This is a quick analysis on how people tipped NYC cabs in 2013. The data was retrieved from Chris Whong’s blog. It’s a sizeable dataset and an amazing visualization came out of it called NYC Taxis: A Day in the Life. I got interested in the dataset and wanted to look at how people tip in NYC cabs.

# Tips by payment type

After cleaning up the dataset, I first noticed that more than half the rides were paid by credit card.

When people pay by cash, it turns out tip is not recorded for 99% of the rides. I’m not sure if this is because drivers can’t record tip into their meter or if they just don’t bother. One thing I do know is that there are at least 6639 cases of tip recorded when paid in cash.

When people pay by card, the tip is automatically recorded. and it shows a nice distribution of tip percentage with about 40% of passengers paying tips of 20%.

# Tips by group size

The main question I wanted to examine was whether people’s tip change as a function of group size.

This is a question that one can easily see going two ways. I might be inclined to tip more in a group than alone because 1) I feel social pressure, or 2) I feel like I owe more service because the driver dealt with more people.

Alternatively, I might tip less in a larger group because 1) diffusion of responsibility, or 2) I feel less connected to the driver since the service is distributed to the group than concentrated on me.

Here is the result:

Consistent with the second idea, tip percentage tends to decrease as group size gets larger.

This is consistent with findings in restaurants, where larger groups tend to tip less than smaller groups.

But why is this the case? It’s difficult to look at the social computations going on in an individual’s head in this kind of data. One possibility cited in the above article referencing Elman(1976) is that larger groups tend to have larger bills where the financial burden increases holding tip percentages. I find that this is not the case below.

# Conclusions

1. People in larger groups tip less.
2. Larger groups don’t necessarily spend more.
3. Tips tend to decrease as total fare increases but drops more heavily near fares close to ending in 0s.