Our empirical strategy exploits the randomized variation that occurred as a result of the Vietnam draft lottery and compares children of fathers who were eligible for the draft with children of fathers who were not. Given the random nature of the lottery, draft‐ and non‐draft‐eligible fathers were comparable except that draft‐eligible fathers were called to report for potential induction into the military. We exploit this randomized variation and the information on both respondents (i.e., children) and parents provided in the National Longitudinal Survey of Youth 1997 to estimate the effect of draft eligibility on children’s risky behaviors, as defined by substance use and delinquent acts.
Our findings indicate that paternal draft eligibility had the following effects on children’s risky behaviors. First, we show that draft eligibility increased the probability to have consumed marijuana by age 18, decreased the age of marijuana initiation, and increased measures of marijuana consumption. Second, draft eligibility reduced the age of cigarette initiation. Third, draft eligibility increased measures of hard drugs consumption (though these effects are less precisely estimated). Fourth, draft eligibility increased the probability of engaging in delinquent behaviors.
There are many potential ways in which fathers’ draft eligibility could increase children’s propensity to engage in risky behaviors. Draft eligibility could lead to military service, and military service could negatively affect veterans by increasing their opioid use; increasing the likelihood of psychiatric conditions, such as PTSD; increasing their propensity to commit violent crimes and become incarcerated; increasing the incidences of domestic violence against the partner and children; or by lowering the socioeconomic status of the household by precluding soldiers from labor market experience, which ultimately decreases wages.
Draft eligibility could also lead to draft avoidance behaviors. Engaging in delinquent activity and crime could lead to draft avoidance, as having a criminal record would lead to failing the moral evaluation of the preinduction exam one must pass to be drafted. Additionally, refusing to serve after receiving a low lottery number could lead to convictions and prison sentences according to the draft law. These direct negative effects of the draft on eligible fathers could, in turn, indirectly affect their children, making them more likely to engage in risky behaviors. On the other hand, the draft could also have positive effects on those fathers affected by it, which could, in turn, affect their children positively. In particular, military service could affect veterans positively by providing training and imparting discipline or by allowing former soldiers to benefit from the GI Bill, which ultimately increases education. Additionally, enrolling in universities to avoid the draft—also referred to as educational deferment—would increase parental college attendance and college retention among those avoiding service, thereby positively affecting the fathers’ labor market outcomes and wages.
While we do not have information on many of these potential direct effects of the lottery draft on the generation of fathers affected by it (such as drug consumption, suffering of PTSD, or incarceration records), we find that parenting styles, attitudes toward the children, and the environments where children were raised differ by fathers’ draft eligibility, potentially driving children of draft‐eligible fathers to engage more in risky behaviors. More specifically, we first find strong evidence that paternal draft eligibility negatively affects potential determinants of father‐children relationships, such as parenting styles and attitudes from the father toward the children. Draft eligibility affects parenting styles of both parents by increasing the likelihood that they are unresponsive to the needs of the children. Additionally, draft eligibility affects the attitude fathers have toward their children, as fathers are less likely to help them and more likely to cancel plans on them (as reported by the children).
Second, we find that children of draft‐eligible fathers grow up in environments more conducive to engaging in risky behavior, as evidenced by, for example, interviewees being less likely to report feeling safe in the residence or neighborhood of draft‐eligible fathers and children being exposed to school peers that are more likely to engage in risky health behaviors (e.g., smoking cigarettes, getting drunk, using drugs, and having sex). These findings are consistent with previous literature that indicates that sons of draft‐eligible fathers reside in lower‐mobility counties and lower‐income zip codes.
Third, we find that children’s scores on aptitude tests do not seem to differ by fathers’ draft eligibility. Additionally, paternal draft eligibility does not seem to affect predetermined maternal characteristics through assortative mating, as measured by the probability that the respondent’s mother was living with her biological parents by age 14 and the level of education of the maternal grandparents. Thus, neither lower school performance nor differences in mothers’ characteristics seem to be behind the higher probability of children engaging in risky behaviors. Finally, we also present evidence that paternal draft eligibility affects maternal health negatively. While the data do not allow us to determine whether the difference in maternal health between draft‐ and non‐draft‐eligible men is driven by selection (i.e., draft‐eligible men marrying unhealthier women) or by a direct effect of draft eligibility (i.e., the mental health status of the husband affecting the mental and physical health status of the wife), the evidence against paternal draft eligibility affecting predetermined maternal characteristics seems to point to a direct effect of draft eligibility.
Because of the changes that have occurred over time, it is important to discuss the extent to which our findings apply to today’s environment and military context. Most importantly, today’s military service is based on volunteering. Men who volunteer to serve could be different in several unobservable ways from men who serve because a lottery pushed them to. Additionally, the current system based on volunteering eliminates incentives to change behaviors to avoid serving. Thus the extrapolation of our results to the current environment of voluntary enlistments should be done with care. That said, under the current context, it is still relevant to understand the unintended consequences—whether through draft avoidance or military service—of a lottery draft system designed to increase the number of individuals available to serve during times of war, as a lottery system similar to the one applied during the Vietnam War is expected to be resumed in times of national emergency, as reported by the Selective Service System. This highlights the relevance of this study, not only from a historical perspective, but also for future reference.
NOTE
This research brief is based Monica Deza and Alvaro Mezza, “The Intergenerational Effects of the Vietnam Draft on Risky Behaviors,” NBER Working Paper no. 27830, September 2020, http://www.nber.org/papers/w27830.
