In the first few months of Donald Trump’s administration, Congress has passed and the president has signed a record 14 Congressional Review Act (CRA) resolutions of disapproval, withdrawing rules implemented by Barack Obama’s administration in its final months in office. These CRA actions will save a total of $1.1 billion in annual compliance costs. In addition, Trump issued an executive order calling for the repeal or amendment of two existing rules for each new rule an agency implements.

Some advocates of limited government have complained that, so far, the administration’s regulatory accomplishments have largely been limited to those CRA votes. They hope Congress will undertake expansive deregulation of financial services, health care, and energy in the coming months.

But one feature of Trump’s regulatory policy is being overlooked: the decline in the issuance of new rules. There has been a massive slowdown in regulatory output, which Trump’s supporters should take as proof that his administration is serious about regulatory reform.

We examined the data and found ample evidence to support the perspective that rulemaking has slowed dramatically since Inauguration Day. Through its first five months, the Trump administration has imposed just 1.9% of the average number of rulemakings for that same length of time since 1994.

Methodology / To carry out our analysis, we compiled data from the Office of Information and Regulatory Affairs (OIRA) from 1994 to present. We used 1994 as our start year because it was the first full year after President Bill Clinton issued Executive Order 12,866, which mandated that all major regulations—that is, those that have a compliance cost of $100 million or more—undergo cost-benefit analysis. The order dramatically narrowed and focused the scope of OIRA’s scrutiny. In 1992 OIRA reviewed 2,285 rulemakings, but by 1994 the number of reviews declined to 831.

We also obtained data on the total rulemakings as well as the number of economically significant rulemakings OIRA reviewed, and the number of rules that OIRA rejected in some ways or the issuing agency retracted because of objections from other agencies. This information allowed us to calculate the percentage of economically significant rules reviewed, the percentage of rules withdrawn, the number of approved rules, and the number of approved economically significant rules.

Finally, with data from the American Action Forum (publically available at www​.regrodeo​.com), we calculated the final regulatory costs for executive branch agencies (excluding independent agencies like the Securities and Exchange Commission or the Federal Communications Commission, which are exempt from EO 12,866) from 2005 to present, years in which data were available. With this information, we could observe a few broad trends in OIRA activity for the past 23 years as well as regulatory cost data for the past 12.

Historic slowdown / So far, the Trump administration’s regulatory output is historically low, we find. The average number of OIRA-reviewed rulemakings over a five-month period during the 23 years we examined was 235, but there have been just 53 reviews in 2017, which is just 22% of the historical average.

When it comes to withdrawn rulemakings, previous administrations averaged about 19 over their first five months, but the Trump administration withdrew 27. The withdrawn measures were typically legacy rules of the previous administration, and withdrawing rules is common during presidential transitions regardless of party. For example, during the period studied, there were 130 withdrawn rulemakings in 2001, the first year of the George W. Bush administration, and 37 such withdrawals in 2009, the first year of the Obama administration.

The Trump administration has focused more of its attention on economically significant measures. Some 34% of OIRA’s work under this administration has involved economically significant rulemakings, as compared to the Clinton and Bush administrations, where economically significant rules comprised less than half that proportion. However, at this early stage of the Trump administration, we are not yet ready to pronounce this a definite change of emphasis; given its two-in/one-out edict, the administration may simply be focusing on larger rules.

During the Obama administration, the proportion of rules that were economically significant rose to nearly 23%, and climbed to a record 30% in its last year. That administration published 118 major rules in 2016, 18% more than in any other year, but the total number of rulemakings was not that much higher than in Obama’s previous years of governing.

On average, cabinet-level agencies finalized rules costing $30.1 billion for the first five months of the year. The Trump total was just $593 million.

During presidential transitions, the number of withdrawn rulemakings can cloud the OIRA review data because so many rules are withdrawn when a new president enters. But the datum that is most important to business—and, we argue, economic growth—is the number of new rules approved. The Trump administration has been positively stingy by this measure, approving just 26 measures during the period we examined, or 12% of the average for that length of time. By contrast, the historical average is 216 measures.

The Trump administration has approved 12 economically significant measures, 32% of the historical average. Even that small number overstates things; three of the 12 were formal delays of previous rules, and one reduced health care compliance costs. Removing those four leaves just eight significant new regulations approved, or 22% of the historical average.

Although counting up the sheer number of rules reviewed is instructive, it does not fully capture the extent of the burden of rulemaking because it fails to capture the magnitude of each rule. For instance, the Clean Power Plan imposed carbon emissions standards on power plants and was far more economically significant than the routine migratory bird hunting regulations that are approved annually.

Using data from the American Action Forum on regulatory cost estimates, we can determine the regulatory burden each year from 2005 to present. On average, the cabinet-level agencies together finalized rules totaling $30.1 billion in net present value costs in the first five months of each year from 2005 through 2017. The high was in 2010, when the Obama administration published rules with $97 billion in costs between January and June.

The Trump administration figure for that period is just $593 million, or 1.9% of the historical average, and $312 million of that figure represents routine airworthiness directives from the Federal Aviation Administration. In addition, a legacy rule from the Obama administration’s Department of Labor that was finally published on January 23rd imposed $345 million in costs—a cost that goes under the Trump administration, but the rule is hardly a Trump product. Excluding those rules means the current administration has actually cut total regulatory costs by $64 million, and that figure doesn’t include the cost savings from the CRA votes.

Can Trump keep it up? / While publishing few new regulations may be considered a good start for those who believe the regulatory state has grown beyond its usefulness, it is simultaneously not enough and also unsustainable. Just saying “no” to regulations for four or eight years—good and bad alike—will be politically difficult to maintain, even with a Republican Congress. Also, some believe that stopping new rules is only a first step to repealing costly, ineffective rules already in place. The Trump administration has made only tentative steps in that direction thus far, no doubt in part because of a lack of personnel in place.

The Trump administration will need help from Congress if it wants to repeal portions of the Affordable Care Act, the Dodd–Frank financial legislation, and many of the Obama administration’s energy rules. The public got its first glimpse of the broader Trump regulatory agenda when it published its Unified Agenda of Regulatory and Deregulatory Actions. Unlike previous agendas, this one focused more on deregulation.

It is easy to stop issuing new regulations, at least for a short period of time. But pulling back existing rules requires legislative and administrative proficiency that has yet to be demonstrated. Based on public information, not every agency has complied with EO 13,777, issued by Trump in February, which establishes regulatory reform officers and task forces within each agency. The administration cannot deregulate if it does not have political appointees in place to identify regulations for repeal.

What’s more, the courts will likely stand in the way, or at least delay, many of the controversial deregulatory moves of the Trump administration. While many small-government supporters may agitate for the repeal of the Clean Power Plan or the Department of Labor’s Fiduciary Rule, the reality is that the federal courts will likely determine their fate. Occasionally, courts have stood in the way of onerous new rules, but there is little doubt unions and environmental groups will spend most of the next three years in court litigating every aspect of deregulation. If President Trump’s regulatory vision is to succeed, he will have to rely on deference from the judicial branch, otherwise the on-paper savings of regulation will dissipate with each new court opinion.

Another point that we have made previously in these pages is that rules that have already been implemented for some time, and that affected firms have already spent money and resources in order to ensure compliance, may not save anyone all that much money if they were to be repealed. Compliance costs typically increase fixed costs more than marginal, ongoing costs.

For instance, many coal-fired plants have already been shuttered in recent years, both in anticipation of more stringent emissions regulations and also because the sustained low price of natural gas has made much coal-fueled generation uneconomical. Repealing the rule will not reduce compliance costs for coal plants that have already spent money to adhere to the new regulations, and many of the shuttered plants have been dismantled. Those that are still in existence may remain cost-ineffective even with the rule repealed.

A successful deregulatory agenda will need to focus on the repeal of regulations that truly hold the promise of generating substantial cost savings compared to the loss of whatever regulatory gains are sacrificed.

Nothing is not enough / If libertarians were told in 2016 that the following year’s regulatory output would be a tiny fraction of the historical average, they probably would have cheered. However, those who agitate for less regulation ultimately want to achieve more freedom, greater labor force participation, and fewer barriers to market entry, each of which should help boost economic growth. To truly boost economic freedom, it will take more than a regulatory slowdown and a few CRA resolutions of disapproval.

This has happened before. The deregulation of the late 1970s and early 1980s helped to transform the economy and was accomplished on a bipartisan basis. What’s more, it was congressional legislation that spurred the deregulatory agenda, with the encouragement and assistance of Jimmy Carter’s administration.

Whether the same sort of success can be accomplished through executive action alone is a question for the courts. But it is unclear whether Congress has the votes or the White House has the consensus-building talents for the kind of durable, bipartisan efforts that could replicate the successes achieved by Presidents Carter and Ronald Reagan.