(This is an edited excerpt from an article that originally appeared on Scott Bradford: Off on a Tangent in 2013.)
One of the perennial problems with our system of government is that our laws, once passed, are rarely reevaluated. When Congress considers legislation, it is usually attempting to solve a problem…but it almost never comes back five or ten years later to consider whether the law actually solved that problem, and whether it had any major unintended consequences.
There are some exceptions. If a law’s consequences are negative enough, or if the problem it was supposed to solve remains a serious problem in the public consciousness, Congress will sometimes circle back and try to make improvements…but this only happens with a very small percentage of the laws we enact. Also, if a law is controversial enough when it first comes up, Congress will sometimes include a sunset date, at which point the law must either be reenacted by Congress or else expire.
Sunset provisions are better than nothing. One was included on the Federal Assault Weapons Ban, which President Bill Clinton signed into law in 1994. The law had no discernible impact on gun violence, mostly because it was based on how guns look, not on how they operate or who uses them. Congress allowed it to expire in 2004, and there has been no subsequent bloodbath…on the contrary, our violent crime rate continues to drop. It was nice to see a useless, ineffective law drop off the books, instead of languishing there permanently.
The tax cuts signed into law by President George W. Bush also had a sunset provision, but Congress renewed the vast majority of those cuts with President Barack Obama’s support. The data shows that those reductions in the tax rates contributed to an eight billion dollar increase in annual federal tax revenue over the 2003-2007 period. So Congress reenacted the bulk of the largely-successful law.
But these are the exceptions, not the rule, and the decisions to renew or not renew these laws were based less on their effectiveness than on the stale political posturing of our two major political parties. Both sides fell into their respective predefined grooves in each debate. It was a good thing that these laws came back up for reevaluation, but the debates had little to do with an honest, data-driven analysis of what impact (if any) the laws had.
So what’s the solution? I propose an approach to legislation where most laws include clearly-stated, measurable goals, and automatically expire if those goals are not met or if the law has significant unintended consequences. We have hundreds of federal agencies collecting statistics on everything imaginable; let’s put those statistics to use in evaluating the success or failure of major legislation.
A Hypothetical Example
Let’s create a hypothetical example. Imagine that there is a nationwide outbreak of vandalism using eggs. Millions of teenagers are buying raw eggs at grocery stores and throwing them at cars, crushing them in mailboxes, and hurling them at houses. It’s a serious problem. It is a public nuisance, yes, but it is also forcing people to spend millions of dollars on egg cleanup, and it’s taking law enforcement time away from more serious public safety activities. We all agree, something has to be done.
Some prominent members of Congress believe that the best way to solve the problem is with a number of legal limits on eggs. They propose that we require people to show identification in order to buy eggs, and that we limit people to one dozen eggs per-purchase. This is not very far-fetched; these are similar to the restrictions that have actually been placed on the purchase of pseudoephedrine decongestants in an effort to limit methamphetamine abuse — a well-intentioned legislative effort to solve a real social problem.
The opposition party, however, thinks that these restrictions will not solve the problem. People will just steal the eggs, or buy one dozen at a time from multiple stores. Worse, they fear that this law will discourage and penalize innocent egg purchasers without having enough benefit to justify the inconvenience. It might reduce egg vandalism, but at what cost to all the other egg consumers and egg producers? This group of legislators would prefer tougher penalties on those who use eggs in acts of vandalism, and better enforcement of existing anti-vandalism laws.
Who is right? Well, you and I can form opinions based on the arguments, but we can’t know for sure. Perhaps the proposals for egg purchase restrictions, although inconvenient for the average citizen, will be so effective at eliminating egg vandalism that it would be worth the cost. Perhaps the opposition is right and it will have no effect at all. Everybody picks their side, the legislation comes up for a vote, and it either passes or fails…and that’s the end of it. If the purchase restrictions become law, you and I will spend the rest of our lives having to show ID to buy eggs (and cold medicine)…whether or not those restrictions actually solved the problem.
And what if they put a sunset provision in the law? Well, in ten years it’ll come up for a vote again. Most members of the party that supported it will still support it, even if the evidence shows it has been a resounding failure. Most members of the party that opposed it will still oppose it, even if the evidence shows it has been a resounding success. Whether it gets renewed will likely be a function of what party is in power that year, not the real outcomes. There has to be a better way.
Instead of a sunset based on an arbitrary period of time, let’s write a sunset clause based on measurable outcomes. Let’s add a clause on the end of the Egg Vandalism Restriction Act saying what we expect the law to do, what the possible unintended consequences might be, and under what conditions the law will either stay in effect or sunset itself. Congress would be forced to consider what it really hopes to accomplish, articulate those expectations in the law itself, and provide provisions for the law’s automatic repeal if it proves ineffective or damaging after a given time.
So proponents of the law could write into it that the expected outcome is a fifty percent or greater reduction in egg-related vandalism, and that a reduction of at least twenty percent would still be considered a success. If there isn’t a reduction in egg-related vandalism of at least twenty percent over five years, directly attributable to the provisions of the law, the law will automatically sunset. Opponents might also contribute a sunset provision, stating that, regardless of the reduction in egg-related vandalism, if the law results in a greater than ten percent reduction in total national egg sales, or if more than 100,000 innocent citizens are improperly prohibited from buying eggs over that five year period, the negative consequences of the law will be considered to outweigh its positives and the law will also sunset itself automatically.
Consider the benefits of this kind of results-based sunset provision. Some opponents of the law might decide to support it, since it has provisions for sunsetting itself if it proves ineffective or damaging. Proponents will be saved the embarrassment of having to come back and repeal it later if it completely crashes and burns, and they’ll be able to say to their constituents (truthfully) that they foresaw the possibility of its failing and built appropriate protections into the law. Best of all, laws that don’t work or have serious unintended consequences will drop off the books all by themselves without need for a protracted, highly-politicized fight over whether or not to keep them.
This is, of course, not a perfect solution. We’d still be relying on Congress to set reasonable benchmarks for success and to spend some time considering the possible unintended consequences of their legislation…and we know that at least one prominent member of Congress can’t even be bothered to articulate the constitutional basis for major bills or read them before passing them.
Not only that, but we would also have to rely on success metrics that, in some cases, might be easily manipulated. For example, the Federal Reserve consistently “underestimates” inflation by excluding food, fuel, and some other costs from the consumer price index. And data about defensive firearm uses per-year in the United States famously ranges from as low as 55,000 (cf. Hemenway) up to a whopping 2.5 million (cf. Kleck and Gertz), depending on who executes the study, how they execute it, and how they define their terms.
So the people, and the press, would still have to actively police these things. But as a matter of principle, we shouldn’t have ineffective laws sitting permanently on the books. If they don’t work, we should get rid of them. The more we can automate that process and remove it from partisan politicking, the better.