Tuition increases at community colleges might lead to more student debt, more federal-aid disbursements, and ... more high-school kids smoking and flirting in the 7-Eleven parking lot?
That is the implication of a paper published online this month by the Economics of Education Review. (An earlier, ungated version is available here.) Benjamin W. Cowan, an assistant professor of economics at Washington State University, found that 17-year-olds are more likely to smoke tobacco or marijuana or to have sex if they live in states where community colleges are relatively expensive. The effects are especially strong among students whose high-school grades and other characteristics suggest that they would fit two-year (but not four-year) colleges.
Mr. Cowan looked at data from nearly 5,000 people who have participated in the National Longitudinal Survey of Youth 1997, which has followed Americans who were between 12 and 16 years old in 1996. He compared the participants’ survey responses when they were 17 years old (meaning that these particular surveys occurred between 1997 and 2001) to the average community-college tuitions in their home states at the time when they were surveyed.
In states where community colleges were more expensive, he found, 17-year-olds were less likely to say that they expected to attend college, and they were more likely to report smoking and sexual activity. (They were no more likely to report binge-drinking, however.) These students were willing to take more risks, Mr. Cowan surmises, because they had lower expectations of enrolling in college and therefore felt that they had less to lose.
The effects were not enormous, but they were statistically significant. Mr. Cowan estimates that a $1,000 increase in community-college tuition is associated with 0.89 additional monthly days of smoking cigarettes, an additional 0.4 monthly days of smoking marijuana, and an additional 0.3 self-reported sexual partners per year.
A study like Mr. Cowan’s needs to overcome two objections. First, high-school students are notorious for misestimating their odds of attending college—so why should their perceptions be meaningful in this case?
In the paper, Mr. Cowan notes that even though the survey participants generally overestimated their likelihood of being enrolled in college the following year, their instincts were not totally off the mark. A one-percentage-point increase in the students’ subjective expectations of enrolling in college was associated with a half-percentage-point increase in their actual college-going behavior. Students in states with higher tuitions correctly saw that they were less likely to enroll than were those in less-expensive states.
The second concern is that the state-by-state variations in adolescent risk-taking might actually be driven by some factor other than community-college tuition. Maybe the states with lower tuitions just happen to be states with more-effective antismoking campaigns. Mr. Cowan attempted to deal with that problem by controlling for states’ median incomes, unemployment rates, spending on K-12 education, and cigarette and beer taxes.
Public-health advocates might look at this study and start to fight for lower college tuitions. But Mr. Cowan stops short of that argument. He concludes his paper by noting that college subsidies are expensive, and “other policies that influence teenagers’ college expectations may provide less expensive ways of achieving a reduction in risky behavior.”
Selective Colleges and Healthy Habits
Another forthcoming Economics of Education Review paper looks at whether attending a more-selective college is more likely to lead to a life of health and sobriety than attending a less-selective college.
Jason M. Fletcher, an assistant professor of public health at Yale University and a scholar at the Columbia University Population Research Center, and David E. Frisvold, an assistant professor of economics at Emory University, looked at nearly 4,000 people who have participated in the National Longitudinal Study of Adolescent Health. They found that people who attended selective colleges (defined as those whose median SAT scores were in the top quartile) were less likely to smoke, to be obese, and to eat fast food than were those who attended less-selective colleges, both while they were in college and when they were in their mid-20s. (To those who might object that students at selective colleges are more likely to have been socialized into healthy habits as children, Mr. Fletcher and Mr. Frisvold controlled for a long list of individual variables, including smoking and obesity during high school. They believe that they have isolated the effects of the colleges themselves.)
As in Mr. Cowan’s study, binge drinking was an exception here. Selective colleges seem, if anything, to be associated with more drinking.
Mr. Fletcher and Mr. Frisvold are most interested in discovering the mechanisms through which education might influence health behaviors. Their new paper did not find any clear patterns, but they suggest that “health literacy and health knowledge as well as peer effects during college” are the most likely channels.
Michael Grossman, a professor of economics at the Graduate Center of the City University of New York who has been studying the economics of health for nearly four decades, offers praise for both papers. In an e-mail message to The Chronicle, he says that both papers advance the understanding of how college shapes behavior. “Returns to investments in education extend far beyond higher earnings,” he says.
One selective college has been at this kind of study for a long time. In 1912, the officers of the Princeton University Class of 1906 published a survey of the class’s indulgences several years after graduation. Most claimed to be more or less temperate. “College, contrary to the belief and desire of the critic fanatics who have pictured it the black hole of vice, plays a comparatively small part in the development of either drinking or smoking,” the officers wrote.
