In 2011, university and high school students in Chile undertook a massive protest against the government (known as the 'Chilean winter'). Thousands of students participated in these strikes, refusing to attend classes or even taking over school buildings, and demanded a new framework for education. Given the widespread nature of the strikes, and the fact that they would affect non-participating students as well (who wouldn't be able to attend a school that had been taken over by protestors), it is worth asking what the consequences of the strike were.
That is the question that this new article by Pablo Celhay, Emilio Depetris-Chauvin (both Pontificia Universidad Católica de Chile), and Cristina Riquelme (University of Maryland, College Park), published in the Journal of Development Economics (ungated earlier version here), seeks to answer. Specifically, they focus on teenage pregnancy as an outcome, defined as a birth to a 15-17-year-old woman, backdated based on gestational age of the baby on the day of birth.
Celhay et al. don't have individual-level data though, so they look at the relationship between the number of births, at the municipality level, and a measure of 'strike intensity', which measures how likely a female student living in a given municipality was exposed to the student strike. It is based on the proportion of schools in a municipality that were on strike, and the length of the strikes overall. In Chile at that time, births were a good measure of conceptions, because abortion was illegal.
Celhay et al. find that:
A municipality with an additional exposure of 10 percentage points, signifying a ten percentage point increase in the number of resident high school female students attending schools on strike, witnessed a monthly rise in conceptions during the strike period ranging from 10% to 11%. For a more straightforward interpretation, consider that a municipality with an average proportion of students on strike (26% according to the combined measure) experienced a 2.7% increase in teenage pregnancies during the strike period.
I guess that tells us what some students were spending at least some of their spare time doing during the strike period? Importantly, this effect was concentrated among women aged 15-17 years, with no statistically significant impact on other age groups. That provides some confidence that the results are specific to school-aged women, and not a general increase in pregnancy and births across the whole population.
However, perhaps more importantly, Celhay et al. then find a more enduring effect on human capital development. Specifically:
Before the strike, schools that eventually experienced strikes were similar to non-striking schools regarding dropout rates and college admission test take-up. However, a significant increase in dropout rates and a decrease in college admission test take-up is observed in the year when the strike occurred. The effects are similar if we disaggregate outcomes by gender. In particular, the schools that took up strikes experienced an increase of 0.7 percentage points in their dropout rate. This represents a 20% increase in dropout rates in comparison to the average level of dropouts in the year 2010 (3.4%)...
The results show a drop of approximately 20% in the number of students taking the test to be admitted to college during the strike year.. Furthermore, our study reveals that it takes approximately two to three years for dropout rates and college admission test take-up to return to pre-strike levels. This indicates a gradual recovery process after the disruption caused by the strike, as the educational system and student engagement stabilize over time.
I guess one way to interpret those results is that, when students don't attend school, they tend to drop out and don't sit the college admission test. That seems somewhat obvious. What is less obvious is that this impact remains for more than two years after the strikes subsided, and that should be a worry. It is also somewhat consistent with the long recovery of the education system, and student engagement, following the pandemic.
Celhay et al. interpret their results as showing:
...the potential benefits of policy interventions such as sexual education and counseling within schools, as well as initiatives that promote access to contraception among teenagers.
I think that interpretation oversteps, as they didn't actually look at the impact of sex education or counselling, or initiatives that promote contraception. It would be interesting if there was variation in those activities between schools, because then they could have looked at the impacts, but didn't. Nevertheless, I do agree with Celhay et al. that:
...we can interpret the observed effects as primarily related to reduced time spent under adult supervision.
No comments:
Post a Comment