Use and Expand Subfederal Power

For the past few years, the US conversation on industrial policy has been dominated by big numbers and big legislations: the Inflation Reduction Act, the CHIPS and Science Act, the Bipartisan Infrastructure Law. These involved standing up new federal offices, hiring new federal employees, and expanding federal funding streams. The last year and a half has highlighted the fragility of that approach, as priorities have abruptly changed and several of these initiatives have been curtailed. The excessive focus on headline numbers and policy maneuvers in Washington, DC, also obfuscates the invisible but critical architecture that intermediates the on-the-ground process of converting ambitious goals into deliverables. Looking ahead to 2029, then, it is worth scoping out strategies that are less vulnerable to the vicissitudes of federal policy by empowering the hidden developmental state that exists outside of DC.

The Existing Stock of Place-Based Strategies and Institutions

Historically, federal responses to regional decline have relied on two imperfect tools: individual transfers and fragmented place-based programs. Means-tested transfers to individuals and households—such as unemployment insurance, disability, and Social Security—are essential safety nets.¹ But they do not rebuild productive capacity or institutional ecosystems. When support arrives only as a check, it places the burden on the recipient to find their way; it does not change the broader economic conditions within which they operate. Without concomitant interventions on the labor demand side, even workforce development strategies that are directed at skilling or reskilling workers fail to move the needle on economic opportunity.²

Over the past few decades, individual transfers have consistently dwarfed place-based transfers.³ The goal of the latter is to upgrade the productive capacities of regions through investments in various domains, including job creation, workforce development, and research and development (R&D).⁴ Yet while they are substantially smaller in magnitude, federal place-based programs have existed in the US for much of the past century. One type of programming includes commissions and agencies like the Appalachian Regional Commission and the Tennessee Valley Authority, both created by the federal government to address regional distress and disparities.

Another type of place-based programming includes federal programs that fund economic development, workforce, and innovation programs in specific regions. During World War II, for instance, the Office of Scientific Research and Development provided funding to a range of existing and new research laboratories, including the MIT Radiation Lab and Caltech’s Jet Propulsion Lab, for applied research. This funding had long-term impacts on regional employment, innovation, and economic indicators.⁵ Postwar technological development advanced this framework, creating federally funded R&D centers across the country. Similarly, the Department of Labor provides funding to community colleges, the Economic Development Administration (EDA) to multicounty economic development districts, and the Small Business Administration to Small Business Development Centers.

The product of all this is what Gordon Hanson, Dani Rodrik, and I have characterized as a “place-based policy supply chain.”⁶ Funding—as loans or grants—flows from federal departments to local actors to implement specific programs in specific places. However, because these intermediary institutions have been put in place through policies over time and by programs spread across federal departments, this supply chain is fragmented. The federal government has typically operated in a siloed manner, which means the burden of coordination is transferred to the local level. Consider, for instance, the extent of fragmentation that rural regions must contend with: 400 economic and community development programs across 13 departments, 10 independent agencies, 50 offices and subagencies, and 14 legislative committees.⁷

While local actors are expected to implement programs and create coherence across these domains, they seldom receive support to build their organizational capabilities. Government programs are often project oriented, funding training initiatives and facilities instead of the broader institutional infrastructure needed to strategize and orchestrate long-term planning. They also take the form of episodic competitions with tight timelines, favoring regions that already have grant-writing and convening capacity. Local philanthropy often steps in to fill the gap, funding nonprofits and public-private organizations. But philanthropic institutions underinvest in rural and distressed regions, which exacerbates regional inequities.⁸

The federal government has also been slow to adapt new local practices. For instance, regional cluster-based approaches became prominent in the 1980s and 1990s, stemming from a combination of Cold War R&D funding and patent-sharing regulations in the 1980s. The cluster approach famously became codified as a field of knowledge by Michael Porter in the mid-1990s. But in federal policy, clusters began to gain prominence only in the 2010s, as a strategy outlined in President Barack Obama’s 2009 Strategy for American Innovation. Since then, this local consortium approach has underpinned a number of programs—including, most notably, several Biden-era programs, like the Build Back Better Regional Challenge, Tech Hubs, and Recompete. But unless this approach is sustained beyond special one-off competitions, it may well be ephemeral.

What a Decentered Approach Would Look Like

Running alongside this story of fragmented federal programming is another, more hopeful one: the emergence of local “systems hubs.” Instead of being intermediaries for federal programming, these institutions act as entrepreneurial orchestrators of regional economic development. In fossil fuel regions, organizations like Coalfield Development in West Virginia and Shaping Our Appalachian Region in Eastern Kentucky have played key orchestration roles, combining training, wraparound services, small business support, and land reuse into locally grounded diversification strategies. Similarly, in Oklahoma, Tulsa Innovation Labs works across a range of partners to synchronize R&D, commercialization, capital, and talent, with the goal of making the region a hub for energy tech start-ups and the clean energy transition. The importance of these hubs stems from their role as embedded but autonomous⁹ interlocutors that can help state and federal agencies learn about local constraints without becoming captured.¹⁰ In places where they are effective, they also play a coordination function, connecting workforce development, infrastructure, finance, and land use.¹¹

If industrial policy is about constructing place-bound capabilities, then building the apparatus to do that work must be understood as a central goal, not an afterthought. The green energy transition will have vastly different impacts across local labor markets. Regions with renewable potential aren’t the same as those that will lose fossil jobs.¹² Fossil regions need deep diversification, workforce development, and often new fiscal arrangements. Even regions that appear like they can make an immaculate transition to renewables are constrained by economic, governance, and political barriers.¹³ Such regions need integrated transition plans linking coal retirement, renewables deployment, workforce transitions, and fiscal reform. And finally, regions with high renewable potential must turn resource potential into investable projects and industrial demand. This requires customized and context-specific action by empowered local actors.¹⁴ One-size-fits-all policies will misfire, especially in regions facing long-term distress and joblessness.

Moving beyond one-off competitions, federal programs should provide multiyear, flexible operating support to systems hubs. These funds should underwrite the convening, planning, and coordination work that rarely fits neatly into project budgets, such as coalition building, employer outreach, community engagement, and local experimentation. Biden-era competition-based programs were a valuable first step, but with limited funding and short time horizons, they should be seen as pilots for further expansion. Several elements of Recompete should become the norm for the EDA’s standard programming: a specific program objective of addressing prime-age employment gaps, strategy and implementation supports, waiving of local match requirements, a broader definition of economic development, and diversity of coalition leaders across applicants.

The federal government needs a more robust institutional mechanism to coordinate both horizontally—across the portfolio of place-bound federal investments from different departments—and vertically, with state and regional programming. This is a role that could be played by the EDA, which along with its regional Economic Development Representatives should serve as the data and strategy clearinghouse for local orchestrators. But to do all this, the EDA needs to be a more empowered federal agency, which means its authorization and appropriations cannot perpetually hang in the balance.

With funding of several initiatives severely curtailed over the last year, future federal funding needs to follow “risk” by appropriately identifying communities that might fall short of realizing their goals, especially in cases where state and philanthropic funding did not make up the deficit. Federal policy often faces a credibility deficit in being appropriately implemented on the ground, especially in energy communities.¹⁵ Investing in this place-based apparatus embedded in communities can serve as a valuable opportunity to build credibility and depoliticize energy transition goals.

The integration of workforce development and economic development is needed across both fossil-dependent and renewable-potential communities.No federal green industrial strategy will be complete without attendant workforce programming, for both distressed communities and those with renewable innovation potential. Workforce development has traditionally operated with different incentives than economic development—prioritizing enrollment instead of broader labor market outcomes. Recent state-level reforms like California’s dual enrollment programs, which enable partnerships between high schools and local community colleges, and Texas’s outcomes-based funding to community colleges could serve as valuable models for federal policy. The CHIPS Incentives Program’s strategy of conditioning subsidies on firms’ investment in local workforce institutions could be applied more broadly. At the same time, community colleges require broader general funds to help build their capacities, particularly to collaborate more effectively with local employers.

Finally, we need to evaluate industrial policies where they land. Building the data infrastructure to answer these questions at the local labor market level is itself a form of industrial policy: It equips both local and federal actors to learn and adjust in real time.¹⁶ We currently track dollars obligated and national aggregates but rarely ask whether the economic structure and livability of Pike County, Kentucky, or Roswell, New Mexico, have changed in ways residents can feel. We also have limited ways to conceptualize and measure local capacity. That disconnect between what federal agencies measure and what communities care about is itself a source of mistrust.

Conclusion

While the US often touts its decentralized governance approach, this ends up being a double-edged sword for local institutions. The decentralization of the practice of economic development has been less a strategy and more a consequence of federal retreat and neglect, starting in the 1970s, after a few decades of strong federal investments in the postwar years. What we need is a new centrifugal arrangement, with strong, consistent, and flexible federal support for sustained local action.

The success of any future incarnation of green industrial policy will not be judged in aggregate emissions graphs alone. It will be evaluated in coal towns deciding whether their children should stay, in rural communities weighing new investments, and in renewable-rich regions considering whether clean-energy projects repeat previous extractive practices or serve as reliable pathways to prosperity. The statecraft required for such policy, therefore, cannot be envisioned through a narrow aperture of federal appropriations and legislative processes. It has to engage with the quotidian exercise of building and coordinating local capabilities, led by local institutions that can braid national goals with local realities.