Three Times Interventionists Moved the Goal Posts
At the Independent Institute’s blog I have 3 posts up:
In the third installment, make sure you don’t miss how slippery Ezra Klein was. For example:
Cass recently summarized the Oregon results for his readers by writing, “In a randomized trial in Oregon that gave some individuals Medicaid while leaving others uninsured, recipients gained no statistically significant improvement in physical health after two years.”
Cass’s language was quite precise and accurate. This is how the Oregon researchers themselves(including Jonathan Gruber) summarize their results: “In the first one to two years of coverage, Medicaid improved self-reported health and reduced depression, but had no statistically significant effect on several measures of physical health.”
And yet, Ezra Klein argued that Oren Cass was wrong in his claims about the Oregon experiment. Thus, Klein is implicitly arguing that the Oregon researchers themselves didn’t understand their results as well as Klein did. (Also note that Klein edited his original article, so now you have to scroll to the bottom to see him talking about Oren Cass.)
When it comes to winning votes, the self-reported feel good factor is more important than actual physical health.
Did you see his update on the article?
Did you see this in the quote I gave?
“(Also note that Klein edited his original article, so now you have to scroll to the bottom to see him talking about Oren Cass.)”
Nice series of posts!
I am not sure the climate one is terribly convincing. ““The scientific rationale for the 2°C target is not really very scientific.” In a world of perfect enforcement by governments around the world, Nordhaus thinks an optimal carbon tax would end up allowing some 2.3°C of warming…”
Really, 2.3 degrees is not a snappy enough target for a political campaign, which is what this is. I happen to agree with you in practical terms – 2.3 degrees sounds right to me. However, it simply does not work as a target, and it is in the nature of targets to be missed in any case.
However, you article is not about the rights and wrongs of the target, but the amount of goalpost shifting. If you put forward a 2.3°C target instead, do you get much kick-back? I imagine that there would be very little dissent, but I have not actually seen anyone propose this limit.
Also, I think that your definition of a denier is wrong. Anyone should be able to challenge the orthodox views as long as they are honest in their sources. For example, googling “Bob Murphy climate denier” did not get me any hits. Googling “Pat Michaels climate denier” got a few on page 1. That seems to be because you generally restrict your comments to economics rather than the science of the climate and cite IPCC data to support your arguments, whereas Pat Michaels makes claims about climate that have no supporting evidence or are just wrong. This makes me think that my idea about what makes a “denier” is reasonable, but I guess not everybody will share my view.
Also, as far as goal post shifting goes, you have to admit the skeptics are also just as guilty. There have been many claims that there are calls for cuts in carbon that will send us back to the stone age. Instead, we get modest proposals to tax carbon that are predicted to have less effect on the economy than the sort of capital interest rate changes we see on a regular basis, or comparable to a trade deal that can be pulled at the last moment without the economy collapsing. Far from apocalyptic or catastrophic, these proposed changes are modest. The real alarmists are those that are suggesting the carbon restricting measures will be a disaster, rather than a moderate set-back at worst. The discussions are surely worth having, but let’s keep the rhetoric reasonable on both sides.
Harold wrote:
Really, 2.3 degrees is not a snappy enough target for a political campaign, which is what this is. I happen to agree with you in practical terms – 2.3 degrees sounds right to me. However, it simply does not work as a target, and it is in the nature of targets to be missed in any case.
However, you article is not about the rights and wrongs of the target, but the amount of goalpost shifting. If you put forward a 2.3°C target instead, do you get much kick-back? I imagine that there would be very little dissent, but I have not actually seen anyone propose this limit.
No, Harold, this is wrong, and it alarms me that you are this unfamiliar with the terms of the climate debate–given how confidently you “correct” me all the time on it.
The Paris agreement doesn’t say, “We should shoot to limit warming to 2C, because that is the point that optimizes the tradeoffs between lower economic growth and reduced climate change damage.”
No, the Paris agreement says that the 2C ceiling is an absolute necessity, and ideally countries should shoot for closer to 1.5C of total warming.
Then Nordhaus comes along and says if all governments in the world acted perfectly, then the optimal target would be 2.3C. But since we know governments *won’t* act perfectly, actual “optimal” target is closer to 4C.
So no, if you are summarizing this as “Paris says 2C is optimal, and they are just rounding from Nordhaus’ 2.3C,” then you are totally misunderstanding what each participant is claiming.
The Oregon study I see as over-hyping the first results. Always wait until the results are in before drawing conclusions. It turns out that the study did not show a statistical improvement in the physical parameters measured. It did show improvement in subjective assessment and depression.
Too much was made of the initial results. It is very hard to pick out improvement on a population level (rather than people who are sick). Most people on the study would not have needed to change their blood pressure, for example.
In the absence of the over-hype by people going off half-cocked in the middle of the study, we could have had a very reasonable discussion about what these results mean. As it is everyone ends up digging in to try to defend their prior position and the information is to a large extent not communicated to the public.
The key question for me is whether health care does any good. It was very clear that the study showed increased us of health care. Yet the results were positive only for depression. Are the physical factors they measured the best ones? Mammograms were increased, for example, but this would not show up in blood pressure statistics.
So has the study shown that health care is less effective than we thought, or has it shown that we don’t know if it is as good as we thought? It is very interesting, but I think the pool has been muddies too much to get a decent answer.
On further reading, an important point.
“Cass recently summarized the Oregon results for his readers by writing, “In a randomized trial in Oregon that gave some individuals Medicaid while leaving others uninsured, recipients gained no statistically significant improvement in physical health after two years.”
Cass’s language was quite precise and accurate. ”
Not quite. Cass says there was no statistically significant improvement in physical health after 2 years. However, from the authors comment we see that physical health was not measured, only several measures of physical health. This is not a nit-pick, but a very different thing. It is quite possible that there were statistically significant improvements in physical health on parameters that were not measured. The study did not conclude that there was no improvement in physical health.
This is a very different conclusion, and one the study does not support. It is possible that there was no improvement in physical health, but we cannot conclude this from the study.
It is also incomplete because it misses out depression, which is quite a significant problem. One could say it is both inaccurate and misleading by wrongly stating the conclusions about physical health and ignoring the improvement in mental health.
Harold wrote:
“Not quite. Cass says there was no statistically significant improvement in physical health after 2 years. However, from the authors comment we see that physical health was not measured, only several measures of physical health.
This is not a nit-pick, but a very different thing. It is quite possible that there were statistically significant improvements in physical health on parameters that were not measured. The study did not conclude that there was no improvement in physical health.
OK fair enough Harold, but by your criterion no one could ever say anything about “physical health.” I could assess the effects of shooting people in the forehead with a .45, and technically no study, no matter how many parameters it measured, could ever conclude, “This policy of shooting people in the head with a .45 reduced their physical health.” Do you want to go there?
“It is also incomplete because it misses out depression, which is quite a significant problem. One could say it is both inaccurate and misleading by wrongly stating the conclusions about physical health and ignoring the improvement in mental health.”
No, here I don’t think you are even technically correct, let alone right “in spirit.” The researchers themselves were the ones to contrast the depression scores with “measures of physical health.” So *they* introduced that distinction, not Cass. That’s what I meant when I said Ezra Klein (and now you, Harold) are saying the researchers screwed up when summarizing their own study.
EDIT to add: Harold, I am OK if you say his summary was “misleading” for leaving out the depression stuff, but it wasn’t “inaccurate.” He was literally copying the language used by the researchers.
On physical health parameters. It is important to report the findings accurately. They measured certain parameters. They did not measure “physical health”. It is reasonable to assume they picked those parameters because they thought they thought they told us something about “physical health” but the authors did not claim there was no improvement in physical health. They did not measure, for example, breast cancer diagnoses, or even mortality. Possibly because they did not expect the study to have sufficient power to spot any differences.
“I could assess the effects of shooting people in the forehead with a .45, and technically no study, no matter how many parameters it measured, could ever conclude, “This policy of shooting people in the head with a .45 reduced their physical health.” Do you want to go there?”
I am very happy to go there. If we used the same parameters as the study we would conclude directly that there had been a marked fall in the physical health as measured by the parameters. With the blood pressure alone we could conclude reduction in “physical health” per se, because a reduction in blood pressure to zero is always accompanied by a reduction in physical health, i.e. death. So using the results of the study and combining it with our wider knowledge would allow us to draw the conclusion you say we could not draw. There is no problem there.
In this study we do not have that wider knowledge to draw on. We cannot say that there there is *never* an improvement in physical health if blood pressure etc. are not altered, so we cannot conclude there is no increase (or decrease) in physical health. We can conclude that this study provides no evidence for such an increase, but that is a very different thing.
This is an important point. When conducting a trial it is vitally important to say what you are going to measure before you start. There is a big difference between “physical health” and various blood pressure parameters used hear.
On the depression, the distinction is OK, as introduced by the authors. Missing it out is not inaccurate because it is OK to single out one aspect to discuss, if that is what you want to do. I had not intended to say that the depression thing was inaccurate, but that the “physical health” thing was inaccurate, although I do not believe it was intentional.
Apart from these details I basically agree that the initial results were over-hyped, which has required the commenters to back-track – shift the goalposts if you like. Looking at Ezra Klein’s article it reported on increases in use of healthcare services and did not report much on the health outcomes. It would have been better to wait for the results rather than assume they would show what you expected, which is what Klein did.
Harold, you’re missing my point about the distinction you drew between “physical health” and “measures of physical health.” Suppose 5 measures of physical health show deterioration, but 63 other measures show improvement. Can we conclude that physical health has deteriorated?
So even in gunshot case, for all we know there are 63 other measures that improved, but were omitted from the study.
“Suppose 5 measures of physical health show deterioration, but 63 other measures show improvement. Can we conclude that physical health has deteriorated?”
No. Can we conclude physical health has improved? No. Without knowing more, the 63 might be trivial and the 5 very important. What should we report? That the 5 measures improved and the 63 declined.
“So even in gunshot case, for all we know there are 63 other measures that improved, but were omitted from the study.”
That is not the case, because death always results in all parameters of physical health declining – particularly over a two year period. Whatever we chose to measure we would find a decline. However perhaps that is not the best example because it is so extreme.
Another example, we find that there were significant changes in all the chosen measures of physical health including blood pressure (systolic or diastolic), cholesterol (HDL or total), glycated hemoglobin, or a measure of 10-year cardiovascular risk that combined several of these risk factors. We then find all patients died 6 months later of malignant tumors. Would we say there had been an improvement in physical health? I would say not.
The authors say
“…but had no statistically significant effect on several measures of physical health.”
They did not say
“recipients gained no statistically significant improvement in physical health after two years.”
The two statements are different in an important way. It is this important distinction I am pointing out.