A Modest Proposal on Experimentalism in Politics

Aug 25 2009 Published by under General Politics, Science Communication

The Revolutionary Minds Think Tank blog ( which I mentioned over at A Vote for Science yesterday) is a new pet project to transfer a Seed project over to the ScienceBlogs side of the Seed Media Group. In the initial post the editors asked the following question, to be followed by answers from, one assumes, revolutionary minds.

The boundaries of science are continually expanding as scientists become increasingly integral to finding solutions for larger social issues, such as poverty, conflict, financial crises, etc. On what specific issue/problem do you feel we need to bring the scientific lens to bear?

I am not a revolutionary mind but I have a minor thought on the application of the scientific lens to the political process. The key concepts are "experimentalism" and "sunset provision".


One of the more frustrating and annoying parts of the political process to my perspective is the unfalsifiable nature of political assertions and hypotheses.
The fundamental reasons for enacting a bit of legislation is because it is proposed to cause some change, yes? A change of the economy, a change in State or personal wealth, health or satisfaction. An alteration in personal behavior which, on a broad scale, will result in a proposed good.
When it comes to negotiating over the specific language the predictions and hypotheses fly thick and fierce in the argument over what the proposed law should or should not contain. Since, minor phrasing is going to have a significant impact on the power of that law to effect....change.
Disappointingly from my view, these hypotheses seem only infrequently to be evaluated after a law has been enacted and, most importantly, falsified where appropriate. Laws persist either unexamined or with proponents and opponents sticking to their original hypotheses, uninfluenced by the actual effects of their law.
This is where I believe the scientific lens of experimentalism can be productively brought to bear. Accompany these laws with mechanisms to test the hypotheses advanced. Think a mandate to vaccinate girls against the human papilloma virus will increase sexual activity? Monitor it! Argue for increased investment associated with yet another capital gains tax decrease? Build in an easy before-after comparison. Insist that decriminalization of recreational drugs will not change addiction rates? We can test that.
The point will be first, to structure the arguments for/against a proposed piece of legislation so as to permit an actual test. I realize politicians will be against this since they are not in the business of testing their beliefs. Still, something to work for. Second, it is necessary to build in explicit test benchmarks which everyone agrees will test their opposing proposals. At the outset and as part and parcel of the legislation itself. This will not prevent the usual political nonsense of pretending you didn't predict what you clearly did two years earlier...but it will shine a light on the hypocrisy. Make it more difficult to move goalposts and whatnot.
Buh-bye Law!
nov_17_1466_sunset.jpg
Source
An important tool used far too little in the political process is the so-called "sunset legislation". This refers to writing the legislation in such a way that it becomes ineffective at some time in the future. Five years, say, the law becomes null and void. This would be a great addition to the experimentalism proposal. A hard version of sunset would be essential- by which I mean something that managed to force full consideration of the effects of the policy and essentially a re-argument of the law. (This contrasts with many processes by which supposed sunset legislation is renewed by rubber stamp with minimal discussion and consideration.) This would be an open invitation to falsify prior hypotheses advanced in support of a law or in opposition to it. And this would put a clear focus on goal post moving, sand shifting, bad faith arguments and the like. That would seem to be a good thing to me.

11 responses so far

  • Why are you hating America and making war on Christianity?

  • Jérôme ^ says:

    I don't know how it is in America, but here in France, there is actually a law that says just about what you mean - that most laws must be evaluated by « independent » commissions (no prealably agreed protocol, though, which is a pity) at some fixed point in the future.
    Needless to say, the right-wing majority has silently been ignoring this law since it has the power. (But then, it's the same guys who make about 6.023x10^23 laws on security, each one voted before the previous five or six have taken effect).

  • MattXIV says:

    The problem is that unless you run two parallel groups that face different policies in the same location and time frame, it's almost impossible to make a comparison between two policy options. Any metric of public interest is going to be determined by a number of factors, many of which will change over time, location, and scale.
    Economists have been trying to do these kinds of studies for a long time. What's happened is that it's very difficult at best and on many questions probably impossible to get clear unequivocable results, results from the past do not necessarily apply in the future because of the constant flux of human civilization, and when you actually do get clear results, nobody listens anyway. And this is dealing with issues where there should be measurable quantitative impacts and the data reporting is good.

  • DrugMonkey says:

    Economists have been trying to do these kinds of studies for a long time.
    But economists don't actually make changes and test outcome, right? They take what the policy makers give them, no? Unless you are talking about economists in regulatory apparatus like the Fed?

  • Govt. Bureaucrat says:

    Mr. Monkey,
    You have soul-mates here...
    http://coalition4evidence.org/wordpress/
    and a few at OMB too.
    GB

  • bsci says:

    But economists don't actually make changes and test outcome, right?
    Neither do astronomers. They use the available data and take new measurements to test hypotheses. Sure astronomers and economists might run small scale controlled studies, but the core topic of research doesn't allow making changes.
    Good economic research can identify small things that change to develop control populations. For example, people doing similar work near states' borders when one state raises minimum wage. Other examples are when looking at people with slight differences in ages when a policy has an age-related cutoff. The work is messy and imprecise, but so is addiction research.

  • bioephemera says:

    I object to your claim that you are not a "revolutionary mind." You are clearly a radical militant who wants to overthrow our political system and replace it with one based on logic and the importance measurable outcomes. Fie! Fie! It's unAmerican!

  • Dude, what kind of actual "revolutionary mind" is gonna visit, let alone comment on, a motherfucking blog called "Revolutionary Minds Think Tank"?

  • natural cynic says:

    One could argue that experiments are how federalism was originally intended. One state would try something and others wouldn't have to follow. This ws the ideal, but it also has run up against political realities and the 14th Amendment.
    For instance, if marijuana were legalized and taxed in Oregon and Hawaii, and this was part of a Federal Law. what would happen to someone busted for marijuana possession/intent to sell in South Dakota? Couldn't that person invoke the 14th Amendment? It appears to be a clear violation of equal protection.
    If your type of proposals were to be established, how can the lucky[unlucky] state be picked? Sort of reminds me of the old Life commercials, "you try it ... no, you try it ... hey,let's get Mikey to try it ...yeah ... he likes it! he likes it!!"

  • MattXIV says:

    They do to some extent in regulator and advisory roles, but in most cases, it's no possible to set up a good control case even when directing policy. The Fed definetly devotes a good deal of effort to characterising the impacts of monetary policy but even then you can't set up a case where you can compare two different monentary policy strategies head to head since it would require two whole copies of the country, so you still will get 5 different answers if you ask 4 different economists how the Fed should run monetary policy.

  • D. C. Sessions says:

    For instance, if marijuana were legalized and taxed in Oregon and Hawaii, and this was part of a Federal Law. what would happen to someone busted for marijuana possession/intent to sell in South Dakota?

    Doesn't seem to be an obstacle on a whole host of issues [1]. I suspect you're reading "equal protection under the law" without any of the words in the middle.
    [1] "Dry" states to name just one.

Leave a Reply