The extreme efficiency of environmental regulation—and the extreme inefficiency of war

Apr 8 JDN 2458217

Insofar as there has been any coherent policy strategy for the Trump administration, it has largely involved three things:

  1. Increase investment in military, incarceration, and immigration enforcement
  2. Redistribute wealth from the poor and middle class to the rich
  3. Remove regulations that affect business, particularly environmental regulations

The human cost of such a policy strategy is difficult to overstate. Literally millions of people will die around the world if such policies continue. This is almost the exact opposite of what our government should be doing.

This is because military is one of the most wasteful and destructive forms of government investment, while environmental regulation is one of the most efficient and beneficial. The magnitude of these differences is staggering.

First of all, it is not clear that the majority of US military spending provides any marginal benefit. It could quite literally be zero. The US spends more on military than the next ten countries combined.

I think it’s quite reasonable to say that the additional defense benefit becomes negligible once you exceed the sum of spending from all plausible enemies. China, Russia, and Saudi Arabia together add up to about $350 billion per year. Current US spending is $610 billion per year. (And this calculation, by the way, requires them all to band together, while simultaneously all our NATO allies completely abandon us.) That means we could probably cut $260 billion per year without losing anything.

What about the remaining $350 billion? I could be extremely generous here, and assume that nuclear weapons, alliances, economic ties, and diplomacy all have absolutely no effect, so that without our military spending we would be invaded and immediately lose, and that if we did lose a war with China or Russia it would be utterly catastrophic and result in the deaths of 10% of the US population. Since in this hypothetical scenario we are only preventing the war by the barest margin, each year of spending only adds 1 year to the lives of the war’s potential victims. That means we are paying some $350 billion per year to add 1 year to the lives of 32 million people. That is a cost of about $11,000 per QALY. If it really is saving us from being invaded, that doesn’t sound all that unreasonable. And indeed, I don’t favor eliminating all military spending.

Of course, the marginal benefit of additional spending is still negligible—and UN peacekeeping is about twice as cost-effective as US military action, even if we had to foot the entire bill ourselves.

Alternatively, I could consider only the actual, documented results of our recent military action, which has resulted in over 280,000 deaths in Iraq and 110,000 in Afghanistan, all for little or no apparent gain. Life expectancy in these countries is about 70 in Iraq and 60 in Afghanistan. Quality of life there is pretty awful, but people are also greatly harmed by war without actually dying in it, so I think a fair conversion factor is about 60 QALY per death. That’s a loss of 23.4 MQALY. The cost of the Iraq War was about $1.1 trillion, while the cost of the Afghanistan War was about a further $1.1 trillion. This means that we paid $94,000 per lost QALY. If this is right, we paid enormous amounts to destroy lives and accomplished nothing at all.

Somewhere in between, we could assume that cutting the military budget greatly would result in the US being harmed in a manner similar to World War 2, which killed about 500,000 Americans. Paying $350 billion per year to gain 500,000 QALY per year is a price of $700,000 per QALY. I think this is about right; we are getting some benefit, but we are spending an enormous amount to get it.

Now let’s compare that to the cost-effectiveness of environmental regulation.

Since 1990, the total cost of implementing the regulations in the Clean Air Act was about $65 billion. That’s over 28 years, so less than $2.5 billion per year. Compare that to the $610 billion per year we spend on the military.

Yet the Clean Air Act saves over 160,000 lives every single year. And these aren’t lives extended one more year as they were in the hypothetical scenario where we are just barely preventing a catastrophic war; most of these people are old, but go on to live another 20 years or more. That means we are gaining 3.2 MQALY for a price of $2.5 billion. This is a price of only $800 per QALY.

From 1970 to 1990, the Clean Air Act cost more to implement: about $520 billion (so, you know, less than one year of military spending). But its estimated benefit was to save over 180,000 lives per year, and its estimated economic benefit was $22 trillion.

Look at those figures again, please. Even under very pessimistic assumptions where we would be on the verge of war if not for our enormous spending, we’re spending at least $11,000 and probably more like $700,000 on the military for each QALY gained. But environmental regulation only costs us about $800 per QALY. That’s a factor of at least 14 and more likely 1000. Environmental regulation is probably about one thousand times as cost-effective as military spending.

And I haven’t even included the fact that there is a direct substitution here: Climate change is predicted to trigger thousands if not millions of deaths due to military conflict. Even if national security were literally the only thing we cared about, it would probably still be more cost-effective to invest in carbon emission reduction rather than building yet another aircraft carrier. And if, like me, you think that a child who dies from asthma is just as important as one who gets bombed by China, then the cost-benefit analysis is absolutely overwhelming; every $60,000 spent on war instead of environmental protection is a statistical murder.

This is not even particularly controversial among economists. There is disagreement about specific environmental regulations, but the general benefits of fighting climate change and keeping air and water clean are universally acknowledged. There is disagreement about exactly how much military spending is necessary, but you’d be hard-pressed to find an economist who doesn’t think we could cut our military substantially with little or no risk to security.

In defense of slacktivism

Jan 22, JDN 2457776

It’s one of those awkward portmanteaus that people often make to try to express a concept in fewer syllables, while also implicitly saying that the phenomenon is specific enough to deserve its own word: “Slacktivism”, made of “slacker” and “activism”, not unlike “mansplain” is made of “man” and “explain” or “edutainment” was made of “education” and “entertainment”—or indeed “gerrymander” was made of “Elbridge Gerry” and “salamander”. The term seems to be particularly popular on Huffington Post, which has a whole category on slacktivism. There is a particular subcategory of slacktivism that is ironically against other slacktivism, which has been dubbed “snarktivism”.

It’s almost always used as a pejorative; very few people self-identify as “slacktivists” (though once I get through this post, you may see why I’m considering it myself). “Slacktivism” is activism that “isn’t real” somehow, activism that “doesn’t count”.

Of course, that raises the question: What “counts” as legitimate activism? Is it only protest marches and sit-ins? Then very few people have ever been or will ever be activists. Surely donations should count, at least? Those have a direct, measurable impact. What about calling your Congressman, or letter-writing campaigns? These have been staples of activism for decades.
If the term “slacktivism” means anything at all, it seems to point to activities surrounding raising awareness, where the goal is not to enact a particular policy or support a particular NGO but to simply get as much public attention to a topic as possible. It seems to be particularly targeted at blogging and social media—and that’s important, for reasons I’ll get to shortly. If you gather a group of people in your community and give a speech about LGBT rights, you’re an activist. If you send out the exact same speech on Facebook, you’re a slacktivist.

One of the arguments against “slacktivism” is that it can be used to funnel resources at the wrong things; this blog post makes a good point that the Kony 2012 campaign doesn’t appear to have actually accomplished anything except profits for the filmmakers behind it. (Then again: A blog post against slacktivism? Are you sure you’re not doing right now the thing you think you are against?) But is this problem unique to slacktivism, or is it a more general phenomenon that people simply aren’t all that informed about how to have the most impact? There are an awful lot of inefficient charities out there, and in fact the most important waste of charitable funds involves people giving to their local churches. Fortunately, this is changing, as people become more secularized; churches used to account for over half of US donations, and now they only account for less than a third. (Naturally, Christian organizations are pulling out their hair over this.) The 60 million Americans who voted for Trump made a horrible mistake and will cause enormous global damage; but they weren’t slacktivists, were they?

Studies do suggest that traditionally “slacktivist” activities like Facebook likes aren’t a very strong predictor of future, larger actions, and more private modes of support (like donations and calling your Congressman) tend to be stronger predictors. But so what? In order for slacktivism to be a bad thing, they would have to be a negative predictor. They would have to substitute for more effective activism, and there’s no evidence that this happens.

In fact, there’s even some evidence that slacktivism has a positive effect (normally I wouldn’t cite Fox News, but I think in this case we should expect a bias in the opposite direction, and you can read the full Georgetown study if you want):

A study from Georgetown University in November entitled “Dynamics of Cause Engagement” looked how Americans learned about and interacted with causes and other social issues, and discovered some surprising findings on Slacktivism.

While the traditional forms of activism like donating money or volunteering far outpaces slacktivism, those who engage in social issues online are twice as likely as their traditional counterparts to volunteer and participate in events. In other words, slacktivists often graduate to full-blown activism.

At worst, most slacktivists are doing nothing for positive social change, and that’s what the vast majority of people have been doing for the entirety of human history. We can bemoan this fact, but that won’t change it. Most people are simply too uniformed to know what’s going on in the world, and too broke and too busy to do anything about it.

Indeed, slacktivism may be the one thing they can do—which is why I think it’s worth defending.

From an economist’s perspective, there’s something quite odd about how people’s objections to slacktivism are almost always formulated. The rational, sensible objection would be to their small benefits—this isn’t accomplishing enough, you should do something more effective. But in fact, almost all the objections to slacktivism I have ever read focus on their small costs—you’re not a “real activist” because you don’t make sacrifices like I do.

Yet it is a basic principle of economic rationality that, all other things equal, lower cost is better. Indeed, this is one of the few principles of economic rationality that I really do think is unassailable; perfect information is unrealistic and total selfishness makes no sense at all. But cost minimization is really very hard to argue with—why pay more, when you can pay less and get the same benefit?

From an economist’s perspective, the most important thing about an activity is its cost-effectiveness, measured either by net benefitbenefit minus cost—or rate of returnbenefit divided by cost. But in both cases, a lower cost is always better; and in fact slacktivism has an astonishing rate of return, precisely because its cost is so small.

Suppose that a campaign of 10 million Facebook likes actually does have a 1% chance of changing a policy in a way that would save 10,000 lives, with a life expectancy of 50 years each. Surely this is conservative, right? I’m only giving it a 1% chance of success, on a policy with a relatively small impact (10,000 lives could be a single clause in an EPA regulatory standard), with a large number of slacktivist participants (10 million is more people than the entire population of Switzerland). Yet because clicking “like” and “share” only costs you maybe 10 seconds, we’re talking about an expected cost of (10 million)(10/86,400/365) = 0.32 QALY for an expected benefit of (10,000)(0.01)(50) = 5000 QALY. That is a rate of return of 1,500,000%—that’s 1.5 million percent.

Let’s compare this to the rate of return on donating to a top charity like UNICEF, Oxfam, the Against Malaria Foundation, or the Schistomoniasis Control Initiative, for which donating about $300 would save the life of 1 child, adding about 50 QALY. That $300 most likely cost you about 0.01 QALY (assuming an annual income of $30,000), so we’re looking at a return of 500,000%. Now, keep in mind that this is a huge rate of return, far beyond what you can ordinarily achieve, that donating $300 to UNICEF is probably one of the best things you could possibly be doing with that money—and yet slacktivism may still exceed it in efficiency. Maybe slacktivism doesn’t sound so bad after all?

Of course, the net benefit of your participation is higher in the case of donation; you yourself contribute 50 QALY instead of only contributing 0.0005 QALY. Ultimately net benefit is what matters; rate of return is a way of estimating what the net benefit would be when comparing different ways of spending the same amount of time or money. But from the figures I just calculated, it begins to seem like maybe the very best thing you could do with your time is clicking “like” and “share” on Facebook posts that will raise awareness of policies of global importance. Now, you have to include all that extra time spent poring through other Facebook posts, and consider that you may not be qualified to assess the most important issues, and there’s a lot of uncertainty involved in what sort of impact you yourself will have… but it’s almost certainly not the worst thing you could be doing with your time, and frankly running these numbers has made me feel a lot better about all the hours I have actually spent doing this sort of thing. It’s a small benefit, yes—but it’s an even smaller cost.

Indeed, the fact that so many people treat low cost as bad, when it is almost by definition good, and the fact that they also target their ire so heavily at blogging and social media, says to me that what they are really trying to accomplish here has nothing to do with actually helping people in the most efficient way possible.

Rather, it’s two things.

The obvious one is generational—it’s yet another chorus in the unending refrain that is “kids these days”. Facebook is new, therefore it is suspicious. Adults have been complaining about their descendants since time immemorial; some of the oldest written works we have are of ancient Babylonians complaining that their kids are lazy and selfish. Either human beings have been getting lazier and more selfish for thousands of years, or, you know, kids are always a bit more lazy and selfish than their parents or at least seem so from afar.

The one that’s more interesting for an economist is signaling. By complaining that other people aren’t paying enough cost for something, what you’re really doing is complaining that they aren’t signaling like you are. The costly signal has been made too cheap, so now it’s no good as a signal anymore.

“Anyone can click a button!” you say. Yes, and? Isn’t it wonderful that now anyone with a smartphone (and there are more people with access to smartphones than toilets, because #WeLiveInTheFuture) can contribute, at least in some small way, to improving the world? But if anyone can do it, then you can’t signal your status by doing it. If your goal was to make yourself look better, I can see why this would bother you; all these other people doing things that look just as good as what you do! How will you ever distinguish yourself from the riffraff now?

This is also likely what’s going on as people fret that “a college degree’s not worth anything anymore” because so many people are getting them now; well, as a signal, maybe not. But if it’s just a signal, why are we spending so much money on it? Surely we can find a more efficient way to rank people by their intellect. I thought it was supposed to be an education—in which case the meteoric rise in global college enrollments should be cause for celebration. (In reality of course a college degree can serve both roles, and it remains an open question among labor economists as to which effect is stronger and by how much. But the signaling role is almost pure waste from the perspective of social welfare; we should be trying to maximize the proportion of real value added.)

For this reason, I think I’m actually prepared to call myself a slacktivist. I aim for cost-effective awareness-raising; I want to spread the best ideas to the most people for the lowest cost. Why, would you prefer I waste more effort, to signal my own righteousness?

Experimentally testing categorical prospect theory

Dec 4, JDN 2457727

In last week’s post I presented a new theory of probability judgments, which doesn’t rely upon people performing complicated math even subconsciously. Instead, I hypothesize that people try to assign categories to their subjective probabilities, and throw away all the information that wasn’t used to assign that category.

The way to most clearly distinguish this from cumulative prospect theory is to show discontinuity. Kahneman’s smooth, continuous function places fairly strong bounds on just how much a shift from 0% to 0.000001% can really affect your behavior. In particular, if you want to explain the fact that people do seem to behave differently around 10% compared to 1% probabilities, you can’t allow the slope of the smooth function to get much higher than 10 at any point, even near 0 and 1. (It does depend on the precise form of the function, but the more complicated you make it, the more free parameters you add to the model. In the most parsimonious form, which is a cubic polynomial, the maximum slope is actually much smaller than this—only 2.)

If that’s the case, then switching from 0.% to 0.0001% should have no more effect in reality than a switch from 0% to 0.00001% would to a rational expected utility optimizer. But in fact I think I can set up scenarios where it would have a larger effect than a switch from 0.001% to 0.01%.

Indeed, these games are already quite profitable for the majority of US states, and they are called lotteries.

Rationally, it should make very little difference to you whether your odds of winning the Powerball are 0 (you bought no ticket) or 0.000000001% (you bought a ticket), even when the prize is $100 million. This is because your utility of $100 million is nowhere near 100 million times as large as your marginal utility of $1. A good guess would be that your lifetime income is about $2 million, your utility is logarithmic, the units of utility are hectoQALY, and the baseline level is about 100,000.

I apologize for the extremely large number of decimals, but I had to do that in order to show any difference at all. I have bolded where the decimals first deviate from the baseline.

Your utility if you don’t have a ticket is ln(20) = 2.9957322736 hQALY.

Your utility if you have a ticket is (1-10^-9) ln(20) + 10^-9 ln(1020) = 2.9957322775 hQALY.

You gain a whopping 40 microQALY over your whole lifetime. I highly doubt you could even perceive such a difference.

And yet, people are willing to pay nontrivial sums for the chance to play such lotteries. Powerball tickets sell for about $2 each, and some people buy tickets every week. If you do that and live to be 80, you will spend some $8,000 on lottery tickets during your lifetime, which results in this expected utility: (1-4*10^-6) ln(20-0.08) + 4*10^-6 ln(1020) = 2.9917399955 hQALY.
You have now sacrificed 0.004 hectoQALY, which is to say 0.4 QALY—that’s months of happiness you’ve given up to play this stupid pointless game.

Which shouldn’t be surprising, as (with 99.9996% probability) you have given up four months of your lifetime income with nothing to show for it. Lifetime income of $2 million / lifespan of 80 years = $25,000 per year; $8,000 / $25,000 = 0.32. You’ve actually sacrificed slightly more than this, which comes from your risk aversion.

Why would anyone do such a thing? Because while the difference between 0 and 10^-9 may be trivial, the difference between “impossible” and “almost impossible” feels enormous. “You can’t win if you don’t play!” they say, but they might as well say “You can’t win if you do play either.” Indeed, the probability of winning without playing isn’t zero; you could find a winning ticket lying on the ground, or win due to an error that is then upheld in court, or be given the winnings bequeathed by a dying family member or gifted by an anonymous donor. These are of course vanishingly unlikely—but so was winning in the first place. You’re talking about the difference between 10^-9 and 10^-12, which in proportional terms sounds like a lot—but in absolute terms is nothing. If you drive to a drug store every week to buy a ticket, you are more likely to die in a car accident on the way to the drug store than you are to win the lottery.

Of course, these are not experimental conditions. So I need to devise a similar game, with smaller stakes but still large enough for people’s brains to care about the “almost impossible” category; maybe thousands? It’s not uncommon for an economics experiment to cost thousands, it’s just usually paid out to many people instead of randomly to one person or nobody. Conducting the experiment in an underdeveloped country like India would also effectively amplify the amounts paid, but at the fixed cost of transporting the research team to India.

But I think in general terms the experiment could look something like this. You are given $20 for participating in the experiment (we treat it as already given to you, to maximize your loss aversion and endowment effect and thereby give us more bang for our buck). You then have a chance to play a game, where you pay $X to get a P probability of $Y*X, and we vary these numbers.

The actual participants wouldn’t see the variables, just the numbers and possibly the rules: “You can pay $2 for a 1% chance of winning $200. You can also play multiple times if you wish.” “You can pay $10 for a 5% chance of winning $250. You can only play once or not at all.”

So I think the first step is to find some dilemmas, cases where people feel ambivalent, and different people differ in their choices. That’s a good role for a pilot study.

Then we take these dilemmas and start varying their probabilities slightly.

In particular, we try to vary them at the edge of where people have mental categories. If subjective probability is continuous, a slight change in actual probability should never result in a large change in behavior, and furthermore the effect of a change shouldn’t vary too much depending on where the change starts.

But if subjective probability is categorical, these categories should have edges. Then, when I present you with two dilemmas that are on opposite sides of one of the edges, your behavior should radically shift; while if I change it in a different way, I can make a large change without changing the result.

Based solely on my own intuition, I guessed that the categories roughly follow this pattern:

Impossible: 0%

Almost impossible: 0.1%

Very unlikely: 1%

Unlikely: 10%

Fairly unlikely: 20%

Roughly even odds: 50%

Fairly likely: 80%

Likely: 90%

Very likely: 99%

Almost certain: 99.9%

Certain: 100%

So for example, if I switch from 0%% to 0.01%, it should have a very large effect, because I’ve moved you out of your “impossible” category (indeed, I think the “impossible” category is almost completely sharp; literally anything above zero seems to be enough for most people, even 10^-9 or 10^-10). But if I move from 1% to 2%, it should have a small effect, because I’m still well within the “very unlikely” category. Yet the latter change is literally one hundred times larger than the former. It is possible to define continuous functions that would behave this way to an arbitrary level of approximation—but they get a lot less parsimonious very fast.

Now, immediately I run into a problem, because I’m not even sure those are my categories, much less that they are everyone else’s. If I knew precisely which categories to look for, I could tell whether or not I had found it. But the process of both finding the categories and determining if their edges are truly sharp is much more complicated, and requires a lot more statistical degrees of freedom to get beyond the noise.

One thing I’m considering is assigning these values as a prior, and then conducting a series of experiments which would adjust that prior. In effect I would be using optimal Bayesian probability reasoning to show that human beings do not use optimal Bayesian probability reasoning. Still, I think that actually pinning down the categories would require a large number of participants or a long series of experiments (in frequentist statistics this distinction is vital; in Bayesian statistics it is basically irrelevant—one of the simplest reasons to be Bayesian is that it no longer bothers you whether someone did 2 experiments of 100 people or 1 experiment of 200 people, provided they were the same experiment of course). And of course there’s always the possibility that my theory is totally off-base, and I find nothing; a dissertation replicating cumulative prospect theory is a lot less exciting (and, sadly, less publishable) than one refuting it.

Still, I think something like this is worth exploring. I highly doubt that people are doing very much math when they make most probabilistic judgments, and using categories would provide a very good way for people to make judgments usefully with no math at all.

How do people think about probability?

Nov 27, JDN 2457690

(This topic was chosen by vote of my Patreons.)

In neoclassical theory, it is assumed (explicitly or implicitly) that human beings judge probability in something like the optimal Bayesian way: We assign prior probabilities to events, and then when confronted with evidence we infer using the observed data to update our prior probabilities to posterior probabilities. Then, when we have to make decisions, we maximize our expected utility subject to our posterior probabilities.

This, of course, is nothing like how human beings actually think. Even very intelligent, rational, numerate people only engage in a vague approximation of this behavior, and only when dealing with major decisions likely to affect the course of their lives. (Yes, I literally decide which universities to attend based upon formal expected utility models. Thus far, I’ve never been dissatisfied with a decision made that way.) No one decides what to eat for lunch or what to do this weekend based on formal expected utility models—or at least I hope they don’t, because that point the computational cost far exceeds the expected benefit.

So how do human beings actually think about probability? Well, a good place to start is to look at ways in which we systematically deviate from expected utility theory.

A classic example is the Allais paradox. See if it applies to you.

In game A, you get $1 million dollars, guaranteed.
In game B, you have a 10% chance of getting $5 million, an 89% chance of getting $1 million, but now you have a 1% chance of getting nothing.

Which do you prefer, game A or game B?

In game C, you have an 11% chance of getting $1 million, and an 89% chance of getting nothing.

In game D, you have a 10% chance of getting $5 million, and a 90% chance of getting nothing.

Which do you prefer, game C or game D?

I have to think about it for a little bit and do some calculations, and it’s still very hard because it depends crucially on my projected lifetime income (which could easily exceed $3 million with a PhD, especially in economics) and the precise form of my marginal utility (I think I have constant relative risk aversion, but I’m not sure what parameter to use precisely), but in general I think I want to choose game A and game C, but I actually feel really ambivalent, because it’s not hard to find plausible parameters for my utility where I should go for the gamble.

But if you’re like most people, you choose game A and game D.

There is no coherent expected utility by which you would do this.

Why? Either a 10% chance of $5 million instead of $1 million is worth risking a 1% chance of nothing, or it isn’t. If it is, you should play B and D. If it’s not, you should play A and C. I can’t tell you for sure whether it is worth it—I can’t even fully decide for myself—but it either is or it isn’t.

Yet most people have a strong intuition that they should take game A but game D. Why? What does this say about how we judge probability?
The leading theory in behavioral economics right now is cumulative prospect theory, developed by the great Kahneman and Tversky, who essentially founded the field of behavioral economics. It’s quite intimidating to try to go up against them—which is probably why we should force ourselves to do it. Fear of challenging the favorite theories of the great scientists before us is how science stagnates.

I wrote about it more in a previous post, but as a brief review, cumulative prospect theory says that instead of judging based on a well-defined utility function, we instead consider gains and losses as fundamentally different sorts of thing, and in three specific ways:

First, we are loss-averse; we feel a loss about twice as intensely as a gain of the same amount.

Second, we are risk-averse for gains, but risk-seeking for losses; we assume that gaining twice as much isn’t actually twice as good (which is almost certainly true), but we also assume that losing twice as much isn’t actually twice as bad (which is almost certainly false and indeed contradictory with the previous).

Third, we judge probabilities as more important when they are close to certainty. We make a large distinction between a 0% probability and a 0.0000001% probability, but almost no distinction at all between a 41% probability and a 43% probability.

That last part is what I want to focus on for today. In Kahneman’s model, this is a continuous, monotonoic function that maps 0 to 0 and 1 to 1, but systematically overestimates probabilities below but near 1/2 and systematically underestimates probabilities above but near 1/2.

It looks something like this, where red is true probability and blue is subjective probability:

cumulative_prospect
I don’t believe this is actually how humans think, for two reasons:

  1. It’s too hard. Humans are astonishingly innumerate creatures, given the enormous processing power of our brains. It’s true that we have some intuitive capacity for “solving” very complex equations, but that’s almost all within our motor system—we can “solve a differential equation” when we catch a ball, but we have no idea how we’re doing it. But probability judgments are often made consciously, especially in experiments like the Allais paradox; and the conscious brain is terrible at math. It’s actually really amazing how bad we are at math. Any model of normal human judgment should assume from the start that we will not do complicated math at any point in the process. Maybe you can hypothesize that we do so subconsciously, but you’d better have a good reason for assuming that.
  2. There is no reason to do this. Why in the world would any kind of optimization system function this way? You start with perfectly good probabilities, and then instead of using them, you subject them to some bizarre, unmotivated transformation that makes them less accurate and costs computing power? You may as well hit yourself in the head with a brick.

So, why might it look like we are doing this? Well, my proposal, admittedly still rather half-baked, is that human beings don’t assign probabilities numerically at all; we assign them categorically.

You may call this, for lack of a better term, categorical prospect theory.

My theory is that people don’t actually have in their head “there is an 11% chance of rain today” (unless they specifically heard that from a weather report this morning); they have in their head “it’s fairly unlikely that it will rain today”.

That is, we assign some small number of discrete categories of probability, and fit things into them. I’m not sure what exactly the categories are, and part of what makes my job difficult here is that they may be fuzzy-edged and vary from person to person, but roughly speaking, I think they correspond to the sort of things psychologists usually put on Likert scales in surveys: Impossible, almost impossible, very unlikely, unlikely, fairly unlikely, roughly even odds, fairly likely, likely, very likely, almost certain, certain. If I’m putting numbers on these probability categories, they go something like this: 0, 0.001, 0.01, 0.10, 0.20, 0.50, 0.8, 0.9, 0.99, 0.999, 1.

Notice that this would preserve the same basic effect as cumulative prospect theory: You care a lot more about differences in probability when they are near 0 or 1, because those are much more likely to actually shift your category. Indeed, as written, you wouldn’t care about a shift from 0.4 to 0.6 at all, despite caring a great deal about a shift from 0.001 to 0.01.

How does this solve the above problems?

  1. It’s easy. Not only don’t you compute a probability and then recompute it for no reason; you never even have to compute it precisely. Just get it within some vague error bounds and that will tell you what box it goes in. Instead of computing an approximation to a continuous function, you just slot things into a small number of discrete boxes, a dozen at the most.
  2. That explains why we would do it: It’s easy. Our brains need to conserve their capacity, and they did especially in our ancestral environment when we struggled to survive. Rather than having to iterate your approximation to arbitrary precision, you just get within 0.1 or so and call it a day. That saves time and computing power, which saves energy, which could save your life.

What new problems have I introduced?

  1. It’s very hard to know exactly where people’s categories are, if they vary between individuals or even between situations, and whether they are fuzzy-edged.
  2. If you take the model I just gave literally, even quite large probability changes will have absolutely no effect as long as they remain within a category such as “roughly even odds”.

With regard to 2, I think Kahneman may himself be able to save me, with his dual process theory concept of System 1 and System 2. What I’m really asserting is that System 1, the fast, intuitive judgment system, operates on these categories. System 2, on the other hand, the careful, rational thought system, can actually make use of proper numerical probabilities; it’s just very costly to boot up System 2 in the first place, much less ensure that it actually gets the right answer.

How might we test this? Well, I think that people are more likely to use System 1 when any of the following are true:

  1. They are under harsh time-pressure
  2. The decision isn’t very important
  3. The intuitive judgment is fast and obvious

And conversely they are likely to use System 2 when the following are true:

  1. They have plenty of time to think
  2. The decision is very important
  3. The intuitive judgment is difficult or unclear

So, it should be possible to arrange an experiment varying these parameters, such that in one treatment people almost always use System 1, and in another they almost always use System 2. And then, my prediction is that in the System 1 treatment, people will in fact not change their behavior at all when you change the probability from 15% to 25% (fairly unlikely) or 40% to 60% (roughly even odds).

To be clear, you can’t just present people with this choice between game E and game F:

Game E: You get a 60% chance of $50, and a 40% chance of nothing.

Game F: You get a 40% chance of $50, and a 60% chance of nothing.

People will obviously choose game E. If you can directly compare the numbers and one game is strictly better in every way, I think even without much effort people will be able to choose correctly.

Instead, what I’m saying is that if you make the following offers to two completely different sets of people, you will observe little difference in their choices, even though under expected utility theory you should.
Group I receives a choice between game E and game G:

Game E: You get a 60% chance of $50, and a 40% chance of nothing.

Game G: You get a 100% chance of $20.

Group II receives a choice between game F and game G:

Game F: You get a 40% chance of $50, and a 60% chance of nothing.

Game G: You get a 100% chance of $20.

Under two very plausible assumptions about marginal utility of wealth, I can fix what the rational judgment should be in each game.

The first assumption is that marginal utility of wealth is decreasing, so people are risk-averse (at least for gains, which these are). The second assumption is that most people’s lifetime income is at least two orders of magnitude higher than $50.

By the first assumption, group II should choose game G. The expected income is precisely the same, and being even ever so slightly risk-averse should make you go for the guaranteed $20.

By the second assumption, group I should choose game E. Yes, there is some risk, but because $50 should not be a huge sum to you, your risk aversion should be small and the higher expected income of $30 should sway you.

But I predict that most people will choose game G in both cases, and (within statistical error) the same proportion will choose F as chose E—thus showing that the difference between a 40% chance and a 60% chance was in fact negligible to their intuitive judgments.

However, this doesn’t actually disprove Kahneman’s theory; perhaps that part of the subjective probability function is just that flat. For that, I need to set up an experiment where I show discontinuity. I need to find the edge of a category and get people to switch categories sharply. Next week I’ll talk about how we might pull that off.

Do we always want to internalize externalities?

JDN 2457437

I often talk about the importance of externalitiesa full discussion in this earlier post, and one of their important implications, the tragedy of the commons, in another. Briefly, externalities are consequences of actions incurred upon people who did not perform those actions. Anything I do affecting you that you had no say in, is an externality.

Usually I’m talking about how we want to internalize externalities, meaning that we set up a system of incentives to make it so that the consequences fall upon the people who chose the actions instead of anyone else. If you pollute a river, you should have to pay to clean it up. If you assault someone, you should serve jail time as punishment. If you invent a new technology, you should be rewarded for it. These are all attempts to internalize externalities.

But today I’m going to push back a little, and ask whether we really always want to internalize externalities. If you think carefully, it’s not hard to come up with scenarios where it actually seems fairer to leave the externality in place, or perhaps reduce it somewhat without eliminating it.

For example, suppose indeed that someone invents a great new technology. To be specific, let’s think about Jonas Salk, inventing the polio vaccine. This vaccine saved the lives of thousands of people and saved millions more from pain and suffering. Its value to society is enormous, and of course Salk deserved to be rewarded for it.

But we did not actually fully internalize the externality. If we had, every family whose child was saved from polio would have had to pay Jonas Salk an amount equal to what they saved on medical treatments as a result, or even an amount somehow equal to the value of their child’s life (imagine how offended people would get if you asked that on a survey!). Those millions of people spared from suffering would need to each pay, at minimum, thousands of dollars to Jonas Salk, making him of course a billionaire.

And indeed this is more or less what would have happened, if he had been willing and able to enforce a patent on the vaccine. The inability of some to pay for the vaccine at its monopoly prices would add some deadweight loss, but even that could be removed if Salk Industries had found a way to offer targeted price vouchers that let them precisely price-discriminate so that every single customer paid exactly what they could afford to pay. If that had happened, we would have fully internalized the externality and therefore maximized economic efficiency.

But doesn’t that sound awful? Doesn’t it sound much worse than what we actually did, where Jonas Salk received a great deal of funding and support from governments and universities, and lived out his life comfortably upper-middle class as a tenured university professor?

Now, perhaps he should have been awarded a Nobel Prize—I take that back, there’s no “perhaps” about it, he definitely should have been awarded a Nobel Prize in Medicine, it’s absurd that he did not—which means that I at least do feel the externality should have been internalized a bit more than it was. But a Nobel Prize is only 10 million SEK, about $1.1 million. That’s about enough to be independently wealthy and live comfortably for the rest of your life; but it’s a small fraction of the roughly $7 billion he could have gotten if he had patented the vaccine. Yet while the possible world in which he wins a Nobel is better than this one, I’m fairly well convinced that the possible world in which he patents the vaccine and becomes a billionaire is considerably worse.

Internalizing externalities makes sense if your goal is to maximize total surplus (a concept I explain further in the linked post), but total surplus is actually a terrible measure of human welfare.

Total surplus counts every dollar of willingness-to-pay exactly the same across different people, regardless of whether they live on $400 per year or $4 billion.

It also takes no account whatsoever of how wealth is distributed. Suppose a new technology adds $10 billion in wealth to the world. As far as total surplus, it makes no difference whether that $10 billion is spread evenly across the entire planet, distributed among a city of a million people, concentrated in a small town of 2,000, or even held entirely in the bank account of a single man.

Particularly a propos of the Salk example, total surplus makes no distinction between these two scenarios: a perfectly-competitive market where everything is sold at a fair price, and a perfectly price-discriminating monopoly, where everything is sold at the very highest possible price each person would be willing to pay.

This is a perfectly-competitive market, where the benefits are more or less equally (in this case exactly equally, but that need not be true in real life) between sellers and buyers:

elastic_supply_competitive_labeled

This is a perfectly price-discriminating monopoly, where the benefits accrue entirely to the corporation selling the good:

elastic_supply_price_discrimination

In the former case, the company profits, consumers are better off, everyone is happy. In the latter case, the company reaps all the benefits and everyone else is left exactly as they were. In real terms those are obviously very different outcomes—the former being what we want, the latter being the cyberpunk dystopia we seem to be hurtling mercilessly toward. But in terms of total surplus, and therefore the kind of “efficiency” that is maximize by internalizing all externalities, they are indistinguishable.

In fact (as I hope to publish a paper about at some point), the way willingness-to-pay works, it weights rich people more. Redistributing goods from the poor to the rich will typically increase total surplus.

Here’s an example. Suppose there is a cake, which is sufficiently delicious that it offers 2 milliQALY in utility to whoever consumes it (this is a truly fabulous cake). Suppose there are two people to whom we might give this cake: Richie, who has $10 million in annual income, and Hungry, who has only $1,000 in annual income. How much will each of them be willing to pay?

Well, assuming logarithmic marginal utility of wealth (which is itself probably biasing slightly in favor of the rich), 1 milliQALY is about $1 to Hungry, so Hungry will be willing to pay $2 for the cake. To Richie, however, 1 milliQALY is about $10,000; so he will be willing to pay a whopping $20,000 for this cake.

What this means is that the cake will almost certainly be sold to Richie; and if we proposed a policy to redistribute the cake from Richie to Hungry, economists would emerge to tell us that we have just reduced total surplus by $19,998 and thereby committed a great sin against economic efficiency. They will cajole us into returning the cake to Richie and thus raising total surplus by $19,998 once more.

This despite the fact that I stipulated that the cake is worth just as much in real terms to Hungry as it is to Richie; the difference is due to their wildly differing marginal utility of wealth.

Indeed, it gets worse, because even if we suppose that the cake is worth much more in real utility to Hungry—because he is in fact hungry—it can still easily turn out that Richie’s willingness-to-pay is substantially higher. Suppose that Hungry actually gets 20 milliQALY out of eating the cake, while Richie still only gets 2 milliQALY. Hungry’s willingness-to-pay is now $20, but Richie is still going to end up with the cake.

Now, if your thought is, “Why would Richie pay $20,000, when he can go to another store and get another cake that’s just as good for $20?” Well, he wouldn’t—but in the sense we mean for total surplus, willingness-to-pay isn’t just what you’d actually be willing to pay given the actual prices of the goods, but the absolute maximum price you’d be willing to pay to get that good under any circumstances. It is instead the marginal utility of the good divided by your marginal utility of wealth. In this sense the cake is “worth” $20,000 to Richie, and “worth” substantially less to Hungry—but not because it’s actually worth less in real terms, but simply because Richie has so much more money.

Even economists often equate these two, implicitly assuming that we are spending our money up to the point where our marginal willingness-to-pay is the actual price we choose to pay; but in general our willingness-to-pay is higher than the price if we are willing to buy the good at all. The consumer surplus we get from goods is in fact equal to the difference between willingness-to-pay and actual price paid, summed up over all the goods we have purchased.

Internalizing all externalities would definitely maximize total surplus—but would it actually maximize happiness? Probably not.

If you asked most people what their marginal utility of wealth is, they’d have no idea what you’re talking about. But most people do actually have an intuitive sense that a dollar is worth more to a homeless person than it is to a millionaire, and that’s really all we mean by diminishing marginal utility of wealth.

I think the reason we’re uncomfortable with the idea of Jonas Salk getting $7 billion from selling the polio vaccine, rather than the same number of people getting the polio vaccine and Jonas Salk only getting the $1.1 million from a Nobel Prize, is that we intuitively grasp that after that $1.1 million makes him independently wealthy, the rest of the money is just going to sit in some stock account and continue making even more money, while if we’d let the families keep it they would have put it to much better use raising their children who are now protected from polio. We do want to reward Salk for his great accomplishment, but we don’t see why we should keep throwing cash at him when it could obviously be spent in better ways.

And indeed I think this intuition is correct; great accomplishments—which is to say, large positive externalities—should be rewarded, but not in direct proportion. Maybe there should be some threshold above which we say, “You know what? You’re rich enough now; we can stop giving you money.” Or maybe it should simply damp down very quickly, so that a contribution which is worth $10 billion to the world pays only slightly more than one that is worth $100 million, but a contribution that is worth $100,000 pays considerably more than one which is only worth $10,000.

What it ultimately comes down to is that if we make all the benefits incur to the person who did it, there aren’t any benefits anymore. The whole point of Jonas Salk inventing the polio vaccine (or Einstein discovering relativity, or Darwin figuring out natural selection, or any great achievement) is that it will benefit the rest of humanity, preferably on to future generations. If you managed to fully internalize that externality, this would no longer be true; Salk and Einstein and Darwin would have become fabulously wealthy, and then somehow we’d all have to continue paying into their estates or something an amount equal to the benefits we received from their discoveries. (Every time you use your GPS, pay a royalty to the Einsteins. Every time you take a pill, pay a royalty to the Darwins.) At some point we’d probably get fed up and decide we’re no better off with them than without them—which is exactly by construction how we should feel if the externality were fully internalized.

Internalizing negative externalities is much less problematic—it’s your mess, clean it up. We don’t want other people to be harmed by your actions, and if we can pull that off that’s fantastic. (In reality, we usually can’t fully internalize negative externalities, but we can at least try.)

But maybe internalizing positive externalities really isn’t so great after all.

Bet five dollars for maximum performance

JDN 2457433

One of the more surprising findings from the study of human behavior under stress is the Yerkes-Dodson curve:

OriginalYerkesDodson
This curve shows how well humans perform at a given task, as a function of how high the stakes are on whether or not they do it properly.

For simple tasks, it says what most people intuitively expect—and what neoclassical economists appear to believe: As the stakes rise, the more highly incentivized you are to do it, and the better you do it.

But for complex tasks, it says something quite different: While increased stakes do raise performance to a point—with nothing at stake at all, people hardly work at all—it is possible to become too incentivized. Formally we say the curve is not monotonic; it has a local maximum.

This is one of many reasons why it’s ridiculous to say that top CEOs should make tens of millions of dollars a year on the rise and fall of their company’s stock price (as a great many economists do in fact say). Even if I believed that stock prices accurately reflect the company’s viability (they do not), and believed that the CEO has a great deal to do with the company’s success, it would still be a case of overincentivizing. When a million dollars rides on a decision, that decision is going to be worse than if the stakes had only been $100. With this in mind, it’s really not surprising that higher CEO pay is correlated with worse company performance. Stock options are terrible motivators, but do offer a subtle way of making wages adjust to the business cycle.

The reason for this is that as the stakes get higher, we become stressed, and that stress response inhibits our ability to use higher cognitive functions. The sympathetic nervous system evolved to make us very good at fighting or running away in the face of danger, which works well should you ever be attacked by a tiger. It did not evolve to make us good at complex tasks under high stakes, the sort of skill we’d need when calculating the trajectory of an errant spacecraft or disarming a nuclear warhead.

To be fair, most of us never have to worry about piloting errant spacecraft or disarming nuclear warheads—indeed, you’re about as likely to get attacked by a tiger even in today’s world. (The rate of tiger attacks in the US is just under 2 per year, and the rate of manned space launches in the US was about 5 per year until the Space Shuttle was terminated.)

There are certain professions, such as pilots and surgeons, where performing complex tasks under life-or-death pressure is commonplace, but only a small fraction of people take such professions for precisely that reason. And if you’ve ever wondered why we use checklists for pilots and there is discussion of also using checklists for surgeons, this is why—checklists convert a single complex task into many simple tasks, allowing high performance even at extreme stakes.

But we do have to do a fair number of quite complex tasks with stakes that are, if not urgent life-or-death scenarios, then at least actions that affect our long-term life prospects substantially. In my tutoring business I encounter one in particular quite frequently: Standardized tests.

Tests like the SAT, ACT, GRE, LSAT, GMAT, and other assorted acronyms are not literally life-or-death, but they often feel that way to students because they really do have a powerful impact on where you’ll end up in life. Will you get into a good college? Will you get into grad school? Will you get the job you want? Even subtle deviations from the path of optimal academic success can make it much harder to achieve career success in the future.

Of course, these are hardly the only examples. Many jobs require us to complete tasks properly on tight deadlines, or else risk being fired. Working in academia infamously requires publishing in journals in time to rise up the tenure track, or else falling off the track entirely. (This incentivizes the production of huge numbers of papers, whether they’re worth writing or not; yes, the number of papers published goes down after tenure, but is that a bad thing? What we need to know is whether the number of good papers goes down. My suspicion is that most if not all of the reduction in publications is due to not publishing things that weren’t worth publishing.)

So if you are faced with this sort of task, what can you do? If you realize that you are faced with a high-stakes complex task, you know your performance will be bad—which only makes your stress worse!

My advice is to pretend you’re betting five dollars on the outcome.

Ignore all other stakes, and pretend you’re betting five dollars. $5.00 USD. Do it right and you get a Lincoln; do it wrong and you lose one.
What this does is ensures that you care enough—you don’t want to lose $5 for no reason—but not too much—if you do lose $5, you don’t feel like your life is ending. We want to put you near that peak of the Yerkes-Dodson curve.

The great irony here is that you most want to do this when it is most untrue. If you actually do have a task for which you’ve bet $5 and nothing else rides on it, you don’t need this technique, and any technique to improve your performance is not particularly worthwhile. It’s when you have a standardized test to pass that you really want to use this—and part of me even hopes that people know to do this whenever they have nuclear warheads to disarm. It is precisely when the stakes are highest that you must put those stakes out of your mind.

Why five dollars? Well, the exact amount is arbitrary, but this is at least about the right order of magnitude for most First World individuals. If you really want to get precise, I think the optimal stakes level for maximum performance is something like 100 microQALY per task, and assuming logarithmic utility of wealth, $5 at the US median household income of $53,600 is approximately 100 microQALY. If you have a particularly low or high income, feel free to adjust accordingly. Literally you should be prepared to bet about an hour of your life; but we are not accustomed to thinking that way, so use $5. (I think most people, if asked outright, would radically overestimate what an hour of life is worth to them. “I wouldn’t give up an hour of my life for $1,000!” Then why do you work at $20 an hour?)

It’s a simple heuristic, easy to remember, and sometimes effective. Give it a try.

The Tragedy of the Commons

JDN 2457387

In a previous post I talked about one of the most fundamental—perhaps the most fundamental—problem in game theory, the Prisoner’s Dilemma, and how neoclassical economic theory totally fails to explain actual human behavior when faced with this problem in both experiments and the real world.

As a brief review, the essence of the game is that both players can either cooperate or defect; if they both cooperate, the outcome is best overall; but it is always in each player’s interest to defect. So a neoclassically “rational” player would always defect—resulting in a bad outcome for everyone. But real human beings typically cooperate, and thus do better. The “paradox” of the Prisoner’s Dilemma is that being “rational” results in making less money at the end.

Obviously, this is not actually a good definition of rational behavior. Being short-sighted and ignoring the impact of your behavior on others doesn’t actually produce good outcomes for anybody, including yourself.

But the Prisoner’s Dilemma only has two players. If we expand to a larger number of players, the expanded game is called a Tragedy of the Commons.

When we do this, something quite surprising happens: As you add more people, their behavior starts converging toward the neoclassical solution, in which everyone defects and we get a bad outcome for everyone.

Indeed, people in general become less cooperative, less courageous, and more apathetic the more of them you put together. K was quite apt when he said, “A person is smart; people are dumb, panicky, dangerous animals and you know it.” There are ways to counteract this effect, as I’ll get to in a moment—but there is a strong effect that needs to be counteracted.

We see this most vividly in the bystander effect. If someone is walking down the street and sees someone fall and injure themselves, there is about a 70% chance that they will go try to help the person who fell—humans are altruistic. But if there are a dozen people walking down the street who all witness the same event, there is only a 40% chance that any of them will help—humans are irrational.

The primary reason appears to be diffusion of responsibility. When we are alone, we are the only one could help, so we feel responsible for helping. But when there are others around, we assume that someone else could take care of it for us, so if it isn’t done that’s not our fault.

There also appears to be a conformity effect: We want to conform our behavior to social norms (as I said, to a first approximation, all human behavior is social norms). The mere fact that there are other people who could have helped but didn’t suggests the presence of an implicit social norm that we aren’t supposed to help this person for some reason. It never occurs to most people to ask why such a norm would exist or whether it’s a good one—it simply never occurs to most people to ask those questions about any social norms. In this case, by hesitating to act, people actually end up creating the very norm they think they are obeying.

This can lead to what’s called an Abilene Paradox, in which people simultaneously try to follow what they think everyone else wants and also try to second-guess what everyone else wants based on what they do, and therefore end up doing something that none of them actually wanted. I think a lot of the weird things humans do can actually be attributed to some form of the Abilene Paradox. (“Why are we sacrificing this goat?” “I don’t know, I thought you wanted to!”)

Autistic people are not as good at following social norms (though some psychologists believe this is simply because our social norms are optimized for the neurotypical population). My suspicion is that autistic people are therefore less likely to suffer from the bystander effect, and more likely to intervene to help someone even if they are surrounded by passive onlookers. (Unfortunately I wasn’t able to find any good empirical data on that—it appears no one has ever thought to check before.) I’m quite certain that autistic people are less likely to suffer from the Abilene Paradox—if they don’t want to do something, they’ll tell you so (which sometimes gets them in trouble).

Because of these psychological effects that blunt our rationality, in large groups human beings often do end up behaving in a way that appears selfish and short-sighted.

Nowhere is this more apparent than in ecology. Recycling, becoming vegetarian, driving less, buying more energy-efficient appliances, insulating buildings better, installing solar panels—none of these things are particularly difficult or expensive to do, especially when weighed against the tens of millions of people who will die if climate change continues unabated. Every recyclable can we throw in the trash is a silent vote for a global holocaust.

But as it no doubt immediately occurred to you to respond: No single one of us is responsible for all that. There’s no way I myself could possibly save enough carbon emissions to significantly reduce climate change—indeed, probably not even enough to save a single human life (though maybe). This is certainly true; the error lies in thinking that this somehow absolves us of the responsibility to do our share.

I think part of what makes the Tragedy of the Commons so different from the Prisoner’s Dilemma, at least psychologically, is that the latter has an identifiable victimwe know we are specifically hurting that person more than we are helping ourselves. We may even know their name (and if we don’t, we’re more likely to defect—simply being on the Internet makes people more aggressive because they don’t interact face-to-face). In the Tragedy of the Commons, it is often the case that we don’t know who any of our victims are; moreover, it’s quite likely that we harm each one less than we benefit ourselves—even though we harm everyone overall more.

Suppose that driving a gas-guzzling car gives me 1 milliQALY of happiness, but takes away an average of 1 nanoQALY from everyone else in the world. A nanoQALY is tiny! Negligible, even, right? One billionth of a year, a mere 30 milliseconds! Literally less than the blink of an eye. But take away 30 milliseconds from everyone on Earth and you have taken away 7 years of human life overall. Do that 10 times, and statistically one more person is dead because of you. And you have gained only 10 milliQALY, roughly the value of $300 to a typical American. Would you kill someone for $300?

Peter Singer has argued that we should in fact think of it this way—when we cause a statistical death by our inaction, we should call it murder, just as if we had left a child to drown to keep our clothes from getting wet. I can’t agree with that. When you think seriously about the scale and uncertainty involved, it would be impossible to live at all if we were constantly trying to assess whether every action would lead to statistically more or less happiness to the aggregate of all human beings through all time. We would agonize over every cup of coffee, every new video game. In fact, the global economy would probably collapse because none of us would be able to work or willing to buy anything for fear of the consequences—and then whom would we be helping?

That uncertainty matters. Even the fact that there are other people who could do the job matters. If a child is drowning and there is a trained lifeguard right next to you, the lifeguard should go save the child, and if they don’t it’s their responsibility, not yours. Maybe if they don’t you should try; but really they should have been the one to do it.
But we must also not allow ourselves to simply fall into apathy, to do nothing simply because we cannot do everything. We cannot assess the consequences of every specific action into the indefinite future, but we can find general rules and patterns that govern the consequences of actions we might take. (This is the difference between act utilitarianism, which is unrealistic, and rule utilitarianism, which I believe is the proper foundation for moral understanding.)

Thus, I believe the solution to the Tragedy of the Commons is policy. It is to coordinate our actions together, and create enforcement mechanisms to ensure compliance with that coordinated effort. We don’t look at acts in isolation, but at policy systems holistically. The proper question is not “What should I do?” but “How should we live?”

In the short run, this can lead to results that seem deeply suboptimal—but in the long run, policy answers lead to sustainable solutions rather than quick-fixes.

People are starving! Why don’t we just steal money from the rich and use it to feed people? Well, think about what would happen if we said that the property system can simply be unilaterally undermined if someone believes they are achieving good by doing so. The property system would essentially collapse, along with the economy as we know it. A policy answer to that same question might involve progressive taxation enacted by a democratic legislature—we agree, as a society, that it is justified to redistribute wealth from those who have much more than they need to those who have much less.

Our government is corrupt! We should launch a revolution! Think about how many people die when you launch a revolution. Think about past revolutions. While some did succeed in bringing about more just governments (e.g. the French Revolution, the American Revolution), they did so only after a long period of strife; and other revolutions (e.g. the Russian Revolution, the Iranian Revolution) have made things even worse. Revolution is extremely costly and highly unpredictable; we must use it only as a last resort against truly intractable tyranny. The policy answer is of course democracy; we establish a system of government that elects leaders based on votes, and then if they become corrupt we vote to remove them. (Sadly, we don’t seem so good about that second part—the US Congress has a 14% approval rating but a 95% re-election rate.)

And in terms of ecology, this means that berating ourselves for our sinfulness in forgetting to recycle or not buying a hybrid car does not solve the problem. (Not that it’s bad to recycle, drive a hybrid car, and eat vegetarian—by all means, do these things. But it’s not enough.) We need a policy solution, something like a carbon tax or cap-and-trade that will enforce incentives against excessive carbon emissions.

In case you don’t think politics makes a difference, all of the Democrat candidates for President have proposed such plans—Bernie Sanders favors a carbon tax, Martin O’Malley supports an aggressive cap-and-trade plan, and Hillary Clinton favors heavily subsidizing wind and solar power. The Republican candidates on the other hand? Most of them don’t even believe in climate change. Chris Christie and Carly Fiorina at least accept the basic scientific facts, but (1) they are very unlikely to win at this point and (2) even they haven’t announced any specific policy proposals for dealing with it.

This is why voting is so important. We can’t do enough on our own; the coordination problem is too large. We need to elect politicians who will make policy. We need to use the systems of coordination enforcement that we have built over generations—and that is fundamentally what a government is, a system of coordination enforcement. Only then can we overcome the tendency among human beings to become apathetic and short-sighted when faced with a Tragedy of the Commons.