“Private-ization” and School Effects: Time to Revisit the Public-Private School Question?
by Christopher A. Lubienski & Sarah Theule Lubienski - September 09, 2005
The authors argue that the policy context of school choice and NCLB highlight common assumptions about the superiority of private schools for enhancing academic achievement. However, emerging evidence suggests the need to re-examine basic assumptions about the relative effectiveness of public and private schools.
The private-school effect is part of common wisdom in the United States and is widely accepted in the policy-making and research communities. The influence of this idea can be seen in current school-reform policies. Indeed, the school choice is premised on the belief that private-style organizational forms such as private schools or charter schools have organizational attributes that make them superior to government-run entities in terms of efficiency and effectiveness. This thinking is also reflected in the provisions of the No Child Left Behind Act of 2001 that uses choice sanctions and charter status to prod low-performing public schools toward higher achievement. We might see these reforms as examples of private-ization in the sense that, although ownership of public schools is not being transferred to private hands (what is typically labeled privatization), private sector models are being imposed on public education (Lubienski, in press). Under the assumption that organizational structure and institutional environment shape the internal processes of schools, reformers believe that the use of mechanisms such as consumer choice and competition between providers associated with the private school sector will force public schools to improve.
And yet there are reasons to question this thinking. For instance, research in other nations appears to undercut the assumption that private organizational forms for schools are necessarily superior (McEwan, 2000; Somers et al., 2004). Additionally, for some time, a few US observers have found claims of private school effectiveness to be overstated or misguided (Alexander & Pallas, 1985; Benveniste et al., 2003; Cookson, 1993). Furthermore, much of the empirical data upon which our current thinking is based is becoming dated. The watershed studies on this topic examined students who began school over a generation ago.1 School demographics in the public and private sectors have changed significantly since then (Cooper, 1988; McLaughlin & Broughman, 1997). Given that some of the most influential current reforms are built upon the belief that private and private-style organizations exhibit attributes that lead to greater achievement, it is important to re-examine the question as to whether private schools today are, in fact, more effective than public schools.
Current evidence clearly indicates that private school students do indeed score higher, on average, than public school students do on achievement tests.2 However, while undertaking a study of mathematics instruction and equity, Lubienski, Camburn & Shelley (2004) found an unexpected occurrence in their examination of the 2000 National Assessment of Educational Progress data: when controlling for private school status and student background variables in statistical models, they noted that mathematics achievement in public schools was higher than in private schools. We considered this finding to be quite curious, in light of the evidence from previous studies that suggested the opposite. Hence, we decided to conduct a more focused comparison of private and public school achievement with the 2000 NAEP data.
Looking at Public and Private School Data
We analyzed mathematics achievement data from the 2000 NAEP, collected from over 28,000 fourth- and eighth-grade students randomly selected from over 1,300 public and private schools. In order to control more adequately for student background characteristics, we used a more comprehensive measure of student socio-economic status (SES) than the traditional reliance on free/reduced lunch eligibility. This included several home variables such as available reading materials, computer/internet access, parents education levels, and the extent to which a family discusses a childs studies, along with school-reported free/reduced lunch and Title 1 eligibility. To create a school-level SES variable, we combined this student SES information with school-wide percentages of students who qualify for free/reduced lunch and Title 1.
First, we compared overall mathematics achievement means without controlling for any demographic differences between public and private schools. As would be expected, private schools mean mathematics scores tended to be higher than those of public schools (by 6-7 points, or approximately .3 standard deviations). However, as is widely known, private schools typically serve students from more advantaged backgrounds: whereas fewer than 40% of public schools were high-SES (above the median on the school SES measure), over 80% of private schools were high-SES.
Therefore, we sought to get a more accurate understanding of public and private school achievement, controlling for background factors, in two ways. First, we broke the schools into SES quartiles and compared public and private school achievement within each of the quartiles (Lubienski & Lubienski, 2005b). Within each SES quartile, the public school mean was actually higher (between 1 and 9 points) than that of the corresponding private school mean at both grades 4 and 8. This situation nicely illustrates Simpsons Paradox although within each subgroup, public school means were higher than private school means, the overall private school means were higher than public school means because of the larger proportion of higher-SES students in those schools.
The second way we compared achievement employed Hierarchical Linear Modeling, which allowed us to control for additional demographic differences while comparing student achievement in public, Catholic, and other private schools (Lubienski & Lubienski, 2005a). Given the differences in enrollment in public and private schools, we controlled for student characteristics such as race/ethnicity and disability in our models. Again, after accounting for these factors, the coefficients reversed, with public school means being equal or higher (by .1-.2 standard deviations) than private school means. However, interaction effects in our models also indicated that SES-related gaps were smaller in private schools than public schools at grade 4, while the Hispanic-White gap was smaller in Catholic schools than others in grade 8 (consistent with the findings of Bryk, Lee & Holland, 1993).
Our examinations of the 2000 NAEP mathematics data indicate that the achievement disparity favoring private school students was eliminated, and even reversed to some extent, after taking demographic differences into account. We might speculate on a number of explanations for this, such as the possibility that public schools employ better-educated or more experienced teachers, or teach a curriculum more aligned with the NAEP mathematics test. It is also possible that a substantial number of parents enroll their students in private schools after having encountered academic or other difficulties in public schools. Moreover, public-private school differences in the reporting of free/reduced lunch eligibility may have contributed to the patterns seen in this study. It is important to note that NAEP is not designed specifically for comparisons of school effectiveness. The data are not longitudinal, so they cannot demonstrate that differences in achievement are caused by a particular school type. Furthermore, these data are from national samples, and do not account for the fact that private schools are unevenly distributed across the US.
Hence, by no means do we consider these results to be a definitive statement on the issue of public and private school effectiveness. Any time that research such as this runs counter to scholarly consensus and common wisdom, caution is warranted. However, in view of the changing nature of public and private schools in the US, and given that important policy decisions are being made on the basis of current beliefs about private school superiority, we believe that this study underlines the need to refocus research in this area.
We are currently undertaking one such study with the 2003 NAEP data set, which has much larger sample sizes (150,000 students per grade) than ever before, as well as data on charter schools and a variety of types of private schools. This study will more thoroughly examine which particular aspects of schools (e.g., climate, teacher qualifications) and students (e.g., particularly SES-related variables) might account for achievement disparities between school sectors. Furthermore, it is important to make distinctions between different types of private (and charter) schools in order to better understand the impact of particular organizational attributes on achievement. Of course, additional studies are needed that give attention to other subject areas and grade levels, and that includes longitudinal data with controls for prior achievement.
If such studies reveal that public schools students are, in fact, outscoring comparable students in private schools, this would suggest that widespread beliefs underlying much of the current private-ization reform agenda need to be reconsidered.
1 Belief in the superiority of private schools is based, in part, on past studies involving the 1980 High School and Beyond dataset. These studies generally found that private schools were more effective than public schools at boosting student achievement, even after accounting for socioeconomic differences between students in public and private schools (Chubb & Moe, 1990; Coleman & Hoffer, 1987; Coleman et al., 1982). Some research on this data pointed to particular aspects of private schools that appeared beneficial, such as the unified purpose among constituents more often evident in Catholic schools (Bryk et al., 1993). Prominent theorists understood achievement patterns to result from the organizational attributes of schools, since public schools are controlled by bureaucracies, and seek to please diverse and often conflicting constituencies, whereas private schools enjoy more operational autonomy (Chubb & Moe, 1990).
2 See, for instance, http://nces.ed.gov/nationsreportcard/mathematics/results2003/schooltype.asp
Alexander, K. L., & Pallas, A. M. (1985). School sector and cognitive performance: When is a little a little? Sociology of Education, 58, 115-128.
Benveniste, L., Carnoy, M., & Rothstein, R. (2003). All else equal: Are public and private schools different? New York: RoutledgeFalmer.
Bryk, A. S., Lee, V. E., & Holland, P. B. (1993). Catholic schools and the common good. Cambridge, MA: Harvard University Press.
Chubb, J. E., & Moe, T. M. (1990). Politics, markets, and Americas schools. Washington, D.C.: Brookings Institution Press.
Coleman, J. S., & Hoffer, T. (1987). Public and private high schools: The impact of communities. New York: Basic Books.
Coleman, J. S., Hoffer, T., & Kilgore, S. (1982). High school achievement: Public, Catholic, and private schools compared. New York: Basic Books.
Cookson, P. W. (1993). Assessing private school effects: Implications for school choice. In M. E. Rasell & R. Rothstein (Eds.), School choice: Examining the evidence (pp. 173-184). Washington, DC: Economic Policy Institute.
Cooper, B. S. (1988). The changing universe of us private schools. In T. James & H. M. Levin (Eds.), Comparing public and private schools: Institutions and organizations (Vol. 1, pp. 18-45). New York: Falmer Press.
Lubienski, C. (in press). Public schools in marketized environments: Shifting incentives and unintended consequences of competition-based educational reforms. American Journal of Education.
Lubienski, C., & Lubienski, S. T. (2005a). Re-examining a primary premise of market theory: An analysis of NAEP data on achievement in public and private schools (No. 102). New York: National Center for the Study of Privatization in Education.
Lubienski, S. T., & Lubienski, C. (2005b, May). A new look at public and private schools: Student background and mathematics achievement. Phi Delta Kappan, 86, 696-699.
Lubienski, S. T., Camburn, E., & Shelley, M. C. (2004). Reform-oriented mathematics instruction, achievement, and equity: Examinations of race and SES in 2000 main NAEP data. Report submitted to the National Center for Education Statistics.
McEwan, P. J. (2000). Private and public schooling in the southern cone: A comparative analysis of Argentina and Chile (Occasional Paper). New York: National Center for the Study of Privatization in Education.
McLaughlin, D. H., & Broughman, S. (1997). Private schools in the United States: A statistical profile, 1993-1994 (No. 97-459). Washington, D.C.: National Center for Education Statistics.
Somers, M.-A., McEwan, P. J., & Willms, J. D. (2004). How effective are private schools in Latin America? Comparative Education Review, 48(1), 48-69.