In these papers I discuss the causal estimation of the effects of fertility and fertility planning developments on mother and child outcomes. A number of concerns are raised with existing identification techniques, and alternative methodologies to consistently estimate the effect of interest are proposed. These concerns and new techniques are illustrated using microdata on slightly more than 43,000,000 births ocurring between 1972 and 2013. In the first substantive chapter (written with Sonia Bhalotra), we discuss the validity of the use of twin births in fertility research. We demonstrate that twin births are not random. Successfully taking twins to term depends upon positive maternal health behaviours and investments in the periods preceding birth. We show that this is of considerable concern for estimation techniques which rely on twin births being (conditionally) randomly assigned to identify causal effects. To illustrate, we consider the estimation of the child quantity-quality (QQ) trade-off, and show that existing instrumental variable estimates are inconsistent in the contexts examined. Upon partially correcting for the fact that twin births are not random, a statistically significant QQ trade-off begins to emerge. We close by examining a number of partial identification techniques to bound the true effect of fertility on child outcomes. In the second substantive chapter, I examine the effect of fertility control policies on the fertility decisions and outcomes of women. I consider the case of the emergency contraceptive pill in Chile. The staggered arrival of this technology to Chile over the last decade has resulted in the availability of the first safe and legal post-coital birth control policies. In a context of high teenage pregnancy rates, difference-in-difference (DD) style estimates suggest that this policy has accounted for reductions in short-term teen childbearing by as much as 7%, an effect similar to the arrival of abortion in the USA. This policy is also shown to reduce fetal deaths reported in early gestation with no similar reduction in late gestation: suggestive evidence that an alternative fertility control policy may reduce costly and dangerous illegal abortions. Finally, I turn to the use of DD estimators as a policy-analysis tool. I discuss how such estimators perform in the case of reforms which may not be sharply demarcated to treatment and control clusters, but rather subject to local spillovers or externalities. I propose an extension of the typical DD estimator: a spillover-robust DD estimator. This methodology is applied to estimate the effect of two localised fertility control reforms in Mexico and Chile, where women close to treatment clusters who were not themselves subject to the reform may nonetheless travel to access treatment.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:711911 |
| Date | January 2015 |
| Creators | Clarke, Damian |
| Contributors | Stevens, Margaret Jane |
| Publisher | University of Oxford |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://ora.ox.ac.uk/objects/uuid:94016283-a3dd-4b6a-8427-373b49a491be |
Page generated in 0.0969 seconds