Yesterday, we conducted a poll asking how often our readers would give money to truly talented street performers. The poll was in response to a Washington Post experiment where world-famous violinist Joshua Bell performed in a subway station for 45 minutes and only earned $32, an amount that would pay for less than a third of a seat at one of his concerts.

One objection to the Post experiment is that Bell wasn’t in a very good location: he was at the entrance to the station, during morning rush hour, when people have the least time to stop and listen. Our poll tried to determine what might have happened had he been in a spot where people are more likely to have time to listen: we asked how often readers give at least $1 to truly talented street performers, when they’ve got at least 5 minutes to listen to them. This graph shows our results:

Almost 50 percent of respondents said they’d give money at least 50 percent of the time! Our 176 respondents would have given at least $80.80 — that’s a lot more than the $32 Bell actually earned during his performance, when over 1,000 Washingtonians passed by. But there are some problems with our “study”‘s method.

First, people might claim to be more generous than they really are. I’d say that’s a significant problem, but it’s at least partially countered by the fact that there is no opportunity for respondents to say how much they give — surely some people will give even more than $1 for a truly impressive performance. If these two factors balance each other out, then when you scale up our response to match the 1079 passersby during Bell’s actual performance, he would earn a whopping $495.36!

That’s still not much compared to what Bell earns for a real concert, but it’s definitely enough to keep him in caviar for a few days.

But there’s another problem with our results: the response rate. On a typical day, Cognitive Daily is visited by around 4,000 different people — but only 176 responded to our poll. Perhaps the people who responded to the poll are disproportionately generous to street performers. Now, the 4,000 number is a bit misleading, because most of those visitors are following links from other sites — which lead directly to other articles. Fortunately, we also have statistics for the number who viewed the main page yesterday, and for those who went directly to the poll post. Combine those two figures, and a total of 836 readers saw the poll, which means 660 people didn’t bother to respond.

I think you can make a good case that these people are so ambivalent about street performers that they probably fall in the “never give money” camp. If you add these responses into the data and recalculate, Bell’s theoretical earnings from 1079 passersby drops to $104.28, or an hourly rate of $138.70 — not bad, but Bell may want to switch to domestic caviar, and he’ll almost certainly have to pawn his $3.5 million Stradivarius.