For what reasons (legal or practical) are there not (or not more) empirically conditional laws to attempt possible solutions to perceived problems? I'm mainly thinking about in the United States.

It seems true to me that laws could have an easier time being passed if they took a form where we could look at some measure (say, government income), try a suspected solution (say, reduce taxes), and if the desired result occurs (government income increases as is suggested by certain groups based on ideas of the Laffer curve), keep the solution, and if it doesn't, the law automatically phases the "solution" back out. This is as opposed to having long political squabbles over whether some such strategy would be effective or not.

In an extreme case, could there be a law that attempts to set, say, maximum sentence severity based on, say, the violent crime rate? What I have in mind as the most extreme form would be a sort of PID based on specific societal ills and laws known or reasonably suspected to somehow impact those ills.

  • 1
    You are confusing law with justice,. And government with duty.
    – Richard
    Commented Dec 29, 2018 at 2:08
  • You say try a suspected solution. What if trying this solution harms people irreversibly (so phasing the law out doesn't "unharm" them)? What if there are multiple possible solutions? Which one do we try first and why? At best, your solution is "try, try again", but, at worst it's government experimentation on people. The more general question would be "why don't we pass laws based on fact, not opinion", which I think has been answered (sounds good in theory, absolutely unworkable in practice)
    – user2565
    Commented Dec 31, 2018 at 20:27
  • @barrycarter I agree, those are possible outcomes of certain specific possible laws that could be passed under this model, but the issues would be with those specific possible laws, not problems with the model of law itself. I simply feel that it would be easier to convince people in a democratic environment to agree to "try this and if it doesn't work (based on such and such terms), reverse course" as opposed to "try this" with no recourse or check on success or failure. My intuition is the prior case could be more palatable to getting a majority of votes, so I'd expect to see more of it. Commented Dec 31, 2018 at 20:33
  • You're expressing the concept of "sunset dates" on laws. I think the problem you'll run into is: why not try my solution first, and, if it doesn't work, we'll back out of it. Both sides will want their solution tried first, because this methodology could be said to favor the first solution tried.
    – user2565
    Commented Dec 31, 2018 at 20:36
  • 1
    @barrycarter Again I agree, that would occur. But that also essentially occurs without this model of laws with the laws we pass now, so I don't see how that applies as a problem to the specific model I'm supposing. Again, my question is basically, "my intuition is there should be more of these, but there aren't (at least apparently). Why not?" Commented Dec 31, 2018 at 20:42

1 Answer 1


There are many reasons, some political, some logical. Too many to answer concisely, but let's pick a couple.

  • One of the easiest to look at is that in almost all areas of human endeavor that have controversies about law-making, making causative attrbution is very difficult if not outright impossible.

    Barring two identical countries, one with a law passed, one without, you don't have a control group to verify whether a crime rate drop/rise is attributable to some law you passed, or a whole host of other reasons (or no causative reason whatsoever - we all know that when ice cream sales increase, the rate of homicides also increases).

  • As a further wrinkle, your two variables relationship may not be linear - as a random made up example, even assuming maximum sentence affects crime rate, the change from sentence of X to Y could reduce crime rate whereas from X to Z could further increase it instead.

  • Additionally, you would likely miss unexpected second order effects.

    You may look at a 5 year study that investigates relationship of minimal wage to employment. It may tell you something. What it will NOT tell you - and thus you won't account for - is the effect of raising minimal wage has on career prospects of today's teenagers 15 years from now.

  • Attribution isn't such a big problem as you make it seem though, there are many pretty clear cases where laws have made an impact on things. It seems disingenuous for many people to point to the various clear impacts laws have on society and then simultaneously claim we cannot measure their impact. And the second point is merely an optimization problem, the sort of thing something like a PID control would be well suited and something we'd already trust for guiding things like a commercial airplane. Commented Dec 29, 2018 at 4:01
  • 1
    @magnus.orion - rule of thumb: "clear" and "obvious" are usually words people use when they don't have peer reviewed solid study to back up their opinion :)
    – user4012
    Commented Dec 29, 2018 at 18:48
  • Are you seriously trying to tell me there exists no peer reviewed research concerning the impact of laws on society? I'm pretty confident, not only does such peer reviewed research exist, there are entire fields that study exactly that. Commented Dec 29, 2018 at 19:12
  • 1
    @magnus.orion The obvious answer :ok, provide some references. However, do these sources show a consistent effect or different effects? How do you know effects in the past will translate to effects in the future (past performance does not imply future performance). You're ultimately trying to predict the future, which is difficult or impossible in society.
    – user2565
    Commented Dec 31, 2018 at 20:34
  • 2
    Yes, but there's no real consistency there. I agree academics will publish all sorts of theories, but they often disagree with each other. The abstract for your link even says Two findings at variance with conventional predictions, meaning they're disagreeing with existing views. Academic papers in the social sciences can be useful summaries of what's happened historically, but don't accurately predict the future.
    – user2565
    Commented Dec 31, 2018 at 20:48

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .