Experimental Evidence from Chile
The objective of the experiment was to help students choose study options to obtain positive returns from their professional career paths. Information was provided to students finishing high school in low-income areas so they could choose post-secondary studies and the type of institution where to study - from among universities, professional institutes, and vocational technical institutes.
In Chile, universities with merit-based selection processes are generally reserved for students with better performance on admission tests and higher school grades. This leaves little room for low-income students, usually with worse school performance, to gain access to elite universities with selection processes.
In this context, the experiment provided information for students who do not enter one of the universities with selective processes in order to give them sufficient knowledge of other options for post-secondary studies in institutions without selection processes.
To conduct the experiment, the researchers worked with the Chilean NGO, Por una Carrera, and attended 300 career guidance fairs in the city of Santiago and the surrounding area from July to November 2013. In the fairs, around 10,000 emails were collected from students in the last year of high school, together with their preferences for studies and institutions. They contacted the students and invited them to participate in the experiment by means of an electronic survey. The initial survey was answered by 1,727 students over a two-week period.
Later, in November and December 2013, the students were randomly divided into a control group and an information treatment group. All students (control and treatment) received an e-mail inviting them to look at the website www.eligecarrera.cl which contains general information on course types, loans and scholarships available in different institutions.
Control group: 556 students received only the link to the website www.eligecarrera.cl with general information.
Information treatment group with respect to financial assistance: 553 students received general information along with personalized information on other specific sources of funding according to their preferences and contexts. This included information on scholarships or loans offered by private schools, firms, municipalities, and by specific programs in the course. With this type of information, the aim was to reduce the cost of finding this information which is very dispersed. This information was compiled with the help of volunteers from the Pontifical Catholic University of Chile.
Information treatment group with respect to returns on education: 559 students received information from the MiFuturo database (www.MiFuturo.cl), which reports on monthly earnings and employability rates for recent graduates from certain university programs. Students received information on the returns from the courses of their choice, and, in addition, about schools in four alternative categories that offered those courses. This included up to three school options in institutions with selective processes and three options in others without such processes (but were not highlighted as such). Students also received information about the quality of each of the options presented.
The idea was to motivate students to try others, if they did not obtain entry into their first option for university.
Students of last grade of secondary in low-income areas in Santiago, Chile.
The aim of the experiment is that the students choose careers with positive returns.
Results
The researchers found that:
1) There is no evidence that providing personalized information about financial assistance and returns changes high school students’ decisions about applying or enrolling in selective institutions.
2) Students who received information on the return of their courses chose schools suggested in the e-mail they received and were also encouraged to apply to suggested schools of higher quality than their baseline preferences. In the treatment group on returns, the fact of being exposed to this information raised by this percentage by 6 points.
3) The results, although not statistically significant, show that the students in treatment groups tended to change their decision to go to a for-profit university and choose a professional technical institution.
For reasons of times between the experiment and the moment when the students had to make a decision on their higher education, as well as statistical issues due to the sample size, the experiment did not show significant changes in student behavior. However, it did show that such low-cost and scalable interventions could be used by the government to help high school students in marginalized areas understand costs and benefits in relation to attending schools with fewer selection processes.