This post reviews R. Pennings, et al., “A Rising Tide Lifts all Boats: Measuring Non-Government School Effects in Service of the Canadian Public Good” (Hamilton, Ontario, Canada, 2012). Available for free download here.
Back in 2011 I reviewed the first Cardus Survey, which provided rare randomly sampled data about young adults who had been homeschooled in the United States. This new study does the same for Canada. In the first study the Cardus researchers uncovered some fascinating information about adult homeschoolers, some of which proved rather controversial because it was not very flattering toward homeschooling.
This new study’s results are fairly similar. The study is about much more than homeschooling, but since this is a blog about homeschooling research I will limit my comments to the homeschooling findings.
First let me explain the methodology. Unlike the widely cited studies that Brian Ray has done over the years that have consistently made homeschooling look very good, Cardus doesn’t recruit homeschoolers from key organizations. Rather, it takes a random and representative sample of the entire population, in this case the population of Canada, in the process catching a few homeschoolers. This way they’re not getting a self-selecting group of volunteers but a representative group. Cardus’ overall sample was of 2,054 Canadian young adults, aged 24 to 39 years. Of that overall group, 41 had been homeschooled in a religious context and 17 in a non-religious context. The way they determined the context was whether or not the mother attended religious services weekly. Since the non-religious group was so low, they didn’t include it in the analysis (they made the same move in the 2011 U.S. study).
For important reasons they also separated out the subjects from Quebec for a separate analysis. This brought the number of religiously homeschooled young adults in the survey down to 34. So everything that I’m about to summarize is based on the responses to survey questions of 34 previously homeschooled, non-Quebecois Canadian young adults. While that is a fairly small sample size, the key is that it was a randomly generated sample. This means that we can compare this group’s responses to those of subjects in the sample who attended government-run schools and other kinds of private schools. Regular readers of this blog have heard me complain over and over for years about how it is illegitimate to compare the results of Ray’s studies on academic achievement in homeschoolers to public schools. This Cardus data is one of the very few studies where we can actually make such comparisons.
Here are the findings:
Demographically, the religiously homeschooled non-Quebecois Canadian young adults (which from here on out I’ll label RHCs) make less money on average than graduates of public schools. But then they started out poorer overall, so this difference is not due to homeschooling. These formerly homeschooled adults are much more likely to be married than other Canadians, to marry at earlier ages, and to have more children. Interestingly, while the U.S. data from 2011 found higher rates of divorce among homeschoolers, RHCs seem to divorce a little less than their government school peers.
How about volunteering? RHCs are very involved in their religious congregations, volunteering at far higher rates than other groups here. But they volunteer less than government school grads in other venues, especially arts, cultural, or political organizations. On the other hand, RHCs are more involved than others in “non-partisan civic organizations,” especially youth and sports organizations (recall that they have more children, so this makes sense). There is a little evidence that RHCs tend to be parts of such organizations at slightly higher rates but less likely to be officers or committee members in charge of such groups.
How about giving? RHCs give at very high rates to charities, but the Cardus data here didn’t specify what kind of charity. It seemed to me from reading the text that these charities are very likely to be church-related. The bottom line seems to be that RHCs give more overall than government school grads, but even so they still give a lot less to secular charities and political causes. It’s a more focused, in-group giving.
Occupation? RHCs area little less likely to hold professional and managerial occupations. They’re also more likely to be unemployed or to work part time. You’d think this would be explained by the fact that they marry earlier and have more kids, but in fact even when you control for these factors you still find higher rates of underemployment. Why? As we’ll see soon, a lot of it has to do with educational attainment. Also interesting is the fact that despite having poorer job prospects, RHCs are far more likely to report feeling that God has called them to their particular jobs. They also have higher expectations that their jobs should allow for creativity and should fulfill a religious calling. They are significantly less likely, however, to look for friendships with coworkers. They report being less concerned about the ethics of their coworkers. The Cardus authors suggest that “The relative lack of concern for relationships at work perhaps reflects a relationship focus within existing family, church, and community circles.” (p. 33)
Political engagement? RHCs vote a little less than government school grads. They are less likely to get information about politics from the media and they seem much less informed about politics generally. The Cardus authors’ gloss on this interesting finding is worth quoting:
Christian school and religious home-educated graduates are less likely than their government school peers to be politically informed…they may not be thinking through political issues as deeply as the government school graduates and perhaps are more superficially involved politically. on the one hand, this is not desirable from a civic perspective; on the other hand, the finding also counters the perception that graduates of non-government religious schools are highly politicized. (p. 34)
Though they seem less informed and less connected, RHCs report talking about politics more as families and in their churches. They report being more likely to campaign and to boycott, though the survey doesn’t ask for what causes. My guess would be that RHC’s get very involved politically with issues of direct personal concern but not so much otherwise. Strangely, this group also reports having higher rates of connection to elected officials, corporation executives, and community leaders. I have no idea why that is. All other groups of privately educated adults were less likely to be connected to such leaders than government school grads. But homeschoolers seem to at least think that they have friends in high places.
Culturally, RHCs are far less likely to read the newspaper, much more likely to read the Bible, and a little more likely to read non-religious literature. They are also significantly more likely to play a musical instrument or sing in a group.
Trust: RHCs are more likely to trust members of their own congregations but less likely to trust other people in general than Government school grads. They are especially distrustful of government school personnel, scientists, atheists, and the media. Why? The Cardus authors suggest that this distrust is “consistent with their sense that they are living in an environment hostile to Christian faith.” (p. 37) Having said all of that, it is then reported that RHCs are “strikingly more likely” to have faith in the prime minister and cabinet. Maybe a Canadian reader can help me out here, for it seems strange to me that a group that is profoundly distrustful of government would nevertheless be so pro-Prime Minister. Is the current PM, like George W. Bush was in the United States, someone with whom religious homeschoolers would perhaps identify strongly?
In terms of political views, RHCs are further on the scale toward libertarian than any other group of Canadians. On every measure asked about they consistently wanted less government involvement.
Environmentalism is a particularly interesting issue. RHCs react negatively to the label of “environmentalist” and tend to actually do less for the environment than other groups, but they nevertheless express a strong “sense of moral or religious obligation to care for the environment.” (p. 39) It’s sort of like the political involvement issue. RHCs feel like they should be involved, but they don’t live it out that much, possibly because “both their families and their churches occupy more of their time” (p. 41)
How about Life Satisfaction? This was one of the controversial findings in the 2011 U.S. data, where it was found that homeschooled adults felt helpless and that their lives lacked goals. The Canadian data finds that RHCs are also more likely to feel helpless in dealing with life’s problems and lack clear goals or direction, but not to the extent of the 2011 U.S. data. In fact, the difference here was small and disappeared when family background variables were controlled (recall that RHCs grew up poorer on average than other groups). RHCs were also less likely to be satisfied with their family life, with the town they lived in, and with their health. You get the sense that RHCs are just a little less happy overall.
Interestingly, despite the fact that RHCs have on average lower-paying jobs and less satisfaction with their current lot in life, they are much more likely to claim that their secondary education prepared them well. The only thing they didn’t like about it was the limited athletic opportunities.
How about Academic Achievement? The results here are what the authors call “bimodal.” Basically, RHCs were on the whole less likely to go to University or to get advanced degrees than any other group, but for those few who did go on, they tended to “go all the way” and get a Ph.D. (p. 44) This finding suggests the wide range of experiences that can happen under the homeschooling umbrella.
How about Religion? Unsurprisingly, RHCs are far more likely to consider religion to play an important role in their lives. They are more respectful of Church authority figures, less likely to doubt their faith, more likely to share their faith with others, and more likely to want religious commitment to play a role in public life, convinced as they are that the dominant Canadian society is hostile to their values. They also are less likely to think that society should tolerate non-Christian religions. They are much, much more likely to believe that God created the world in six 24 hour days and that the Bible is infallible, and they also don’t think that the Bible and science contradict one another. They pray more and attend more services. They are more likely to believe that it’s morally wrong to live together outside of marriage (though 41% of them have done so), and to be against extra-marital sex and especially same-sex marriage.
And there you have it. What’s especially interesting to me is how the tone of the Cardus report tries to spin all of this data. If one reads only the executive summary or the conclusions one would easily conclude that religiously homeschooled Canadians are doing great–even better than graduates of government schools. One would also think that the definition of the “Public Good” with which the Cardus Survey begins was being met. But when one actually reads through the details of the report the picture is far less rosy. I don’t know why exactly the authors chose to white-wash their own findings in this way.
The stated goal of the survey was to see to what degree the various forms of private education in Canada contribute to a “common good that values not only economic prosperity, but also healthy citizens, democratic participation, pluralism, tolerance, and cohesion.” (p. 13) If those are the goals, then it seems to me that according to Cardus’ own data homeschooling is not doing as good a job as government schooling at fostering the public good. As we’ve seen, RHCs underperform economically, report lower health satisfaction, vote less, are less tolerant of other religious and moral views, and tend to stick with their own rather than befriending or volunteering with those outside their family and religious circles. On every single measure mentioned in the quote I just gave homeschoolers are doing worse. On average it’s not much worse, just a little. But that is the clear finding from this data.
Finally, I’d like to repeat the gripe I made about the 2011 data concerning non-religious homeschoolers. It’s completely understandable that Cardus decided that a sample size of 17 was insufficient to merit inclusion of this group into their overall analysis. But since we know so little about this vastly under-reported homeschooling demographic I’d still love to have the data. How do these 17 non-religiously homeschooled young adults compare to their religiously homeschooled peers? The Cardus data wouldn’t be definitive here, but it would still be very interesting to at least see what’s there.