Homeschooling in Wisconsin

Record: Denton Marks & David M. Welsch,  “Homeschooling Choice and Timing:
An Examination of Socioeconomic and Policy Influences in Wisconsin.” Journal of School Choice, 13, no. 1 (2019): 33-57 [Abstract Here]

Summary: Marks and Welsch were both economics professors at the University of Wisconsin-Whitewater at the time this article was published.  Marks seems no longer to be in the system.  Marks’ publications range widely, but he seems to have a special interest in economic issues related to wine.  Welsch has published several articles on educational issues, especially concerning Wisconsin.  Here they collaborate to provide a rare quantitative gem that mines Wisconsin’s excellent homeschooler registration data to make some interesting generalizations.They begin with a literature review of the few quantitative studies that have been done on homeschooling, especially the studies of Isenberg (one reviewed here).  Like Isenberg, they use Wisconsin data because it’s better than that of almost any other state, but given the recent nature of this article they are able to use more data, and they engage in more sophisticated statistical manipulations to extract more conclusions.  Winsconsin’s data is so good because its procedure for registration is “simple and unobstrusive,” (p. 34) requiring only that parents report the number of students in the household being homeschooled and their grade level, note the school district of residence, and agree that they comply with relevant state statutes.  No evaluation or oversight is imposed, so, think the authors, there is little incentive not to report.

Marks and Welsch are able to compare the district-by district homeschooling counts with other variables such as percentage of district students qualifying for free or reduced lunch (a proxy for SES), ethnic diversity, total population, and number of students in private schools, as well as more general demographic characteristics of the district (median income, education levels, age distribution, percent urban, political affiliation, religious characteristics).  Running statistical analyses of all of this data leads to some very intriguing findings:

1. Probably their most significant finding is a consistent trend from 2004 to 2011 of dramatic declines in numbers of homeschoolers in 11th and 12th grades.  When you include the not insubstantial decrease in 10th grade as well, 1/4 or more of Wisconsin homeschoolers stop homeschooling sometime in their last 3 years of high school.  This is an important finding on its own, but it also suggests, the authors note, that a lot of students who were homeschooled for most of their educations are not counted as such when they take SATs and ACTs or when they enter college, because they finished up in the public school system.  This means that studies of college students comparing public schooled and homeschooled children may be flawed.  It is very possible that homeschoolers who were not doing very well shifted over to public schools to try to improve their prospects of getting into college.  Such students are then mislabeled by researchers, thus making “homeschoolers look better than they are.” (p. 52)

Does the data tell us anything more about which upper grade children transition to schools?  First, the more urban the district, the sharper the drop off.  Second, the better the test scores in a public school, the less likely homeschoolers are to rejoin public education.  It is not entirely clear what that means, but Marks and Welsch speculate that perhaps families situated in a suburban district with good schools don’t want their children attending because it will be harder for them to compete with the excellent students already attending such schools in terms of class rank and GPA. (p. 51)

That was the lead finding of this study.  Here are some more:

2. Districts with a higher percentage of poor students (those qualifying for free or reduced lunch) also have higher percentages of homeschoolers.

3. Districts with more Catholics have fewer homeschoolers.  The authors make much of this finding, suggesting that “it is important to distinguish between Catholics and other Christians since it appears that they may have differing views on homeschooling.” (p. 48)

4. Districts with higher per-pupil public school expenditures have fewer homeschoolers.

5. Though it’s not as big an influence as percentage of Catholics, the higher public school students in a district score on state tests, the lower that district’s percentage of homeschooling.  Math scores are more salient here than reading scores.

5. Counter-intuitively, the higher the percentage of public school teachers in a district with Masters or higher degrees, the higher the percentage of homeschoolers.

6. As Isenberg found, so Marks and Welsch find that the perfect storm for a homeschooling decision is a high socioeconomic status family living in a lower socioeconomic status district.

Appraisal: The most important thing to say about this study is that it is a wonderful addition to the literature.  Very little homeschooling research is quantitative, and what quantitative studies exist often draw on convenience samples.  This study uses rigorous statistical methodology to tease conclusions out of very good state-wide data.  The conclusions, moreover, pass the smell test, especially the lead finding.  We can now say that there is solid empirical evidence that a quarter or more of homeschoolers transition to brick-and-mortar schools in the last years of their secondary education.  I have taught many students at my college for whom this was the case.  It’s nice to have a major study to buttress this anecdotal intuition.

When the authors depart from the data, however, I am not so sure about their conclusions.  The data don’t tell us why certain families choose to put their older homeschooled children into schools.  These economics professors do what economics professors tend to do, which is to make real people into rational actors always motivated by and acting upon their own economic interests. As a growing chorus has been arguing for years now, this is not how people really work.  It is plausible I suppose to think that suburban homeschooling families keep their oldest children at home longer because they fear that the excellent students at Suburban High would be tough competition.  But it’s equally plausible that suburban families who homeschool are already self-selected to be the sort that will go the distance, probably because of religious motivations.  More urban homeschoolers, in contrast, might be homeschooling for more pragmatic than religious reasons and thus might be more likely to abandon the practice when it no longer suits.  I have no more evidence for that than do Marks and Welsch for their rational choice explanation, which is where qualitative research should come in.  Now that we have good data finding that a large chunk of homeschoolers go to school in high school, future researchers should ask parents and/or older children why they chose to stop or to keep homeschooling.

Finally, a word about the Catholic claim.  Wisconsin’s example might not be generalizable because of the unique situation in that state.  Wisconsin has a lot of Catholics, and it also has robust and growing public funding mechanisms for families who want to send their children to religious schools.  It may indeed be the case that Winsconsin Catholics are not heavy backers of homeschooling.  But Catholics in other states with fewer co-religionists and less public funding for Catholic schooling might homeschool in greater numbers.  That would be another wonderful topic for future research to investigate.

Milton Gaither

Messiah College