.

We chose the name Factory Settings because so much of what we experienced came down to the defaults — the inherited ways the government is wired to operate. Those defaults often make it harder, not easier, to deliver. Every law, every regulation, every process, every oversight mechanism was built for a reason, but together they’ve created a system that wasn’t built for the challenge of 21st-century governance.

This project is about examining those defaults — the “factory settings” of government — and asking how they can be improved. If we want successful execution to become routine rather than exceptional, we can’t just design better programs; we have to update the operating system that runs them.

We invite you to follow along!