Record: Jeremy E. Uecker and Jonathan P. Hill, “Religious Schools, Home Schools, and the Timing of First Marriage and First Birth” in Review of Religious Research 56, no. 2 (June 2014): 189-218. [Abstract Here]
Summary: Uecker, a sociology professor at Baylor University, and Hill, a sociology professor at Calvin College, are both familiar names to readers of these reviews. In a 2008 article Uecker found (among other things) that there was no difference in levels of adult religious commitment between graduates of public or home schools. Parent religiosity, not school type, made all the difference. In a 2013 article Hill found that homeschooled young adults were less likely to engage in volunteer activities than demographically equivalent graduates of public schools. Both of these articles had drawn from the National Study of Youth and Religion (NSYR), a remarkably ambitious project that has borne great fruit in understanding the religious and political lives of young adults in the United States.
In the present article Uecker and Hill turn to the other significant recent data set that, like the NSYR, uses random sampling to provide us with generalizable data about homeschooled young adults in the population. This is the Cardus Education Survey, whose results have also been frequently mentioned on this blog. For this article Uecker and Hill wanted to know the degree to which attendance at various kinds of private schools or homeschooling impacted the age of marriage and of first childbirth among young adults. The NYSR did not follow its subjects far enough in their life course to answer this question. The Cardus Survey, which conducted in-depth interviews with 1,496 randomly selected young adults age 24-39 (including an oversample of private and homeschoolers), gives them the data they need.
Uecker and Hill begin by laying out several hypotheses they will test. They expected that Evangelical, Catholic, and homeschoolers would marry earlier and have a first child earlier than public school graduates, and that this would be explained in the data by levels of religiosity of parents and of the subjects themselves, and also by their more conventional beliefs about cohabitation and gender roles.
What they found, however, surprised them quite a bit. Graduates of Catholic schools ended up looking more like graduates of nonreligious (usually elite) private schools. Catholic schoolers married later than public schoolers and had a first child later as well. By their 30s the overall rates of marriage and childbearing among Catholic graduates were higher than among public schoolers, but it took them a long time to get there. Uecker and Hill speculate that their data may confirm the observation frequently made that Catholic schools now function largely as elite prep schools and that graduates of Catholic schools are marked more by their social class than by their religion’s pro-natalist Theology when it comes to marriage and childbearing patterns.
Evangelical private schools provided the authors with their most significant findings. While Catholic school graduates were a little different from public schoolers, Evangelical protestant school graduates were a lot different. They were much less likely than public schoolers to be married or to have children as teenagers, but they were much, much more likely to marry in their early twenties and to have a child in their mid to late 20s than all other groups. When they ran their controls Uecker and Hill were surprised to find that level of parental or subject religiosity did not really account for much of this difference, nor did beliefs about gender roles. There seemed to be something about attending a private Protestant school that made graduates more likely to get married young and start having children soon thereafter. Uecker and Hill speculate that perhaps the school itself may be something of a marriage market or that perhaps the school reflects a larger culture of close networks that facilitate matching and reinforce the value of marriage and childbearing.
The homeschooling finding was interesting only in that it was not very interesting. Of all groups studied homeschoolers looked most like public schoolers. Unlike graduates of private schools of all types, a significant minority of homeschoolers, like public schoolers, got married very young and/or had a child very young. Also like public schoolers, a significant portion never married and never had a child by age 39. So at age 20 homeschoolers (and public schoolers) are more likely to be married than Evangelical and Catholic schoolers, but by age 24 the Evangelicals have caught up and thereafter surpass them, and by age 27 the Catholics have too. The Cardus Survey did distinguish between religious and non-religious homeschoolers, and when Uecker and Hill ran the numbers they did find that the religious homeschoolers tended to marry earlier, but this difference was explained entirely by parental religiosity, not by homeschooling (unlike the Evangelical situation, where levels of religiosity did not account for the difference). Even among the very religious, however, age of first birth was no different than for public schoolers.
Appraisal: The first thing to say about this article is that it adds yet another helpful stroke to our increasingly vivid portrait of homeschool graduates. I have a couple of comments relative to what this data means.
First, regarding the dramatic Evangelical school findings, as I was reading through the article I has a thought that Uecker and Hill mention very near the end of their text. It could be that it is not the K-12 school itself but the colleges to which graduates are sent that is most responsible for Evangelical early marriage and childbearing. I have taught for many years at a Christian college, many of whose students are the graduates of Evangelical Protestant day schools. Our alums do marry one another a lot, often shortly after graduation. It is venerable chestnut at Christian colleges that the female students come to get their “Mrs. Degree.” Students frequently discuss the panic that sets in if a senior woman does not have a “ring by spring.” While these are only jokes, they do point to a general culture that affirms and encourages marriage in a way one does not find at a secular college or university.
Second, regarding homeschoolers, Uecker and Hill themselves offer no explanation for why they do not look any different than public schoolers. The closest they come is to speculate that perhaps there is “some unobserved characteristics of homeschoolers that make them less likely to marry than Evangelical schoolers or Catholic schoolers,” calling for more research into the phenomenon (p. 214). Concurring in the call for more research, let me get the ball rolling with two possible speculative causes of the lower overall rate of homeschooler marriage (and the parallel small percentage of very early marriage that one finds in public schooling as well):
- first, for all its nuance, the Cardus data does not, at least as presented by Uecker and Hill, control for SES. The closest variable is parent education level. For homeschooling this might not be enough. It could be the case that the reason homeschoolers and public schoolers look a lot alike in their marriage and childbearing patterns is because they have a similar SES profile. Recent work by authors like Naomi Cahn and June Carbone has shown, and current work is continuing to show that American marriage patterns are profoundly shaped by social class, with wealthy Americans increasingly marrying later but in high percentages, while poorer Americans increasingly forego marriage because they can’t afford it. It is speculative but plausible that the Cardus survey’s homeschooled graduates had parents who had relatively high levels of education themselves but not enough money to send their children to private Christian schools, and thus graduates’ social class ended up trumping their religious training when it came to their children’s future decisions about marriage and family.
- second, and also entirely speculatively, it could be the case that there is something about the kind of families who choose homeschooling or about homeschooling itself that makes graduates less likely to marry. Contenders here might include the relative isolation of homeschooled young adults from eligible marriage partners, patriarchal attitudes among some homeschoolers toward dating and women’s roles that might limit these women’s marriage prospects, the experiences of some older homeschooled children (particularly girls) who spend their teen years caring for younger siblings and thus are not keen to jump right into motherhood themselves when they reach adulthood, and the possibility that the stereotypical plainness and awkwardness of some homeschoolers might make them less likely to find a spouse.
Whether either or both of these speculations explain Uecker and Hill’s finding will be for future research to determine. The present article has done enough by showing that on yet another variable the results of homeschooling are not much different than those of other forms of education.
Milton Gaither, Messiah College