There are probably lots of examples of coffee-table “damned leeches, living off my tax-dollars” conversation getting proven wrong, or at least grossly simplified, but this is probably one of the most damning examples.
There really ought to be a means of conveying information like this to the public, en masse. News outlets have stories like this on their websites, though I haven’t found any live TV reports. Just imagine how any political debate down in the United States would go if every rhetorical comment and argument was shot down with articles like this, on both sides. In fact, there ought to be a repository for scientific findings that have immediate political and sociological consequences.
In fact, there ought to be a repository for scientific findings that have immediate political and sociological consequences.
A few years back, Aaron Swartz & Peter Eckersley tried this with a website called Science That Matters (now down), which revealed a couple of potential problems with this idea.
The first is pretty minor: the people running the repository might stop updating it after a while. STM got updated quite a lot through 2007, a couple of times in 2008, and never again after that.
The second is that scientific findings might not be as weighty as they sometimes look. They turn out to be wrong, get over-interpreted, get under-played, or prove less relevant than they first appear.
The last post on STM was titled “There is No Satisfactory Form of Utilitarianism”, based on a paper arguing that any sane way of aggregating the welfare of many into a single measure leads to at least one of “The Repugnant Conclusion”, “The Sadistic Conclusion”, and “The Very Anti-Egalitarian Conclusion”. This appears to force utilitarianism into an agonizing trilemma, as each of these three conclusions seems appalling. But the Repugnant Conclusion is arguably not so Repugnant after all, in which case the trilemma loses its force and the paper loses its urgency.
Another STM post discussed the RAND Health Insurance Experiment. The post first went one way, exaggerating the study’s results with the over-the-top title “Is the net effect of health care zero?”, and then swung the other way, warning readers in an update that JohnNyman highlighted a “terrible flaw” in the experiment that “would seem to severely throw these findings into question”. (Families given free healthcare were less likely to drop out of the experiment than families that had to pay some healthcare costs, introducing a bias.) Half a decade after Nyman, it turns out that some of the experiment’s results still stand after accounting for the flaw.
A similar issue crops up for the proposed lead-and-crime link you mention. Although the effect of high lead exposure on mortality and mental test performance is well established (quasi-)experimentally, the effect of low lead exposure on violent behaviour isn’t, so it’s not clear how much we should trust the quantitative estimates of the latter.
If this was intended as a response to me, I don’t see the relevance to what I said.
As for the lead story, that’s been going around for a long time, and don’t really see the relevance to most US political discussions. Childhood lead exposure can cause developmental problems leading to adult behavioral changes. Believable to me.
Sorry about that, it wasn’t. I misclicked, and the retract button’s line crossing seemed even worse than just leaving it there.
The relevance to US political discussion is in policy decisions. With the information we have in this article, it’s a better long term investment to get rid of lead usage in industrial and commercial settings that risk such exposure, if one is trying to reduce crime in cities, instead of building more prisons, “cracking down on crime”, etc.
Quite recently under Obama the EPA did make a decision to limit mercury polution which got opposed by the Republicans.
If you accept that lead should be regulated then why not mercury? The EPA did a pretty good calculation that estimated the costs and benefits of mercury regulation.
There are probably lots of examples of coffee-table “damned leeches, living off my tax-dollars” conversation getting proven wrong, or at least grossly simplified, but this is probably one of the most damning examples.
There really ought to be a means of conveying information like this to the public, en masse. News outlets have stories like this on their websites, though I haven’t found any live TV reports. Just imagine how any political debate down in the United States would go if every rhetorical comment and argument was shot down with articles like this, on both sides. In fact, there ought to be a repository for scientific findings that have immediate political and sociological consequences.
A few years back, Aaron Swartz & Peter Eckersley tried this with a website called Science That Matters (now down), which revealed a couple of potential problems with this idea.
The first is pretty minor: the people running the repository might stop updating it after a while. STM got updated quite a lot through 2007, a couple of times in 2008, and never again after that.
The second is that scientific findings might not be as weighty as they sometimes look. They turn out to be wrong, get over-interpreted, get under-played, or prove less relevant than they first appear.
The last post on STM was titled “There is No Satisfactory Form of Utilitarianism”, based on a paper arguing that any sane way of aggregating the welfare of many into a single measure leads to at least one of “The Repugnant Conclusion”, “The Sadistic Conclusion”, and “The Very Anti-Egalitarian Conclusion”. This appears to force utilitarianism into an agonizing trilemma, as each of these three conclusions seems appalling. But the Repugnant Conclusion is arguably not so Repugnant after all, in which case the trilemma loses its force and the paper loses its urgency.
Another STM post discussed the RAND Health Insurance Experiment. The post first went one way, exaggerating the study’s results with the over-the-top title “Is the net effect of health care zero?”, and then swung the other way, warning readers in an update that John Nyman highlighted a “terrible flaw” in the experiment that “would seem to severely throw these findings into question”. (Families given free healthcare were less likely to drop out of the experiment than families that had to pay some healthcare costs, introducing a bias.) Half a decade after Nyman, it turns out that some of the experiment’s results still stand after accounting for the flaw.
A similar issue crops up for the proposed lead-and-crime link you mention. Although the effect of high lead exposure on mortality and mental test performance is well established (quasi-)experimentally, the effect of low lead exposure on violent behaviour isn’t, so it’s not clear how much we should trust the quantitative estimates of the latter.
If this was intended as a response to me, I don’t see the relevance to what I said.
As for the lead story, that’s been going around for a long time, and don’t really see the relevance to most US political discussions. Childhood lead exposure can cause developmental problems leading to adult behavioral changes. Believable to me.
Sorry about that, it wasn’t. I misclicked, and the retract button’s line crossing seemed even worse than just leaving it there.
The relevance to US political discussion is in policy decisions. With the information we have in this article, it’s a better long term investment to get rid of lead usage in industrial and commercial settings that risk such exposure, if one is trying to reduce crime in cities, instead of building more prisons, “cracking down on crime”, etc.
Quite recently under Obama the EPA did make a decision to limit mercury polution which got opposed by the Republicans.
If you accept that lead should be regulated then why not mercury? The EPA did a pretty good calculation that estimated the costs and benefits of mercury regulation.