Association for Public Policy Analysis & Management

Choosing a college major is one of the most influential, long-lasting, and complicated decisions a student makes. In addition to large differential returns in earnings for college majors, a college major can affect a student's career choice, geographic location, and lifestyle. Despite potentially large long-term effects of a major, small changes to when majors are experienced during a college career may impact students’ choice of major. In particular, if students are uncertain about the value of majors, have limited information about certain majors, or recall and evaluate experiences differently depending on when they occur, then minor changes to the timing of students' college experiences may influence their choice of major and therefore the direction of their lives.

We test whether students' college major choices are influenced by changes in the timing of when students take courses by exploiting random variation in student schedules at the United States Military Academy (USMA) at West Point. Specifically, we use the random assignment of USMA students to certain required courses during or after the semester in which they must select a college major. We find that students assigned to a course in the semester when initial major decisions are made (the first semester of sophomore year) are 109 percent (2.6 percentage points) more likely to choose a major that corresponds to the course than students who are assigned the same course in the following semester (the second semester of sophomore year). This result is robust to a number of specifications, including those that use faculty fixed effects and fixed effects for the complete roster of scheduled sophomore courses.

We explore several potential mechanisms for this result, including a rational response to new information, exposure effects, and models of salience and availability. Although our findings are difficult to reconcile fully with rational responses to new information or exposure effects, they are consistent with a simple framework we develop that incorporates salience and availability.

We also find that a large portion of the effect of semester order persists through graduation. Students are 39 percent (1.4 percentage points) more likely to graduate in a subject they were assigned during the first semester of their sophomore year (the semester of initial major choice) than a subject they were assigned in the second semester of their sophomore year. This occurs despite there being no clear penalty or significant cost to switching majors during sophomore year; students in the sample do not start major-related courses until their junior year and only need two signatures to switch majors. These results are again difficult to reconcile fully with standard economic models, and they are more consistent with status quo bias.

Taken together, our findings show that exposing students to a subject in the semester they choose a major increases the probability that they initially choose and eventually graduate in a major related to that subject. Generally, these results suggest that small differences in the timing and order of experiences can meaningfully influence important decisions.