High school graduation rates among children of same-sex households

This is a really interesting study into the effect of gay parenting on children because the sample is not one of convenience but comes straight off the Canadian Census returns. You can read the full paper here.

Let’s look at the model that was built.


allen-table6I’ll let Allen explain the results.

Table 5 reports on three logit regressions, where the dependent variable equals 1 if the child has graduated from high school. All of the regressions in this table control for whether or not the family moved with the past year. Table 8 in the appendix reports on another three logit regressions with the same dependent variable and the same right hand side variables, except for the variable used to control for family mobility—it uses the mobility measure ‘‘did child move within past 5 years.’’ There is no qualitative difference in the estimates when using the different mobility controls.

Table 5 only reports the logit coefficient, its standard error, the odds ratio, and marginal effects, for the household type variables in order to keep the tables a reasonable size. The log of the odds ratio is the logit, which is a linear combination of the parameters and exogenous variables. The odds ratio is found by taking the exponential of both sides of the logit equation. The odds ratio has the nice property of an easy interpretation. The marginal effect equals (qy/qx), where y is the graduation rate and x is one of the right hand side variables. The different columns result from different types of controls. Column (1) includes controls for child characteristics, and these include: province, visible minority, disabled, mobility, urban, age, family size, family income, female, and same race. Column (2) adds the parental education controls: did the mother/father graduate from high school, and did the mother/father have a graduate degree. Column (3) adds the parental marital status variables found in Table 2. Before examining the results of Tables 5, some comment on the unreported results is warranted. Among the child characteristics, being disabled or having moved in the recent past reduces the odds of graduation (on average, to about 50 and 75 % respectively). Living in an urban area, being female, and having all family members the same race raises the odds of graduation (on average, by 30, 60, and 35 % respectively). Parental education matters a great deal: if the parents have graduated from high school, the child is almost twice as likely to do so. Finally, marital history has the expected effects. Any marital disruption reduces the odds of a child graduating from high school. For any given household type variable the odds ratio and level of statistical significance is generally robust to the different specifications. That is, changing the controls does not change the parameter estimates for the association of graduation rates and household type in large ways.

Table 5 shows the associations between family type and child graduation. In all cases, the odds of a child with gay or lesbian parents completing high school are lower, by a considerable margin, compared to children of married opposite sex parents. For gay and lesbian households, adding the parental education controls to the base controls lowers the odds of a child graduating for same-sex families. This is because gay and lesbian homes are characterized by high levels of parent education which contributes to child graduation, and so conditional on this the odds of a child graduating are even lower. When all controls are used, including those for parental marital status, the conditional graduation rate odds ratios are reasonably similar between the two types of same-sex couples: 0.69 for gays and 0.60 for lesbians using the 1 year mobility measure. The difference between the two point estimates for gay and lesbian parents in column (3) is not significant. To put this in another context, the marginal effect on the probability of graduating for children of samesex homes is a reduction of approximately 6–9 % points. The point estimates for gay households are always statistically significant at the 5 % level, but the estimate for lesbian households in column (3) is not.39

Table 6 repeats the column (2) and (3) regressions of Table 5, but this time separating girls and boys. This table shows that the particular gender mix of a same sex household has a dramatic difference in the association with child graduation. Consider the case of girls first. Regardless of the controls and whether or not girls are currently living in a gay or lesbian household, the odds of graduating from high school are considerably lower than any other household type. Indeed, girls living in gay households are only 15 % as likely to graduate compared to girls from opposite sex married homes. In all cases for girls the estimates are measured with precision. The point estimates for boys are considerably different. Looking at equation (4) in Table 6, boys in lesbian homes are 76 % as likely to graduate, while boys in gay homes are 61 % more likely to graduate compared to boys in opposite sex married homes. However, none of these estimates are statistically significant. The results from Table 5 mask this gender difference, and the significant effect found in Table 5 column (3) for gay households is clearly being driven by the strong daughter effect.

The different results for the household gender mix are fascinating, especially since this difference is not found in single parent households. Table 6 shows that boys do better than girls in single parent homes, but the difference is not nearly as pronounced as in same sex households. Looking at the unconditional graduation rates (with standard errors in parentheses) for gay households, sons achieve 0.72 (0.074), while daughters achieve 0.43 (0.090). For lesbian households, son’s graduation rates are 0.48 (0.060), and daughter’s have 0.55 (0.055). Based simply on these unconditional measures, sons do better with fathers, and daughters do better with mothers. At this state, such a result is an interesting empirical finding, and one worthy of further investigation. On the one hand, it seems this result is inconsistent with any type of discrimination theory for the lower graduation rates among children of same-sex households. Or, a discrimination theory would have to be modified to include the household gender mix. Within the child development literature and pop culture, there is the belief that mothers and fathers provide different parenting inputs  that are not perfectly substitutable. These results would be consistent with this notion, but further research is necessary to show any causality.

It’s worth pointing out that Potter did some similar research and also looked into family transitions. He found that once you included those, same-sex parenting became less significant in explaining differing educational outcomes of children. That said, Allen’s research is hugely significant because it has an enormous same-sex parenting sample (over 8,000 compared to 248 for Regenerus) and also the nature of that sample avoids many of the previous problems for research in this area.

Currently, the 2006 Canada census has several strengths compared to any other data set. First, it uses information from a country where same-sex couples have enjoyed all taxation and government benefits since 1997, and legal same-sex marriage since 2005. As Biblarz and Savci note, such legalization reduces the stress and stigma of homosexuality, and encourages honest participation in census questions. Second, not only does the census provide a large random sample, but married and common law same-sex couples and their children are self identified. This is an important advantage over the US census. Third, because the child and parent records are linked together, the marital status and educational levels of the parents can be controlled for when analyzing child performance. Finally, because of the relatively large sample size, there is enough power to not only separate gay from lesbian households, but also enough to examine the gender mix of same-sex households.

Simply put, it’s hard to argue with this research without impugning the whole Canadian Census programme. From that census we have a robust piece of research suggesting that children of same-sex parents are c. 10% less likely to graduate from high school than children of married couples.

Let the debate begin…