Post-modernists often claim that science reflects cultural biases, and that scientific objectivity is a myth. Conservatives have traditionally been quite hostile to this message. But at least some conservatives are beginning to see that “science studies” may have some merit. Here’s a Razib Khan tweet, quoted by Chad Orzel:
one change btwn 2000 and 2020 is that i respect ‘science studies’ a lot more. in fields i keep close track of i’ve seen wholesale reinterpretation of the same data which reflects cultural shifts, not underlying data.
The solution is not to blindly accept left-wing post-modernism, which has its own biases (humorously exposed in the Alan Sokal hoax.) Rather, we need to guard against blindly accepting any pronouncements from scientists, and instead think about how those claims might be shaped by various cultural forces. Here’s Razib Khan:
In November I did a quick interview with a producer for a Spanish television program. He reached out for comment because so many scientists who off the record, would credit the idea of lab escape wouldn’t speak on the record. He related how demoralizing he was finding the difference in how scientists talk off-camera versus what they were willing to say on the record. He had grown much more skeptical of scientific pronouncements in public as a consequence.
I wish I had a more positive spin on this. Much of what you see pronounced with certainty about COVID-19 is likely wrong. SARS-Cov-2 is a new virus, and COVID-19 is a new disease. We don’t have all the details. The experts have a great deal of experience with pathogens and pandemics in general, but there remain limitations to their accrued wisdom in novel situations like this. As a consumer of news and analysis, you have to be personally critical-rational. The consequences of your conclusions, of your actions, are going to impact you, as well as society. Buying fully into the consensus du jour may have more than abstract consequences. Remember washing produce with baking soda? Or the conventional wisdom that droplets, not aerosols were the major vector? No one was lying. But experts were not nearly as certain about the truth as they signaled to the public.
Some people have too much faith in the scientific consensus, believing that scientists hold some sort of “objective truth” above and beyond their empirical observations. But as Tyler Cowen suggests, when you pull back the curtain there is no objective observer:
The most striking thing about the Biden administration shift to a version of “First Doses First” is how little protest there has been. Given how many public health experts were upset about the idea only a few days ago, you might expect them to organize a Wall Street Journal petition from hundreds of their colleagues: “Biden administration proposal endangers the lives of millions of Americans.” . . .
as I had explained, sins of omission are treated as far less significant than sins of commission. Now that a version of “First Doses First” is on the verge of becoming policy, to do nothing about that is only a sin of omission, and thus not so bad. Remarkable! Status quo bias really matters here.
I haven’t seen a single peep on Twitter opposing the new policy.
Just keep all this in mind the next time you see a debate over public health policy. There is often less behind the curtain than you might think.
Just a bunch of empirical observations. The first dose is estimated as 80% effective. A new and more virulent strain is on the way. Manufacturing capacity is limited and hard to expand rapidly. An inadequate dose of vaccine risks creating a mutant form of the virus. Lots of facts, which point in different directions. But these facts don’t by themselves solve complex public policy problems, and we should not expect the scientists who study these facts to be good at solving complex public policy problems. Their expertise is in ascertaining empirical facts and building useful models of causality from those facts.
Speaking for myself, I mostly tend to defer to scientists on empirical questions, and also questions of how to model those empirical facts (except on controversial political issues). But once I’ve learned their views, I bring my own area of expertise (economics) into the picture before forming my view on how scientific data should inform public policy. That’s why I favor kidney markets and price gouging for masks.
When I first read yesterday’s story about the Biden administration’s decision to go with “first dose first”, I thought to myself, “Wow, this is going to be incredibly controversial.” After all, when I suggested the idea to a professional vaccine researcher than I knew, she said I didn’t know what I was talking about. And yet apparently the decision is not controversial at all. I learned something today.
We need to keep our eyes open and pay attention to what the world is telling us. I wonder how many people read Tyler’s post today without changing their views of how the world works. If so, they made a mistake.
People often respond, “Of course I know that’s how the world works”. But do they? I once raised the puzzling fact that roughly 100% of the people claiming the election was stolen were Trump supporters, and zero percent were Biden supporters. Why should there have been any correlation at all? It doesn’t seem rational. Of course commenters rolled their eyes at my naiveté. They told me that there was nothing surprising about this; this is how we all form beliefs.
I had two reactions. First, I’m not sure it is how everyone forms beliefs. And second, if the commenter were correct then I should never read another comment from that individual, as he has admitted that is unable to think rationally. It’s like when an economist tells me that he can’t understand why people leave tips at a restaurant that they never plan to revisit. I think to myself, “I’ll never trust that guy.”
I’d encourage people to think a bit harder about the question of “bias”. How do we recognize it in others? How do we recognize it in ourselves? Am I special? I don’t believe the answers are as simple as lots of people assume.
PS. Speaking of unconscious bias, check out this amazing video.
READER COMMENTS
Mark Z
Jan 9 2021 at 4:07pm
I think one of Khan’s points was also that “public” scientific consensus is not the same as actual scientific consensus. Khan of course has written a lot about topics on which in private, many scientist express or admit the plausibility of one position, but publicly adhere to a different position, even as a rigid orthodoxy. What we call ‘scientific consensus’ is partly (largely) the consensus of non-scientist gatekeepers like journalists, administrators, funding agencies, and journal editors who wield tremendous power in shaping public perception of science and even what science gets done. Of course it doesn’t seem like a silent majority of relevant experts secretly supported ‘first doses first,’ but they do seem to fall in line behind the top institutions in their field, the FDA and CDC. Just as it’s sometimes said the Fed exerts a similar influence on the macroeconomics community.
Btw, your third sentence confuses me. Is it supposed to be ‘[the critique of] “scientific studies” may have some merit?’ Hostility to the rejection if scientific objectivity and seeing scientific studies as having merit don’t seem like contrasting positions. There seems to be a word missing.
Scott Sumner
Jan 9 2021 at 5:27pm
Good points.
My understanding is that Khan uses the term ‘science studies’ to look at studies of science from a sociological perspective, including post-modern critiques of science.
Mark Z
Jan 9 2021 at 7:13pm
Ah ok, I was too hasty at the keyboard.
john hare
Jan 9 2021 at 4:09pm
I didn’t hear anything. Apparently some others were able to hear “green needle”. By comments, others couldn’t hear anything either. So I don’t get it either.
Scott Sumner
Jan 9 2021 at 5:28pm
It worked for me. I heard two completely different words, depending on what word I was looking at. Funny how some brains (like mine) are wired to deceive us.
john hare
Jan 10 2021 at 4:22am
My brain has deceived me from time to time, so I am quite aware of my lack of immunity. Don’t know if it is relevant. I haven’t had a television in the house in twelve years and am not on Facebook or Twitter.
Henri Hein
Jan 9 2021 at 11:51pm
I had a different experience. I read and heard alternating words a few times, just as the text described, but then I kept hearing alternate words regardless which one I was reading. I think I programmed my brain to expect the word on the soundtrack to alternate.
Jon Murphy
Jan 9 2021 at 4:35pm
Great post. I agree with your statement:
To recognize that scientists and experts have our own biases, judgements, observations, and the like shouldn’t lead to a sort of scientific nihilism. Rather, I think, such diversity should be celebrated; a division of labor gives rise to wealth, and likewise a division of knowledge gives rise to wealth of science.
But, likewise, the division of knowledge and labor also suggests we cannot simply “follow the experts.” Both in our role of experts and our role of consumers of expert opinion, we need to have epistemological humility
Scott Sumner
Jan 9 2021 at 5:32pm
I agree.
TMC
Jan 9 2021 at 4:53pm
Science itself is objective. Scientists are human, so will always add a bit of their beliefs into their interpretation of the results. Even 20 years ago, conservatives thought scientists were corruptible. See Steve McIntyre from ClimateAudit, not particularly political, get removed from the IPCC committee for pointing out that the IPCC summaries, that everyone relied on and quoted, didn’t reflect the content of the IPCC report itself. The summaries had a rather hard leftist, and often false, interpretation of the science in the report. This is how cultural biases integrate with science, but aren’t science in itself.
More recently we’ve been told that 2+2 does not always equal 4. Some professional mathematicians even tried to back up the claim, beclowning themselves. “Math is racist”. Cultural biases will always be a part of human nature, but science is a process to truth. When it is followed, and transparency exists, science will not be biased.
BTW, polls have that about 25-30% of Democrats also believe the vote was corrupted enough to change the results. (25 from Reuters, and 30 from Rasmussen).
+1 to you on tipping.
Scott Sumner
Jan 10 2021 at 12:11am
TMC, Are you saying that some Democrats think Trump actually won and the election was stolen? I’m not saying you are wrong, I just haven’t seen any examples.
I don’t think “Science is objective” has any clear meaning. What could that sentence mean? Science includes people, ideas, and the subject matter that is being studied. What part of it is “objective”? I’m sincerely confused.
TMC
Jan 10 2021 at 11:20am
This was the poll I saw : https://twitter.com/Rasmussen_Poll/status/1329741803025801217
There also was a Reuters poll with it at 25%
Rasmussen has it down quite a bit in it’s most recent poll though, down to 12% of Dems thinking they cheated.
https://twitter.com/Rasmussen_Poll/status/1347526231630209024/photo/1
As for objective science, I guess I think of science as a process, so when biases get involved, it’s no longer science. There may be a ‘no true Scottsman’ fallacy in there though I guess.
Jon Murphy
Jan 10 2021 at 2:09pm
This sentence doesn’t really mean anything. Elements of science are objective. Elements of science are subjective. I know you try to rescue your claim by making a distinction between scientists and science, but that’s not really a meaningful distinction. Science is unalienable from scientists.
TMC
Jan 10 2021 at 6:13pm
You are confusing the players with the game. Tom Brady is a great quarterback, but he is not ‘football’. Science is a process. That process is objective even though it can be corrupted by those practicing it. Falsifiable experiments produce objective results.
Jon Murphy
Jan 11 2021 at 9:16am
But likewise, there is no football without players. The two are unalienable from one another.
That science is a process is exactly my point. A process needs people. People make judgements. So the process cannot be wholly objective. There will always be a subjective element.
Jerry Brown
Jan 9 2021 at 5:12pm
This is a really good post. But I didn’t think that policy shift would be very controversial at all. The policy of holding back half the shots to ensure supply for the second shot- now that is pretty controversial. I mean we are dealing with an item that has limited shelf life and are in the middle of an epidemic where hundreds of thousands of people are getting infected every day. Getting that first shot out as soon as possible seems very reasonable. If we can do it- and that isn’t all that clear so far. But I expect we can soon.
Scott Sumner
Jan 9 2021 at 5:32pm
Keep in mind that this should allow us to vaccinate twice as many people by the end of February as expected (and claimed by people like Fauci) just a few weeks ago. So it seems like a pretty big deal. I recognize that the current limitation is not production, but surely the delivery bottleneck can’t last much longer.
BTW, We’ve been told for 6 months that the pace would depend on manufacturing; I don’t know why the public is not more up in arms over the slow delivery.
Liam
Jan 10 2021 at 2:29am
I don’t know if the Coronavirus subreddit counts as the public, but it has 2 million subscribers and every post about vaccines is almost entirely filled with criticism of how slow dose administration is.
Scott Sumner
Jan 10 2021 at 11:47am
Fair point. And perhaps the Trump controversies are dominating news coverage.
Fazal Majid
Jan 9 2021 at 5:48pm
I have a lot more confidence in scientists on public policy than in economists.
There’s a well-worn trope where physicists with the attitude of “physics is hard, therefore everything else must be easy” barge arrogantly into other disciplines like biology to show how it should be done, and generally make fools of themselves.
Now, as the 2008 debacle showed, economists are incapable of providing useful predictions or guidance even in their own discipline, let alone others.
We have absolutely no idea of the consequences of changing the vaccination protocols, the Oxford vaccine trials showed completely unintuitive results (half first dose yielding better results) and it is reckless to gamble on an untested first dose first approach, a perfect illustration of Dunning-Kruger in action.
Jerry Brown
Jan 9 2021 at 8:18pm
But the trials did show significant protective effects after just one dose starting at 10 days. And the Biden plan is not that people would not get a second shot- it is that those shots that have already been delivered would get used because this is an emergency and that we trust that more vaccine and more capability to administer it will be there to provide the second shot in a reasonably close timeframe.
Jon Murphy
Jan 9 2021 at 11:39pm
Question: are scientists any better? Evidence suggests no.
(Also, as a point of fact, some economists like myself did indeed call the 2008 recession. I am quite proud of it given I was 17 at the time. We also predicted a lot of the policy failures that have occurred over the past year.)
Jon Murphy
Jan 9 2021 at 11:42pm
By the way, I do agree with you that there are lots of forecasting problems. But, for reasons I state above, I think that is because people have too high expectations about what models can reasonably achieve, and likewise scientists of all disciplines oversell.
As scientists and experts, we must be more honest about what we actually know. We need to be honest with those who we advise and ourselves.
We need to discuss the fact that models, even hard scientific models, are assumption-dependent. There is much judgement involved.
Scott Sumner
Jan 10 2021 at 12:14am
Talk to a physicist about economics sometime and I assure you that you’ll quickly change your mind. Some of the wackiest ideas I’ve ever heard in my life have come from the mouths of physicists.
I have enormous respect for physicists, most of whom are smarter than me. But please keep them away from public policy at all costs.
Alan Goldhammer
Jan 10 2021 at 8:49am
…and some of the wackiest ideas about vaccines and COVID-19 treatments have come from economists! Case in point, Peter Navarro who is still clinging to hydroxychloroquine.
BTW, one should not rely on Twitter for policy making. There have been a number of articles and papers against the the First Doses concept that I’ve read. Personally, I am not in favor of it because it will disadvantage seniors who likely will need the first dose and booster to get satisfactory immunity. Going forward we may be able to use antibody profiling to demonstrate vaccine efficacy but there is not enough data from the clinical trials to make that leap of faith (seasonal flu vaccine is approved based on historical data and a defined manufacturing process; subsequent mRNA vaccines may fall into this category).
I’ve not seen a number this high mentioned. It’s also age dependent as noted above.
Jon Murphy
Jan 10 2021 at 11:00am
Which leads to the question: why should only a singular group of experts decide policy? Even if we want to (incorrectly) say that the pandemic is primarily a public health problem and only secondarily an economic problem, given your statement here we should be skeptical of any single group of experts making policy. They may end up just being very silly as soon as things veer from their field of expertise (eg, Anthony Fauci playing psychologist and saying there is no need for masks as a means of keeping panic down.)
Scott Sumner
Jan 10 2021 at 11:48am
That’s an extremely misleading comment. Navarro is almost universally despised by other economists. He’s not typical of economists at all.
Mark Barbieri
Jan 9 2021 at 6:04pm
This! I find it very frustrating when people tell me I need to listen to the scientists when those scientists are leaping from science to policy. No offense, but academics in general don’t have the best record with policy design.
My related pet peeve is the typical argument in the form “Science says X is a problem and so we propose policy Y to address that problem. If you aren’t for Y, you must not understand that X is a problem.” I often find that I agree with the scientists on the science but disagree on the policy. That usually annoys my friends on both sides because people seem to be more tribal and like things to have 2 clear sides. They aren’t fond of people that only partially agree with them.
robc
Jan 10 2021 at 2:26pm
Where do Jorgenson supporters who think the election was stolen fit in?
Fred_in_PA
Jan 11 2021 at 2:47pm
You say, “I once raised the puzzling fact that roughly 100% of the people claiming the election was stolen were Trump supporters, and zero percent were Biden supporters. Why should there have been any correlation at all?”
If this were an objective event (i.e., the properties reside in the object and will be independent of who observes it), yes, one would expect different observers to see the same thing. And yet “election was stolen” appears to be rather subjective in that it’s entangled with judgements about the goodness or wickedness of the outcome.
Suppose instead of an election we were discussing a mantel clock (perhaps in the disposition of grandpa’s estate) and we said “The mantel clock rightfully should have gone to X, but instead was carried away by Y.” Here the goodness or wickedness of the outcome seems very much to depend upon who’s doing the looking (the subject of the sentence). Those who see reality from X’s perspective will likely judge that the mantel clock was stolen. Those who see it from Y’s will likely judge the outcome to be as it should have been.
Fred_in_PA
Jan 11 2021 at 3:10pm
I should add that one can be quite certain the thing was stolen without having any clear idea of how the thief pulled it off. (Think of a cultural masterpiece that has disappeared from a well-guarded museum.)
In the case of the just-past election, how did those vote totals turn out that way?
My own guess is that the Democrats were much smarter about playing upon the weaknesses of (small d) democracy. Earlier voting (so more opportunities to get your voting done) and no-excuse-needed absentee voting (so easier to vote) drew more people into the process who otherwise would not have bothered. But the “can’t be bothered” crowd is less likely to put a lot of thought into what they’re doing: More likely to vote on whim or emotion. And God knows Donald Trump evokes a lot of emotional response — most of it negative.
(And while it may have been different elsewhere, here in Pennsylvania every sample ballot I saw on TV had Biden/Harris at the top with Trump/Pence in second position. I think it’s pretty well established in political science that first position on the ballot is good for an extra 1% or 2% of the vote. Which in this case was the margin of victory.)
Comments are closed.