Children skip school when temperatures rise - and risk lagging behind in academic performance

When temperatures climb above 26 degrees Celsius, high school students in Costa Rica are more likely to not attend lectures. The warmer it gets, the higher is the absentee rate.
‘These are important findings since non-attendance is associated with worse academic performance’, says Laura Villalobos-Fiatt a doctoral candidate in economics at the School of Business, Economics and Law, University of Gothenburg.

What made you interested in the topic school attendance and weather patterns in the first place?
There are three things that motivated me to study the relationship between weather and school attendance. Firstly, I noticed a striking pattern in Costa Rica, in which schools with the lowest levels of academic achievement tend to be located in warmer and more humid areas. Because several factors can explain this pattern, my goal was to isolate the effect of weather on education outcomes, to see whether there is a causal relationship that can explain this pattern.

Secondly, recent evidence shows that higher temperatures cause fatigue, cognitive impairment, and lower human productivity. This is easy to relate to, as weather conditions certainly affect my behavior and my mood. I wanted to take this question to the context of schools in a tropical middle-income country, where infrastructure often is not adequate to mute the effects of weather.

Thirdly, in the debate about the effects of climate change, people often talk about the damage to infrastructure caused by the sea level rising, or the effects on agriculture. But missing in the discussion is the direct effects that weather can have on humans. I wanted to start a research agenda in this direction.

What Makes Costa Rica an interesting case to study?
Whereas primary school coverage is quite good in Costa Rica, there are major challenges for secondary school. The number of people who complete high school is very low here. Only half the population in their early twenties finished high school. This is a low figure compared to the rest of Latin America (53%) , and even lower compared to OECD countries (75%). Also, at least half the schools report inadequate infrastructure. In 2011, one third of all public high schools had at least 50% of their classrooms in poor or mediocre condition. These results suggest that we need to look more closely at just how infrastructure failure could be contributing to weather-linked non-attendance.

What is the most important result in your thesis, and why?
I show that when it is warmer than usual and when it rains, students are more likely to skip lectures. I estimate that on average one additional student will be absent in any given lecture at 34 Celsius degrees compared to a lecture at 20 Celsius degrees. I also show that the more lectures a student skips, the worse their academic performance. This is important because that weather conditions that are an external factor, out of the control of the students, and yet they can affect their human capital accumulation.

What impact does your work have on research and society?
This is a contribution to the scientific evidence showing that warmer temperatures have a direct effect on human capital. This is important in a context of climate change, where we expect higher temperatures and more erratic weather in general. This type of research shows that we should worry not only about sea level rising and damage to infrastructure and crops, but also to human productivity, and in particular human capital formation. By learning more about this relationship, we are in a better position to lead adaptation strategies.

What is new in your study?
To the best of my knowledge, this is the first study that analyzes attendance patterns at the individual level lecture level throughout an entire schooling year. For example, I find that the average absenteeism rate is around 15%. This means that in a group of 30 students, around four will be absent in any given lecture on average. I think this is new information for decision makers in my country, as these statistics are not usually analyzed at the national level. By collecting and analyzing this information, and combining it with information from the local weather stations, I show that weather conditions within a short time window of the lecture can influence whether or not the students decide to go to class.

Did anything in your results surprise you?
More than surprised, I was excited to see that with completely new data in a new context, my findings are very much in line with other studies that find similar patterns in which weather affects workers’ productivity in manufacturing sectors, and human mortality for example. I think it is becoming a well-established and robust result that affects human capital.

Do you suggest any policy recommendations based on your findings?
From the perspective of climate change, there is room for adaptation. The question is to what extent this is possible, how costly this will be, and what is the best way of doing it.

The first policy recommendation is to document and acknowledge this relationship between weather and school attendance. Then, we must explore policies that could mute the negative effects of weather on school outcomes. For example, by investing in air conditioning, or simply by a better design of class schedules. For instance, it seems that it is a bad idea to schedule math classes for around noon, when temperatures are at their maximum. What is the better way of addressing this concern is still an open question.

Can the study be replicated elsewhere and are there any plans to do so?
I would like to see if I can replicate these results in all tropical EfD countries, and have applied for funding to do this research. It would also be valuable to look into what policies can be successful in offsetting these weather-related school attendance trends. The feasibility will depend on how detailed the weather and attendance data are for the regions in question, as these are the main inputs for such an analysis. I think the question is important and deserves to be scaled up and studied in more detail.

Laura Villalobos-Fiatt defended her thesis "Essays on forest conservation policies, weather and school attendance", 24 May 2017 at the School of Business, Economics and Law, University of Gothenburg. Laura Villalobos-Fiatt has been part of the Sida funded doctoral programme in Environmental Economics at the Department of Economics.

Link to thesis: http://hdl.handle.net/2077/52254