Loading summary
Transcript40 lines
- [00:02]
A
Get ready to take a flamethrower to the official narrative and learn what the elites don't want you to know. You're listening to the Tom Woods Show.
- [00:19]
B
Hi, everybody. Tom woods here. It's episode 2761 of the Tom woods show, and we're joined today by Aaron Brown, author of the new book Wrong how to Extract Truth From a Blizzard of Quantitative Disinformation. And, Aaron, I don't ordinarily do this, but I was looking through your credentials, and I think of the great many authors I've had on the show over the years, you are one of the most qualified to write your book. So I'd actually like it if you could start by sharing a bit of your background with us rather than my barking it at everybody.
- [00:54]
C
Sure. Yes, I am extraordinarily well qualified for writing this book. Unfortunately, I'm very well qualified in a negative way. I'm not a great researcher who discovers new things. I'm a professionally qualified skeptic for debunking other people's claims. But it's a valuable thing to do, I guess. I got my start in all of this. I trace it all the way back to when I'm a freshman in College. This is 1975, and Fisher Black got me a job on the national standby gas rationing plan. And by the way, he was extraordinarily skeptical about the plan, and he put me on it. I don't think there was any big endorsement of what he thought of me. It was comment on what he thought of the plan. And I was assigned to figure out how we were going to ration gasoline. If we had to do that back in 1975, that was a real possibility. And I was assigned to worry about agriculture. You know, could we harvest the crops? Because we didn't want everybody to starve to death because farmers couldn't get fuel. And I'm a city boy, and so I didn't know whether tractors use gasoline or diesel fuel. So I call around and I got Department of Agriculture. And they said, yeah, yeah, there's a guy who knows that, but he wasn't in. And I called Ford because Ford makes most of the tractors. And I got passed around. We got to a guy said, oh, yeah, I just finished a comprehensive economic model, and 75% of the tractors in this country run on diesel fuel. Great, thanks. You know. Well, we went through. I got a lot of other detail from him, and then the Department of Ag guy calls back, and I didn't want to, you know, blow him off because you know, he was nice. He was returning my call. So I say, sure. We said, oh, 75% of the tractors in the country run on gasoline. I said, oh, you mean diesel, right? He said, no, no gasoline. I said, well, Ford just told me it was diesel. He said, oh, you know Ford, he's talking about new tractors manufactured. He's not talking about the existing stock. Well, that just wasn't true. I mean, I'd spent a lot of time with the Ford guy, and I knew exactly what he had done. And when I went back to the Ford guy, he said, oh, yeah, you know, the agriculture, they're counting tractors that are rusting in barns or were shipped to Mexico years ago. You know, they just have no idea what's actually in use today. Also not true. You know, both of them had these models. And what kind of hit me is both of these people wanted to teach, neither one wanted to learn. Neither one was willing to engage in reconciling with the other with, you know, seriously doing. They had their model, they had their number, and they were sticking to it. But what really hit me was so I talked about, you know, some of the senior economists on the project, and I told them this story. Every single one had the same reaction. They laughed. And they topped it with a better story of experts who gave completely different answers to questions that were both simple and necessary for our project. So I went away thinking, there's just absolutely no way this project's going to succeed. It's just a joke. And they refused. I said, you know, we should talk to people from the Soviet Union and Gosplan. And, you know, they do this for a living, and they've been doing it for 50 years. Maybe they know a few things we. We don't know. That was. No, that was shot down. Of course not. You know, Soviets are idiots. Communists don't know anything. And I said, well, at least let's talk to the people who were involved in World War II rationing. A lot of them are still alive, you know, and they probably have some useful information. No, no, no. Those guys, you know, we're smart, modern economists. Those guys didn't have computers. They were just, you know, amateurs. And so I kind of took this with me, that experts will tell you things very confidently. People who should know, people who are paid to know, people who are in positions of author, know one guy is advising Ford on its marketing, and other guys, you know, running Department of Agriculture Policy on this issue, you know, they'll put out numbers that they have absolutely no basis for. So this Kind of followed me along in my entire career, both academic and in professional practice, mostly in finance and risk management. You know, financial risk management is basically the idea of ferreting out all the wrong numbers that might trip you up. So then that's kind of how I got to maybe 15 years ago or so when I started writing some articles for reason and eventually got a video series called Wrong Number. That's proven surprising, surprisingly, to me, anyway, successful, and I've become kind of a professional debunker of wrong numbers.
- [05:20]
B
How would you describe the thesis of your book? Because it's not just the commonplace that, well, people can lie with statistics. There's something more institutional going on. There's something worse than that, in a way, going on. Some of these numbers that we're more or less taught to accept unthinkingly have institutional backing to them and have a certain prestige to them. So it's understandable why a lot of people might unthinkingly just accept them.
- [05:48]
C
Yes, and there is. The book does talk about sort of some of the institutional reasons why people produce these and why people believe them. But I do want to be very careful about something. The opening anecdote of the book. This was actually my first Wrong Number article for reason. This was before we were doing videos, was the Chinatown Bus survey. So in 1996, a Chinese immigrant noodle delivery guy started a business called Fung Hua that ran curbside buses. So he picked up people in Chinatown and New York, took them to Boston for 10 bucks. You know, it was $25 on a Greyhounder Peter Pan. And it was a lot more convenient. It was, you know, the buses came to a convenient place. You didn't have to hike to Port Authority, which in those days was not anything anybody did voluntarily. And they never ran out of seats. So, you know, on Thanksgiving, they would keep just running bus. They just hire a new bus and run it, whereas all the major bus companies were all sold out. So, you know, for the college students going home Thanksgiving, it was just tremendously better thing. And congressmen from New York, Chuck Schumer, Nadia Velasquez were trying to shut this down. You know, the big carriers were all against it, and they finally got their chance. There was a horrific accident in 2011. It was not a curbside bus. It was a casino bus that was run by the Mohegan Sun. You know, they just chartered a bus to take their customers to and from New York City. But it had a horrific accident and there were a lot of people killed. So Schumer and Velasquez told the National Transportation Safety Board to do a study on curbside buses, even though this wasn't a curbside bus. And they dictated the conclusion. They said, and it's going to show that these unregulated things are way too unsafe. And so the NTSB complied and came out with a study that said they're seven times more likely to be involved in a fatal accident. A curbside bus versus a traditional carrier. And the study just had all kinds of statistical problems I won't go into at the moment. But so that's why I got involved with Reason, wrote a Reason article about this. And by the way, it was used to shut down Fenghua, which had an excellent safety record, and 26 other immigrant owned, Chinese owned bus companies that were curbside, but not to shut down any of the major carriers. You know, Greyhound and Peter Pan and all those were running their own curbside carriers. Those did not get shut down. And then Jim Epstein, my producer for Reason videos, he discovered by accident after a had to file a freedom of information request that got denied. And for six months he fought it still was denied. So the NTSB is telling us some of these bus companies are seven times more dangerous than others, but it's a secret which ones they are. I mean, does this make any sense at all to anybody? But when Jim finally did figure out by accident, he discovered that 30 of the 37 fatal accidents that were ascribed to the curbside companies were actually by Greyhound and Peter Pan, the terminal carriers. So they were just stuffing the ballot box. Now here's the important point though. I mean, this story is going to be a lot familiar to your listeners. You know, all those libertarians. Yeah, yeah, this is what happened. Sure, you got some tame congressmen pleased their donors by shutting down competition and they pressure regulators to issue false reports and shut people down. And okay, that's the conspiracy theory, but it just doesn't fit all the facts because this story, the fatal accident story, was widely covered by reporters and journalists who did incisive, deep reporting on the issue. And they all took this seven times study, ntsb, they all put it in headlines, they reported it uncritically, nobody asked about any statistical issues that were being raised, anything like that. Everybody trusted it. And then when Reason came out with the revelation, not just that the study was flawed statistically, but that they'd stuffed the ballot box, nobody ran a retraction, nobody ran a correction. The only story about it from a major media outlet was Bloomberg. And Bloomberg did a he said, she said, you know, squabbling statisticians. They didn't say, hey, wait, this study was clearly fraudulent. So, you know, these reporters aren't in on the conspiracy, if there is one. I mean, these are people, they want to report the truth. They want to run on things. And yet they didn't seem mad that they had been lied to. They didn't seem like, you know, we should investigate this or at least we should retract our previous studies. So you can't say this is all conspiracy. You gotta put in some room for incompetence. But incompetence isn't a great explanation either, because an incompetent person would make lots of errors like this. The errors all went in the same direction. Every error they made made the curbside bus carriers look worse. So this is kind of what I think of as the central question of the book that I don't have a great answer for is what combination of conspiracy and incompetence can explain this tremendous issue we have with wrong numbers getting out there, affecting policy, affecting public opinion, and nobody seems to care?
- [10:57]
B
Well, on some level, people would be likely to be cynical or skeptical if, say, a study about the health effects of cigarettes came out from a tobacco company, they would know right away their radar would go up. Maybe I should be suspicious of this, but people have just as much incentive, even though it may not be financial, they have ideological incentives to want certain things to be true. I want this to be true. So therefore, as soon as I hear about it, I'm going to promote it. And, you know, let's face it, almost everybody listening to this has been guilty. You hear something that tears down the side you disagree with, and you retweet it right away without pausing to take a look. So some of it is we tend to want to amplify what we want to believe in. And is that a phenomenon that's just somehow getting worse? Is social media maybe making it worse?
- [11:47]
C
I think it's always been with us. And maybe I'll get a little uncomfortable jabbing some of our listeners here. I bet there are listeners out there who, you know, heard about Sam Peltzman's result that seatbelt regulation, seatbelt requirements kill people. And by the way, I took all of Sam Peltzman courses when I was in University of Chicago, and he's a great guy and he did great research. But, you know, the actual evidence about whether seatbelt regulations save or cost lives is pretty mixed. I mean, there is a lot of subtlety to it, but I know there's a lot of libertarians I know who think of it as gospel, like it's carved in stone because it's what they prefer to believe, or that the FDA kills more people than it saves. Again, you know, very mixed evidence on that. Minimum wage, you know, a lot of libertarians believe, you know, minimum wage laws absolutely hurt low skill workers. And of course, you know, there's a solid kind of microeconomic argument that that would be true. But again, there's a lot of nuance when you actually start sifting through the data and looking at things. So I suspect most of our listeners, and certainly me personally, and perhaps you have adopted some wrong numbers uncritically because we like what they say. One of the things I say in the book is try to make it nonpartisan and. But I don't have a lot of libertarian examples because the libertarian examples don't get amplified by the press, you know, don't get amplified by mainstream institutions the way some of these other ones do. And so wrong Number kind of comes out as a populist book. It's like, you know, debunking all the things, you know, that experts say. And while that's fair, because those are the wrong numbers that drive policy that are widely accepted by, you know, majority of people. But every field's got its wrong numbers and you have to be just as alert to it for yourself. You can't change the world. You can't make everybody smart, you can't make everybody skeptical, but you can avoid fooling yourself. As Richard Feynman said, the most important thing is don't fool yourself because you're the easiest person to fool.
- [13:53]
B
Hey, gay. You know how much the medical bills came to for the rather challenging pregnancy that yielded us the wonderful little Henry woods, who's just about five months old right now. You know how much that came to? You're not going to believe it. About $200,000. Yeah, $200,000. And yet crowd health took care of it all. No problem. A year or two earlier, the woods family had seceded from the deeply broken American Health Insurance system and embraced crowd health, which is an alternative to that system. And it's really simple. You cover smaller expenses yourself and a very small portion of larger ones. Crowd health negotiates those bills down and then crowdfunds the rest among the membership. It works fantastically well. Even when we were afraid our bill was surely going to break the system. Nope. Worked like a charm. A guy in my elite mastermind, his wife got both hips replaced. Grand total $500. My business partner, Paul Counts, had knee surgery $63,000, but he paid $500. Meanwhile, instead of the woods family's thousands of dollars in monthly premiums, we pay crowd health a small fraction of that. It's brilliant and amazing, and you should be part of it. Get started today for just $99 per month for your first three months. When you go to JoinCrowdHealth.com and use code Woods, Crowd health is not insurance. Opt out. Take your power back. This is how we win. Join CrowdHealth.com, code Woods. Let's do a few examples, then. So the first one, and this caught my eye, not just because it was the first one, but because was all over the news, the USAID matter, because when funding started to be pulled from that, the response was really hysterical, that if you do this, all these terrible things are going to follow. We were hearing extraordinary claims for how many deaths would be caused if these programs were terminated or reduced in some way. Incidentally, it reminded me of back when Obamacare was being debated. There were arguments being made about the tens of thousands of deaths that there would be if Obamacare were repealed. So this must have been probably maybe 2015, 2016. And that number was just more or less invented. It was like the number of homeless. That Mitch Schneider, he just came up with a number and he gave it to the media and they just accepted it. So what's the situation with the USAID stuff?
- [16:19]
C
Okay, well, the reason that was the first substantive example in the book is it came up in the Lancet, and the Lancet is the most prestigious and oldest medical journal in the world. And they have the most, the boldest claims about how carefully they vet everything in there. They find skeptical reviewers. They have professional data. People go over, you know, they do all this stuff to do it, do it, do it. So they published this article that USAID funding from 2002 to 2020, I think it was, or 2021, saved 91 million lives. And, you know, if you know anything about global mortality statistics, you don't have to be an expert. You know, well, gee, the total decline in global mortality over those years was 78 million. So somehow USAID is not only responsible for the entire decline in global mortality. Forget technological improvements, forget advances in public health, forget economic growth, forget all the other aid programs. It's responsible at 114%, the entire thing. Everything else killed people. Everything else we thought was reducing global mortality was actually killing people. Or another way to look at it is USAID is something on the order of 8% of all foreign aid in the world. And if that saved 91 million lives and all the others were equally effective, nobody should have died since 2002. So this is just a number that just on its face is clearly ridiculous. Anybody reporting on any reporter who reported on it should have been able to see, wait a minute. I mean, I assume if you're, you know, a scientific reporter for New York Times or NPR or organization like that, you have some familiarity with global mortality statistics and can see immediately that this number just cannot be right. Now, I drill down in the paper and I show exactly how they got this wrong result. You know, fairly elementary statistical error. They did and they buried it. I don't believe this was an accident, that they just somehow did it because they didn't. They were deceptive about the methodology they used. So all of this stuff about the Lancet, you know, doing careful vetting of the things, a peer review, the methodology review, the idea that science reporters for major outlets have any idea what they're reporting, are they just parroting numbers? And none of them, you know, since this was written in a Reason article and now in my book, you know, nobody's gone back and reviewed it or questioned it. And in fact, the Lancet came out with a new version with an even higher number. And you're right to compare it to the Obamacare claims. And those Obamacare claims are still around, but they're recycled for the Trump proposals to roll back the 2020 era Medicaid expansion. And again, we get these incredible figures that it's going to, you know, cost hundreds of thousands of lives if we do this. And I have another chapter on that.
- [19:14]
B
Can we move to gun control here? Because the email from Gene Epstein, Jim's father, that got my attention about your book had a subject line involving gun control papers. I don't remember the number, but it was some number in the five figures of how many gun control papers there are. And the subject line was something like such and such number of gun control papers can't be wrong or could they? And I thought, oh, what's old Gene got up his sleeve here? So it was about had to do with your book. So there is something that a lot of these gun control papers apparently have in common and there are very, very few of them that overcome this issue. So what exactly is that?
- [19:52]
C
Okay, well, the opening for the chapter is actually a study by the RAND Corporation. You know, RAND Corporation, fairly neutral, maybe slightly left leaning. I don't know how you want to characterize it, but one thing they did is they cataloged 27,000 gun control papers. And they discovered only a handful, you know, 40 or 50 were not so flawed methodologically to be completely worthless. And even the ones that weren't, weren't very strong, were pretty weak. And the only thing you could find some very, very thin evidence for is that it's possible that child safety regulations, gun regulations, prevent, you know, two or three children's accidental deaths in a year. You know, not very strong evidence, not very good study, but it does kind of suggest that. And the other one I think was, is possible that two week waiting periods have a very slight reduction in number of gun suicides. Not necessarily total suicides, but gun suicide. So. And again, both of those things kind of make sense. I mean, if you're going to, you know, child safety rules for guns might prevent a few accidents. A two week waiting period might stop one or two people from killing themselves with a gun, what else they do? And who knows, but it might do it. But this is the point isn't what these few studies show, it's that people are perfectly happy to put out 27,000 papers, you know, notching publications, getting grants, getting promotions academically and so on without telling us anything meaningful or useful about gun control. And by the way, I'm Gene Epstein has me I'm doing a SoHo debate in August.
- [21:32]
B
Oh, on what?
- [21:33]
C
On gun control.
- [21:34]
B
Oh, gun control, okay.
- [21:35]
C
Yeah. But the point is we could start studying this seriously. But you know, the first thing we do is we'd figure out how many guns are there. I mean, people disagree by more than a factor of two. We'd have to study defensive gun use. We'd have to learn a lot more about crime and violence because really, guns is a funny way to start approaching the subject. What we really care about is crime and violence and you know, guns are a part of that. But if we don't know why people kill each other, we can't really say much about what guns do. And this is a serious 10 year difficult project. But if we were devoted, you know, if we took all the money we spent on those 27,000 studies and started it on this subject, at the end of it, we might know something. And it might be something uncomfortable for libertarians. It might be something that says, boy, if we just, you know, had a national gun registry and we restricted gun ownership to people who could prove they needed a gun and were competent in using it or something like that, we would save a thousand lives a year. And then you'd have to say, okay, you know, which is more important, human rights or those thousand Lives, greatest good or the greatest number. And then we have an ideological problem, but at least we all agree on, okay, these things are effective, but we just don't know. And by the way, I will say there are a lot of libertarians out there, I think, who are put too much credence in the lot study. I'm not, you know, it's got its points, it found some useful things, but it also, there's a lot of debate on the other side. I think gun control. We genuinely do not know if the US adopted Canada's gun laws. Could it get common as gun death rate? I think that's an open question. It might be true. I still wouldn't support it. I wouldn't want to be Canadian, even if it made me a little bit less likely to get shot to death. But I understand most people disagree with me. Most people aren't libertarian. And most people would say, yeah, if you can save a thousand lives, it's worth tossing the second amendment in the dumpster.
- [23:27]
B
Let me think of what there is, the California minimum wage example, but let's see, is that a good one? I'm trying to think of what would be the best one if I want to do one or two more before we get into more general things like a book.
- [23:38]
C
Well, the calculation minimum wage is just, it's not even a good study. The chapters I care most about are where there's an interesting study, you know, that does tell you something useful, but it got kind of distorted and pulled out. For example, one that's much less important than California minimum wage. But the Nobel Prize study, I don't know if you saw that one. So it said there, there's a study out, it says that, you know, Nobel Prizes are dominated by people who are born in the top 10% socioeconomic class. And so science is biased and we have this tremendous problem, whatever. And when you actually drill down on how they did it, you find out science is incredibly meritocratic. That what they did is they looked at father's profession and from father's profession they inferred socioeconomic status. And from socioeconomic status, they said that's the reason. But, you know, so children of scientists have 300 times the chance of winning a no science Nobel Prize. They excluded economics and peace and literature. But, you know, okay, so that sounds, wow, you know, scientists are dominating this. And it is true. Having I had a scientist father myself. So I, you know, where's my Nobel Prize? But you look at it, you know, Hollywood Oscar winners, you look at presidents, you look at, you know, leaders in almost every field, you find a lot of successful people have children who follow them. And in fact, it's less true in science. Science has a lot more people who were born poor, born from, you know, fathers who weren't anything like, you know, anything you would expect for a Nobel prize winner. And also, many of the people they call, you know, top 10% socioeconomically were not. Some cases, you know, the father died very young and the child was raised in poverty. Other cases they were, you know, ministers or rabbis, which, you know, they code as top 10%. But that often isn't true. I mean, yeah, if you're a minister or rabbi in a very rich parish or something like that, you're probably pretty well off. But most are not. You know, some of these are in tiny churches or, you know, don't even get a appointment of some sort.
- [25:49]
B
Isn't peer review supposed to protect us from at least bad numbers that are coming out of professional journals?
- [25:57]
C
Yeah, okay, that's the claim. And actually what peer review does is it enforces conformity. So peer review weeds out the people who are going against the fields claims. It probably does have some kind of quality control. I mean, it. But you'd be much better off taking a statistician outside the field just to opine on the methodology. First of all, peer reviewers aren't paid. So, you know, these journals take papers written with public money, charge public institutions, you know, millions of dollars to get access to the research, and they don't pay their reviewers or their authors. In fact, the authors typically have to pay to get published. So these peer reviewers, to truly peer review a paper takes several days of work. You know, I mean, I try to. What I do is I try to replicate the paper. I say, okay, you know, look at, you know, I never get the same results as the authors, although in fairness, sometimes that's because I didn't understand exactly what they did. And when I. Some authors, when you communicate with them, will help you match their paper results, most of them won't. Most of them aren't interested in that. But if I can't replicate the result, I don't understand it. So these peer reviewers are not just doing basic things. And I document this over and over in the book. People have a key point in the paper that the entire paper depends on. I have a chapter on does, you know, smoking marijuana increase heart attacks? And the key thing, these authors were basing their all their data on a telephone survey. And they say these telephone survey results are accurate. And, you know, that's demonstrated in the literature. And they cite two papers, both of which say these data are completely unreliable and can't be used for anything. So none of the peer reviewers ever checked the literature to see if the claims are correct. It's just a click. That's all you have to do. I mean, it's a link in the paper. You just click on it, you read the abstract and say, no, this is the opposite. Moreover, any peer reviewer of this paper should be aware that telephone survey data are not reliable for this kind of work. And they're certainly not doing sophisticated, you know, analysis where the statistical techniques used. Right. You know, is the data correct? They aren't redoing the calculations to see that they're right. What they are doing is saying, yes, this paper is an important original contribution. And that's entirely a matter of opinion of, you know, does it advance the field as I see it, which basically, you know, does it make my research more valuable? Does it, you know, support my idea about the field, or does it, you know, contradict it? So peer review is much worse than useless.
- [28:33]
B
How do government agencies stack up against other institutions in terms of the. The quality of the numbers that they put out?
- [28:41]
C
Well, overall, I'd say better, mostly because most of the stuff the government puts out is intent. First of all, these days with, you know, the National Transportation Safety Board was an exception. Typically, everything about their methodology is public. Lots of researchers use it, and those researchers are in different fields. They come at it different ways. And so on most, at least, government websites, I tend to use, you know, I go to Fred, I go to cdc, Wonder, you know, I mean, I go to these places. They have enormously useful data sorting and selection tools for researchers. So they really are kind of a public utility. And the problems they have. And, you know, there are problems with any data set. They tend to be very well documented. And because different researchers in different fields with different biases are looking at them and talking about them. So overall, I'd say that's better. Unfortunately, you know, Donald Trump is trying to politicize it. We're also seeing a lot of things from the left where they're trying to replace numbers they don't like with their own numbers they do like. But by and large, a government number is more reliable, I would say, than the average academic publication. But only when they're just telling you facts. You know, if they want to tell you the number of people who died from X, Y, z cause, or they're telling you, you know, the number of people in a survey who said they're unemployed and looking for work. When they move from that to interpretation, they're no better than academics.
- [30:12]
B
Well, I can recall a mask study during COVID that was supposed to demonstrate the effectiveness of mask mandates because it was looking at counties in Kansas that were next to each other and presumably demographically identical. They were saying that, look, this one had this mask mandate up till this point or, you know, and this one didn't and that so on and so forth. And they say, look, look at the numbers. I mean, it's just what you would expect. The masks helped and the not masks was very bad decision. But then they just cut the numbers off right there. If you were to be impertinent enough to continue out the numbers, you find that within weeks exactly the opposite occurs, which is hard for them to explain given their worldview. So it's hard to explain. We just don't include it. They're not technically lying. The charts they produced were accurate. There was definitely some of that.
- [31:03]
C
That's absolutely true. And that's again, if the CDC put out a data set and the data set said, you know, county masking rate and county COVID fatality rate, I would trust that data set. And I would. I mean, not 100%, but it's, you know, it's about as good data as you're going to get on something like that. Similarly, you know, when NASA, NASA publishes the NASA National Oceanic Atmosphere Administration and the Space Administration publish global average mean temperatures and you know, I mean, there's a lot of issues with that number sort of conceptually and how you measure whatever, but they do a well documented job. Here's how we calculated that number. When they move to things like masks work or masks next not that's where they tend to be as politicized as any or even more so than an academic. An academic kind of has to worry about the granting agencies and his colleagues and the university promotion committees and places like that. A government agency has to worry about Congress, what they're going to think about what they put out. Still, the world is a better place because we have government statistics. The specific case I talk about in my book that's related to your masking study is the president's Commission on Opioid Use, which just, you know, they, they. The basic situation, the statistical situation is deaths of despair. So we're talking about suicide deaths from alcoholism. Deaths from drug addiction have been increasing since the 1970s, and they increased pretty sharply from the mid-90s to about 2010. And in that exact same period, we got A bunch of new synthetic opioids that were much better than the old ones. They gave more pain relief with less addiction potential. And so they were prescribed a lot. And so the two curves went up at the same time. And in the classic, you know, correlation means causation fallacy, people said, oh, it's the opioids are killing all these people. But then the FDA cracked down on opioids and opioid prescriptions fell dramatically and the deaths from despair kept going up. So in 2017, the President's Commission did a study and they just ignored all the data from the previous seven years that showed, no, these two things are not related. You know, opioids can't be the explanation. And they just relied on the evidence from 1996 to 2010, which did look, if you graphed it, it looked pretty convincing. Yeah, these two things are going up around the same, you know, same time. So again, this is when you ask, you appoint a president's commission, its job is going to be to figure out that, you know, opioids are the cause of all of these problems and it's going to do it. And if that means throwing out all the data that disproves their point, that's what they're going to do.
- [33:55]
B
Hey everybody, Tom woods here with a quick tip for small business owners. If your business isn't showing up online, your competitors are getting the leads and you're missing out. That's where Persist SEO comes in. For over 15 years, they've been helping local businesses grow through SEO, paid ads and the latest in AI powered search optimization so you stay visible and competitive in the digital age. Whether you're in home services, legal or healthcare. Persist SEO delivers real results without locking you into long term contracts or overwhelming you with tech jargon. Visit Ineedseo Help or call 770-580-3736 to schedule your free consultation. That's Ineedseo Help. Easy to remember, powerful for your small business with economics. There's a tradition in economics that views it as a deductive science. And there's another tradition that was amplified by a famous paper by Milton Friedman that argued that it's fundamentally empirical and that we gather data and we draw conclusions. But yet the same people who say that if you devised a study that told you that price controls actually led to greater and greater abundance of goods as opposed to shortages, I don't think they would say, well, the empirical data has overturned the foundations of economics. They'd say, go do that study again. There's something wrong with that study. Now on some level, I think our fundamental understanding of how the world works can be a legitimate check to make us say, wait a minute, maybe there's something wrong with this study. But on the other hand, we have to be careful that we're not actually letting our worldview overrule a good study. You know what I mean? So I mean, it's like I can imagine there are times when it actually has been helpful to say wait a minute, that seems absurd. Maybe there's a mistake. On the other hand, maybe the absurd result is correct, but we're going to have to go back and give it a second look. So in other words, not all second looks, I think are ideologically self interested. They could be legitimately, I'm pretty sure I understand the world and this doesn't make any sense at all. So that's okay. It's more a matter of, I think what's going on is that we probably have a lot of organizations that do studies, the results don't come out the way they want, so we just never hear about the study. That's the end of it.
- [36:15]
C
Yeah, and one way this is getting a little nerdy here, but you know, there's something called a P value and I'm not going to discuss it here. Although you can.
- [36:23]
B
I almost mentioned P values and I thought too nerdy even for me. So go, go for it.
- [36:29]
C
Well, here's the thing though about them. So, so there's a convention in most fields that a 5% P value, if you can get a P value below 5%, you can publish, it's a result. But if you're making an extraordinary claim, like price controls, you know, improve economic growth, you want an extraordinary evidence. So you'd want a p value of 0.01% or something like that. And that's perfectly legitimate. If you're making a perfectly reasonable claim that people are, you know, raising the minimum wage is going to increase unemployment among low skill workers. Even if you can't get a p value below 5%, you're probably, you know, willing to take that. But I do have a very important cautionary tale about that. My dad was a graduate student in physics at the University of Chicago and he had a summer job repairing cosmic ray detectors. I mean, so you drive up to these remote mountaintops where because you have to have the detector, you know, as high as you can. What would happen with the detectors is the, this is back in early 50s, 1950s, the contacts would get dirty or you know, condensation or Something and they would give just incredibly high readings. It would just, you know, do nothing but click. So he took them apart, he'd clean the contacts and put them back together, and they'd get their normal readings again. Well, one day during the semester, he was no longer doing this. The researcher he had been working for was doing it himself. Went up to one of these detectors that was running hot, took it apart and cleaned it, put it back together. It was still running hot. So he did it a second time more carefully, put it back together, and caught the tail end of the biggest cosmic ray event in history. So sometimes when you throw out data that you don't like, you miss the most important thing. There's a good, great Isaac Asimov quote. Isaac Asimov said, you know, the great discoveries in science are not heralded by Eureka. They're heralded by. That's funny. So when you see this funny result, you see the price control seem to be working in some place. You know, and you see, you go to Sweden and you say, well, you know, they have a big social welfare state and people don't seem too miserable about it. You know, these are the things that are worth investigating. You know, you don't turn around suddenly and say, us should start imitating Sweden and everything it does, but it is worth looking at. What do they do? Why do they do it? Why do some things seem to work in Sweden but are disasters in Italy?
- [38:52]
B
What do we say to the average person now who, let's say, reads your book and feels like just despairing like now at this point, how do I know what to believe? So are there pieces of advice you can give to a reasonably intelligent person? Maybe red flags that they might spot or they obviously can't. Nor are you asking them to dismiss out of hand everything they see in the headlines, but at the same time, they have to be cautious. What do you recommend that we do?
- [39:23]
C
Number one is everything we know is Ancient Greek. Xenophanes said is a woven web of conjecture that if you come across a study that isn't woven into a fabric in a field where you can trust the results, it's just something released directly to the media that you're supposed to believe is a fact independent of some field of knowledge that is just automatically suspicious, there's a reason this person hasn't tried to weave their finding into an accepted field. So, you know, if a physicist tells you something about physics, well, you can put a fair amount of trust in that because the field is well woven. And if they're saying something that it's not just their work that they're talking about. They're talking about. This is. Fits into a pattern of knowledge. But if somebody says, you know, USAID saved 91 million lives, well, that's one study. It's not woven into a general theory about how foreign aid saves lives. And it's released direct to the newspapers. It was not vetted or, you know, built into some body of knowledge and some fields, you know, you know, the hard sciences are more trusting. Not medicine, but, you know, physics, chemistry, biology, you know, there's a lot of fraud there, there's a lot of problems, a lot of bad studies, but by and large, there's more good than bad. You start talking about sociology, an awful lot of economics, just ideologically motivated. So what you really want to know is, you know, does this make sense? Does this fit into a larger body of knowledge? But here's where you have to be careful, you know, so you're a libertarian. You kind of have these libertarian mindsets that somebody comes along and says something. Yeah, this, this fits into the general libertarian view. Price controls herd economic growth. Okay, but that doesn't mean it's true, you know, means perhaps you don't have to be as skeptical as if you found the opposite. Another thing is very important, one that I think gets a lot. We hear a lot of studies that talk about statistical lives saved. So that's the USAID and that's the Medicare cuts, things like that. Well, think about all the things that didn't kill you today. You know, there are a lot of them. If you add up the total number of statistical lives saved just by government policies, forget about everything else. Forget about every breath you took, every time, your heartbeat, it adds up to much more than the total number of lives there are. So if somebody tells you that something saved 91 million lives, you ask, compared to what? Well, everybody dies once. Nothing can save lives unless somebody comes up with immortality. So what you're really saying is 91 million people that might have died from this thing died of something else instead. Was that better? Was that worse? I don't know. You know, we have to go through and figure out. And they always tell you one side of the ledger. They say, you know, okay, so if we spend $9 billion of Medicaid increases and you're going to save 10,000 lives, well, what if we'd left that $9 billion in the hands of the taxpayers? You know, would it have saved more lives, would have done more good? And maybe that's the most important thing. Always ask Compared to what? You know, these studies are always a lot better on. Here's what you know, this policy will do. They don't tell you what it will do compared to not doing the policy.
- [42:43]
B
Well, how about this then? Let's say not censorship, of course, but let's say just as a professional norm in science journalism, you could impose one professional norm that you thought the observance of, which would give us better, more reliable journalism and fewer wrong numbers, so to speak, being propagated. What would that be?
- [43:06]
C
It would be review by independent people outside the field that you have. People who publish should not be the people who are deciding what things get published. Every article should go to profession. And these would be statistical professionals. Now here I'm talking my book because, you know, people hire more statisticians. That's good for me, good for my students, and they should be knowledgeable about a field. So I'm a. If I'm a medical reviewer of statistics, you know, I'm a statistician who reviews medical studies, I have to know something about medicine as well. But I don't have any vested interest. And I just, I do this for a salary. I look at the papers, I get paid the same whether I accept them or reject them. I would even, and this is a little bit more radical, I would separate the theoreticians from the empiricists. Kind of like, you know, in physics, you have theoreticians and you have bench scientists and they're not the same. So I want to write a paper proving that Medicaid expansion saves lives. I come up with the claim, I come up with why I think, you know, write all the literature about it, but then I hand it off for somebody else who goes out and tests it, and that person has no vested interest in the result being true or false. They just gather the data and they come back. And they might say the data strongly supported the data, weakly supported the data, contradict it, whatever. But that is more radical. Sort of the idea of the gentleman scientist who comes up with ideas and tests them personally is so deeply ingrained in a lot of fields that we forget they're entirely different skills. And basically what we're doing is we're having a lot of amateurs who have vested interests, who have ideological biases, and they're doing the technical work that really should be done by data professionals.
- [44:50]
B
Well, that seems to make pretty good sense to me. You would think in the age of social media that these errors would be caught and corrected faster. Do you think that's happening?
- [45:03]
C
I actually do. But it sort of depends on how, you know, if you go on Twitter or even worse, blueski or something like that and sort of read the great mass of posts, you wouldn't think that's. You think it's the opposite, if anything else. But I think today, especially if you're using, you know, an AI filter or something like that, if, you know, you get on Claude, I happen to like Claude. But whatever, whichever one you like, and you just ask it something, or you go to Wikipedia, okay, there's a lot of bias there. There's a lot of things you gotta watch out for. But you can get a much better summary of opinion than, you know, 10 years ago and certainly more than 20 or 30 years ago you could have gotten. And you do see the criticisms, you do see the debate about these things. And I think that there are a lot of people working on this. Fortunately, what they're doing is they're trying to fix something that's broken. It would be much better to attack this problem at the source and not publish a lot of bad research in the first place. But given that's true, there are people trying to sift something. And by the way, we're much more hit in finance, in quant finance, when I work for aqr, we're much, much better at this because it's dollars and cents to us. We have to filter out the bad stuff. And we use data science, we use AI, we use things like that to filter it out. And so we can get much closer to the truth than we they put in the past. But if you're not careful about it, if you just have a few favorite, you know, websites or people you trust, it's probably worse now than it ever was.
- [46:35]
B
By the way, that's exactly been my feeling is that even though it's very fashionable to stick your nose up at social media because, you know, we know all the downsides to it, it is sometimes a way to spread information very quickly. And that can also include debunkings of things that need to be debunked. So it's not all bad. There's still some good out there. Well, all the same, I really enjoyed this and I knew about your series and of course that it makes perfect sense to make a book out of it, especially when the examples chosen are so wide ranging across so many fields and yet the same unfortunate phenomena just rear their heads time and time again. And to write about this and identify this problem and urge people to be alert against it is a contribution to society. Aaron Brown so I thank you for that and for your book Wrong how to Extract Truth from a Blizzard of Quantitative Disinformation. I'll have it linked@tomwoods.com 2761 Aaron thanks again.
- [47:39]
C
One more thing. Could you also stick in there a link to the reasoned Wrong Number video page? So even absolutely want to shell out the money for the book, you can get a lot of it for free.
- [47:49]
B
Okay, well I hope folks will still shell out the money, but I will put a link to those videos also@tomwoods.com 2761 all right, thanks a millionaire and we appreciate it.
- [47:58]
C
Thank you very much Tom. Bye bye.
- [48:00]
B
Thank you, ladies and gentlemen.
- [48:04]
A
Make yourself and those you love less vulnerable to the regime, both mentally and physically. Get more forbidden information at Tom's for and be sure to subscribe to the show wherever you listen. See you next time.
- [48:26]
B
Like the sound of the Tom woods show, my audio production is provided by Podsworth Media. Check them out@podsworth.com Enter code WOODS50 to get 50 off your first order. If your recording sounds rough, the Podsworth app can make it not only listenable, but professional. Remember, when you use code WOODS50, you'll get half off your first order and you'll also be supporting this show.