Panhandlers and Price Theory

Do panhandlers behave as price theory would predict?

In my previous two posts (here and here), I focused on the role of price theory in explaining drug use. One reason that I wrote those posts is that I wanted to illustrate that price theory has explanatory power beyond the standard examples. I’m a firm believer that price theory is useful for understanding any decision that involves costs (which is essentially all decision-making). This doesn’t mean that there aren’t other useful tools or even that price theory is the most important tool in all circumstances. Nonetheless, price theory often has something to add when it comes to understanding decision-making.

I also like examining topics like this precisely because they highlight what price theory is and is not. Price theory is a framework for understanding, explaining, and predicting human behavior. Price theory is not a theory of mind.

A different way of thinking about this is as follows. In the early days of behavioral economics, a number of researchers focused on finance. The reason that they focused on finance is that financial markets tend to be populated by professional finance people who have an incentive to get things right and acquire information. Financial markets are also characterized by frequent trading and highly liquid markets. Given those characteristics, the thinking was that if behavioral assumptions could survive financial markets and not simply be arbitraged away, this would be strong evidence in favor of these behavioral assumptions. We can debate whether that turned out to be true. Nonetheless, the sentiment itself is useful.

I take a similar view of using price theory to analyze things like drug use. The stereotypical view is that drug users don’t behave in predictable ways. If price theory can explain how people respond to changes in drug policy or even what types of drugs that people use, then this would seem to be a pretty strong indicator of the usefulness and broad applicability to price theory.

With that in mind, I would like to continue with this theme of exploring price theory in a context where one might otherwise find it lacking: panhandling.

Competition, Entry, and Profits

Price theory, at least the Chicago brand, tends to focus a lot of its attention on the so-called competitive model. A basic idea of the competitive model is that competition will tend to drive economic profit to zero. This does not mean that competitive firms don’t earn a profit. Instead, what this means is that the accounting profit earned is equal to the opportunity cost to the owner(s) of operating the firm.

In finance, there is the idea that variation in returns of different assets can be explained by the risk characteristics of the different assets. A higher expected return is simply compensation for taking on more risk. A useful concept that follows from this logic is that equilibrium is characterized by the absence of risk-free arbitrage. Again, competition should see to it that this is true. If one can earn income from risk-free arbitrage, then we should expect people to enter the market to capture some of those profits. Furthermore, we should expect that people will continue to enter the market until the only arbitrage opportunities left require taking on some degree of risk.

This is not to say that these outcomes will always prevail. The basic assumption lurking in the background, but unacknowledged explicitly to this point, is that these outcomes are predicated on free entry. The presence of positive economic profit or risk-free arbitrage could be the result of the fact that it is costly to enter the market. This could be true for a variety of reasons. There could be legal or regulatory barriers to entry. There could be significant fixed costs associated with entering the market. In that case, one has to make sure that the profits from day-to-day operations are large enough to offset that initial fixed cost.

Suppose that you wanted to test this idea that, all else equal, competition tends to equalize rates of return on particular types of investments or activities. How could you do it?

One way to test the theory would be to estimate the rate of return on investment in various industries. In a competitive market with free entry, one would expect that the rate of return on investment would equalize across industries. Any failure to equalize these rates of return should be explained by some sort of barrier to entry.

This is easier said than done. For one thing, there is no reason to think that at any particular moment in time, we are in equilibrium. This is a challenge in finance. Sure, there are a lot of papers that seem to show evidence of clear arbitrage opportunities. However, the fact that trading is happening by the second in financial markets seems to indicate that there are a lot of people who believe they have an arbitrage opportunity. Even if one were to observe that an arbitrage opportunity exists at a moment in time, this would not tell us whether the absence of risk-free arbitrage is a useful equilibrium concept because we might simply be on the path back to equilibrium.

Although I’m not sure that David Laidler ever wrote this down, I’ve heard him say that every economist should have to revisit their dissertation at the end of their career and see if it holds up. In our context, this would mean that if you had evidence of risk-free arbitrage or the failure to equalize returns across industries, you would have to replicate your result after approximately 30 additional years of observations. If you could replicate it, that would be pretty strong evidence that you were correct.

But all is not lost. We need not wait 30 years for people to update their data. There are alternative ways to test the implications of the competitive model.

For example, think of a situation in which there is (a) free entry, but (b) some reason to believe that the competitive model might not apply. If one finds evidence in favor of the competitive model under those conditions, then that is pretty strong evidence in support of the model.

One such example is that of panhandlers. It seems pretty clear that there are places in which panhandling is perfectly legal and there is no restriction on panhandling. The absence of those restrictions suggests that anyone can show up and panhandle. As a result, if panhandling turns out to be quite lucrative, we would expect to see people switch to panhandling from some alternative use of their time. In fact, competition in panhandling should drive down the rate of return on panhandling until it reaches the opportunity cost of the marginal panhandler.

At the same time, casual observation suggests that panhandlers might suffer from the same erratic behavior attributed to drug users. This shouldn’t be surprising since many panhandlers tend to be drug users. Given their erratic behavior, a common retort would be that panhandling isn’t the best example. Sure, there is free entry. However, we are not dealing with rational actors. We are dealing with people who are on drugs or desperate. They aren’t calculating where to panhandle to get the highest rate of return. They’re panhandling because they need all the help that they can get, possibly to buy a meal or perhaps even to get their drug fix.

It is not despite those objections, but precisely because of those objections that this is a useful testing ground. There is free entry, but panhandlers aren’t crude, calculating maximizers. So what happens?

Evidence on Panhandlers

Fortunately for you, dear reader, there are people who get paid to think about such things. A recent paper by Peter Leeson, August Hardy, and Paola Suarez used data they obtained by observing panhandlers in Washington, D.C. to test this very hypothesis.

Specifically, they observed panhandlers outside the D.C. Metrorail stations. The places where the panhandling takes place around these stations appears to be legally permitted based on local statutes. Regardless, at the very least, it is de facto permitted. They were able to observe the panhandlers and get information about how much they collected by asking the panhandlers to count their earnings in exchange for five dollars. They are also able to get data on the number of customers that pass through that area using publicly available data from the Metro system. They also used a standardized process to examine the friendliness of passersby.

One possible barrier to entry is the ability to get to a particular Metro station. Not all homeless are panhandlers, but most panhandlers are homeless. If that is the case, then a basic barrier to entry might be the ability to get to a particular Metro station.

Setting aside whether competition equalizes rates of return across different stations, one simple test would be to see if panhandlers follow basic economics incentives. For example, one would expect that there would be more panhandlers at the busier stations and the friendlier stations. One would also expect that there would be more panhandlers where the barriers to entry are low, such as at stations where there is a shuttle stop for the homeless nearby.

This is precisely what they find for the full sample.

Nonetheless, it is important to remember that the main thing that they want to test is whether the rates of return equalize. Yes, all else equal, one should expect busier stations to have more panhandlers. When Willie Sutton was asked why he robbed banks, he replied “because that’s where the money is.” The same principle applies here. Busier stations mean more people to ask for money. The more people one asks, the more likely one is to receive some money from a passerby. At the same time, this tendency will cause rate of return equalization between Metro stations. Busier stations have more competition, which drives down returns. Less busy stations will have less competition, which boosts returns.

Because they were able to obtain data on how much the panhandlers earned per hour directly from the panhandlers themselves (and verify the information by having the panhandler count for them), they are able to test whether the mean, median, and variance of the returns are similar across stations. Using the data that they obtained, they cannot reject the null hypothesis that the mean, median, and variance of returns across the stations are equal.

This is especially strong support for the competitive model. Panhandlers are much more likely to suffer from mental health disorders and substance abuse problems than the rest of the population. In addition, panhandlers are often thought to have less self-control and be deemed erratic or irrational. Yet, the prediction of the competitive model prevails even in those circumstances.

Conclusion

I bring up examples like this because it allows me to address criticisms of price theory that largely miss the mark. Critics of economics generally, and price theory in particular, tend to argue that we assume that everyone is a hyper-maximizer, only concerned with self-interest and that the world is more complicated than that. People aren’t walking around all day solving utility- and profit-maximizing problems in their head. Not everyone is a rational calculator.

I think that we can reject these criticisms. I am not saying that we should reject them on the grounds that they are false characterizations of the real world, but rather that they are a false characterization of price theory. As I wrote in my previous posts, price theory is about providing rational frameworks to understand, explain, and predict human behavior. It is about rational frameworks, not rational people.