To expedite this effort, Trump’s “DOGE” taskforce employed artificial intelligence (AI) to find deregulation targets and prepare the necessary administrative paperwork. Whether this effort proves effective and desirable depends on how the task is framed and the scope of what is being delegated.

The economic criteria that are appropriate for deregulation parallel the tests that should be applied before enacting regulations in the first place. To be desirable, the benefits of a regulation must exceed its costs, and ideally policies should maximize the spread between those benefits and costs. For new regulations, this assessment comes before any rules have been enacted, while for deregulation efforts the vantage point should consider the cost of ending the regulation versus the ongoing benefits of its demise. Even if the initial costs are significant, the ongoing savings may ultimately exceed the transition costs.

Rather than taking a comprehensive benefit–cost approach, the DOGE guidance adopted a cost-focused strategy, drawing on a variety of different cost analyses. Cost categories include compliance costs, displaced investments, decreased sales, and federal costs. Setting aside issues of possible overlaps across categories, a more fundamental concern is there are no mentions of benefits or possible measures of cost-effectiveness.

The potential targets of reform include the estimated 50 percent of all regulations that are not statutorily mandated or needed by any agency. The task for AI is to identify which regulations can potentially be eliminated and then prepare the necessary paperwork to do so. Omitted from the process is any reference to the preparation of a regulatory impact analysis (RIA) and overcoming the standard hurdles for regulatory actions. Eliminating regulations by fiat will not be successful because the Administrative Procedure Act requires that regulatory policies—whether enacting or repealing a regulation—not be arbitrary and capricious.

Eventually, regulatory agencies and the Office of Management and Budget (OMB) will have to justify regulatory actions, including deregulatory efforts. The payoff may not be the sweeping deregulation effort that is anticipated: Some may regard having to do a RIA as sludge, but it is “good sludge” that will lead to better regulation. While there has been some initial action on the RIA front, it is well behind the promised schedule. The Environmental Protection Agency’s RIA supporting an endangerment finding regarding greenhouse gas emissions underwent judicial review, where it was upheld. The briefer Trump administration RIA seeking to overturn the endangerment finding and the associated climate change regulations surely also will receive judicial scrutiny, and it is not clear how it will play out.

Obstacles

The impetus to undertake sweeping deregulation actions is not new. President Ronald Reagan launched a major deregulatory effort that targeted many expensive regulations. As part of this initiative, the Occupational Safety and Health Administration (OSHA) undertook a retrospective assessment of the controversial cotton dust standard, which limits worker exposure in textile operations. The textile industry challenged the regulation and eventually lost a Supreme Court case, American Textile Manufacturers v. Donovan (1981).

Earlier, regulatory oversight economists in the Carter White House, including Council of Economic Advisers chair Charles Schulze, had opposed the regulation because of a benefit–cost imbalance. Given this, making cotton dust a potential deregulation target seemed very promising. OSHA commissioned a RIA, for which one of us (Viscusi) prepared the benefit assessment for regulatory alternatives. From a benefit–cost standpoint, relaxing the standard would have been clearly desirable, but the more technologically advanced firms that had already complied with the regulation opposed any relaxation, which brought that deregulation initiative to a halt. The Trump administration’s EPA effort to terminate the Energy Star program encountered similar opposition as both appliance manufacturers and retailers opposed the initiative.

Post-Regulation Evaluation

The most straightforward way for regulation to gain public support is for regulators to routinely monitor regulations post-implementation to empirically verify the intended net benefits. This would require making public the data behind the calculations of the costs and benefits, for analysis by researchers both inside and outside the government. A successful example comes from individual income tax policy, where most of the post-implementation outcomes research is done by academics with public data on income taxation’s behavioral and economic well-being consequences. Publicly funded data gathering would be money well-spent in monitoring regulatory policy outcomes.

Takeaways

It is reasonable that the regulatory process should continue after a regulation is approved and implemented, and that part of the process could be considered “good sludge.” Among other things, the regulatory evaluation procedure used should itself be cost-effective. We propose that both regulations and taxes be put on parallel policy paths where post-implementation evaluation is regularly done with publicly available data.

Readings

• Sunstein, Cass R., 2021, Sludge: What Stops Us from Getting Things Done and What to Do about It, MIT Press.
• Viscusi, W. Kip, 1992, Fatal Tradeoffs: Public and Private Responsibilities for Risk, Oxford University Press.