The High Cost of Political Capture

Politics doesn’t just make strange bedfellows; it can also make self-dealing ones. A new study of the economic impact of the $787 billion 2009 American Recovery and Reinvestment Act (ARRA) is trying to tell us about the need to close the gap between intentions and outcomes by moving decision-making and accountability closer to the communities the federal government is trying to assist.

In their recent paper, Joonkyu Choi, Veronika Penciakova, and Felipe Saffie found that politics significantly hindered ARRA’s effectiveness. Their study reveals that firms that contributed to winning political candidates were more likely to receive ARRA grants and these firms subsequently created fewer jobs. Specifically, a 10 percentage point increase in the law’s politically connected spending resulted in a 33 percent drop in the job creation at the state level. In other words, instead of being driven solely by economic need and oriented toward employment outcomes, the distribution of stimulus funds was often skewed by firms’ political ties, leading to less efficient outcomes.

The findings underscore the high cost of political influence in government spending and grant-making. That political influence can so significantly distort the efficient allocation of resources serves as a critical reminder of the importance of transparency, accountability, and evidence-based policy. Such a reminder is all the more relevant in the face of massive new federal spending designed to restore key industries and boost employment and economic growth.

Over the past few years, on a bipartisan basis, the federal government has begun funneling hundreds of billions of dollars into new, highly targeted initiatives intended to boost energy efficiency, reshore manufacturing, and increase employment in areas of the country with below-average labor force participation. Through these employment initiatives, the current administration has emphasized the need for “quality” jobs (those with good wages and benefits) and is seeking to help disadvantaged workers gain skills, overcome barriers, and get to work.

Many of these programs are just now reaching their operational phases. The scale of the investments makes them a proving-ground for whether place-based economic and workforce development programs, if properly aligned and energetically implemented, can restore vibrancy in communities hit by the effects of automation and trade.

The jury will be out on whether these programs are successful for at least several years. There are, however, already some signs for concern. Recent reporting on two key Biden initiatives—broadband deployment and electric car charging stations—points out how plans and good intentions struggle to translate into positive outcomes.

In the case of broadband, despite a massive $43 billion appropriation to expand access to high-speed internet, not a single business or home has yet been connected via the program. As multiple people on X pointed out, that’s enough money to buy everyone who lacks an internet connection a Starlink subscription, thus bypassing the time-consuming process of laying cable. It’s a similar story with electric car chargers. The federally-led effort to establish a national network of charging stations is proceeding at a painfully slow pace with just seven new stations installed out of the 500,000 planned by 2030. As challenging as these initiatives are to implement, they are child’s play compared to the complexities involved in helping low-skilled, disadvantaged individuals succeed in training and employment programs.

As the ARRA study suggests, this tendency to over-promise and under-deliver is aggravated by political capture. Place-based economic development is tough because the processes involved in securing federal grants favor larger organizations that have applied before and have good grant writers and managers. Larger groups also have critical social connections that are essential to the organizational partnerships required for competitive applications.

Meanwhile, small organizations, which often have a better understanding of the neighborhoods, families, and individuals being served, struggle to compete for federal dollars and are not in the mental (or actual) rolodexes of the large organization leaders. As a result, the social capital and fine-grained knowledge needed to address the needs of disadvantaged workers get left out of the equation. Under these conditions—challenging program objectives combined with a lack of “high-touch” partners at the local level—poor performance seems almost predestined or, at the very least, unsurprising.

When it comes to job training and social policy interventions, small and personalized is sometimes beautiful. For instance, one of the most effective job training models on offer is “sector-based” employment programs: small and mid-sized, locally-driven initiatives that help under-skilled workers gain training in higher-salary sectors like information technology or healthcare. Sector-based programs train for both noncognitive (“soft”) skills and technical skills. The returns to such training in placements and wages are impressive. While replicating and scaling them effectively remains difficult, we at least have evidence that when done properly, they work.

In other areas, evidence suggests that emphasizing personal agency may be the fast-track to improving life outcomes for disadvantaged individuals and families. One evaluation found that homeless families who received a voucher for housing did better on almost every measure than those who received vouchers and services. Individual training accounts that provide money for training without intensive support services have also been successful in improving employment and wages. Another study on the Bush administration’s Moving to Opportunity Program, which provided resources to low-income families that wanted to relocate to more stable neighborhoods, showed strong long-term impacts on children. The common element to each of these programs is that they focus on leveraging the personal knowledge and agency of individuals rather than shifting responsibility for planning to a case worker.

While none of these findings are definitive, they do strongly suggest we expand the use of such models—small, local, and focused on personal agency—alongside traditional top-down, centralized approaches to improving social and economic outcomes and then evaluate the relative outcomes. Let the more successful model win.

What would such an approach look like when it comes to place-based economic development funded through mechanisms like ARRA and the more recent Biden administration programs? At the state and local level, it might look a lot like the “revenue sharing” programs of the Nixon administration. These were essentially federal grants to units of local government that came with few if any strings attached. Revenue sharing created pools of resources that could be used at the discretion of local governments to fund projects that boosted local well-being. Such grants could be spent on infrastructure, parks, and economic development initiatives the local jurisdiction believed would provide the greatest benefit to their communities. Planning and implementation—and, crucially, accountability—would also be in the hands of local government. This wouldn’t fundamentally alter the conditions that cause political capture, but at least local voters would be able to ask questions and demand answers when it arose. Of course, there’s one thing revenue sharing arrangements cannot do that categorical grants can: provide photo-ops and credit-taking opportunities for the elected officials who vote on individual pieces of legislation—even when they actually voted against them. Unfortunately, federal spending on economic development is an act of political capture all the way down. It is time to get out of the trap.