The Chip Stops Here: A Call for a Truman Committee for CHIPS Act Oversight

When recalling Harry S. Truman’s legacy in World War II, we tend to think more about the end of the war than the beginning. But in 1940, Truman began receiving letters about a new Army fort being constructed in Pulaski County, in the Ozark hills of Missouri, the state he represented in the U.S. Senate. Something just “wasn’t right,” the letters insisted, and so he set out on a road trip to have a look for himself.¹ What he found there sickened him. In Truman’s own words, he saw “There were buildings being built . . . and they were cost­ing three to four times what they should have . . . material of all kind, that was out in the snow and the rain, getting ruined . . . [and] hundreds of men, just standing around and collecting their pay, doing nothing.” After that, he started driving. He drove all over the country, simply walking into the sites at which army camps were being constructed, and he saw the same thing everywhere he went: “millions of dollars were being wasted.”²

While Harry Truman would be far less likely to be able to drive up and walk into the semiconductor fabs that will be built with the incen­tives created by the chips and Science Act, the risk still stands. The passage of this bill prompted a dramatic expansion of spending on science, deep-tech industry, and regional commercialization policies. Arguably the legislative crown jewel of what might eventually become known as the “Bidenomics” Congress, it nevertheless benefited from relatively broad, bipartisan support—a situation not unlike the dramatic expansion of military spending in 1940–41. The bill’s remarkable success is, perhaps, a recognition of the importance of our moment in history. We sit either at the zenith of U.S. global power and influence, after which all declines, or the dawn of a new age of more inclusive and widespread American growth. The goal of chips is to ensure the latter outcome, by accelerating a new golden age of American industry through investment in science, entrepreneurship, and more equitable distribution of resources.

Key to this tipping point is the question of whether chips will be administered effectively. For a spending package of epic proportions, the chips package gave almost no thought to oversight. The bill itself directs the Government Accountability Office (GAO) to conduct a number of program-level evaluations, specific funding reviews, and reviews to ensure that intellectual property generated with federal funding doesn’t end up in the hands of the Chinese Communist Party. Outside of those limited reports, the bill is silent when it comes to accountability, over­sight, and evaluation.

This means chips could easily become either an instrument of reckless spending or a partisan cudgel. But alternately, it could become the point at which bipartisanship wins the day—with Congress, agencies, and industry banding together to choose their better angels (showing an admittedly uncharacteristic level of restraint), providing real, substantive oversight that aims at continuous improvement and holds to the lofty goals of the chips Act that benefit us all.

Independent, external oversight and the accountability it provides need not be feared by the Biden administration—in fact, it might even be an asset. Historical examples such as the Truman Committee—and others—can provide blueprints for how to get it right at this critical inflection point for the country.

Making Sure Oversight Isn’t an Oversight

Two kinds of governance of the chips expenditures are necessary to ensure the bill reaches its full potential: operational oversight and impact evaluation. While these two capabilities are often lumped together in terms of accountability (such as under the umbrella of the GAO, or agency-level inspectors general), they answer very different questions, and thus should be considered two separate domains of focus. Operational oversight answers the question: is the relevant program or expenditure managed and executed well? (“Well managed and executed,” of course, covers a range of descriptors, from lawful and ethical to efficient, consistent, and well-communicated.) Impact evaluation, meanwhile, an­swers the questions: Is the program or expenditure having the impact it was intended to have? How do we know if it is on the right track or the wrong track? Both oversight and evaluation of the chips programs and incentives will be complex. In addition, there may be strategic questions about the decisions that agency leaders or the White House make that should be examined or at least subject to debate in a public forum.

Thus far, the White House and federal agencies have been extremely forth­coming about their planned approach, issuing an executive order to clarify implementation priorities, creating a cross-agency implementation committee, creating and staffing a chips program office, and creating an online hub for chips-related information. On that point, the administration should be commended, and there is no evidence to suggest that this posture of open communication will change in the future. It must be noted, however, that every aspect of oversight that the Biden administration has put in place—tasking OMB with cross-agency oversight and underlining the roles of agency inspectors general—is entirely internal.

While the administration’s plan includes an appropriate number of advisory groups and significant industry input, there is no requirement that the administration take any of that advice, nor is there opportunity for more public expert debate outside of those closed circles. The administration should revise this plan and dismiss any thought that adjusting it would be some admission of weakness or poor leadership. Instead, appropriate revisions would send an important signal that the administration plans to take every reasonable measure to avoid even the appearance of conflict of interest. Acceptance of (or better yet advocacy for) independent oversight functions external to the administration is the best possible signal that one has nothing to hide. And realistically, the administration can choose external oversight now, and have a hand in the creation of a reasonable and serious panel. Or it can leave Congress to its own devices, likely resulting in a less balanced, less mature, and less practical process.

Operational Oversight: Is Work Done Well and Honestly?

In terms of operational oversight, the bill creates massive incentives and programs that will be operationalized by a wide swath of federal agencies. These incentives are the first of their kind and, as such, present a number of challenges, from the creation of diligence standards to ensuring that they are properly and consistently applied. The programs are also new, and in some cases, confusingly similar. And with so many agencies in the mix, how can taxpayers ensure that a few well-informed and sophisticated consumers of government funding aren’t running the tables at every agency, pitching, essentially, the same projects in slightly different contexts? This requires a proactive approach to oversight at a cross-agency level, which incorporates the following domains:

Fraud, waste, abuse, and cross-agency audit. Are existing ethics standards and accounting practices being followed? Is there evidence of outright fraud, abuse, corruption, or self-dealing? Is there evidence of redundant or inefficient spending across agencies?

Cross-agency traffic management, lane-keeping, and communication. Are agencies making good faith efforts to work together to ensure those best equipped to manage programs are involved in program operations (e.g., is the SBA involved in the management of all direct lending programs)? Do programs targeted to similar audiences (such as place-based innovation programs) have end-user accessibility in mind (or are the deadlines to major programs stacked ineffectively)? How is the ability to differentiate across all of these programs communicated, as a whole-of-government offering, to those in the field who would apply?

Operational efficiency. Are diligence standards created equitably, communicated widely, and applied uniformly? Are programs meeting the timelines that they set forth to the public (or otherwise reasonable timelines) for review, execution of awards, and announcement?

Clear guidance around novel processes. Some provisions created by the Act extend beyond even the bounds of GAAP accounting principles—there is simply no accounting standard for how to recognize government support given directly to a for-profit business. Is it revenue? Is it a gift? Can a business be held to account for improper use of government funds when no such guidance exists?

Equitable access. Are criteria and processes constructed with a reasonable degree of care for people, companies, and organizations that have experienced systemic disinvestment? Are those people, companies, and organizations without access to generational wealth held to the same financial standards as others? Are small businesses considered for incentives alongside large companies? Are the influence and sophistication associated with incumbency mitigated to create a more level playing field?

Impact Evaluation: Is the Work Meeting Our Goals?

In terms of impact evaluation, the sheer size of the bill (over one thousand pages) and the complexity of its “theory of change” present a challenge. Inevitably, this is the case when a broad coalition informs the “why” behind a particular effort or action. As a result, the chief goal of the chips Act might differ, depending on whom you ask.

Talk to House Republicans and they will likely talk about the need to counter China’s global influence. Ask members of the Biden admin­istration, and they will likely talk about goals of bringing manufacturing back to the United States or offering more equitable economic opportunity. Ask NSF staffers, and they will discuss the reinvigoration of our research in critical technology areas. Ask economic developers or eco­system builders, and they’ll talk about place-based innovation funding and the need to support innovation everywhere. All of these things can be true at the same time. But when in conflict, which of these goals is the most important? And how will we know if the decisions we make in the interest of national security are negatively impacting our efforts to improve racial or geographic equity? These questions are at the heart of developing theories of change and impact evaluation strategies.

In addition, it is worth noting that the program evaluations completed by the federal government today, while somewhat common, are typically conducted by the agencies themselves on a timeline that is more likely to be informed by reauthorization or political change than by the cadence of the program itself. On other occasions, Congress might request that the GAO undertake a program evaluation by either authorizing it as part of a legislative effort (as it does sporadically throughout the chips Act) or by making a specific request. Both of those circumstances require Congress becoming focused on a specific aspect of oversight when there is no actual requirement that lawmakers do so.

Ongoing evaluation strategies, developed alongside the programs themselves, are a better option. They provide the best view into what goals each program was intended to achieve in the first place and whether or not they are on the right track. The alternative is the typical federal approach to program management, which leaves the door open to political manipulation and ex post facto changes to assumptions about the program’s goals. Clearly, the best approach is to set and stick to evaluation strategies that last from the beginning of the program to the end, and to treat impact assessment as a learning exercise, not a political football.

But the complexity doesn’t end there. There are a number of dimensions on which chips work might be considered a success or a failure. Part of developing an evaluation strategy for this work will be anticipating the questions that may not arise until five or ten years in the future, upon each of these dimensions. Failing to create a comprehensive evaluation plan (one that incorporates measures to judge success on each of these potential outcomes) is not only methodologically short-sighted, it opens the door for politically motivated attacks on a program.

Surely, taxpayers today and in the future deserve to know whether or not this new approach to industrial strategy has been successful. An exhaustive evaluation strategy should explore the following dimen­sions of impact:

But-for impact. Today, the primary judgment of the success or failure of economic development incentives stems from the “but-for” argument. Would this development, relocation, or expansion have happened without (or but for) the inclusion of the government incentive in ques­tion? This justification is used often at the state and local level to justify the relocation and retention incentives at the center of the twentieth-century approach to economic development.

While conceptually sound, the practice of this approach is often flawed; when companies on the receiving end of these incentives are asked this ques­tion, their answers typically go unexamined. Clearly, chips incen­tives should be assessed rigorously on a but-for basis to ensure that they amount to real job creation and new economic activi­ty—not corporate graft for plants and investments that otherwise would have been made without intervention (or with a smaller price tag).

Community benefit agreement transparency. One tool currently being used to govern chips incentives is community benefit agreements (CBAs), which are negotiated between the parties expanding development and the communities in which they are located. Oftentimes, these agreements include commitments around affordable housing, job quali­ty, or workforce development. While CBAs provide a useful platform for communities and businesses to negotiate what local benefits are most needed to support massive investments, they have a few drawbacks.

First and most importantly, it has been deemed unconstitutional for government at the local, state, or federal level to be party to CBAs. This means that the government cannot impose penalties or incentive claw­backs related to CBA violation. Second, the only parties that can enforce CBAs are those that are party to the agreement—that means that com­munities must depend on grassroots groups, which are not typically well‑equipped to take on legal battles with a Fortune 50 company to enforce these agreements. In reality, this likely means that many will never be enforced.

That doesn’t have to be the end of the process, however. Government’s role at the bully pulpit can also be important—and it need not be party to a CBA to reap the benefits of a community‑informed approach to incentive clawbacks. All it needs to do is ensure that a CBA exists as a condition of closing its incentive deals, require that companies complete regular reporting aligned with the CBA, and then transparently share that information. Enforcement, then, becomes a matter of capacity-build­ing for local parties to CBAs and media attention—much easier problems to solve.

Impact evaluation aligned with congressional intent. When it comes to program evaluations, Congress has an asset in the GAO. Founded in 1921, the GAO has been bringing accountability to government spend­ing for over a hundred years. Throughout that period, it has grown from an accounting organization to become an auditor, an evaluator, and an investigator operating at the direction of Congress.

Today, the chips Act calls for a few very limited evaluations to be conducted by the GAO: study the impact of biometric systems on his­torically marginalized communities (Section 10226); review NIST se­curity practices to guard against foreign interference (Section 10247); evaluate the capacity of the National Center for Science and Engineering Statis­tics (NCSES) to meet current and future needs for data on the STEM workforce (Section 10314); conduct an assessment of precision agriculture technologies (Section 10361); report to Congress an inven­tory of federal research agency competitive funding programs targeted to minority-serving institutions (MSIs), assess Federal research agency outreach to MSIs, and make recommendations for steps agencies can take to increase the participation and competitiveness of MSIs in such programs (Section 10521); conduct a study on federal research funding made available to foreign entities of concern (Section 10635); and evaluate potential government steps to avoid semiconductor shortages, to describe steps taken to hire individuals from disadvantaged populations into the semiconductor workforce, and to detail how funded projects support the needs of critical infrastructure industries (Section 105).

Even if the GAO were to complete each of these tasks perfectly, Congress would still be limited in its ability to understand whether or not the package of incentives and programs funded by chips is meeting its goal of improving American competitiveness and onshoring critical supply chains. To answer this question, one or more members of Congress would have to request such an evaluation or report from the GAO—and for the report to be prioritized, that member likely would have to be relatively senior. Again, we can leave the process of understanding whether the congressional intent of the chips Act has been met to chance, or we can be proactive in studying why it has or hasn’t been achieved.

Impact evaluation aligned with presidential intent. Congress and the administration supported the chips effort for slightly different reasons, and these differing assumptions are borne out in some of the early conflicts around chips implementation. For instance, when the Biden administration created a requirement that companies seeking more than $150 million in incentives need to provide a plan for affordable, high quality childcare for workers, critics crowed. But the fact that the Biden administration included such a provision in what was previously dis­cussed as a national security bill should not have been a surprise to observers. The administration was clear that among its intentions for the bill were goals like creating more equitable access to STEM innovation and investment in regional development and good jobs.

These three aspects alone—equitable access to innovation, creation of good jobs, and more regionally inclusive development strategies—represent three additional and distinct theories of change. Achievement against each of these yardsticks, as well as the priorities set forth by Congress, must be measured separately, because it is possible to succeed in one but not another. Sometimes, these competing goals might even be in tension with each other.

Without the GSA at his disposal, the president has taken steps to integrate and streamline chips implementation across agencies. He has put in place a well-qualified and respected chips czar in Ronnie Chat­terjee. He has created a chips steering council that includes senior members of the administration and incorporates the voice of industry through an Industrial Advisory Committee. He has tasked OMB with implementation oversight. He can also rely on agency-level protections and experts, like agency inspectors general and the chips Office created at the Department of Commerce. Finally, at the agency level, program managers can budget for and undertake program-level evaluation.

And yet, every one of these solutions is internal to the Biden administration. This arrangement simply does not go far enough to avoid the appearance of conflict of interest in how incentives are allocat­ed and programs are managed. This argument should not be considered a disparagement of the experts and long-serving public servants who hold these roles today—in fact many of them are experts and advisers of the highest order. But even good, smart, and well-meaning people some­times make mistakes and often make choices between competing priorities and values. Certainly, given that these choices involve the commitment of taxpayer dollars, external oversight and public debate are not only prudent, but must be considered an important and natural part of the process.

As agencies work to launch and grow programs authorized by chips, another important question is how these programs, many of which target similar applicant audiences, should work together to achieve larger goals as parts of a whole. In this author’s experience, these questions are particularly important in the context of place-based inno­vation ecosystems and cluster development, which many new programs are designed to address. This slate of programs includes those managed by the Department of Commerce’s Economic Development Administration (EDA), like Tech Hubs and Recompetes, the National Science Foundation’s Technology, Innovation, and Partnerships Directorate (nsf tip), like Regional Innovation Engines and Accelerating Research Translation, and a bevy of place-based innovation programs at the Department of Energy.

How should these programs interact? Should they work with existing programs to form a pipeline that serves communities over time, or should they concentrate on a smaller number of regions, working together to deliver different kinds of support? Should program timelines be sequenced or stacked in a certain way? How should they integrate with and leverage long-standing SBA programs, despite the fact that the SBA was not included in the chips Act? Which programs are best aligned with the administration’s goals of equity, inclusive wealth crea­tion, and good jobs? Which are best aligned with congressional desires to make sure China doesn’t benefit from U.S. investments in R&D? Should every program speak to both? All of these questions have yet to be answered.

Developing one overarching effort to clarify and categorize the many theories of change at play in chips is critical to ensuring that each of its programs can claim success in the coming years. Just as important is a shared interagency effort to clearly define for “customers” (applicants in communities across the country) which programs are best suited to their needs. Some elements of this might be achieved through internal means—an interagency working group, a formal interagency agreement (like the one signed by the EDA and nsf tip in July), or just a directive to coordinate. But yet again, external oversight can ensure that these activities are prioritized and their outcomes are communicated clearly.

Oversight in a Changing Business Environment

As the Truman Commission found, with respect to military contractor waste, “There was no ‘unpatriotic motive’ involved here, concluded the committee in its report. It was only ‘big business playing the game according to the rules,’ with a heavy price, ‘to be borne by the entire nation.’”³

Today, an argument can be made that oversight of the chips Act is more important than ever before, primarily due to the evolution of business ethics and the unique structural makeup of markets at the furthest fringes of critical technology—like semiconductor fabrication, synthetic biology, and material science. Aggressive oversight is necessary first and foremost because access to real-time business insights is asymmetrical; companies have information about their needs, goals, profit margins, and more that exceeds that which the government can access. Public companies have new information and business projections that might not be included in tax filings and annual reports. Privately held companies are famously difficult to track and analyze because they are not required to issue annual reports.

In addition, the broad use of customized incentives funded in chips requires that government actors make judgments about the size, scope, and conditions of their use. To do so in a way that is truly in line with a “but-for” principle requires access to that asymmetrical information. At their best, tax incentives like those included in chips should be carefully calibrated to fill financial gaps and make doing business in the United States a palatable option for U.S. companies. In the reality of the post-neoliberal era, that means that these incentives must make doing busi­ness in the United States the most profitable option. At the same time, incentives funded by the taxpayers should be carefully diligenced, and deals should be structured to drive the desired investment outcome with the most efficient use of government resources. Ethical incentives must therefore be a product of transparent information sharing between companies and government.

But again, the modern, shareholder-primacy-based framework of fiduciary duty implies that all companies will seek government funding, whether or not they actually need the money. The best-resourced among them will be able to make the best arguments and hire the best lobbyists. As Scott Lincicome, director of general economics at the libertarian Cato Institute, has said directly, “This is a low-risk, high-reward maneuver. . . . It would almost be corporate malpractice to not go after that cash.”

Yet not even the Cato Institute can argue against the bill’s intention—choosing instead to launch criticism about the potential for lobbying and the abuses of political largesse. Such criticisms are not wrong—certainly the power, influence, and sophistication of incumbency make it easier for established companies to take advantage of the complex require­ments, bidding processes, and relational cues needed to navigate the process of securing chips incentives. That these incentives generally disadvantage small and start-up businesses is clear. They cannot afford to hire the lobbyists, attorneys, and compliance staff needed to navigate these programs.

Taken together, this means that companies seeking incentives have competing interests when it comes to business ethics. Being transparent and forthcoming with the government when seeking an incentive is ethical. But what about when there is no legal compulsion to do so, especially when fiduciary duties to shareholders might discourage such transparency? Clearly the temptation to be less-than-forthcoming exists, and for that reason, multiple layers of oversight, including both internal and external factors, can be an important tool. Such oversight will not only directly ensure that taxpayer dollars are used as wisely as possible, but the threat of detailed examination also has the potential to indirectly motivate corporate sector transparency in a way that other factors cannot. After all, no semiconductor company wants to become the next Solyndra.

The only thing that is certain about the chips incentives is that both good and bad actors will seek them. It will be the responsibility of those managing incentive programs, as well as those that provide oversight, to distinguish between those companies with the genuine ability to meet America’s needs, and those that are wolves, wrapped in sheep’s clothing, waving an American flag. In that endeavor, certainly, proactive oversight is far more valuable to taxpayers than congressional grandstanding after the fact.

Those Who Cannot Remember the Past Are Condemned to Repeat It

Reflecting on his time as chairman of the Senate Special Committee to Investigate the National Defense Program, Truman said, “I knew that [following World War I] . . . there had been a hundred and sixteen investigating committees after the fact, and I felt that one committee before the fact would prevent a lot of waste and maybe save some lives.”⁴ Truman’s committee was formed at a time when government spending in the United States had exploded—defense expenditures were budgeted at $13 billion in 1941 alone, and that was prior to the passage of the $7 billion Lend-Lease Act. This $20 billion price tag would be akin to $420 billion today.⁵ While cynics might chuckle at the comparison between that number and our real fiscal 2024 defense spending (projected to be more than $750 billion), it represented a dramatic expansion in terms of GDP—from 1.7 percent of GDP in 1940 to 5.1 percent in 1941, and much greater than the 2.7 percent of GDP that is projected for 2024.

In forming his committee, Truman, ever a great student of history, looked to examples further afield for a lesson in “what not to do.” He spent long nights at the Library of Congress, reading fragile copies of testimony before the Civil War–era Committee on the Conduct of War. The negative example of this obstructionist body informed Truman’s thinking on how to structure his own. He later reflected, “Its members wanted to cause trouble for Lincoln, which they succeeded in doing. [Robert E.] Lee later said that it was worth at least two divisions to the Confederate cause, and I think he was right. Except I wouldn’t doubt that the committee did more harm than that.”

Through his research, Truman distilled some of the best practices of oversight committees from historical records. Upon doing so, he set out with the unanimous mandate of the Senate to create a bipartisan com­mittee. The Truman Committee lasted from 1941 to 1948, and throughout that tenure, its goals were to ensure operational efficiency between a mess of federal agencies created to helm the war effort, eradicate fraud, waste, and abuse, and ensure the inclusion of small businesses. Members of the committee were selected based on their reputations for willingness to work across the aisle and their work ethic.⁶ In addition, the culture of the committee reinforced expectations of bipartisan cooperation—every report that it released was unanimously approved, witnesses were treated respectfully and fairly, and hearings were open to all senators.⁷ Finally, in order to allow for prepared responses rather than “gotcha” moments, the committee sent advance copies of reports to the president and all named parties.

Over the course of seven years, many of the committee’s most important tips came from the letters of everyday people—workers and line managers in plants, wives who overheard their discussions, and small business owners who resented missing out on opportunities to supply lumber or construction workers for local bases. As Truman worked himself to exhaustion, traveling the country to investigate claims, he found the voices of everyday citizens to be his most reliable sources. Truman realized that most taxpayers can see when their money is being misused—and considered them to be among his most important assets. Additionally, Truman staffed the committee with the day’s top prosecutors and auditors, and their first task was reviewing and responding to every single letter that they received.

That the Truman Committee delivered cost savings and saved lives is indisputable. Cost savings were safely estimated at “billions” of dollars, and reporters estimated that the committee’s work saved the lives of a few thousand American soldiers.⁸ In addition, the committee put the fear of God into those receiving government contracts, as well as the bureaucrats and military men signing them. It helped smooth and integrate the mobilization effort at a time when President Roosevelt’s “alphabet soup” of agencies was absorbed in turf battles and duplicative efforts.⁹

The Chip Stops Here

Certainly, the Truman Committee is just one of many examples that architects of chips oversight could turn to for inspiration. Over the years, the Church Committee assembled to investigate intelligence abuses, the Warren Commission created to investigate the assassination of President Kennedy, the Iran-Contra Committee, and Senator Eliza­beth Warren’s TARP Oversight Panel each add to our understanding of which tactics and strategies provide the most effective oversight of gov­ernment efforts—as well as what doesn’t work. From these examples, we can see the wisdom of engaging industry partners to ensure that the proper expertise is around the table, and including panelists from out­side the legislative branch as a means of ensuring that partisanship does not eclipse the committee’s mission.

Despite the differences between 1941 and 2023, the experience and lessons of the Truman Committee remain applicable today. We seek to build up our own industrial efforts, grounded in a need to defend the sphere of American commercial and cultural interests amid a shifting landscape of global powers. We want to make smart investments in the industries that are critical to our national security, and to ensure that these investments are held to high standards of integrity and effectiveness. We want, through these efforts, not only to grow American influ­ence and commerce, but to promote American values like freedom, democracy, and self-determination.

To meet this challenge, and to protect the American people from our geopolitical rivals, today’s leaders should follow Truman’s example in advocating for independent oversight of the chips Act. Here are some key principles from the Truman Committee that should be adopted: (1) Better one committee in search of the big picture than many efforts missing pieces of the puzzle. (2) External oversight is critical because it ensures that political convenience does not impact investigation or prioritization, and because it supports the administration in avoiding even the appearance of conflicts of interest. Friendly oversight is flexible; oppositional oversight can only obstruct. (3) Bipartisanship must be taken seriously at every step, beginning with selection. (4) Hiring the very best investigational staff ensures that the panel’s return will exceed the expense. (5) Everyday taxpayers, when asked, can become an army of free monitors for waste, abuse, and insider deals.

Admittedly, our Congress is more bitterly partisan and divided today, and centrist views like Truman’s have little space in either party. We are quick to decry and dismiss those whose views differ from our own. We readily disrespect and bully those we have never met, and yet we interact less with those we know, reporting increasing levels of loneliness and isolation. Is it naïve to expect that bipartisanship and a shared purpose might win out, as they did in 1941, holding America’s industrial giants to high standards of ethics and effectiveness? Perhaps.

Yet even if the world is not at war, the stakes are high. What is on the line is not just our commercial competitiveness among allies, but our position of global leadership in critical defense technologies vis-à-vis geopolitical rivals. And the nature of these dual-use technologies makes them central to our daily lives. While the risks of failing to properly police and produce war materials in 1941 fell mainly on the armed services, today they involve countless civilian and commercial applications. The risk is not just engine failure on a test flight, but failure of the automated lane-keeping system on scores of family minivans. Strong, bipartisan public oversight is essential to success, and should be wel­comed. As Truman might say today, if you can’t stand the heat of reasonable scrutiny, perhaps you should get out of the fab.

This article originally appeared in American Affairs Volume VII, Number 3 (Fall 2023): 31–44.

Notes