Incentivizing the School Commute

Photo: Mat_the_W

We’ve written about bribing kids to get better grades. But what about bribing them to walk or ride their bike to school?

A new working paper examines a program in Boulder, Colorado that attempted to incentivize kids to bike or walk to school over a span of several years. The program began with a $10 cash prize for the first two years, but then switched over to a $10 bike store coupon thereafter. One lucky student who rode and walked to school every day during a “prize period” won the coupon.

Even considering the small, non-cash winnings, biking and walking to school increased 16 percent during the prize period.  Here’s the abstract:

We analyze the effects of a school-based incentive program on children’s exercise habits. The program offers children an opportunity to win prizes if they walk or bike to school during prize periods. We use daily child-level data and individual fixed effects models to measure the impact of the prizes by comparing behavior during prize periods with behavior during non-prize periods. Variation in the timing of prize periods across different schools allows us to estimate models with calendar-date fixed effects to control for day-specific attributes, such as weather and proximity to holidays. On average, we find that being in a prize period increases riding behavior by sixteen percent, a large impact given that the prize value is just six cents per participating student. We also find that winning a prize lottery has a positive impact on ridership over subsequent weeks; consider heterogeneity across prize type, gender, age, and calendar month; and explore differential effects on the intensive versus extensive margins.

This bump in exercise is significant, especially considering the overall downturn in children walking to school at all. From 1969 to 2001, the percentage of self-commuters dropped from 41 percent to 13 percent. In his recent article  pointing this out, Freakonomics contributor Eric Morris ended on an inquisitive note:

“So the next question is, what factors cause kids to walk and bike? And how can we use that information to promote active travel?”

In the Boulder experiment, children who won the lottery were excluded from winning future lotteries, but remained actively commuting to school for the next two weeks. After this, however, they went back to normal. Problematically, the 16 percent increase in walking and biking did not exist during non-prize-periods. But this study, at least in part, offers the beginning of an answer.


Isn't the newer incentive somewhat circular? A student who doesn't walk or bike to school has no need for a bike store coupon. Theoretically, this program would only benefit the students who are already biking or walking to school.


Intriguing findings, especially since it seems that walking and riding are directly tied to the prize-periods.

Did the school offer other incentives, because I can't see how ten dollars or a coupon would really motivate someone to ride/walk to school? Were there awareness and education events during the prize-periods that may have influenced decisions to ride/walk?


Why not give each kid 1/bus route capacity* busing expense?


When oh when will we learn that incentives do not work look at the work of Alfie Kohn for the evidence of it. Rewarding or incentivising only gets people to change behaviours in the short term. Afterwards they will entirely predictably return to previous behaviours and then reject any notion of change without increased incentives.


That's a pretty bold claim. Incentives don't need to be money. An incentive is just a factor that motivates or enables some behavior. While incentives might have unintended consequences, they are central to understanding human behavior. I concede that incentive structures are difficult to design, especially with unintended consequences, but that does not mean that "incentives do not work."

Also, because the 16% increase was not maintained throughout non-prize periods, we can't conclude that positive behavioral change didn't occur . Imagine a student who seldom bikes or walks to school, yet responds to the incentive by participating, to realize that he enjoys biking and walking. Thus, he continues to bike/walk part of the time to school or in his free-time. While their study might not have measured this positive change, it would be difficult to argue that the incentive failed or that it only caused short-term change. Unless of course the goal of the policy was specifically to encourage biking and walking to School, and not healthy behavior and exercise in general.

As long as incentive structures are designed properly and given serious thoughts (and consideration of unintended consequences), they can and do work.



To solution is simple, on an individual level, I simply don't offer to drive my kids to school. A 30 minute walk twice a day is good for them (I walk to work too). But then again, this post was about how to change a trend in the population, which is a lot harder. I think every effort to solve this puzzle should be encouraged. Somehow, I still think that parents will be the key, not the kids., and their incentives would probably be more about safety issues than monetary/material rewards.


I agree, we have to focus more on incentivizing the parents than the kids, because ultimately the kids have to find another way to get to school if the parents refuse to drive them. This would work for kids around middle school age and older. But another problem these days is that it's not considered "safe" for younger kids (elementary school age) to walk to school without an adult with them the way it was when I was a kid (in the 1980's). I remember walking to school with other kids my age from the time I was 6, in first grade. As I understand it, the kids at my son's public elementary school are not permitted to leave the school without a parent or other approved adult to walk home alone. I live in a perfectly safe neighborhood with quite streets, and my house is a 10 minute walk from the school, for a child, yet there is a bus that takes him there and back. It would be perfectly safe for him to walk to and from school if there were a bunch of other kids walking at the same time, but there aren't - so he can't.



How do these kids get to school when they don't bike / walk? Do their parents drive them? Because then I don't think they've proved anything other than that you can use a small prize to guilt parents into certain behaviors.


since i convinced our office manager to allow me to store a bicycle in our server room (rather than on the street) everyone (except the office manager) has started to cycle in at least once a month. and 3 of the 10 staff now cycle every day. this is from a baseline of 0 before internal bike storage.

everyone uses public transport when not cycling. cycling is the fastest and most predictable (time wise) way for everyone to get to work.

the key determining factor for the occasional cyclists seems to be the weather.

every member of staff is and was able to ride a bike at all times.

the decrease in cost of cycling was key (lower chance of bike theft) combined with the evidence of its ease and reliability (me doing it every day) seems to have altered the balance enough for almost everyone.

offering very low prizes would have been unlikely to change any of our behaviour.
but for the same cost my employer was able to dramatically lower my personal cost of cycling.



I have kids in Boulder schools. Yes, there was extensive education in addition to the incentives. Our kids schools have a large portion of student from out of the district and very limited school bus service. This creates a rather congested environment at morning drop offs. There was lots of encouragement to drop your kids off at less congested points further away from school. Walking a little way was considered better then nothing.

Since cynicism runs high in my household, the incentives had a negative effect on one child, who took it as an opportunity to defy the school and found every way they could to get rides those weeks.

To answer another question, Boulder student do get subsided passes for public transit. In fact, the school district says this is the intended replacement for school bus service in most areas.


Oh, and there was an incentive to the parents to take part in a survey on walk/biking to school. And I won a $25 gift card to REI. Take that kidos!


Wait -- isn't biking to school child neglect?


Another option:
Eliminating buses


I'm doubting that would work in poor, rural areas where students are sometimes 30+ minutes away from the school and parents lack the resources to provide transportation.


On the contrary, in poor rural areas (or at least the one I grew up in some years ago) a 20-30 minute walk or ride to the school was commonplace.


And students probably walked in the snow, up-hill both ways.

I know many people in Appalachian communities (especially with the school consolidation movement of the 90's) who are 20-30 miles from schools. Add this to the safety hazards created by a lack of lighting, markings, etc. on roads, and abolishing buses sounds like a pretty detrimental policy.


Apples and oranges: 20-30 miles is not a 20-30 minute walk or bike ride. I would suppose that even in those Appalachian communities (not dissimilar to the one I grew up in, except for the folklore) there are a good number of the students who do live within easy walk/bike distance, since the consolidated schools tends to be in the larger towns.

And unfortunately, my walk was pretty flat, though we did have a good bit of snow and sub-zero temperatures.


From a training perspective, these incentives don't help kids change their lifestyle, because the reward is not immediate and the reward is based on a chance that they might happen to get some money. This is why I don't like the incentive programs at school. They are set up so that kids get money at the end of 3-6 month period. This is not an immediate payoff, and the kids loose interest over that time frame. It would best be accomplished if test were conducted on computers, and say $.01 was given for each correct answer. Over a test, you would end up with a bigger amount of money, and over test after test, the money would build up, until that money would be a substantial amount. Then you would see test scores rise. You have to change the motivation for good grades, and to do that, you need the correct tools. Ahh, if more people followed the work of Skinner.


I'd rather see suburban communities go to much, much smaller school sizes, combine grades and have neighborhood schools. The trend, at least in suburban ATL, has been to create schools with 5 or 6 classes of each grade, in one big campus. Thus, they are net farther away from more kids and there are simply fewer opportunities to ride bikes.

The suburban neighborhoods don't help, because it seems so many neighborhoods are now huge groups of cul-de-sacs with only one entry/ exit onto a major road where riding for kids, even on a side walk, would really be dangerous.

I'm a cyclist, and I have tried to ride on the suburban roads near my house, and it is impossible to avoid major roads/ intersections.