Recent research has suggested that the adverse consequences of teenage motherhood are short- lived, and that, after a few years, teenage mothers do better than similar women who do not give birth as teens, In effect, these studies find that teenage motherhood changes the timing of births rather than the number of births and that, for the type of woman who gives birth as a teen, early childbearing is preferable to later childbearing. The recent literature has relied on an instrumental variables technique that essentially compares women who miscarry (have spontaneous abortions) as teens with women who do not miscarry. This proposal points out that the existence and timing of (induced) abortion creates important biases in that approach that may lead to the conclusion that teenage motherhood is benign and proposes two alternate estimators. The first drops from the sample women who aborted and compares those who miscarry with those who gave birth. This estimator is biased away from the benign view. Using the estimators with and without abortions gives us bounds on the effects of teen motherhood. Preliminary estimates indicate that the bounds are sufficiently tight to be informative. The second new estimator is valid if the timing (but not the use) of abortion is unrelated to adult outcomes. In this case, the IV and restricted-sample estimates can be combined to give a consistent estimate of the effects of teen motherhood on adult outcomes. Preliminary estimates support the validity of the assumption. The study will apply these techniques to four waves of the National Survey of Family Growth to ask 1) What is the effect of giving birth as a teen on educational attainment (measured in years and by high school graduation), marital status, employment and hours worked, individual earnings and household income, income relative to a measure of needs and total fertility? 2) How do these effects vary over the life-cycle? and 3) How have these effects changed over time? The appropriate policies for addressing teen motherhood depending critically on its effects. If the recent literature is correct, then either teen motherhood should not be viewed as problematic or it must be addressed by changing the life expectations of potential teen mothers so that they have prospects in life that are sufficiently positive to be jeopardized by early childbirth. If the effects of teen motherhood are adverse even for the population at-risk, then education about the consequences is a potentially effective intervention. ? ? ?
| Ashcraft, Adam; Fernández-Val, Iván; Lang, Kevin (2013) The Consequences of Teenage Childbearing: Consistent Estimates When Abortion Makes Miscarriage Nonrandom. Econ J (London) 123:875-905 |
