Blog

Agencies Write Thousands of Rules a Year While Congress Passes Hundreds of Laws

Compare the Federal Register's final rules with the laws Congress enacts and you can see where lawmaking in America has actually moved, and why it matters.

Count them. That is the whole assignment, and it is the cheapest civics lesson in America. Take the number of laws Congress enacts in a year and set it beside the number of rules the executive agencies publish in the same year. Then ask the question the Constitution asks: who is making the law of the land?

The answer is not the body whose first sentence reads, "All legislative Powers herein granted shall be vested in a Congress of the United States." The answer, by the arithmetic, is a few hundred agencies and offices that nobody elected, working under statutes broad enough to drive a truck through. This post does the counting and then draws the constitutional conclusion.

The terms

A law, in the sense I am using, is a bill passed by both houses of Congress and signed by the President or enacted over his veto. Congress.gov numbers each one as a public law, in the form "Public Law 118-" followed by a sequence number, with the Congress number first.

A rule, or regulation, is a binding legal requirement issued by an executive branch agency under authority Congress has delegated. Agencies publish proposed and final rules in the Federal Register, the government's daily journal, and codify the final ones in the Code of Federal Regulations. A final rule is the finished product: it has the force of law, and a person who violates it can be fined or worse.

Those are the terms. Now the numbers.

The count

The facts sheet I work from gives these figures from the Federal Register: it published 3,248 final rules in 2024 and 2,441 final rules in 2025. Two years, 5,689 final rules, which is 3,248 plus 2,441.

Now the laws. The number of public laws a Congress enacts varies, but it runs in the hundreds for each two-year Congress, not the thousands. The 115th Congress's last law was Public Law 115-441, while the 118th Congress was far less productive, with only 78 enacted by August 2024. I would ask you to check the exact number for any Congress you care about on Congress.gov, which lists every public law by Congress. For the arithmetic I will assume a round 300 public laws for a two-year Congress, which is 150 a year. That is an assumption for illustration, not a count of any particular Congress, and the conclusion does not depend on it.

Divide. 3,248 final rules in 2024 divided by 150 laws a year is about 21.7. 2,441 divided by 150 is about 16.3. So, on that assumption, agencies published roughly sixteen to twenty-two final rules for every law enacted in a year. If the true figure for laws is 200 a year instead of 150, the ratio falls to about twelve to sixteen. Change the assumption any way you like within reason and the answer stays in the double digits. The executive branch issues ten or more final rules for every statute Congress passes.

The strongest objection

Let me do what a fair argument demands and state the other side as its best defenders would.

They say the comparison is unfair, for four reasons. First, the two counts measure different things. A public law can be a thousand-page omnibus statute that governs the whole economy, while a final rule can be a notice that closes a fishing area for a season, sets a drawbridge schedule or updates an airworthiness directive for a particular aircraft. Counting them one-for-one is misleading. Second, many public laws are small: the naming of a post office, the transfer of a parcel of land. Third, a large share of the rules are routine and technical, implement a statute Congress passed and are required by that statute. Fourth, Congress cannot do everything. Modern government is complex and legislators cannot write the specifications for every aircraft engine, drug label or bank capital ratio. Delegating the details to experts is the only way the laws can work.

Every one of those points is correct, and I accept all of them. The routine rules are indeed numerous. The count is a rough instrument and I will not pretend it measures the weight of governance on a family.

But notice what the objection concedes. The defenders do not say the executive is not making law. They say it is making law that is mostly technical. Fine. The Constitution did not carve out an exception for technical law. And the count is not my only evidence.

The numbered argument

  1. The Constitution vests lawmaking in Congress alone. Article I, Section 1: "All legislative Powers herein granted shall be vested in a Congress of the United States, which shall consist of a Senate and House of Representatives." The word "all" is there for a reason. The Founders had just spent a decade under a Crown that made law by proclamation and by the acts of far-off officials.
  2. A rule that binds a citizen with the force of law is legislation in substance. It does not matter what the agency calls it. If it creates an obligation that a person can be penalized for breaking, it is a law in everything but the name of the body that made it. The counts above are counts of obligations.
  3. The volume is the evidence of the shift. The Federal Register first appeared on March 16, 1936. Its annual page count has grown many times over since, and I would ask you to check the page counts for any year you wish, since they are published by the Office of the Federal Register. Pages are an imperfect measure, since the Register includes notices and proposals as well as final rules, so treat the trend as a pointer and the final-rule counts above as the firmer figure.
  4. Legislators who vote on hundreds of laws cannot answer for thousands of rules. A representative can be held accountable for a vote. He cannot be held accountable for a rule he never saw. When a family in Ohio is fined under a regulation, it cannot vote out the official who wrote it. The ballot reaches the member of Congress, who did not write the rule, and not the writer, who cannot be removed.
  5. Therefore the center of gravity of lawmaking has moved to the executive. Not entirely, not in every case, and with Congress's consent. But the proportion is what it is, and the proportion is the problem.

What the Founders said about volume

Madison wrote about this exact danger in Federalist 62, in the passage on the evils of an unstable legislature. He warned that lawmaking by men of the people's own choice does the people little good "if the laws be so voluminous that they cannot be read," and he added incoherence to the list of defects.

Read it twice. The Founders considered a body of law that nobody could read a defect that destroyed the benefit of elected lawmakers. They were thinking about statutes. They would be stunned by a Code of Federal Regulations of that size, revised every year.

A related point comes from Federalist 47, where Madison, writing about Montesquieu's separation-of-powers precept, put the danger this way: "The accumulation of all powers, legislative, executive, and judiciary, in the same hands, whether of one, a few, or many, and whether hereditary, self-appointed, or elective, may justly be pronounced the very definition of tyranny." When an agency writes the rule, enforces it and sits in judgment of the person accused of breaking it, the three powers have met in one office. I have written about that in Madison Called It the Very Definition of Tyranny, and Agencies Hold All Three, and I would urge you to read it with this one.

How the shift happened

It did not happen in a day, and it did not happen by coup. It happened by statute, in steps, each of which seemed reasonable.

In the early republic Congress made the rules in some detail. In the late nineteenth century it began creating commissions, starting with the Interstate Commerce Commission in 1887, and giving them authority to set standards, such as "just and reasonable" rates. In the New Deal, the pace picked up. Statutes began to say, in effect: here is a goal, and the agency will work out how to reach it. In the 1960s and 1970s, a wave of new regulatory statutes covering the environment, workplace safety, consumer products and more handed agencies broad mandates with the details left to rulemaking.

Each statute was passed by a Congress that preferred to announce a goal and leave the hard choices to someone else. That is the key. The delegation was voluntary. A legislator who votes for a bill that says "the agency shall issue regulations to protect the public" gets credit for the goal and blame for none of the costs. The agency takes the heat. Everyone in the system has a reason to keep it going, except the citizen.

The Supreme Court has said that Congress may not delegate its legislative power, and has also said that Congress may seek the assistance of the other branches if it supplies an "intelligible principle." The Court has used the "intelligible principle" test in a very permissive way for most of the last ninety years. In recent years the Court has shown more interest in limiting the practice, including in the "major questions" line of cases and in West Virginia v. EPA in 2022, which held that the Environmental Protection Agency could not claim, on a thin statutory basis, the power to restructure the nation's electricity generation. And in Loper Bright Enterprises v. Raimondo in 2024, the Court overruled the Chevron doctrine, which had told courts to defer to an agency's reasonable reading of an ambiguous statute. Those are significant corrections. They do not shrink the count of rules by themselves.

For the doctrine behind this, see The Nondelegation Doctrine Says Congress Cannot Hand Its Job to Someone Else, and for the Court's recent rulings on big questions, Five Times the Court Told an Agency That Big Questions Belong to Congress.

A note on what the count does not prove

I owe the reader candor. The count does not prove that any particular rule is bad. Many rules are sensible. Some protect people from real harm. A mature economy needs standards for aircraft and banks and food, and I would not abolish them. The count does not prove bad faith, either. Most of the people who write rules are doing a job Congress gave them.

What the count shows is where the lawmaking happens. It happens in the executive. And a republic whose laws are mostly made by officials who cannot be voted out has not kept the bargain it made at the Founding. The argument is not that rules should vanish. It is that the people who make them should answer to the people who live under them.

What would restore the balance

A few reforms follow from the argument, and each has a model in the law.

The case in brief

  1. Article I vests "all" legislative power in Congress, and a binding rule with the force of law is legislation in substance, whoever signs it.
  2. The Federal Register published 3,248 final rules in 2024 and 2,441 in 2025, while Congress enacts public laws in the hundreds per two-year Congress, so the ratio is in the double digits on any reasonable assumption.
  3. The objection that many rules are routine and many laws are large is fair, but it concedes that the executive is making law, only that much of it is technical.
  4. Federalist 62 warned that elected lawmakers are of little avail if the laws are so voluminous that they cannot be read, and Federalist 47 warned against gathering powers in the same hands.
  5. The remedy is for Congress to take back its job: vote on major rules, count the cost and set end dates.

What you can do: Open the Red Tape page, pick one pending rule, and write your member of Congress asking whether the statute that authorizes it should be rewritten so that Congress votes on it.

Latest on this topic