Tim Harford gives his stamp of approval to randomized trials:
What is missing is the political demand for tests of what really works. Too many policies on education, welfare and criminal justice are just so much homeopathy: cute-sounding stories about what works leaning more on faith than on evidence. Politicians and civil servants, faced with some fancy new idea, should get into the habit of asking for a proper randomised trial. And we, as citizens, should be equally demanding….
We’ve had FairTrade coffee – what about FairTest policies? Most voters don’t know much about randomisation or trial protocols, but they’ll know when they see the FairTest logo that a policy has had a proper, scientific test to see if it works.
Next Comments →
Politicians and civil servants, faced with some fancy new idea, should get into the habit of asking for a proper randomised trial.
Aren’t proper randomised trials are the very last things that politicians and civil servants would want? They have no interest in plausible-sounding programs being revealed as ineffective (witness the extreme resistance to evidence that D.A.R.E. is useless and Head Start provides no lasting benefits).
(responding to Slocum)
Depends on the politician; some are more about ideology, some are more about pork, some are more about results. From what I have seen, civil servants usually want to do things that work (gives their work more meaning, if nothing else) but if they aren’t allowed to or if they don’t know what those things are, they can’t.
I think more actual valid studies on policies would be an excellent idea.
“cute-sounding stories about what works leaning more on faith than on evidence.” I thought that WAS AGW.
I thought that WAS AGW.
In your case, “thought” appears to have been deliberately avoided.
You can’t evaluate everything using randomized trials. Sometimes it is not possible (not everything can be randomly assigned, think about characteristics like race and gender), sometimes it is not ethical (withholding a potentially good treatment from a group), and sometimes its just not cost effective.
But I do agree that there should be some standard for ‘FairTest’-ing policies.
I don’t think those who call for a FairTest logo would approve of the policies that got the logo.
If we compare the competitive advantage of different health care systems which are almost a random walk of policies, the one thing common to those with the greatest efficiency – benefit/cost – is universal coverage.
If that wasn’t clear by 1990, it was clear by 2000, and if not clear by 2000, it is clear today. The US health care system is the most inefficient by huge margins when compared to the developed nations with specific universal coverage policies, and even the nations with the worst health outcomes are more efficient with nearly universal suffering.
This is yet another attempt to claim to be objective while rejecting all the objective evidence that supports policies contrary to “theory” more properly called dogma.
Another example is “lower taxes will increase economic growth/employment” when Federal taxes are now 25% lower than in 2000, down from over 20% of GDP to under 15%.
While I’ve been a fan of “rational and/or empirical analysis-based” policymaking for a long time, we have serious issues to overcome before we can even get near that. Namely, that there are so many that aren’t *interested*, in policies that “work best” or are “good” because they care more about:
- their ideology
- their view of the world
- their own power
than the results of any policy, even if could be proven to work 100% AND benefit the opposing people in question. The other thing is diametrically opposed goals. How can you come to a consensus when – even if you agree that a particular policy will work for a certain goal – you think the *means* or the *ends* are bad/immoral/stupid. Let’s take some of the arguments against recreational drug legalization.
For the sake of this argument, let’s say we can
prove beyond a shadow of a doubt that:
1) Recreational drug legalization and regulation would lead to a 90% reduction in drug-related deaths.
2) It would lead to a 90% reduction in the incidence of HIV.
3) It would add 10% to yearly GDP.
4) It would lead to a 35% reduction in overall crime.
These numbers are made up, and for illustration.
Now, take some of the counter-arguments:
“I don’t want drug deaths to go down. I think using drugs is /evil/ even though it doesn’t personally harm me and anyone who decides to use them in a regulated system has to do so by going through a screening process and getting a license. I don’t think people should do things that give themselves pleasure, as that is Hedonism and is evil. In fact, I think we should design policies so that drug deaths go up. I also don’t care if it’s better economically. It’s wrong. I also don’t care if crime goes down. Even though I want crime to go down, I would rather it stay the same or even go up than legitimize evil Hedonism. So I do not care if your policies will work for the goals you set (less crime, etc.), as your policies lead to outcomes/use means I consider wrong.”
This sort of argument could be applied to just about anything. In a sane world, we’d lay out our goals *FIRST*, then pick the best policies to achieve them (there would of course be some constraints for say, human rights), but we’d start from a position of “these are our goals” and work from there using various kinds of analysis (empirical tests, Bayesian reasoning, simulations, prediction markets, whatever) to decide which. In this world, however, we have the twin issues of many people not agreeing on the ends, even in a very basic way and people who aren’t even interested in the ends, per se, but only care about the means (“process legitimizes outcome”).
How exactly do you *argue* with people who do not agree on the very basics? You don’t. You get paralysis, measured over a period of years.
Good idea. But useless. To test Health Care Reform, you’d need a sample population large enough to at least represent cancer, and to see if you affected health care costs by 5%, it would have to represent the rate within 5%. So your sample error requires you to have …
1/sqrt(samp_pop) = 0.05*(cancer_rate)
for a cancer rate ~ 400/100000 (UK), that sample population is about 25 million.
One positive thing is that the UK has been running this experiment for over 60 years with 61 million people and spends half as much as the US does on health care.
“In your case, “thought” appears to have been deliberately avoided.”
Ha! Actually thought was not even needed on my part. Burden of proof is on you, to the tune of an absolute fail.
The idea of having randomized tests for policies on what really works sounds like a great idea on paper but could it really work? How can you have a randomized trial on the new health care policy? I find it hard to believe as a citizen that these tests would actually produce any data that would be helpful. I know Harford says citizens should be equally demanding for these trials but I think citizens just want straight-up answers instead of sheets of statistical data. Who has the time to listen to their news broadcaster spit out statistics from a randomized test, when they could just say the pros and cons that come from the policy by how it has been written? Most policies that come into play have already been in the works in other places and have been tweaked in order to accommodate what our people need at that time and place. In the end though, it would not matter what data was found but what the person voting feels at the time is right for the majority of the people the policy will affect.
AyeCarumba:
What I found fascinating in The Myth of the Rational Voter was the evidence that people stopped favoring such crazy policies if they were put in positions of real power. The same people who favor the drug war now would likely legalize drugs as dictator.
In terms of voting the desire to show you aren’t one of those dirty hippies or that you want to protect our children or that you are a serious person and etc.. is far more influential than any calculation about the actual effects of the drug war. Changing your view won’t actually stop the drug war but it will change how people think of you.
The article written by your very good, I like it very much. I will keep your new article.
Was honored to see such a good article!
The article written by your very good, I like it very much. I will keep your new article.
i think the product is so good ! Dont miss it ! come and buy one for yourself !
i think the product is so good ! Dont miss it ! come and buy one for yourself !
This looks absolutely perfect. All these tinny details are made with lot of background knowledge. I like it a lot. This was a useful post and I think it is rather easy to see from the other comments as well that this post is well written and useful.
Youve got style, class, bravado. I mean it. Please keep it up because without the internet is definitely lacking in intelligence.
These kind of post are always inspiring and I prefer to read quality content so I happy to find many good point here in the post, writing is simply great, thank you for the post!
Im not going to say what everyone else has already said, but I do want to comment on your knowledge of the topic. You’re truly well-informed.
Your blog is my stepping stone, my friend. Thanks for the heads up on this subject.
Your blog is a great one. What really impresses me is that you are correctly mentioned that there are thousands of tools that are available to create a website or launch one but what matters is that you choose the right one, the one that gives you all that is actually needed.
hey,this is one of the best posts that I’ve ever seen; you may include some more ideas in the same theme. I’m still waiting for some interesting thoughts from your side in your next post.
Your blog is a great one. What really impresses me is that you are correctly mentioned that there are thousands of tools that are available to create a website or launch one but what matters is that you choose the right one, the one that gives you all that is actually needed…
doesn’t it was AGW? “cute-sounding stories about what works leaning more on faith than on evidence.”
and you got your point, we need some fair-test standard policies
we need fair-test standard policies, but it still need analyze
Let us always meet each other with a smile,for the smile is the beginning of love,and once we begin to love each other naturally we want to do something. [url=http://www.onlineusb.net]oem usb[/url] Branded usb drives
We need a canada goose jacket in the winter to keep warm.
I was reading something else about this on another blog. Interesting. Your position on it is diametrically contradicted to what I read earlier. I am still contemplating over the opposite points of view, but I’m tipped heavily toward yours. And no matter, that’s what is so great about modernized democracy and the marketplace of thoughts on-line.
other with a smile,for the smile is the beginning of love,and once we begin to love each other naturally
the right one, the one that gives you all that is actually needed.
other with a smile,for the smile
yes baby
seeaww
smile
smilesesee
hehmsees
hehmseeseses
something else about this on another blog. Interesting. Your position on it is diametrically contradicted to what I read earlier.
contradicted to what I read earlier.
sseeeseee
seeeeeeeeeeeeeesss
yes baby
ghfgklhmkglfhlf
gbfxbxdfgdf
kbfgbfldblfd
fvgdflldflzfd
lpfgbfgfc
dsfvdsvçsdz
this is great .
by admin
can relate to that. When I was in my 20′s and living alone, this was a frequent activity of mine. I lived in a very rural area with not much to do other than working, running, and spending time with friends.
Next Comments →
Comments on this entry are closed.