The Power of Small Amounts of New Information

I noted an interesting exercise in how new information, even in small doses, can change peoples’ behavior, during my recent visit to the Southern Economic Association annual meeting.

Along with presenting my own work, I was able to attend several other sessions to learn about other new economic research. My favorite of these was about “Information, Expectations and Education Choices.”

One of the papers, written by Zachary Bleemer and Basit Zafar and recently released as Federal Reserve Bank of New York staff report, was particularly interesting. They found that survey respondents, on average, underestimate the earnings differences between college graduates and non-college graduates, and overestimate the net price of public college. Either of those might cause people to be less likely to expect to send their children to college: The average benefit is bigger than the average person thinks, and the average cost is smaller.

The authors tested how much this mistaken belief matters by correcting it — for some people. They randomly sorted respondents into one of three groups. After being asked about their beliefs about earnings differences and college cost, one group was told the true earnings ratio for college and non-college workers. A second group was told the average net cost of public college. The third group was the control group, who was given no new information. Later in the survey, all respondents were asked again about the probability of sending their child to college, and they were asked again as part of a follow-up survey two months later.

If people update their beliefs about the net gains to college, we would expect the first two groups to say they are more likely to send their child to college in follow-up surveys. If people’s college choice isn’t that sensitive to returns (or if people aren’t taking the survey that seriously), the researchers might find no change. So what happened?

Well, some of both. The first group – who found out that the wage ratio was bigger than most of them had thought – did report a higher probability of sending their child to college, especially the people in that group with lower income or education. About 80 percent said the first time they expected their child to attend college, and about 85 percent did when asked again, now armed with the new information. The effect held on the follow-up survey after two months, so something about the new information stuck with people. The people who were told about costs (as well as the control group) didn’t change their minds.

It is not clear why the two groups responded differently, but I think it’s amazing they found anything at all. The difference may seem small, but keep in mind most people in the survey were planning to send their kids to college already. One line of information reduced the number of kids not being considered as future college students by one-quarter.

Of course, there is a huge difference between saying you plan to send your child to college and actually doing so: The true college enrollment rates in the United States are closer to 60 percent than 80, so we know right away there is a difference between the expectation from the survey and reality.

On the other hand, this was just about the tiniest, cheapest treatment imaginable, and it changed how some people were thinking about as important a decision as sending their kids to college even months after the fact. Other, more extensive work in promoting information about colleges has also been promising in terms of changing enrollment behavior. This study is another piece of evidence that even very small amounts of information can help people think about very big decisions in a new light.