In 1981, President Ronald Reagan instructed federal agencies that they could issue regulations only after demonstrating that their benefits justified their costs. With modest changes, Reagan’s successors, both Republican and Democratic, have continued to impose this requirement on agencies considering rules designed to protect clean air, food safety, workplace safety and much more.

But a crucial question remains: Can the government’s numbers be trusted, or are they just a lot of hot air?

In Washington, there are two radically different answers to this question. They tell us a lot about what is wrong with current political debates, and about what might be a sensible future path.

Within the Republican Party, not to mention the business community and conservative policy groups, many people share a single view: Government agencies are far too optimistic, even self-serving. Many Republicans think agencies systematically inflate benefits and deflate costs, cooking the numbers to make their rules appear far better than they actually are.

Within the Democratic Party, not to mention progressive organizations and liberal research groups, many intelligent people hold precisely the opposite belief. They insist that agencies systematically underestimate benefits and exaggerate costs. They think that in the real world, rules will impose only a small fraction of the costs that agencies project.

It turns out that neither side is right. Both Republicans and Democrats — and industry and the public-interest groups — have a great deal of confidence in evidence-free dogmas.

Noise Rules

To be sure, both camps can cite apparently persuasive examples. In the Barack Obama administration, some proposed rules have been withdrawn because agencies were convinced that the costs would be considerably higher than they originally expected. For example, the Occupational Safety and Health Administration withdrew a controversial rule designed to protect workers from excessive noise, in part on the ground that public officials appeared to have underestimated the costs.

On the other hand, agencies have sometimes offered inflated cost estimates. The Clean Air Act requires companies to reduce acid rain. The costs of this requirement turned out to be a lot lower than the Environmental Protection Agency anticipated. Because industry learns how to do things more efficiently, innovation often drives anticipated costs way down.

But with respect to regulation in general, the Republican and Democratic narratives have little support. There is no basis for the business community’s claim that agencies systematically inflate benefits and low-ball costs. Nor is there reason to accept the progressive claim that agencies typically exaggerate costs and underestimate benefits.

Consider, for example, a careful study by the economist Winston Harrington, who explored 61 rules for which benefit-cost ratios could be compared before and after the fact. He found no systematic bias. In his account, agencies overestimated both benefits and costs with about equal frequency.

To be more specific: In 16 of Harrington’s 61 cases, the ratios were essentially accurate. In 24 cases, the ratio was better than the agency had anticipated. In 21 cases, the ratio was worse than anticipated. Harrington’s general conclusion is that while both costs and benefits tend to be lower than estimated, no consistent bias is found in estimates of benefit- cost ratios.

Other studies reach the same basic conclusion. In 2005, for example, the Office of Management and Budget provided an overview of many retrospective analyses. The overview did find that agencies tended to overestimate the benefit-cost ratio, and in that sense to be a bit too optimistic about the consequences of their rules (a modest point for the Republican view). But it also found that agencies were slightly more likely to overestimate than to underestimate costs (a modest point for the Democratic view).

Phones on Planes

Of course, the existing studies cover only a very small fraction of existing regulations. We need to know much more.

In 2011, Obama took a major step in that direction, calling for a government-wide review of rules now on the books. Agencies were able to identify more than 500 reform proposals and ultimately to produce billions of dollars in savings. As a recent example of the potential gains from reassessing benefits and costs, consider the Federal Aviation Administration’s widely applauded decision to authorize greater use of tablets, computers and phones on planes.

The Obama administration’s “regulatory lookback,” as it is called, should become a regular feature of American government, not least because it can enlist evidence to cut through the competing dogmas, and help us to make progress in resolving disagreements that are otherwise intractable (as well as mind-numbingly predictable).

Some rules achieve much less than anticipated. Others achieve a lot more. Armed with an understanding of how regulations are working in the real world, we should be in a much better position to decide how to proceed — and whether to streamline, improve or even eliminate existing requirements.

(Cass R. Sunstein, the Robert Walmsley University professor at Harvard Law School, is a Bloomberg View columnist. He is a former administrator of the White House Office of Information and Regulatory Affairs, the co-author of “Nudge” and author of “Conspiracies and Other Dangerous Ideas,” forthcoming in March 2014.)

–Editors: Katy Roberts, Max Berley.