The central implication of maximizing behavior amid competition is that rates of return tend toward equality.
We test that implication in a market whose participants have the traits that behavioral economics suggests should make it hardest to find evidence of maximization: the market for panhandling at Metrorail stations in Washington, District of Columbia.
We find that stations with more panhandling opportunities attract more panhandlers and that cross-station differences in hourly panhandling receipts are statistically indistinguishable from zero.
Panhandling rates of return thus tend toward equality. Extreme ‘behavioral’ traits do not prevent maximization in this market.
…We collect data on the number of panhandlers at 26 Metrorail stations in Washington, DC and on hourly panhandling receipts at 5 of those stations. Metrorail is Washington’s public rapid-transit system. Panhandlers solicit passersby outside its station exits. Some Metrorail stations are trafficked by more passersby and thus offer more panhandling opportunities. If panhandlers respond rationally to incentives, such stations should attract more panhandlers. And if panhandlers’ station choices are maximizing, panhandling rates of return across stations should tend toward equality.
We find that stations with more panhandling opportunities attract more panhandlers and that cross-station differences in hourly panhandling receipts are statistically indistinguishable from zero. Panhandling rates of return thus tend toward equality. Extreme ‘behavioral’ traits do not prevent maximization in this market. Panhandlers choose stations as homo economicus would if homo economicus were a street person who solicited passersby at Metrorail.
…We observed hundreds of panhandlers for hundreds of hours over a period of 13 months and did not observe a single panhandler being interfered with by Metro Transit Police or other authorities. Lawful or simply ignored, panhandling in the Metro spaces we study proceeds unmolested.
The market for panhandling in those spaces exhibits free entry and exit. We observed no effort by any panhandler to limit or otherwise control the presence of other panhandlers (or anyone else) at any Metro station. Panhandlers frequently came and went from stations where other panhandlers were present and did so without conflict or even acknowledging one another. We saw no evidence of panhandler property rights to solicit at certain Metro stations. To the contrary, we encountered different panhandlers on different visits to the same stations. Nor did we see evidence of panhandler property rights to occupy certain spots at a given station, save the fact that no panhandler attempted to occupy a spot while it was occupied by another panhandler. Panhandlers do not sleep at Metro stations, so spots are reallocated daily, if not sooner when a panhandler moves on.
…Number of Panhandlers and Passersby: For 10 months in 2016–2017 we visited 25 Metrorail stations and the intersection of Wisconsin Avenue and M Street in Georgetown—a popular shopping corridor—to collect data on the number of panhandlers.5 Appendix A maps Metrorail. Solid circles identify sample stations.6 They cover all 6 Metro lines and serviced nearly half of all Metro riders during our study period.7
We made a total of 242 Metro station visits to collect data on the number of panhandlers. We visited each sample station an average of ~9× over 4 months. On each visit we canvassed a one square-block area around the station exit(s) to count panhandlers.8 Every street person observed soliciting donations from passersby was considered a panhandler. Street people were identified by appearance: the ‘disheveled, [and] apparently destitute’ (1996 [Making Room: The Economics of Homelessness], pg7). Our data contain 258 panhandlers, 218 of whom are unique. We use them to create a variable that measures the number of panhandlers at each Metro station on each visit.
…All subsample stations were visited simultaneously 7:30AM–11:30AM. On each visit we canvassed a one square-block area around the station exit(s) for panhandlers and selected one or more panhandlers for observation. Panhandler selection was guided by the practicality of discreetly observing panhandlers work, and preference was given to panhandlers whose observation would permit simultaneous observation of other, neighbouring panhandlers. Selected panhandlers were observed until they departed the station. After they did, the station was canvassed again and new panhandlers were selected for observation. This procedure was repeated until the collection day ended at 11:30AM We observed a total of 67.6 hours of panhandling work.
Panhandlers were observed at work without their knowledge. We recorded the number of donations each panhandler received and the number of minutes he was observed working. The average panhandler in our subsample was observed working for 51 minutes. When a panhandler departed the station, if he received any donations, we approached and offered him $5 to count in front of us the money he just received. 82% of approached panhandlers accepted our offer.14 After watching a panhandler count his money, we recorded the dollar amount.15
…The average panhandler in our subsample receives 2.8 donations per hour, the value of which is $6.10. That is equal to 46% of the DC minimum wage and to 84% of the federal minimum wage at the time of data collection. The median panhandler in our subsample receives 1.8 donations per hour, the value of which is $1.40. Figure 1 illustrates why average and median hourly panhandling receipts differ. Figure 1(a) depicts the distribution of the number of panhandling donations received per hour at each Metro station. Figure 1(b) depicts the distribution of panhandling dollars received per hour at each station. In both panels, a small fraction of panhandlers receives large receipts, and a large fraction of panhandlers receives none. Panhandling is like fishing: often it is a bust, but occasionally one lands a ‘whale’ that makes the effort worthwhile.
…That extreme ‘behavioral’ traits do not prevent maximization in the market for panhandling begs the question of why modest behavioral traits seemingly often prevent maximization in other markets. One possibility is that the stakes for panhandlers who fail to maximize are more dire than for participants in most other markets because, unlike participants in those markets, panhandlers live at the edge of subsistence. If panhandlers did not maximize they might not survive. Another, non-mutually exclusive, possibility is that modest behavioral traits do not prevent maximization in other markets as often as some have claimed. Maximization may be a more robust foundation for economics than behavioral considerations suggest.