Typically, I run my stuff past a few test readers to see if it will meet the lofty standards of Freakworthiness. Reactions were mixed on my plan to do a series on gender and travel (see the first installments here). Some thought it was bound to be a bore because, duh, of course men’s and women’s travel patterns are going to be very different. Others thought it was bound to be a bore because, duh, of course in this day and age men’s and women’s travel patterns are going to be very similar. There was widespread sentiment that I’m going to bore you — but who was right about why?
I’ve been working with data from the American Time Use Survey (ATUS), which asks people what they’re doing on a minute-by-minute basis. These data are a treasure trove for social scientists and anybody interested in the way Americans live (Levitt begged to have it saved from the chopping block here. Cheery facts abound: for example, Americans with kids report spending about five minutes a day on their children’s education and two hours and twelve minutes a day watching television.
Place your bets: who spends more time traveling, men or women?
According to the ATUS, between 2003 and 2007 adult Americans averaged about one hour and fifteen minutes of travel daily. Men averaged an hour and eighteen minutes: women, an hour and thirteen. This difference is statistically significant, but in all I’d call a five-minute disparity strikingly small.
We can dig deeper using the empirical economist’s best friend, a mathematical technique called regression. Regression is a statistical technique that helps to untangle which factors are truly important for a particular outcome and which factors are relatively trivial.
We need to do regression here because men and women differ, on average, in ways besides mere gender. For example, women tend to live longer than men, so taken as a group they are a bit older. Perhaps observed differences between men and women are actually due to age, not sex.
To construct a regression model, you feed in the variables that you think work in concert to affect your outcome — in this case, daily travel time. I used age; number of children in the household; number of adults in the household; race/ethnicity; living in a center city, suburb, or nonmetropolitan area; family income; hours worked; weekly wages; and travel mode, specifically mass transit ridership.
Which variables matter? The numbers show that age is an important factor in determining how much we travel. Holding all the other variables constant (including, importantly, income), daily travel time is predicted to rise till we are about age 36, peak there, and then begin to decline.
Controlling for the other variables, those who live in suburbs are projected to travel about five minutes more per day than others.
Both family incomes and hours worked make a big difference. All else equal, more of each is associated with more travel.
Perhaps the most stark relationship is between travel time and transit ridership. All else equal, travelers who go by bus or train, even for one minute in the day, are predicted to spend about an hour a day more traveling than those who also travel but do so exclusively by car and/or on foot. This disparity holds up even when controlling for the fact that transit riders also walk a lot, and for the fact that they are more likely to live in New York City. But this is a story for another time.
And what of gender? Controlling for the other factors, men and women are even more similar than the simple averages indicate. In fact, there is very little travel-time difference between the sexes.
Thanks to the magic of regression, we discover that the majority of the small difference between the two genders can be explained by hours worked (the study finds men average 50 percent more than women) and family income (very roughly, the study shows that men have family incomes about $5,000 a year higher than women’s). Controlling for these characteristics, we find that men and women have daily travel times that are within about 1.2 minutes of each other. Given that mean travel times are over 70 minutes, total travel by gender is pretty close to identical.
And lest you doubt my fancy statistical hocus pocus, here’s some corroboration: according to the 2001 National Household Travel Survey, men and women both take an average of about 4.1 trips per day.
Bored yet? You may be, but don’t blame the data: the finding that men and women travel virtually the same amount is dramatic, cast-iron proof that we have reached a golden age of complete gender equality in America.
Or is it? More coming up.
and 
. You can also watch it on 
Wait – so correcting for all the factors that make the difference interesting, there’s no difference?
Men work 50% more hours and have family incomes $5000 greater than women and this accounts for most of the difference in travel time? Duh. People who are engaged in different non-travel activities travel different amounts. Most travel is instrumental to other activities, and is not done for the sake of traveling.
If it is true that men “work” 50% more hours and have family incomes “only” $5000 higher, it suggests very different gender roles. This deserves more than a mention in passing.
I agree with John. Why state that men’s working for 50% more and having bigger incomes does not imply differences between sexes / gender roles?
I’m not as worried about the statistics as Panem et Circanses might be, but the interpretation seems like it could be reconsidered… First, I would have expected something like “the differences–UNadjusted differences, that is–are quite small, and that tiny difference seems really rather shocking considering that men work 50% more than women and have higher incomes.”
But then, delving into the effects of those two variables ((a) time devoted to work & (b) income) muddies the water a lot. Less time at work, eh? Well, of course, only about 20% of all travel is work-related, so then we might start thinking that working less might lead you to travel farther (more time to go to the more distant preferred place rather than the nearby poor substitute) but then again lower incomes should lead to less travel, right? Also, since women still make the majority of childcare and household-serving trips in dual-earner married households, mightn’t we think that these women are traveling more — though then again, we know that they are more likely to take more proximate jobs in this situation. Et cetera et cetera.
@7–Rachel.
I made a mistake. I meant to say “No REAL men do this”.
what does “travel” mean? if I walk a couple of blocks to buy lunch, is that travel? If I walk downstairs to the cafeteria in my building, does that count?
Well I know that’s false, I know a man who has three or four alarms set on his cell phone so he can fool himself into thinking he has more time to sleep.
I suppose the data means that men, in general, travel to and from work and to/from their work/lunch. Women, in general, go to a lot more places (errands).
Correct?