Month: October 2019

Revealed preference: Does the fact that I did it mean I preferred it?

PNRJ critique of neoclassical economics, heuristics and biases , , , , , , ,

Post 312 Oct 27 JDN 2458784

One of the most basic axioms of neoclassical economics is revealed preference: Because we cannot observe preferences directly, we infer them from actions. Whatever you chose must be what you preferred.

Stated so badly, this is obviously not true: We often make decisions that we later come to regret. We may choose under duress, or confusion; we may lack necessary information. We change our minds.

And there really do seem to be economists who use it in this bald way: From the fact that a particular outcome occurred in a free market, they will infer that it must be optimally efficient. (“Freshwater” economists who are dubious of any intervention into markets seem to be most guilty of this.) In the most extreme form, this account would have us believe that people who trip and fall do so on purpose.

I doubt anyone believes that particular version—but there do seem to be people who believe that unemployment is the result of people voluntarily choosing not to work, and revealed preference has also led economists down some strange paths when trying to explain what sure looks like irrational behavior—such as “rational addiction” theory, positing that someone can absolutely become addicted to alcohol or heroin and end up ruining their life all based on completely rational, forward-thinking decision planning.

The theory can be adapted to deal with these issues, by specifying that it’s only choices made with full information and all of our faculties intact that count as revealing our preferences.

But when are we ever in such circumstances? When do we ever really have all the information we need in order to make a rational decision? Just what constitutes intact faculties? No one is perfectly rational—so how rational must we be in order for our decisions to count as revealing our preferences?

Revealed preference theory also quickly becomes tautologous: Why do we choose to do things? Because we prefer them. What do we prefer? What we choose to do. Without some independent account of what our preferences are, we can’t really predict behavior this way.

A standard counter-argument to this is that revealed preference theory imposes certain constraints of consistency and transitivity, so it is not utterly vacuous. The problem with this answer is that human beings don’t obey those constraints. The Allais Paradox, the Ellsberg Paradox, the sunk cost fallacy. It’s even possible to use these inconsistencies to create “money pumps” that will cause people to systematically give you money; this has been done in experiments. While real-world violations seem to be small, they’re definitely present. So insofar as your theory is testable, it’s false.

The good news is that we really don’t need revealed preference theory. We already have ways of telling what human beings prefer that are considerably richer than simply observing what they choose in various scenarios. One very simple but surprisingly powerful method is to ask. In general, if you ask people what they want and they have no reason to distrust you, they will in fact tell you what they want.

We also have our own introspection, as well as our knowledge about millions of years of evolutionary history that shaped our brains. We don’t expect a lot of people to prefer suffering, for instance (even masochists, who might be said to ‘prefer pain’, seem to be experiencing that pain rather differently than the rest of us would). We generally expect that people will prefer to stay alive rather than die. Some may prefer chocolate, others vanilla; but few prefer motor oil. Our preferences may vary, but they do follow consistent patterns; they are not utterly arbitrary and inscrutable.

There is a deeper problem that any account of human desires must face, however: Sometimes we are actually wrong about our own desires. Affective forecasting, the prediction of our own future mental states, is astonishingly unreliable. People often wildly overestimate the emotional effects of both positive and negative outcomes. (Interestingly, people with depression tend not to do this—those with severe depression often underestimate the emotional effects of positive outcomes, while those with mild depression seem to be some of the most accurate forecasters, an example of the depressive realism effect.)

There may be no simple solution to this problem. Human existence is complicated; we spend large portions of our lives trying to figure out what it is we really want.
This means that we should not simply trust that whatever it is happens is what everyone—or even necessarily anyone—wants to happen. People make mistakes, even large, systematic, repeated mistakes. Sometimes what happens is just bad, and we should be trying to change it. Indeed, sometimes people need to be protected from their own bad decisions.

Unsolved problems

PNRJ core principles, critique of neoclassical economics, macroeconomics, public policy, science and academia , , , , , ,

Oct 20 JDN 2458777

The beauty and clearness of the dynamical theory, which asserts heat and light to be modes of motion, is at present obscured by two clouds. The first came into existence with the undulatory theory of light, and was dealt with by Fresnel and Dr. Thomas Young; it involved the question, how could the earth move through an elastic solid, such as essentially is the luminiferous ether? The second is the Maxwell-Boltzmann doctrine regarding the partition of energy.


~ Lord Kelvin, April 27, 1900

The above quote is part of a speech where Kelvin basically says that physics is a completed field, with just these two little problems to clear up, “two clouds” in a vast clear horizon. Those “two clouds” Kelvin talked about, regarding the ‘luminiferous ether’ and the ‘partition of energy’? They are, respectively, relativity and quantum mechanics. Almost 120 years later we still haven’t managed to really solve them, at least not in a way that works consistently as part of one broader theory.

But I’ll give Kelvin this: He knew where the problems were. He vastly underestimated how complex and difficult those problems would be, but he knew where they were.

I’m not sure I can say the same about economists. We don’t seem to have even reached the point where we agree where the problems are. Consider another quotation:

For a long while after the explosion of macroeconomics in the 1970s, the field looked like a battlefield. Over time however, largely because facts do not go away, a largely shared vision both of fluctuations and of methodology has emerged. Not everything is fine. Like all revolutions, this one has come with the destruction of some knowledge, and suffers from extremism and herding. None of this deadly however. The state of macro is good.


~ Oliver Blanchard, 2008

The timing of Blanchard’s remark is particularly ominous: It is much like the turkey who declares, the day before Thanksgiving, that his life is better than ever.

But the content is also important: Blanchard didn’t say that microeconomics is in good shape (which I think one could make a better case for). He didn’t even say that economics, in general, is in good shape. He specifically said, right before the greatest economic collapse since the Great Depression, that macroeconomics was in good shape. He didn’t merely underestimate the difficulty of the problem; he didn’t even see where the problem was.

If you search the Web, you can find a few lists of unsolved problems in economics. Wikipedia has such a list that I find particularly bad; Mike Moffatt offers a better list that still has significant blind spots.

Wikipedia’s list is full of esoteric problems that require deeply faulty assumptions to even exist, like the ‘American option problem’ which assumes that the Black-Scholes model is even remotely an accurate description of how option prices work, or the ‘tatonnement problem’ which ignores the fact that there may be many equilibria and we might never reach one at all, or the problem they list under ‘revealed preferences’ which doesn’t address any of the fundamental reasons why the entire concept of revealed preferences may fail once we apply a realistic account of cognitive science. (I could go pretty far afield with that last one—and perhaps I will in a later post—but for now, suffice it to say that human beings often freely choose to do things that we later go on to regret.) I think the only one that Wikipedia’s list really gets right is ‘Unified models of human biases’. The ‘home bias in trade’ and ‘Feldstein-Horioka Puzzle’ problems are sort of edging toward genuine problems, but they’re bound up in too many false assumptions to really get at the right question, which is actually something like “How do we deal with nationalism?” Referring to the ‘Feldstein-Horioka Puzzle’ misses the forest for the trees. Likewise, the ‘PPP Puzzle’ and the ‘Exchange rate disconnect puzzle’ (and to some extent the ‘equity premium puzzle’ as well) are really side effects of a much deeper problem, which is that financial markets in general are ludicrously volatile and inefficient and we have no idea why.

And Wikipedia’s list doesn’t have some of the largest, most important problems in economics. Moffatt’s list does better, including good choices like “What Caused the Industrial Revolution?”, “What Is the Proper Size and Scope of Government?”, and “What Truly Caused the Great Depression?”, but it also includes some of the more esoteric problems like the ‘equity premium puzzle’ and the ‘endogeneity of money’. The way he states the problem “What Causes the Variation of Income Among Ethnic Groups?” suggests that he doesn’t quite understand what’s going on there either. More importantly, Moffatt still leaves out very obviously important questions like “How do we achieve economic development in poor countries?” (Or as I sometimes put it, “What did South Korea do from 1950 to 2000, and how can we do it again?”), “How do we fix shortages of housing and other necessities?”, “What is causing the global rise of income and wealth inequality?”, “How altruistic are human beings, to whom, and under what conditions?” and “What makes financial markets so unstable?” Ironically, ‘Unified models of human biases’, the one problem that Wikipedia got right, is missing from Moffatt’s list.

And I’m also humble enough to realize that some of the deepest problems in economics may be ones that we don’t even quite know how to formulate yet. We like to pretend that economics is a mature science, almost on the coattails of physics; but it’s really a very young science, more like psychology. We go through these ‘cargo cult science‘ rituals of p-values and econometric hypothesis tests, but there are deep, basic forces we don’t understand. We have precisely prepared all the apparatus for the detection of the phlogiston, and by God, we’ll get that 0.05 however we have to. (Think I’m being too harsh? “Real Business Cycle” theory essentially posits that the Great Depression was caused by everyone deciding that they weren’t going to work for a few years, and as whole countries fell into the abyss from failing financial markets, most economists still clung to the Efficient Market Hypothesis.) Our whole discipline requires major injections of intellectual humility: We not only don’t have all the answers; we’re not even sure we have all the questions.

I think the esoteric nature of questions like ‘the equity premium puzzle’ and the ‘tatonnement problem‘ is precisely the source of their appeal: It’s the sort of thing you can say you’re working on and sound very smart, because the person you’re talking to likely has no idea what you’re talking about. (Or else they are a fellow economist, and thus in on the con.) If you said that you’re trying to explain why poor countries are poor and why rich countries are rich—and if economics isn’t doing that, then what in the world are we doing?—you’d have to admit that we honestly have only the faintest idea, and that millions of people have suffered from bad advice economists gave their governments based on ideas that turned out to be wrong.

It’s really quite problematic how closely economists are tied to policymaking (except when we do really know what we’re talking about?). We’re trying to do engineering without even knowing physics. Maybe there’s no way around it: We have to make some sort of economic policy, and it makes more sense to do it based on half-proven ideas than on completely unfounded ideas. (Engineering without physics worked pretty well for the Romans, after all.) But it seems to me that we could be relying more, at least for the time being, on the experiences and intuitions of the people who have worked on the ground, rather than on sophisticated theoretical models that often turn out to be utterly false. We could eschew ‘shock therapy‘ approaches that try to make large interventions in an economy all at once, in favor of smaller, subtler adjustments whose consequences are more predictable. We could endeavor to focus on the cases where we do have relatively clear knowledge (like rent control) and avoid those where the uncertainty is greatest (like economic development).

At the very least, we could admit what we don’t know, and admit that there is probably a great deal we don’t know that we don’t know.

Mental illness is different from physical illness.

PNRJ cognitive science, personal , , , , , , , , , , , ,

Post 311 Oct 13 JDN 2458770

There’s something I have heard a lot of people say about mental illness that is obviously well-intentioned, but ultimately misguided: “Mental illness is just like physical illness.”

Sometimes they say it explicitly in those terms. Other times they make analogies, like “If you wouldn’t shame someone with diabetes for using insulin, why shame someone with depression for using SSRIs?”

Yet I don’t think this line of argument will ever meaningfully reduce the stigma surrounding mental illness, because, well, it’s obviously not true.

There are some characteristics of mental illness that are analogous to physical illness—but there are some that really are quite different. And these are not just superficial differences, the way that pancreatic disease is different from liver disease. No one would say that liver cancer is exactly the same as pancreatic cancer; but they’re both obviously of the same basic category. There are differences between physical and mental illness which are both obvious, and fundamental.

Here’s the biggest one: Talk therapy works on mental illness.

You can’t talk yourself out of diabetes. You can’t talk yourself out of myocardial infarct. You can’t even talk yourself out of migraine (though I’ll get back to that one in a little bit). But you can, in a very important sense, talk yourself out of depression.

In fact, talk therapy is one of the most effective treatments for most mental disorders. Cognitive behavioral therapy for depression is on its own as effective as most antidepressants (with far fewer harmful side effects), and the two combined are clearly more effective than either alone. Talk therapy is as effective as medication on bipolar disorder, and considerably better on social anxiety disorder.

To be clear: Talk therapy is not just people telling you to cheer up, or saying it’s “all in your head”, or suggesting that you get more exercise or eat some chocolate. Nor does it consist of you ruminating by yourself and trying to talk yourself out of your disorder. Cognitive behavioral therapy is a very complex, sophisticated series of techniques that require years of expert training to master. Yet, at its core, cognitive therapy really is just a very sophisticated form of talking.

The fact that mental disorders can be so strongly affected by talk therapy shows that there really is an important sense in which mental disorders are “all in your head”, and not just the trivial way that an axe wound or even a migraine is all in your head. It isn’t just the fact that it is physically located in your brain that makes a mental disorder different; it’s something deeper than that.

Here’s the best analogy I can come up with: Physical illness is hardware. Mental illness is software.

If a computer breaks after being dropped on the floor, that’s like an axe wound: An obvious, traumatic source of physical damage that is an unambiguous cause of the failure.

If a computer’s CPU starts overheating, that’s like a physical illness, like diabetes: There may be no particular traumatic cause, or even any clear cause at all, but there is obviously something physically wrong that needs physical intervention to correct.

But if a computer is suffering glitches and showing error messages when it tries to run particular programs, that is like mental illness: Something is wrong not on the low-level hardware, but on the high-level software.

These different types of problem require different types of solutions. If your CPU is overheating, you might want to see about replacing your cooling fan or your heat sink. But if your software is glitching while your CPU is otherwise running fine, there’s no point in replacing your fan or heat sink. You need to get a programmer in there to look at the code and find out where it’s going wrong. A talk therapist is like a programmer: The words they say to you are like code scripts they’re trying to get your processor to run correctly.

Of course, our understanding of computers is vastly better than our understanding of human brains, and as a result, programmers tend to get a lot better results than psychotherapists. (Interestingly they do actually get paid about the same, though! Programmers make about 10% more on average than psychotherapists, and both are solidly within the realm of average upper-middle-class service jobs.) But the basic process is the same: Using your expert knowledge of the system, find the right set of inputs that will fix the underlying code and solve the problem. At no point do you physically intervene on the system; you could do it remotely without ever touching it—and indeed, remote talk therapy is a thing.

What about other neurological illnesses, like migraine or fibromyalgia? Well, I think these are somewhere in between. They’re definitely more physical in some sense than a mental disorder like depression. There isn’t any cognitive content to a migraine the way there is to a depressive episode. When I feel depressed or anxious, I feel depressed or anxious about something. But there’s nothing a migraine is about. To use the technical term in cognitive science, neurological disorders lack the intentionality that mental disorders generally have. “What are you depressed about?” is a question you usually can answer. “What are you migrained about?” generally isn’t.

But like mental disorders, neurological disorders are directly linked to the functioning of the brain, and often seem to operate at a higher level of functional abstraction. The brain doesn’t have pain receptors on itself the way most of your body does; getting a migraine behind your left eye doesn’t actually mean that that specific lobe of your brain is what’s malfunctioning. It’s more like a general alert your brain is sending out that something is wrong, somewhere. And fibromyalgia often feels like it’s taking place in your entire body at once. Moreover, most neurological disorders are strongly correlated with mental disorders—indeed, the comorbidity of depression with migraine and fibromyalgia in particular is extremely high.

Which disorder causes the other? That’s a surprisingly difficult question. Intuitively we might expect the “more physical” disorder to be the primary cause, but that’s not always clear. Successful treatment for depression often improves symptoms of migraine and fibromyalgia as well (though the converse is also true). They seem to be mutually reinforcing one another, and it’s not at all clear which came first. I suppose if I had to venture a guess, I’d say the pain disorders probably have causal precedence over the mood disorders, but I don’t actually know that for a fact.

To stretch my analogy a little, it may be like a software problem that ends up causing a hardware problem, or a hardware problem that ends up causing a software problem. There actually have been a few examples of this, like games with graphics so demanding that they caused GPUs to overheat.

The human brain is a lot more complicated than a computer, and the distinction between software and hardware is fuzzier; we don’t actually have “code” that runs on a “processor”. We have synapses that continually fire on and off and rewire each other. The closest thing we have to code that gets processed in sequence would be our genome, and that is several orders of magnitude less complex than the structure of our brains. Aside from simply physically copying the entire brain down to every synapse, it’s not clear that you could ever “download” a mind, science fiction notwithstanding.

Indeed, anything that changes your mind necessarily also changes your brain; the effects of talking are generally subtler than the effects of a drug (and certainly subtler than the effects of an axe wound!), but they are nevertheless real, physical changes. (This is why it is so idiotic whenever the popular science press comes out with: “New study finds that X actually changes your brain!” where X might be anything from drinking coffee to reading romance novels. Of course it does! If it has an effect on your mind, it did so by having an effect on your brain. That’s the Basic Fact of Cognitive Science.) This is not so different from computers, however: Any change in software is also a physical change, in the form of some sequence of electrical charges that were moved from one place to another. Actual physical electrons are a few microns away from where they otherwise would have been because of what was typed into that code.

Of course I want to reduce the stigma surrounding mental illness. (For both selfish and altruistic reasons, really.) But blatantly false assertions don’t seem terribly productive toward that goal. Mental illness is different from physical illness; we can’t treat it the same.

Trump is finally being impeached

PNRJ current events, ethics, public policy , , , , , , , , , , , , ,

Post 310 Oct 6 JDN 2458763

Given that there have been efforts to impeach Trump since before he took office (which is totally unprecedented, by the way; while several others have committed crimes and been impeached while in office, no other US President has gone into office with widespread suspicion of mass criminal activity), it seems odd that it has taken this long to finally actually start formal impeachment hearings.

Why did it take so long? We needed two things to happen: One, absolutely overwhelming evidence of absolutely flagrant crimes, and two, a Democratic majority in the House of Representatives.

This is how divided America has become. If the Republicans were really a mainstream center-right party as they purport to be, they would have supported impeachment just as much as the Democrats, we would have impeached Trump in 2017, and he would have been removed from office by 2018. But in fact they are nothing of the sort. The Republicans no longer believe in democracy. The Democrats are a mainstream center-right party, and the Republicans are far-right White-nationalist crypto-fascists (and less ‘crypto-‘ all the time). After seeing how they reacted to his tax evasion, foreign bribes, national security leaks, human rights violations, obstruction of justice, and overall ubiquitous corruption and incompetence, by this point it’s clear that there is almost nothing that Trump could do which would make either the voter base or the politicians of the Republican Party turn against him—he may literally be correct that he could commit a murder in broad daylight on Fifth Avenue. Maybe if he raised taxes on billionaires or expressed support for Roe v. Wade they would finally revolt.

Even as it stands, there is good reason to fear that the Republican-majority Senate will not confirm the impeachment and remove Trump from office. The political fallout from such a failed impeachment is highly uncertain. So far, markets are taking it in stride; it may even turn out to be good for the economy. (Then again, a good economy may be good for Trump in 2020!) But at this point the evidence is so damning that if we don’t impeach now, we may never impeach again; if this isn’t enough, nothing is. (The Washington Examiner said that months ago, and may already have been right; but the case is even stronger now.)

So, the most likely scenario is that the impeachment goes through the House, but fails in the Senate. The good news is that if the Republicans do block the impeachment, they’ll be publicly admitting that even charges this serious and this substantiated mean nothing to them. Anyone watching who is still on the fence about them will see how corrupt they have become.

After that, this is probably what will happen: The impeachment will be big news for a month or two, then be largely ignored. Trump will probably try to make himself a martyr, talking even louder about ‘witch hunts’. He will lose popularity with a few voters, but his base will continue to support him through thick and thin. (Astonishingly, almost nothing really seems to move his overall approval rating.) The economy will be largely unaffected, or maybe slightly improve. And then we’ll find out in the 2020 election whether the Democrats can mobilize enough opposition to Trump, and—just as importantly—enough support for whoever wins the primaries, to actually win this time around.

If by some miracle enough Republicans find a moral conscience and vote to remove Trump from office, this means that until 2020 we will have President Mike Pence. In a sane world, that in itself would sound like a worst-case scenario; he’s basically a less-sleazy Ted Cruz. He is misogynistic, homophobic, and fanatically religious. He is also a partisan ideologue who toes the party line on basically every issue. Some have even argued that Pence is worse than Trump, because he represents the same ideology but with more subtlety and competence.

But subtlety and competence are important. Indeed, I would much rather have an intelligent, rational, competent ideologue managing our government, leading our military, and controlling our nuclear launch codes than an idiotic, narcissistic, impulsive one. Pence at least can be trusted to be consistent in his actions and diplomatic in his words—two things which Trump has absolutely never been.

Indeed, Pence’s ideological consistency has benefits; unlike Trump, he reliably supports free trade and his fiscal conservatism actually seems genuine for once. Consistency in itself has value: Life is much easier, and the economy is much stronger, when the rules of the game remain the same rather than randomly lurching from one extreme to another.

Pence is also not the pathological liar that Trump is. Yes, Pence has lied many times (only 22% of his statements were evaluated by PolitiFact as “Mostly True” or “True”, and 30% were “False” or “Pants on Fire”). But Trump lies constantly. A mere 14% of Trump’s statements were evaluated by Trump as “Mostly True” or “True”, while 48% were “False” or “Pants on Fire”. For Bernie Sanders, 49% were “Mostly True” or better, and only 11% were “False”, with no “Pants on Fire” at all; for Hillary Clinton, 49% were “Mostly True” or better, and only 10% were “False”, with 3% “Pants on Fire”. People have tried to keep running tallies of Trump’s lies, but it’s a tall order: The Washington Post records over 12,000 lies since he took office less than three years ago. Four thousand lies a year. More than ten every single day. Most people commit lies of omission or say ‘white lies’ several times per day (depending on who you ask, I’ve seen everything from an average of 2 times per day to an average of 100 times per day), but that’s not what we’re talking about here. These are consequential, outright statements of fact that aren’t true. And these are not literally everything he has said that wasn’t true; they are only public lies with relevance to policy or his own personal record. Indeed, Trump lies recklessly, stupidly, pointlessly, nonsensically. He seems like a pathological liar, or someone with dementia who is confabulating to try to fill gaps in his memory. (Indeed, a lot of his behavior is consistent with dementia, and similar to how Reagan acted in the early days of his Alzheimer’s.) At least if Pence takes office, we’ll be able to believe some of what he says.

Of course, Pence won’t be much better on some of the most important issues, such as climate change. When asked how important he thinks climate change is and what should be done about it, Pence always gives mealy-mouthed, evasive responses—but at least he doesn’t make up stories about windmills getting special permits to kill endangered birds.

I admit, choosing Pence over Trump feels like choosing to get shot in the leg instead of the face—but that’s really not a difficult choice, is it?