Economists Do It With Models

I’m guessing many of you are familiar with the trolley problem:

The problem is interesting because there is an objectively “right” answer- absent specific circumstances, one dead person is better than five dead people- but psychology, philosophy, ethics, etc. bring in a whole host of other considerations having to do with intention, fate, and so on. Such considerations result in a problem without a correct answer, and these considerations can’t (and probably shouldn’t) be ignored in a society of human beings and not robots.

Because of this unexpected complexity, the trolley problem has spawned a number of extensions, ranging from the even more nerdy…

…to the snarky and political:

There’s even what I will call the economist version, which incorporates opportunity cost/cost of effort as well as a few other factors…

In any case, we seem to be pretty familiar with the “clear efficient answer under some basic assumptions, but fairness and ethical considerations make things complicated” concept. So allow me to present a more accurate economic version of the trolley problem:

You are currently looking at a crisis area. The status quo is that there is a large shortage of Uber drivers to get people out of the area. Do you 1. Do nothing, or 2. Implement surge pricing?

I think this is a situation that we’ve seen before a number of times. Allow me to explain the similarities to the trolley problem:

“Clear Efficient Answer Under Some Basic Assumptions”

Surge pricing is the obvious answer here, under two assumptions: first, that surge pricing gets more drivers to the area, and second, that how much a person is willing to pay for an Uber is an accurate proxy for how important it is to them. Under these assumptions, shortages are smaller (or nonexistent) under surge pricing than they would be otherwise, and cars go to those who need/want them the most. (In case you’re curious, the first assumption seems to have empirical support even though surge pricing doesn’t appear to always get more drivers on the road overall.)

“But Fairness and Ethical Considerations Make Things Complicated”

I can’t really tell you what’s fair- that’s kind of the deal with value judgments- so I will instead report some common themes that I’ve come across. One is that people should have at least a chance to get an item at the “regular” price, and some people view random rationing as more ethical than price-based rationing when extenuating circumstances are present. (I wonder how this would change if pricing were framed as regular prices and discounts rather than surge pricing.) Another is that willingness to pay is a better proxy for wealth than need/want, in which case surge pricing unfairly rations items to rich people. (This may be true in cases of extreme income inequality, but shouldn’t be the case in a market with more uniformly distributed resources, so this view is somewhat of a fact/opinion hybrid.) Yet another that hadn’t even occurred to me (thanks Internet!) is that it’s unethical to use the promise of money to get largely low-income individuals (the Uber drivers) to take on risk of bodily harm, especially when said risk is incurred during the service of higher-income individuals. (See last point, and note that this is the same logic used to justify outlawing kidney donations and such.) Yet another is what Russ Roberts says. You’ll notice that all of the fairness arguments presented except the last one are against surge pricing.

My point in bringing this up isn’t to have a discussion on fairness or convince you of anything in this regard- like I said, you’re more than welcome to subscribe to one of these viewpoints or come up with your own. My point, instead, is to highlight the role that economics can play in conversations about what is best for society. To that end, here are a few points to keep in mind:

• Economists can tell you what is efficient under certain assumptions, but they can’t definitively tell you what is fair.
• The assumptions used to determine the efficient outcome can and should be examined.
• Just because something is a market outcome doesn’t mean it’s fair. (In fact, the existence of market failures implies that market outcomes aren’t even always efficient.)
• Fairness matters.
• But so does efficiency- it’s reasonable to ask economists to stay in their lane, but not to discount them entirely.

I guess I could do something similar for economists:

• State your assumptions.
• At least try to stay in your lane.
• When you venture into the philosophical lane, make it clear that you are doing so.
• Fairness matters, and people really hate it when you dismiss their value judgments as irrelevant. They’re even likely to reject what you can show them about efficiency if you do so.

Okay? Great- now let’s go have some thoughtful policy discussions.

Technically, “constant returns to scale” describes a production process where you get exactly twice as much stuff out if you put twice as much stuff in. Economists often argue that at least constant returns to scale should be achievable since, worst case scenario, you could just build a second identical factory next to the first one. As such, I want economic instructors to start using this as their example of constant returns to scale.

Alternatively, you could examine the economics of chocolate more directly. Mmmmmmmmm…

Technically, I’m cheating with the “causal Friday” title, since, while regressions do identify associations that exist when controlling for other variables, these associations aren’t always of the causal variety. (This is particularly true when not all relevant factors can be controlled for.) But I choose to not be too persnickety because I think the comic is funny and wanted to share it.

…

…

…

Okay, you should have known better than to believe that I was going to avoid “too persnickety.” Personally, I won’t decide whether I am suspicious of the linear regression until someone tells me whether the slope is statistically significant. Also, if there are multiple explanatory variables that affect an outcome, a scatter plot that only looks at one of them at a time will generally looks like a mess even when all of the variables are individually important. In related news, this is a good opportunity to talk about the distinction between estimated effects (i.e. regression coefficients) and R-squared. (Don’t stop reading if you aren’t super into econometrics, I promise to make this make sense.)

Let’s say an economist is trying to model how much coffee I drink. (In reality, this is not necessary- the regression would just have a really big constant term, but go with me here.) Unfortunately, the only data available to use as an explanatory variable is income. Obviously, there are a lot more factors that affect my coffee consumption than just my income, so it shouldn’t surprise you that if I were to plot coffee consumption as a function of income (where each data point is a month of time, let’s say) I would get something that looks like the scatter plot above.

Let’s say that I’m measuring my income in hundreds of dollars and the estimated slope of the regression line is 0.01. This means that, on average, each hundred dollar increase in income is associated with 0.01 more coffees per month. If the numbers show that this estimate is statistically significant, then it’s pretty unlikely that this association exists in the data by random chance. Let’s also say that the R-squared of the regression is 0.06, like in the picture. This means that changes in my income only explain 6 percent of the variation in my coffee consumption.

My point is that these two conclusions aren’t in conflict with one another- it’s entirely possible for a relationship to both be statistically significant and for it to explain only a small fraction of what is going on. (This happens a lot in finance, actually, and an R-squared of 0.06 wouldn’t generally be seen as a red flag just because there is so much unexplainable noise in the data.) Sure, the result would be more impressive with a higher R-squared, but it’s largely a matter of personal judgment whether explaining, say, 6 percent of a phenomenon is worth talking about. (Not gonna lie- some economics journals vote no on this question.)

That said, I do recommend watching out for a red flag of a slightly different sort- one of the conditions in order for a regression to be valid is that your explanatory variables are uncorrelated with all the relevant stuff that you aren’t controlling for (your error term, in technical terms). In the case of my coffee regression, my result is valid only if my income isn’t correlated with whatever else could be causing variation in my coffee consumption (hours worked, for example). I can tell you personally that that is a lot of stuff.

I’m now tempted to perform a neural net analysis of my coffee consumption in order to see if I could get Rexthor out of it.

Well played, sir. =P

Tired of the disposition effect yet? This one’s short, I promise- just shows how the incentives for tax-motivated selling of losing stocks change over the year and cloud the disposition effect test statistics.

As usual, you can see the whole behavioral economics playlist here in case you want to catch up or need a review.