A Precommitment

Academia has been good to me so far.  For as long as I’ve been a professor, I have tried to speak politely, thoughtfully, and candidly –  privately as well publicly.  From where I’m sitting, the system treats me quite well.

Perhaps, however, I’ve simply been lucky.  Or perhaps the system is swiftly decaying right before my unperceiving eyes.

In case either of these pessimistic scenarios turn out to be correct, I am now making a precommitment.  Namely:

I will never apologize for politely saying or writing anything that seems reasonable to me… except under extreme duress.

So if I ever do so apologize, please assume extreme duress.

P.S. If I politely say or write something that seems reasonable to me, but later turns out to be false, I will acknowledge and correct my mistake.  But as long as you’ve applied this due diligence, error is not blameworthy and warrants no apology.


Read More

Wiblin, Probability, and Nuclear War

Here’s a guest post from the noble Rob Wiblin of 80,000 Hours.  Posted with Rob’s permission.

I’ve periodically read commenters online say that with random unprecedented events (e.g. a total nuclear war) one can’t give meaningful Bayesian probabilities and therefore the probability of e.g. a nuclear war over the next 100 years is 50/50.

Tabarrok got this from many people in response to his series of blog posts on the likelihood of nuclear war. It’s hard to believe these people are serious, but they are and insist on it even when pressed.

I don’t know what university course melted their brain, but evidently one of them did!

The fastest way to show this is wrong is to ask them three probability questions simultaneously:

1. What is the probability of a single total nuclear war over the next 100 years?
2. What is the probability of a single total nuclear war between between 2121 and 2221?
3. What is the probability of one or more total nuclear wars occurring over the next 200 years?

Someone with this philosophy must respond 50/50 to all of them which leads to an internal contradiction.


Read More

Information, behavior, and frictions (short course on economics)

“Economics is really about two stories. One is the story of the old economist and younger economist walking down the street, and the younger economist says, ‘Look, there’s a hundred-dollar bill,’ and the older one says, ‘Nonsense, if it was there somebody would have picked it up already.’ So sometimes you do find hundred-dollar bills lying on the street, but not often—generally people respond to opportunities. The other is the Yogi Berra line ‘Nobody goes to Coney Island anymore; it’s too crowded.’ That’s the idea that things tend to settle into some kind of equilibrium where what people expect is in line with what they actually encounter.”
― Paul Krugman

I love this quote. These two jokes do sort of describe economics, at least what you might call “pure” or celestial economics, where there are no real world frictions to worry about. If we add frictions to the mix we get terrestrial economics, the economics of the real world.

Krugman’s first joke gets at the way economists think about information. It’s essentially epistemology, about whether to believe something is true. The second joke is about behavior, about how we model the response of people to changes in their environment. At the end I’ll add frictions, and try to give you a sense of how economists like me think about the world.  So here’s my 4-minute course on economics.

1. Information:  Imagine a giant encyclopedia, written in Japanese.  It contains a vast amount of information about the world.  But to read it one must first learn Japanese.  By learning economics we are able to read an enormous amount of information from prices, if we assume that people are rational utility maximizers.  Thus if conventional Treasury bonds yield 7% and indexed bonds yield 3%, we can infer that optimal forecast of inflation is 4%.  If that were not true, there would be $100 bills lying on the sidewalk for investors to pick up.

Here’s another example.  Suppose there’s a neighborhood of cookie-cutter homes on the Irvine/Lake Forest boundary, here in Orange County, CA.  If we compare two similar houses on each side the border, we can infer the difference in value that people see in living in each city, capitalized into the home value (actually land value.)  Most likely, this reflects differences in the perceived value of the two school systems, with Irvine viewed as superior.  If the price gap didn’t reflect amenity differences then homebuyers would take advantage of any mis-pricing.

Here’s another example.  Assume that Mexican farm workers in California earn $11/hour on average while Central American farm workers earn $10/hour on average.  We can infer that the Mexican farm workers are probably about 10% more productive, on average, otherwise California farmers would choose to hire Central Americans, not Mexicans.  Again, no $100 bills on the sidewalk.

The economy contains billions of such pieces of information, all embedded in prices, for those who know how to “read economics”.  It’s like a giant encyclopedia.

2. Behavior:  Economists assume that rational people will keep doing X up until the point where the benefit of one more (marginal) unit of X no longer exceeds the marginal cost of X.  This is the “equilibrium”.  X could be any activity: units consumed, hours worked, dollars invested, etc.:

Urban planners often suggest that expanding highways does not reduce traffic congestion.  That’s wrong.  If you widen a highway, more people will travel on the highway.  That part is true.  But traffic congestion will be reduced; indeed it must be in order to induce more people to travel on the highway.  Why do urban planners get this wrong?  Because they noticed that traffic did not seem to improve when places like Orange County built more highways.  But that’s because both lines were shifting at the same time, as Orange County’s population was growing rapidly when it was building new roads.  It is still true that, other things equal, building more highways reduces traffic congestion.  Recently, Orange County’s population stopped growing.  Now if they were to build more highways in OC, it really would reduce traffic congestion.

Or consider how firms respond to a change in the cost of inputs.  Most students understand that firms will respond to cost increases by raising prices, but often fail to see that firms will respond to price decreases by cutting prices.  But the model is symmetrical; a shift up in the MC curve has the opposite effect of a shift downward, even for a 100% monopoly.  In both cases, cost curve shifts move the optimal output point, which requires a price change.  And we know that firms don’t want to leave $100 bills on the sidewalk.

So that’s celestial economics in a nutshell.  It describes a world where inefficiencies should be quickly eliminated, as utility maximizers do deals to improve efficiency and share the gains.  No $100 bills left on the sidewalk.

3.  Frictions:  Here in the real world, things don’t work so smoothly.  One friction is transactions costs.  In principle, inefficiencies related to externalities (pollution, etc.) and monopoly could be eliminated through negotiations.  Pollution victims could bribe factories not to pollute.  Monopolies could negotiate perfect price discrimination with consumers.  But such negotiations are often costly and hard to do, for all sorts of reasons.  Another friction is sticky wages and prices, which result in nominal shocks having real effects.  Bad real effects, such as high unemployment.  Another friction is illiquidity, which explains why the TIPS spread may not perfectly measure the public’s inflation expectations.  TIPS are less liquid, and hence less desirable.  Furthermore, information is costly.  So there may be a few $100 bills on the sidewalk because no one spent the resources to look for those $100 bills.  And there’ll be lots of coins on the sidewalk.

Left leaning economists focus more on frictions.  Right leaning economists focus more on celestial economics.  I’m somewhere in the middle, but definitely leaning to the right.




Read More

Persecution Bet

I’ve already seen George Mason University badly mistreat non-conformist professors.  Should I expect the same to happen to me?  I wouldn’t be surprised, but I’m not greatly worried either.  I’m definitely not going to self-censor to protect myself.  Yet quite a few people tell me that I’m in grave danger.  One of them, Todd Proebsting, a C.S. professor at the University of Arizona, has offered to bet me.

To be clear, Todd wishes me well.  He does not want me to suffer mistreatment; he merely predicts it.  Here are the terms on which we have agreed.

I bet Todd Proebsting $50 at even odds that I will NOT be “clearly mistreated” by George Mason University before January 1, 2031.

If there is a dispute, Phil Magness will have final authority to determine the winner of the bet.

Examples of “clear mistreatment” include but are not limited to:

1. Any public announcement by an agent of my university’s administration that my speech or writings are under investigation, or a subject of further inquiry.

2. Any GMU investigation or inquiry into my speech or writing. This could be done by administrators or an ad hoc committee or HR or whatever.

3. Any official disciplinary action due to my speech or writing, even if limited to a mere “warning”.

4. Cancellation of a public talk by me at GMU.

5. Any negative financial consequences explicitly justified by my speech or writings.

6. “Clear mistreatment” does NOT include any employee of GMU expressing disapproval of the content of my speech or writings, but DOES include any official personal or professional condemnation.  So: Todd doesn’t win for “GMU strongly disagrees with what Prof. Caplan said about pickpockets,” but Todd does win for “GMU strongly condemns Prof. Caplan for saying that about pickpockets.”


Read More

I Win My Trump Bet

Back when Trump was President-Elect Trump, I made the following bet:

If Donald Trump dies in office, resigns, is removed  by the Senate after impeachment, or otherwise is permanently removed as per the the 25th Amendment, or if it never happens that he takes the Oath of office as POTUS on Jan 20, 2017, the BC owes [redacted] $350. Otherwise, [redacted] owes BC $100.

Since Trump has now completed his term, I have won this bet, bringing my cumulative public betting record to 22 wins, 0 losses.  False modesty aside, I take great pride in my track record.

Surprisingly, I could have gotten much better odds if I had made this bet anytime from late 2017 to the present.  Check out the historical odds.

This market wasn’t open when I made the original bet, but I discovered the existence of this Trump betting market by early 2018.  So why didn’t I leap on this opportunity to raise the stakes at much better odds?  If I leapt into the market when it opened, my odds would have been even instead of 7:2 against me.

So what gives?  My honest answer is a combination of (a) I was disturbed to learn that the market was so against me, and (b) I didn’t (and still don’t) know how to place a bet in this market.  In hindsight, of course, I wish I’d bet the farm.  Live and learn.


Read More

Betting vs. the Nuclear Option

I have a long-standing dispute with Tyler Cowen about the epistemic value of betting.  To my mind, my position is modest: Bets advance our knowledge by clarifying contentions and raising the price of error.  While isolated bets don’t “prove” anything, they tip the argumentative scales in favor of the winner.  And a long track record of successful betting is a strong sign of reliable judgment.  Furthermore, the refusal to make bets at all shows that deep down, you know that you don’t know what you’re talking about.

In contrast, Tyler’s position is extreme: Bets have zero epistemic value.

Why would he think this?  While Tyler has multiple arguments against the epistemic value of betting, his “nuclear option” is just “If you’re so foresighted, why aren’t you the proud owner of a successful hedge fund?”  Instead of betting your friends that the unemployment rate will sharply fall in a few years, why haven’t you mortgaged your home to buy the S&P?  (Better yet, buy option contracts that only pay if the S&P sharply rises).  I’ve already heavily criticized this argument, but now I’d like to share yet another reason why Tyler is deeply mistaken.

Suppose I say, “Unemployment will fall to 4% by January 1, 2022.”  You say I’m wrong – and use Tyler’s nuclear option against me.  “If that’s right, you should leverage every penny you can on the S&P.”  Sounds good, right?

Think again.  It’s totally possible that the market already agrees with me!  The current level of the S&P could be exactly the level you’d expect if the unemployment rate were destined to equal 4% on January 1, 2022.  If so, my claims about unemployment and my disinterest in buying stock options are fully consistent.

In sharp contrast, it makes no sense to simultaneously hold my view on the path of unemployment and refuse to bet on unemployment.  Why?  Because as long as I’m right and you disagree, I will profit.

I don’t need to worry about “what the market has already figured out.”

I don’t need to understand a thousand other factors that influence the S&P.

I don’t need a Theory of Everything.

As long as I genuinely know one disputed fact about the future, and my opponents are intellectually honest, they will bet me and I will take their money.

End of story.

P.S. In 19 days, I expect my betting track record will hit 22 for 22.



Read More

Epistemology, Economics, and Conspiracies

Epistemology (the philosophy of knowledge) is important because it underlies the problem of truth in economics and in all other area of rational research and discourse. Epistemology is also relevant to conspiracies theories. As philosopher Robert Nozick pointed out, in the social sciences, invisible-hand explanations are always preferable because otherwise the conclusion is planted in the premises–a vindication of Adam Smith and classical-liberal economics!

The Ptolemaic system of astronomy also faced an epistemological trap in explaining the movement of planets and stars with the help of epicycles (cycles moving on other circles). When an empirical observation contradicted the system’s predictions, the astronomer only needed to add an epicycle to make the theory fit the fact. Similarly, adding one new conspirator or a new conspiratorial component can always explain away, ad infinitum, any objection to a given conspiracy. Only much later, with the work of mathematician Jean-Baptiste Joseph Fourier in the 18th century, did we start understanding that any smooth curve or movement in space can be approximated with a sufficient number of epicycles.

Ptolemy’s theory was more complicated than needed to understand, and to better understand, the movement of planets and stars. Just like Ptolemaic astronomy, conspiracy theories (at least complex ones) violate Occam’s razor, that is, the principle that “pluralitas non est ponenda sine necessitate,” or “plurality should not be posited without necessity.” In other words, of two explanatory theories, the simplest one should be preferred ceteris paribus. Granted that it is not always clear what “the simplest” means.

Conspiracies are not impossible, but the more complex and the less incentive-compatible they are, the lower their probability. (See my post “Why a Vast Election Fraud is Highly Implausible” and its complement, “Implausible Conspiracy and Unfair Election.)

The shaky epistemological status of conspiracy theories can be illustrated by a recent Facebook post of mine and the comment of Professor Sinclair Davidson, an economist at RMIT University in Australia. I posted:

Here is another [I should have written: “the correct”] conspiracy theory: The Deep State approached Trump around 2015 and asked him to run for president, assuring him of their support. “We know how to run elections,” they told him. The Deep State needed some puppet or clown who would make individual liberty (including the 1st and 2nd Amendments) look totally cranky, thereby preparing the terrain for a future dictator. They told Trump that only he, with his genius, his legendary honesty, and his golf game, could play this important role. Alas, Trump fell in love with the job (as he did with the North Korean dictator), the tweets, the honors, the constant attention, and broke with his Deep State handlers. We saw the consequence on November 3.

Sinclair Davidson brillantly commented:

I have a different theory: Deep state approached Trump exactly like you said but lost control of the 2016 election. He was the patsy meant to lose. Now we see 2020.

Conspiracies can explain any event (even in the physical world if the gods, like Greek gods, engage in conspiracies), and a large number of different conspiracies can explain the same event. Hence conspiracy theories are generally useless, at best.


Read More

188 Years After the Death of Jean-Baptiste Say

Sunday November 15 will mark the 188th anniversary of the death of Jean-Baptiste Say (1767-1832), author of the Traité d’économie politique, whose first edition appeared in 1803. The 4th edition (1819) was translated in English and published as A Treatise on Political Economy (1821). I recently directed a Liberty Fund conference of this great economist, mainly known as the discoverer of Say’s Law (supply creates its own demand), against which John Maynard Keynes more or less conceived his 1936 General Theory of Employment, Interest and Money.

Among his many ideas that preceded today’s economics by one or two centuries, Say explained that the middlemen between the producer and the final consumer play an efficient role by moving goods to where the consumer can purchase them. The middlemen create value too. Say also developed an idea that many of our contemporaries—think of the defenders of “price-gouging” laws—still do not grasp. He explained how the speculators and hoarders benefit the consumer:

There is a further branch of commerce, called the trade of speculation, which consists in the purchase of goods at one time, to be re-sold in the same place and condition at another time, when they are expected to be dearer. Even this trade is productive; its utility consists in the employment of capital, warehouses, care in the preservation, in short, human industry in the withdrawing from circulation a commodity depressed in value by temporary superabundance, and thereby reduced in price below the charges of production, so as to discourage its production, with the design and purpose of restoring it to circulation when it shall become more scarce, and when its price shall be raised above the natural price, the charges of production, so as to throw a loss upon the consumers. The evident operation of this kind of trade is to transport commodities in respect of time, instead of locality.

The last sentence is the crux of the matter. Let me also quote these lines in their original French, as they appeared in the 6th edition of the Treatise, the last one in Say’s lifetime. The text is much clearer than the previous translation:

Il y a un commerce qu’on appelle de spéculation, et qui consiste à acheter des marchandises dans un temps pour les revendre au même lieu et intactes, à une époque où l’on suppose qu’elles se vendront plus cher. Ce commerce lui-même est productif ; son utilité consiste à employer des capitaux, des magasins, des soins de conservation, une industrie enfin, pour retirer de la circulation une marchandise lorsque sa surabondance l’avilirait, en ferait tomber le prix au-dessous de ses frais de production, et découragerait par conséquent sa production, pour la revendre lorsqu’elle deviendra trop rare, et que, son prix étant porté au-dessus de son taux naturel des frais de production, elle causerait de la perte à ses consommateurs. Ce commerce tend, comme on voit à transporter, pour ainsi dire, la marchandise d’un temps à un autre, au lieu de la transporter, d’un endroit dans un autre.

The formulation could perhaps have been more general—for the same speculation happens when, say, facemasks are not produced under their cost of production but the speculator foresees that their demand will jump over the current quantity supplied. But remember that Say wrote in the early 19th century, just a quarter of a century after Adam Smith’s The Wealth of Nations, at a time when the first conceptualization of this sort of economic problems was attempted. In fact, such issues are so difficult to understand that many among our intelligent contemporaries still don’t.

On Sunday, raise your glass to Say.

Yet, a historical mystery remains. Why was classical liberalism, which was then on its rise, so rapidly restrained by reactionary opinions? Is the classical-liberal or libertarian ideal a mirage? Jean-Baptiste Say foresaw an explanation:

To speak the truth, it is because the first principles of political economy are as yet but little known; because ingenious systems and reasonings have been built upon hollow foundations, and taken advantage of, on the one hand, by interested rulers, who employ prohibition as a weapon of offence or as an instrument of revenue; and, on the other, by the personal avarice of merchants and manufacturers, who have a private interest in exclusive measures, and take but little pains to inquire, whether their profits arise from actual production, or from a simultaneous loss thrown upon other classes of the community.

A good explanation, no doubt, and which was much improved by the public-choice school of economics that developed a century or more after the Treatise. But is it sufficient?


Read More


I’ve spent over 30 years arguing about ideas.  During those decades, I’ve learned a lot.  I’ve changed my mind.  I’ve changed minds.

Normally, however, arguing about ideas is fruitless.  Tempers fray.  Discussion goes in circles.  Each and every mental corruption that Philip Tetlock has explored rears itself ugly epistemic head.  You even lose friends.

When a conversation goes off the rails, I’m sorely tempted to bluntly assess the other party’s deep intellectual flaws.  (As I repeatedly told my mom when I was a teenager, “When will you get it through your thick skull that…”)  You don’t have to master Dale Carnegie’s How to Win Friends and Influence People to predict the results.  The other party typically has the temerity to bluntly assess my deep intellectual flaws, which in turn sparks an even more unpleasant, fruitless, and potentially friendship-ending exchange.

The wise approach to fruitless argument, rather, is to politely disengage.  Yet how can you do this without counter-productively moving the conversation from bad to worse?

The classic move is to make “one last point,” then terminate the conversation.  Again, you don’t have to master Carnegie to predict the results.  The other side rushes to get in their “one last point” and the cycle of suffering resumes.

A better approach is to meekly announce, “I can’t think of anything else productive to say.”  Alas, this is still red meat in the eyes of many disputants.  “Aha, so you can’t even answer my brilliant arguments.  Typical!”

The best ejector button I’ve discovered so far is a single word: “Impasse.”  You can stretch it out to, “I fear we’ve reached an impasse,” but even that provides a hand-hold for the other party to say, “Oh, we’ve reached an impasse, eh?  Speak for yourself.”  When you say, “Impasse” and stop talking, the conversation swiftly ends.  The other side won’t like it, but at this point you should meet further taunts with a silent shrug.  While this might spawn a grudge, it’s less likely to do so than further wasted words.

Admittedly, if your real goal is to manipulate the other party into purging you, your best bet is probably Agree and Amplify.  But if that’s your goal, you have no need of my help.


Read More

2020 Election Bets

Virtually any intellectual serious writing will be drowned out during Election Week, so I’m pausing until next Monday.  Instead, I’ll just remind you of some bets.

In the 2016 election, I had three outstanding bets.  I won all three. This year, I only made one election bet, which I’ve already won.  I do however have one outstanding election-related bet: That Trump does not leave office early. Specifically:

If Donald Trump dies in office, resigns, is removed  by the Senate after impeachment, or otherwise is permanently removed as per the the 25th Amendment, or if it never happens that he takes the Oath of office as POTUS on Jan 20, 2017, the BC owes [redacted] $350. Otherwise, [redacted] owes BC $100″.

There is considerable speculation that Trump, if defeated, will not gracefully acknowledge defeat.  I considered this a plausible scenario back in this 2016 bet, though I ultimately won the conventional way.

This time around, many suspect Trump will actually try to illegally cling to office.  While you can imagine that this attempted clinging will get him removed early, I seriously doubt it will.  If Trump stubbornly refuses to leave office, the system will just let him run out his clock, ranting all the while, then sideline him on Inauguration Day.

P.S. Remember my Twitter poll from a year ago?  It was really about Bolivia…


Read More