Thesis Defense - Evelina Linnros

Thesis defence

Date: Friday 14 June 2024

Time: 09.00 – 11.00

Location: Hörsal 9, Hus D, Södra Husen, Frescati

Essays on Fertility and Health

Link to thesis

Opponent: Maarten Lindeboom, Vrije Universiteit Amsterdam.

Supervisor: Peter Nilsson, Anna Tompsett.

Infertility Risk and Child Marriage

The high infertility rates observed in some developing countries may have broad societal impacts, for example by people marrying and having children at a young age to increase their chances of reaching their fertility target. I study the link between infertility risk and marriage timing using data from Madagascar. Specifically, I focus on how infertility risk affects the probability of child marriage, a practice associated with adverse outcomes for young brides and their children. I use spatial variation in exposure to the parasite schistosomiasis. The empirical strategy compares two strains of this parasite, similar in their transmission mechanisms and health impacts, except that one of the two strains causes infertility. In my data, exposure to this strain increases the probability that a woman is infertile by 40%. I find that exposure to the infertility-causing strain increases the probability of child marriage and early fertility by 22%.

Maternal Health and Labor Market Outcomes

We study how severe injuries related to childbirth affect mothers' labor market outcomes. 1 in 20 first-time mothers who have a vaginal delivery suffer a severe birth tear, which can have long-lasting adverse impacts on their health and quality of life. Using a difference-in-differences design with a matched control group, we find that severe birth tears lead to a 6% higher earnings loss in the first five years after childbirth compared to the control group. The effect is larger for mothers from low SES backgrounds, while high SES mothers are found to seek more healthcare following their injury.

Alcohol Availability, Prenatal Conditions and Midlife Mental Health

 We examine the long-term mental health effects of an 8.5-month policy experiment that led to a sharp and unexpected increase in alcohol availability, focusing on individuals exposed to the policy in utero. We use administrative healthcare and drug prescription records to identify individuals who have received treatment for a mental health disorder. Prenatal exposure to the policy had a large and persistent effect on mental health: the exposed cohort is 16% more likely to be treated for a mental disorder in midlife. The effect is largest for those exposed from the second trimester and is only partly explained by the lower earnings observed among exposed individuals.

The Value of Monitoring for Disaster Prevention: The Desert Locust

Monitoring systems are meant to detect early signs of potentially disastrous outbreaks of diseases and pests, in time for preventative action. These monitoring systems are costly, and identifying their economic value requires estimating damages from outbreaks in empirical settings where monitoring is neither uniform nor exogenous. We estimate the value of monitoring systems for desert locusts, known to devour entire agricultural fields. We leverage conflict and weather events in breeding areas to detect the effects of monitoring interruptions on swarm outbreaks. We then reconstruct the spatial patterns of locust migrations to propagate these effects on swarm outbreaks beyond breeding areas. Finally, we show that in-utero exposure to a swarm increases the probability of stunting by 16%. These estimates allow us to quantify the effects of a change in monitoring efforts on subsequent locust swarms and on human health.