EMPIRICAL LIMITATIONS TO THOSE CLAIMS
Although homeschooling advocates and allied lawmakers push for policies aligned with the aforementioned conclusions, further investigation reveals that the empirical basis for many of the most profound claims is remarkably questionable. In fact, claims in the areas of (a) outcomes and effects of homeschooling, (b) fiscal advantages to homeschooling, and (c) the benefits of further deregulation are all, on closer inspection, quite problematic.
Claims on Effects
First, probably the most important claim made by and for homeschoolers is that the approach “works,” is “effective,” or “gets results” because it “likely leads to” certain desirable outcomes such as enhancing academic effectiveness, promoting greater civic engagement, increasing participation and success in higher education, and enriching later life and job satisfaction (HSLDA, 2013b; Lips & Feinberg, 2008; Ray, 2009). The focus on academic achievement in particular is a critical issue because hopes of improving the educational experiences and future life opportunities of homeschooled students depend on evidence that the treatment is more effective than other alternatives. It is thus a key argument for expanding the homeschooling movement, as proponents of homeschooling believe it will both help individual students as well as boost the educational productivity of the country in an era of international economic competitiveness. Similarly, other claims about the impact of homeschooling on success in higher education, enriched civic engagement, and future life outcomes are also useful for supporting policies that can further expand the movement.
Any such claims about the effects of homeschooling on desirable outcomes are undercut, however, by basic empirical imperatives. Although there is little dispute that homeschooling children typically attain higher test scores on average, the question is whether homeschooling causes better achievement (or engagement, or higher education participation, etc.). Outcomes such as increased achievement and engagement may simply be a reflection of the advantages that homeschooling families typically bring to their children—advantages that would make it likely that these students would succeed academically and in life even if they were educated in schools.
For instance, according to research results produced by proponents, homeschooled children “score above national averages on standardized achievement tests” (Ray, 2000, p. 74). However, although advocates suggest that above-average test scores demonstrate the success of homeschooling, what is known is that the most important factors influencing student performance rests on socioeconomic predispositions like family income, parental educational attainment, and so on (Coleman et al., 1966; Sacks, 2007; Wrigley, 2011). Accordingly, although there may be a correlation between the act of homeschooling and higher academic outcomes, researchers, and advocates have yet to demonstrate a causal relationship between these two factors. What is more likely is that those parents who choose to homeschool are more invested in the educational outcomes of their children, can afford supplemental materials, have the financial flexibility and benefits to forgo a secondary income, and have higher educational attainment—factors that we know are true of the homeschooling population (Ray, 2010). These factors are likely the explanation for higher test scores rather than the practice of homeschooling itself. That being said, high-achieving students who are homeschooled might very well still reap the benefits of their socioeconomic advantages if they were enrolled in a public school. It follows that the children of parents who homeschool could fare at least as well in public school as they do in the private realm of the home.
Perhaps the best way to test for any causal relationship of homeschooling on these outcomes would be to conduct randomized trials that naturally account for the influence of confounding factors. Yet, by definition, homeschooling families are those that are motivated to self-select into the “treatment” group, making it virtually impossible to construct a useful comparison group with the same attributes and motivations, thereby undercutting the possibility of identifying homeschooling as the causal mechanism in improving outcomes. Alternatively, in trying to isolate the impact of homeschooling itself, researchers could attempt to control for all the other factors known to influence academic outcomes. However, researchers can more easily control for observable factors in comparing groups and treatments; they face significant obstacles with unobservable factors like motivation, initiative, and commitment to education—extremely important predictors of academic success that are known to be well represented in the homeschooling community. Therefore, any attempts to discern the impact of homeschooling are extremely limited, if not fatally flawed.
Nevertheless, homeschooling proponents often point to surveys such as the report from Rudner (1999) in claiming that higher scores for homeschoolers show the effectiveness of the approach. Yet this is akin to arguing that dentists are more effective than emergency room physicians because they see lower mortality rates. The population represented in the sample in the Rudner study is qualitatively different than the larger population to which they were compared, making causal claims unsupportable, as even Rudner noted. In fact, the study drew on a sample of homeschoolers using a testing service offered by a conservative Christian university. Not only is such a sample not representative of the wider homeschooling population (as the author acknowledged), but—even if we were able to overcome the obstacles of controlling for unobservable factors with survey data—it is virtually impossible to construct a sample that is representative of the wider population. This is because basic information about the size and nature of the population that homeschools their children in the United States (an essential prerequisite for making general claims about the treatment) is unknowable due to the substantial degree of under- and nonreporting associated with the movement.
Consequently, efforts to encourage policymakers to expand the homeschooling movement based on claims of the effectiveness of this approach are on extremely tenuous empirical ground. The research used to advocate for such an expansion does not and cannot support claims made regarding the effectiveness of homeschooling as a treatment. In fact, even some of the correlative research produced by homeschooling proponents suggests that it is not the act of homeschooling itself, but instead being the type of family that is interested in homeschooling, that is more closely associated with better outcomes. For instance, NHERI found that there were no statistically significant differences between students who spent varying years being homeschooled, nor any substantial differences in outcomes—less than .5% of the variance—based on the specific approaches used by homeschooling families (Ray, 2009). If there are no significant academic achievement differences between a student who spent his or her entire life being homeschooled and one who spent less, or between homeschooled students subjected to various pedagogical approaches, it would appear that what causes high academic achievement is not homeschooling per se, but the predispositions that most homeschooling families share. Further, it is shown that what causes the largest disparity between scores of homeschooled students are parental education levels, which is also true in the regular schooled population.
In the realm of higher education, Saunders (2009) showed that homeschooled students displayed higher rates of persistence into their sophomore years. However, given that most homeschooled students come from homes with parents holding college degrees, a support structure within those families that aids homeschooled children as they matriculate and persist through college is more likely to be present than in the general population. It very well may be good role models—not homeschooling—that serve as an asset to these students. Thus, rather than encouraging the act of homeschooling, policymakers would be on firmer empirical ground by encouraging all families to be more like homeschooling families: to be highly interested and invested in the education of their children.
Claims on Efficiencies
A second, purportedly research-based claim made about homeschooling deals with the movement's potential cost-saving features and improved efficiencies. Regarding the former, advocates note that homeschooling families pay taxes for public education but do not themselves take advantages of those services, thus providing a financial boon for districts. Furthermore, on the latter issue, they make the argument that homeschooling is more effective by measuring the costs of education in public schools relative to homes, particularly in light of perceived academic outcomes (Ray, 2009). In both of these instances, public schooling is seen as an inefficient alternative to homeschooling because it takes and uses more resources than necessary. Precisely for that reason, such claims can be important in policy discussions because they position homeschooling as a more efficient and effective policy option that should be encouraged and expanded.
As we have shown, such claims appear in the research of advocacy organizations promoting homeschooling. Although there is certainly some truth to the claim that districts are collecting money to educate students who will never set foot in their hallways (as is also true in the case of students attending private schools), the implicit and overt basis of these claims is not as strong as it may initially seem. In fact, many of these arguments break down on closer inspection. For instance, claims on expenditures often rely on inappropriate apples-to-oranges comparisons. Citing Ray (2009), the HSLDA indicates that public school students perform well below the level of homeschooled students, despite the fact that $9,963 is spent on the public school students compared to a median of about $400 to $599 for homeschooled students. Consistent with its insistence that government-run entities are inherently wasteful, the Heritage Foundation claims that homeschooling saves an average of $4.4 billion to $9.9 billion annually (Lips & Feinberg, 2008). These savings, according to the Heritage Foundation, can “be saved or reallocated to other uses” (Lips & Feinberg, 2008). This notion, partnered with claims of higher academic achievement, constitutes a claim of financial and academic efficiency.
But such simplistic comparisons neglect basic social science tenets by comparing one population or process to another despite ignoring well-documented differences between the two groups and the inputs to the productive processes, including family income and education, special education costs, and so forth. Furthermore, these claims appear to ignore the substantial costs of homeschooling to the families that admirably shoulder these burdens, including overhead costs factored into the public school figure, as well as opportunity costs of adults foregoing paid employment or career advancement.
Indeed, the basic claim of efficiency advantages is fundamentally flawed. Efficiency is a question of the ratio of inputs to outcomes, with greater efficiency being a matter of increasing outcomes while holding inputs constant and/or reducing inputs while outcomes do not decline. Yet, despite the appeal of homeschooling for its appearance of getting great academic outcomes with relatively minimal inputs, in fact, neither side of that equation is or can be appropriately specified. As we discussed in the previous section, researchers have yet to demonstrate the actual outcomes of homeschooling itself (controlling for other confounding factors). Furthermore, studies have yet to appropriately account for the inputs necessary for homeschooling, partly because so many of the factors going into any educational effort—motivation, experience, commitment, and so on—are nearly impossible to quantify. Moreover, the collective benefits typically associated with public education, such as increased social tolerance and cohesion, enhanced social capital and economic productivity, reduced fertility, and crime, are also difficult to quantify as direct outcomes of the endeavor.
Yet, even if we accept the unsupported claim that homeschooling embodies efficiencies, it does not then follow that policies encouraging more families to homeschool will result in better circumstances for students remaining in public schools, even if fewer students are then using the resources devoted to public education. The transfer of one child from public school to homeschooling would typically have an insignificant impact on costs to a school or district, and does not really represent any savings because the school would typically still need the same number of teachers, a principal, budget for overhead, and so on.
In fact, to effect significant cost savings for public schools, a critical mass of families—enough to merit reduction in teaching staff, for instance—would have to leave a particular school or district, and its budget would have to remain the same (i.e., the school or district's budget allocation is not determined on a per-pupil basis). Thus, even though the homeschooling movement is substantial across the country, only where it reaches this critical mass in specific localities could it potentially result in real budget savings, at least according to the logic suggested by advocates. However, if such a dramatic shift in student population were to occur, it could represent not only potential savings but also serious threats to those remaining in the schools. The loss of a substantial number of students may mean the loss of political support for local funding of public schools. Furthermore, the exit from public schools of a mass of educated, active families with an interest in education is likely to have detrimental impacts on the school community through a degraded peer effect (on the peer effect, see Epple & Romano, 1998; Hanushek, Markman, Kain, & Rivkin, 2003; Hoxby, 2000; Lubienski, 2003).
A notable irony in all of this is that, even as advocates claim that there are cost savings to the public education system, they are advancing proposals to cut into those supposed savings to encourage the further expansion of homeschooling. For instance, the Heritage Foundation advocates for the expansion of educational tax credits and deductions for homeschooling expenses for families, as well as state tax incentives for other parties that contribute to a child's tax-free education savings account (Lips & Feinberg, 2008).
Claims for Deregulation
Finally, a problematic claim made regarding homeschooling is that further deregulation of the movement is necessary, presumably to improve opportunities and outcomes for more students. This claim is derived from observations about the relative performance of students taught by formally trained or untrained educators but is also situated within the context of homeschooling proponents’ general aversion on the part of leading homeschooling groups to almost any regulation or oversight. Consequently, this type of claim is used to support policies that further reduce public responsibility and involvement in education for homeschooling families. Current and recent proposals seek to accelerate and extend this trend by removing the requirement that parents educating their children at home have a teacher certification or even a college degree, and restricting state oversight of homeschooling, as with the reduction or eradication of mandatory testing for homeschooled children.
These calls are based on a number of assertions from homeschooling advocacy research, particularly the claim that parents with teacher certification are not more effective than noncertified parents (Ray, 2009), or that certification does not have a significant benefit for education in general. This issue of rolling back state requirements that homeschooling parents need to be certified came to the fore when a court ruling in California unexpectedly (and temporarily) upheld such a requirement (Egelko & Tucker, 2008; Home School Legal Defense Association, 2008). Consequently, advocacy groups like the HSLDA point to survey data indicating that students whose parents were not certified actually scored higher on standardized tests, arguing that “critics of home-schooling have long insisted that parents who want to teach their own children should become certified teachers first,” yet their study “found that whether or not parents were teacher-certified had no impact on these high scores” (Ray, 2009; see also Ray, 2010).
Such conclusions with regard to homeschooling are quite dubious. Again, available data and analyses are not able to support the claims made by these organizations. Survey data are not suited to making claims regarding a causal link, or lack thereof, between parents’ teacher certification status and the academic outcomes of their children. In this same vein, there has been a vigorous debate in research circles over the years about the degree to which teacher certification matters not only in homeschooling settings but in public education as well (e.g., Darling-Hammond, Berry, & Thoreson, 2001; Goldhaber & Brewer, 2000; Moe, 2005; Walsh, 2001). Although much of the research has been mixed and contested, a recent large-scale study found that teacher certification was a significant predictor of achievement in school settings (Lubienski & Lubienski, 2013). Whether or not teacher certification is a significant predictor of achievement in home education settings has not been established in the empirical literature, and analyses of survey data simply have not examined the issue with any rigor; however, as we noted earlier, there is reason to think that the socioeconomic and unmeasurable motivational advantage of homeschool families often make up for or mask any deficiencies in pedagogical training (a secondary factor even in school settings) and that further expansion of the homeschooling movement could draw in families where these advantages are not as pronounced, thus diminishing the primacy of family background and enhancing the potential impact of formal training.
EXPLORING THE DIVIDE BETWEEN RESEARCH EVIDENCE AND POLICY ADVOCACY
In an era of school reform characterized by demands for scientifically based interventions, the curious case of homeschooling stands out for its lack of grounding in any sound empirical evidence. Facilitated by policymakers’ efforts to lessen restrictions on its growth, and despite a notable dearth of empirical evidence on its effectiveness, the homeschooling movement has grown by leaps and bounds, even as policymakers require research-based practices, and private funders pursue “effective philanthropy” that shows evidence of the impact of the programs they support. We do not intend to take a stand at this point regarding the overall desirability of homeschooling, nor on specific issues such as the requirement for parents to have a teaching degree. Instead, we simply want to point out the tenuous empirical basis for many claims made to advance the homeschooling movement.
Considering the state of the data available, it is simply not possible to claim that homeschooling “works” and “leads to” desirable outcomes. Those claims might be true but cannot be supported by analyses of extant empirical evidence. Indeed, homeschooling advocates are on much firmer footing simply arguing for greater deregulation and expansion based on other grounds, such as the demonstrable satisfaction of many of those engaging in the practice, or the moral or legal argument that parents have a substantial right to control the education of their children. Still, leading proponents persist in trying to prove the academic impacts of this approach. We contend that this is because evidence of impact is quite persuasive in policymaking arenas that have been so focused on academic effectiveness as evidenced by standardized test scores and that advocates see this as a key element of their efforts to expand and further deregulate the practice.
Despite advocacy organizations clamoring for more change, homeschooling has already been substantially deregulated over the last few decades in the United States, with fewer barriers, restrictions, and points of public accountability. However, we are not aware of any compelling evidence that deregulation to this point has improved the effectiveness of the practice. Indeed, in lieu of any firm evidence that the homeschooling “treatment”—as opposed to home factors—is at all effective, it is far from clear that expanding the movement will increase its impact. In fact, as the practice is likely further expanded due to deregulation, it could be that results will diminish as families with characteristics that are more marginally associated with academic success join the movement.
Although they are often dressed up in a scientific rhetoric of performative measures and results, it appears that calls for further deregulation of homeschooling may be ideological rather than empirical imperative. Rather than showing a strong empirical basis to justify the expansion of homeschooling, the evidence indicates that the movement is growing for other reasons and that empirical claims of its effectiveness are just a very useful marketing mechanism.
The Advocacy Agenda
Of course, calls for further deregulation of homeschooling are taking place in an atmosphere of antigovernment advocacy. Although there are often good justifications for limiting state intrusions in the private sphere, proponents of many efforts to roll back the reach of the government frequently see this as an end in itself, regardless of the actual evidence on the effectiveness of public institutions or their proposed alternatives. Of interest, even as some proponents seek to reduce the government role in education by further deregulating and encouraging homeschooling, they also seek state subsidies to entice families into joining their anti-state agenda (e.g., Lips & Feinberg, 2008).
There has probably been no organization with more success in advocating for homeschooling than the HSLDA. The group has been pushing for an amendment to the U.S. Constitution as well as for state legislation that would affirm parents’ basic legal right to control and direct the education of their children. HSDLA contends that
the Parental Rights amendment to the U.S. Constitution would ensure that parents have a fundamental right to raise, educate, and care for their children. The amendment would also prevent treaties from overruling U.S. law regarding parents and children, such as the UN Convention on the Rights of the Child. (see HSLDA, 2012a, 2012d)
HSLDA actively monitors state and federal legislative proposals and is very effective in mobilizing opposition to any bills it believes could even remotely represent the possibility of impacting the rights of homeschooling parents. For instance, it was famously instrumental in defeating the ratification of an international treaty—based on U.S. law—designed to secure the rights of disabled children because the proposed legal standard of the “best interests of the child” could, according to HSLDA's interpretation, “override the traditional fundamental right of parents to direct the education and upbringing of their child with special needs” (Estrada, 2012b). HSLDA (2011, 2012b) has also opposed numerous legislative actions that proposed to change child abuse reporting standards. Such actions would require all adults to report child abuse. Among the concerns raised by HSLDA are that such moves will allow for an increase in baseless child abuse reports, and some of the proposals allow investigations without evidence of abuse.
In addition, the HSLDA (2012c, 2013a) has been very active in opposing legislation to increase the compulsory school age because that would expand government control over education. The group responded to President Obama's 2012 declaration that states should require all children to stay in school until they graduate or turn 18 by noting that
if there were a federal mandate (either passed by Congress or through regulations) that required the states to keep students in school until they graduate or turn age 18, this could lead to a federal definition of what constitutes “graduation from high school.” Once the federal government creates federal guidelines or definitions in this area, additional and harmful federal regulations on homeschoolers could easily follow. (Estrada, 2012a)
Such strident and preemptive advocacy for this agenda bears similarities to the successful strategies of the National Rifle Association in promoting its interpretation of the Second Amendment in response to any perceived threat to those rights. In these cases, an abstract principle is elevated to a pure, if extreme, interpretation of rights regardless of the real-world consequences.
CONCLUSION
In this analysis we have offered a critique of the empirical arguments made by and for the homeschooling movement. Rather than a critique of homeschooling per se, we have demonstrated that there is essentially no scientific evidence on the effectiveness of homeschooling. This is not to say that the practice is not effective, particularly in every case, but only that multiple research attempts have not yet proven its effectiveness. Despite massive increases in the scale of the practice, moves to further expand and deregulate homeschooling are not supported by empirical evidence.