A Regulatory Budget Would Make Washington Count the Cost of Its Rules

Every family in America has a budget, and the family that ignores it learns the lesson in short order. Every business has one. Most states must balance theirs. The federal government has one too, for spending: Congress appropriates a number, and the agencies cannot spend more.

But when Washington wants to make you spend your own money, through a rule that says you must install this, test for that or file the other, there is no number at all. The agency decides, the agency estimates, and the agency moves on. The money is real and it comes out of your pocket or your employer's, but it never appears in a ledger anyone in Congress votes on. That gap is what the idea of a regulatory budget is meant to close.

What it is, in plain terms

Here is the definition, so we can argue from fixed ground. A regulatory budget is a limit, set each year, on the total cost that federal agencies may impose on the private sector and on state and local governments through new rules. Each agency would receive an allowance. To issue a rule that costs money to comply with, the agency would have to fit it under the allowance, either by staying within the cap or by repealing or modifying older rules to free up room.

The analogy to an ordinary budget is deliberate. An appropriation says: here is the money you may spend this year, and no more. A regulatory budget says: here is the amount you may make others spend this year, and no more. The cost of a rule is a tax in everything but name, a transfer of resources from the people who must comply to a purpose the government has chosen. The Constitution gives Congress the power over taxes and spending, as the post on No Money Shall Be Drawn but by Appropriation, Congress's Power of the Purse explains. The argument for a regulatory budget is that the power over the cost of rules should be treated the same way.

The idea has a history. It has been argued over by economists and policy analysts for decades, and it has surfaced regularly in proposals without becoming a lasting part of federal law.

Why the cost is invisible now

To see why a budget would matter, look at how a rule gets made today.

1. Congress writes a statute, often a general one. It tells the agency to protect, to ensure, to promote or to prevent.
2. The agency drafts a rule. If the rule is expected to have major effects, the agency prepares a Regulatory Impact Analysis estimating costs and benefits.
3. The White House reviews it. Under Executive Order 12866, signed in 1993, the Office of Information and Regulatory Affairs, part of the Office of Management and Budget, reviews "significant" rules, which include those with an annual effect on the economy of $100 million or more. The post on The Small White House Office That Reads Every Major Rule Before You Do describes that office.
4. The public may comment, and the agency must respond.
5. The rule is published in the Federal Register and takes effect.

At no point in that chain does anyone ask: Is there room in the total? Each rule is judged by itself, by the agency that wrote it, on estimates the agency produced. There is no running total that anyone must stay under. The result is predictable. A thousand rules, each of which looks sensible in isolation, add up to a burden that no one chose.

Congress has also asked for an after-the-fact accounting, in the form of a White House report to Congress on the costs and benefits of federal regulation. That report is a rear-view mirror. It tells Congress how much was spent after it is spent, and it sets no limit.

How agencies estimate cost today

The tool the agencies use to count the cost is OMB Circular A-4, first issued on September 17, 2003, which is the guidance on regulatory analysis. It tells agencies how to define the problem, how to examine alternatives, how to estimate costs and benefits and how to discount future effects to present value. The guidance has been revised from time to time, a reminder that the numbers depend on the analysts' choices, and that those choices are made inside the executive branch.

That is the first thing to understand about the measurement objection. Cost estimates are already part of the system. Agencies prepare them. Courts review them. An agency that can say that a rule will save $15 billion can also be asked what it will cost. A regulatory budget does not require inventing a new science. It requires taking the numbers agencies already produce and treating the cost side as a limit instead of a footnote.

The precedent: Executive Order 13771

For one period, the United States actually tried a version of this. On January 30, 2017, President Trump signed Executive Order 13771, titled "Reducing Regulation and Controlling Regulatory Costs." It had two main features.

The two-for-one rule. For every new regulation an agency issued, it had to identify at least two existing regulations to be repealed. The order applied the principle to new regulations generally, unless prohibited by law, and set the ratio the way a household might say: for every new expense we take on, two old ones come off.

The cost cap. The order directed that the total incremental cost of new regulations finalized in fiscal year 2017, including repeals, be no greater than zero, unless required by law or approved in writing by the Director of OMB, and it told the Director to identify to agencies a total amount of incremental costs allowed for each agency in later years. Here is how the arithmetic was meant to work. Suppose an agency's new rules add $3 billion in annual costs and its repeals remove $4 billion. The net is minus $1 billion, which is under the cap. If the new rules add $3 billion and the repeals remove only $2 billion, the net is plus $1 billion, and the agency is over.

The order was revoked on January 20, 2021, by Executive Order 13992, issued on the first day of the Biden administration. That is the most important weakness of the 2017 approach, and it is worth saying plainly: an executive order is only as durable as the President who signed it. A regulatory budget that rests on a pen can be erased by a pen. If it matters, it belongs in a statute, passed by Congress, signed by the President and amendable only by another law. Other countries, including Canada and the United Kingdom, have used rules of the same family, such as "one-for-one" and "one-in, one-out" schemes, though readers should check the details of those programs at the governments' own websites, since the rules have changed over time.

A worked example, labelled as a hypothetical

Let us make the mechanism concrete. This is an invented illustration, not a description of any real rule.

Suppose Congress sets a regulatory budget of $20 billion in new annual compliance costs for a given year. Four agencies each propose a rule. Each agency estimates both costs and benefits:

Add up the costs: 6 + 9 + 4 + 5 = $24 billion. That is $4 billion over the cap, since 24 minus 20 equals 4. Under today's system, there is no cap, and all four rules go forward. Under a regulatory budget, someone must choose.

The ranking by net benefit puts A first (+9), then C (+8), then D (0), then B (-2). Rule A and Rule C together cost 6 + 4 = $10 billion. Adding D brings the total to 10 + 5 = $15 billion, under the cap. Adding B on top would bring the total to $24 billion, over the cap, and B is the rule with the worst ratio: it costs more than it returns, according to its own agency's numbers. B does not get issued unless the agency can find $4 billion of old rules to repeal, which would bring the net back to $20 billion.

Look at what happened. The good rules, A and C, went forward. The rule that fails its own cost-benefit test was the one squeezed out. No one was forbidden to protect anyone. The budget simply forced a choice among priorities, which is what a budget is for. The same forced choice is what every family makes when the bills exceed the paycheck.

The objections, stated at their strongest

A serious proposal faces serious objections. Here they are.

One: benefits are hard to measure. A rule that prevents a rare disaster, protects a species or saves lives has benefits that are hard to put into dollars. A cost cap, the objection goes, would privilege what is measurable over what matters.

Two: costs are hard to measure too. Industry has an incentive to overstate the cost of a rule and agencies have an incentive to understate it, and a cap would turn the whole system into a contest between estimates.

Three: a cap could block a rule that Congress itself required. If a statute orders an agency to issue a rule, the agency cannot refuse because it ran out of budget.

Four: it would let the President or OMB decide which rules go forward, increasing executive power. If a cap is administered by the White House, it hands one more lever to the executive, and perhaps to the wrong side.

Five: a cap on cost is not the same as a measure of burden. Some costs fall hardest on small firms, and an aggregate number hides that.

The answers

On benefits. A cap limits cost, not benefit. It does not tell the agency to discount a life saved. It tells the agency to say how much its rule costs, and then to compete for a limited allowance with other rules, including those with large benefits. A rule with high benefits at modest cost wins the competition easily, as Rule A did above. And a cap does not prevent Congress from declaring that a particular goal is worth the price. It just makes Congress say so.

On cost estimates. The measurement problem exists today, because agencies already estimate costs, and the cap changes only the stakes. The cure is to improve the estimates: independent review, as Congress already has for spending in the Congressional Budget Office, with a standing body that audits agency figures. A cap makes accurate cost accounting more valuable, which is the opposite of what the critics fear.

On statutory mandates. Congress can write the exemption, or the allowance, into the statute itself. Where Congress commands a rule and attaches a price, the budget can be raised to match. That is exactly the point: the cost gets debated at the place where it is decided.

On executive power. The strongest version of the proposal puts the cap in a statute that Congress sets, with Congress choosing the total, which is the opposite of handing the lever to the executive. A budget set by the people's representatives restores their control over the cost of the rules that bind their constituents. This is a point consistent with the arguments in Make Congress Vote on Every Major Rule, and Watch Accountability Return and The Congressional Review Act Lets Congress Veto a Rule, if It Moves in Time.

On small firms. A fair design could set separate sub-caps or require a special accounting of costs to small businesses. The existing record of those burdens is documented in the regulation bill a ten-person company pays that a giant can shrug off, and the pattern by which big firms sometimes welcome costs their rivals cannot bear is traced in Big Companies Can Afford Regulation, Which Is Why Some of Them Ask for More.

What the Founders would say

The Founders did not know the word "regulation" in its modern sense, but they knew the principle. Their grievance was that they were taxed without their consent. Every compliance cost is a charge on a citizen. The Constitution reserves the power to lay taxes to Congress, in Article I, Section 8, and Section 7 requires that revenue bills originate in the House, the body closest to the people.

A rule that imposes billions of dollars in costs without a vote is the same kind of act under another name. The regulatory budget is a way of returning the decision to the body that holds the purse. It would not end regulation. It would end the fiction that regulation is free. For the larger argument about how Congress surrendered this ground, see Agencies Write Thousands of Rules a Year While Congress Passes Hundreds of Laws, and for the daily flow of rules, the red tape page.

A practical word about the numbers: the Federal Register recorded 3,248 final rules in 2024 and 2,441 in 2025, and no one in Congress voted on the total cost of either year's output. That is not an indictment of any particular rule. It is a description of a system in which the sum is nobody's job.

The case in brief

1. A regulatory budget would cap the compliance costs agencies may impose each year, as an appropriation caps spending, and would restore Congress's control over what rules cost.
2. Agencies already estimate costs under OMB Circular A-4, so the measurement already exists, and the budget would turn the cost line from a footnote into a limit.
3. Executive Order 13771 of 2017, with its two-for-one rule and zero-cost target, was a precedent, and its revocation in 2021 shows why such a limit belongs in a statute.
4. The best objections, about hard-to-measure benefits, contested estimates and statutory mandates, are answered by a cap on costs that leaves Congress free to decide that a goal is worth its price.
5. The principle comes from the Founders: the people's representatives should decide what government may take from the people, whether the taking is called a tax or a rule.

What you can do: ask your representative, using the representatives page, whether they would sponsor or support a statute setting an annual cap on the compliance costs of new federal rules.