Is there “white flight” into private schools? New evidence from High School and Beyond

https://doi.org/10.1016/j.econedurev.2008.07.004Get rights and content

Abstract

In the U.S., white parents may choose to send their children to private schools in response to the local concentration of minority schoolchildren, commonly referred to as “white flight”. This paper contributes to the existing literature by providing new evidence on white flight from the data set High School and Beyond (HSB). I find that a one-percentage-point increase in the county (metropolitan area, or MA) level minority share of school-age population increases the private schooling probability of white students by 0.2 (0.27) percentage points. White flight appears to be more sensitive to the concentration of black schoolchildren than to any other minority concentration. White families also appear to respond more strongly to the concentration of poor minorities than to non-poor minorities. Results also suggest substantial variation in the rate of white flight across different areas of the country. Finally, when minority shares are measured at the county level, there is some evidence of increasing rate of white flight from the local concentration of black schoolchildren as the black share increases.

Introduction

One important factor that may influence the private school attendance decisions of white students in the U.S. and has strong policy implications is the possible aversion of white families to the concentration of minority schoolchildren in local public schools. White parents may choose to send their children to private schools in response to minority concentration, commonly referred to as “white flight”. The growing recognition of white flight brings about the fears that its existence may threaten the long-run efficacy of integration policies and lead to virtual resegregation of schools in the U.S. Clotfelter (1976) examines the role of school racial composition in the demand for private schooling by whites. Using a sample of U.S. metropolitan areas (MAs) in 1960 and 1970, he finds that an increase in the proportion of minorities in public schools of 10% is associated with increases in the proportion of whites in private schools of 1–8%. Evidence on white flight can also be found in Conlon and Kimenyi (1991), Lankford and Wyckoff (1992), Lankford, Lee, and Wyckoff (1995), Clotfelter (2001) and Fairlie and Resch (2002).1

This paper contributes to the existing literature by providing new evidence on white flight from the data set High School and Beyond (HSB). The HSB sample allows me to examine white flight for high school sophomores and seniors in 1980. A recent study on white flight by Fairlie and Resch (2002) focuses on high school sophomores in 1990 sampled in the National Education Longitudinal Study (NELS), another survey conducted by the National Center for Education Statistics (NCES), a decade after the HSB survey. Similar to the NELS sample, the HSB sample has an advantage over many national data sets because it includes information on many very useful individual level characteristics, such as religious affiliation and cognitive test score which can be used as a proxy for cognitive ability. Because the NCES conducted the HSB and NELS surveys in a consistent way and intended to allow researchers to make comparisons of U.S. high school students over a decade, it is quite interesting to study the variability in responses of white families to minority concentration over time. An increase in the rate of white flight over the decade may indicate deepened aversion of white parents to minority concentration, whereas a fall may suggest the opposite.

Using the HSB sample, this paper supports the white flight hypothesis by finding that a one-percentage-point increase in the county (MA) level minority share of school-age population increases the white private schooling probability by 0.2 (0.27) percentage points. White flight appears to be more sensitive to the concentration of black schoolchildren than to any other minority concentration. For example, a one-standard-deviation increase in the county (MA) level black share of school-age population increases the white private schooling probability by 7.2 (8.1) percentage points. White families also appear to respond more strongly to the concentration of poor minorities than to non-poor minorities.

Some studies in the private schooling literature use national data sets, such as the first sample in Clotfelter (1976), Lankford et al. (1995), Figlio and Stone (2001), Clotfelter (2001) and Fairlie and Resch (2002), but there are also many studies restricting their attention on a specific state. For example, the second sample in Clotfelter (1976) focuses on Mississippi around 1970, Conlon and Kimenyi (1991) focus on Mississippi in 1980, Lankford and Wyckoff (1992) on New York state in 1980, and Buddin, Cordes, and Kirby (1998) on California in 1990. Using California data, Buddin et al. (1998) find that private schooling rates of all students are rather insensitive to the percentage of minorities in public schools. In contrast, Lankford and Wyckoff (1992) find from New York state data that parents are more likely to send their children to private elementary schools when black students make up a larger proportion of the public school population, although they find the opposite effect for private high school enrollments. Therefore, the effect of minority concentration on private schooling may vary over time or across areas.

The hypothesis is that white parents in some areas of the country may have less deep aversion to minority concentration than in other areas. It is possible that the variation in the rate of white flight across studies may reflect in part the geographical heterogeneity in responses of white families to minority concentration. To capture such heterogeneity, I estimate an additional set of probit regressions, allowing for minority share interacted with region dummies. The results indicate substantial variability in the rate of white flight across different regions. In particular, white flight appears less prominent in the West region but is more pronounced in the Northeast/South/Midwest regions.

Clotfelter, 1976, Clotfelter, 2001 pays special attention to the question of whether there is a threshold or “tipping” point in white flight from public to private schools in response to minority concentration. As a robustness check, I allow for nonlinearity in minority shares of school-age population for various ethnicity groups. Because Fairlie and Resch (2002) are concerned about the arbitrariness in creating the thresholds, I use two different forms of nonlinearity for minority shares, a fourth order polynomial and a semiparametric specification. Results using different nonlinear specifications produce similar results and suggest that, when minority shares are measured at the county level, there is some evidence of increasing rate of white flight from black concentration as the black share increases. The next section describes the HSB sample. Section 3 discusses empirical results and examines white flight, white flight from different minority groups and the geographical heterogeneity and nonlinearity in white flight. The final section concludes.

Section snippets

Data

High School and Beyond is a national survey of U.S. high school sophomores and seniors in 1980 and is administered by the National Center for Education Statistics. To study whether white flight exists, it is natural to restrict the sample to include only white, non-Hispanic students. To check the robustness of my results, I employ two different measures of school and community characteristics, those at the county level and those at the metropolitan area level. I report descriptive statistics of

Empirical results

Before looking into the issue of white flight, it is of interest to compare the private schooling rate of different ethnic groups, as this may shed some light on white flight. According to the HSB sample, 11% of white students attend private schools in 1980, but only 5.8% of minority students attend private schools. The private schooling rates of black, American Indian, Asian and Hispanic students are 3.4%, 4.2%, 10% and 8.7%, respectively. Therefore, in response to minority concentration,

Conclusion

This paper contributes to the existing literature by providing new evidence on white flight from High School and Beyond. I find that a one-percentage-point increase in the county (MA) level minority share of school-age population increases the white private schooling probability by 0.2 (0.27) percentage points. White flight appears to be more sensitive to the concentration of black schoolchildren than to any other minority concentration. For example, a one-standard-deviation increase in the

References (20)

There are more references available in the full text version of this article.

Cited by (22)

  • Private schools: Choice and effects

    2020, The Economics of Education: A Comprehensive Overview
  • Refugees and ‘native flight’ from public to private schools

    2019, Economics Letters
    Citation Excerpt :

    For example, Betts and Fairlie (2003) find that for every 4 immigrant children enrolled in public schools, 1 native child switches to private education. Li (2009) and Farre et al. (2018) document that a 1 percentage-point increase in the share of immigrant/minority students generates 0.2 and 1.2 percentage-point increase in private school attendance probability for natives in the U.S. and Spain, respectively. I also report positive and statistically significant “native flight” estimates, but the magnitudes are much smaller than those reported in the literature.

  • Public subsidies to private schools do make a difference for achievement in mathematics: Longitudinal evidence from Canada

    2011, Economics of Education Review
    Citation Excerpt :

    Second, although some evidence suggests that a more competitive schooling market where parents have a larger set of choices can improve the academic performance of children (Hoxby, 2002), some critics argue that school choice may worsen outcomes of students left behind in public schools because of a cream-skimming effect, assuming that better peers have a positive impact on achievement (Dee & Fu, 2004; Walsh, 2009). Third, are schooling effects associated to families moving in better neighborhoods with desirable schools causal or due to sorting (Levine & Painter, 2008; Li, 2009; Liu, Mroz, & van der Klaauw, 2010). This paper presents new evidence that attendance in non-religious subsidized private schools has a causal and positive impact on achievement in mathematics.

View all citing articles on Scopus
View full text