Does Homeschooling “Work”? A Critique of the Empirical Claims and Agenda of Advocacy Organizations.

Author(s): Christopher Lubienski, Tiffany Puckett, and T. Jameson Brewer

Source: Lubienski, Christoher, Tiffany Puckett, and T. Jameson Brewer. 2013. “Does Homeschooling ‘Work’? A Critique of the Empirical Claims and Agenda of Advocacy Organizations.” Peabody Journal of Education 88(3): 378-392, DOI: 10.1080/0161956X.


Homeschooling proponents advocate for the expansion of financial and ideological support for homeschooling, as well as the loosening of limitations on homeschooling families that they say unnecessarily introduce barriers to entry into homeschooling.  They base their arguments on what they claim is empirical evidence that such environments promote higher academic achievement and post-secondary enrollment and reduce the cost burden to the public versus the inefficiencies of the current public school models in the United States.  Lubienski, Puckett, and Brewer review three dozen academic articles and reports that give empirical results of homeschool – public school comparisons.  They find no empirical evidence to support homeschooling proponents’ claims that homeschooling results in increased academic achievement or a marked cost savings.  In fact, they conclude that primarily because of difficulties in measuring academic success and costs in this sector, very little empirical evidence exists to demonstrate the effectiveness of homeschooling on any measure.

Examining the Evidence
The authors’ primary criticism of the majority of the claims of homeschooling’s greater effectiveness and efficiency is that the conclusions are based on data from samples of homeschoolers that represent only one small segment of the homeschooling population rather than being representative of all homeschoolers.  The authors recognize the extreme difficulty in building a nationally-representative sample given the necessarily decentralized nature of homeschooling, but caution those who attempt to make claims about all homeschoolers based on limited samples, which likely leads to biased estimates and skewed conclusions.

Even if a perfect sample could be constructed, the authors argue that the common problem in education research of separating school sector effects from selection effects is particularly problematic for studies of homeschooling.  That is, because of the nature of homeschooling, which melds the home and learning environments, it is very difficult to ascertain if a homeschooler’s academic achievement or other measure of success (or failure) is due to having participated in homeschooling, or from the home environment itself. 

Lubienski, Puckett, and Brewer argue that parents who choose to homeschool their children very often have higher educational attainment, socioeconomic resources, and/or motivation and interest to participate in their children’s education compared with non-homeschooling parents.  Because all of these factors are strongly and positively associated with academic achievement, one cannot be sure if positive results for homeschoolers are due to the experience of being homeschooled, or to the experience of being raised by parents who select homeschooling in the first place.  Said another way, would homeschooled students do just as well or possibly even better in an institutional school than they did in homeschool environments?

Given this, the authors conclude that when policy makers are asked to make decisions about expanding support for and deregulating homeschooling based on existing empirical evidence, they have incomplete and possibly incorrect information.  Beyond the sample and measurement issues already discussed, the authors contend that results are misleading, confirming an agenda in pursuit of a particular policy decision rather than factual data about homeschooling.

For instance, Lubienski, Puckett, and Brewer argue that although the per-pupil cost for a homeschool student might be substantially lower than the per-pupil costs for public or private school, no cost savings are realized when only a few families opt in to homeschooling.  That is, only a massive shift from public schooling to homeschooling would result in a noticeable, significant cost savings, as institutional costs of running a full classroom and a full classroom minus one or two students will be nearly identical.  They particularly question claims about public cost savings that also call for tax credits for homeschooling families, which would seem to eliminate any potential cost savings.

Lubienski, Puckett, and Brewer also note that homeschool advocates call for deregulation, particularly regarding the requirement that homeschooling parents hold teaching certificates, but do not provide evidence that non-certified homeschool parent/teachers are as effective as those who are certified.  Such evidence would be particularly necessary in order to make a convincing argument since certification of public and private school teachers is positively associated with student achievement.

The authors conclude their review of the homeschooling literature regarding academic outcomes, cost reductions associated with the shift to homeschooling, and potential effects of deregulation with the assertion that there is no empirical evidence that homeschooling “works,” that deregulation might lead to improved effectiveness, or that anything less than a massive shift to homeschooling would result in a significant cost savings.

Ideas to Consider
Lubienski, Puckett, and Brewer correctly identify the difficult challenges facing those who wish to study homeschooling and its participants.  Not only is adequate sample selection problematic given the decentralized nature of homeschooling, but the task of parsing school effects from home effects is tricky given the context in which schooling takes place inside the home.

These authors address the claims often discussed by homeschooling advocates in terms of academic achievement, post-secondary enrollment, and the use of resources in order to garner additional public support.  Such an examination, however, is insufficient to capture the outcomes that are likely of most importance to families who opt-in to homeschooling, such as the potential to preserve and strengthen family relationships, the ability to control the security and tone of the learning environment, and the ability to provide specific cultural and/or religious instruction.  Although these concerns were outside of the scope of Lubienski, Puckett, and Brewer’s argument, future studies – particularly those that attempt to build a sample that addresses the challenges these authors suggest – should measure outcomes more broadly in order to gain a thorough understanding of how homeschooling graduates compare with graduates of more traditional schooling models.  

Keywords: homeschooling, empirical claims, sample selection, deregulation, policy

Sector: Homeschool

Outcomes: Academic,  Civil and Political

Date Posted: 2015-11-19