Association for Public Policy Analysis & Management

 ABSTRACT

 Over the last several decades, inequality in family incomes has increased substantially (Duncan and Murnane, 2014).  Inequality of educational outcomes between U.S. children from low- and higher-income families has also increased markedly during this period (Reardon, 2011; Bailey and Dynarski, 2011). There are several mechanisms through which the increase in family income inequality may have contributed causally to the increased inequality in educational outcomes.  First, income-based inequality in family spending on child enrichment has increased substantially (Kornrich and Furstenberg, 2013). Second, increasing residential segregation by income for families with school-aged children led to increased public school segregation by income (Owens, 2016; Altonji and Mansfield, 2011). In this paper we examine the evidence bearing on a potential third mechanism, namely, that very high-income families were increasingly likely to send their children to private schools.  This could have resulted in increasing inequality in educational outcomes by providing high-income families with a high quality education and by reducing financial and political support for public schools.

The research reported in this paper examines trends in private elementary and secondary school enrollments by family income and by private school type over the last four decades. We use several nationally representative data sources, including education supplements to the Current Population Survey for the years 1968-2013, and U.S. Department of Education Longitudinal Surveys including ECLS-K (two cohorts), ECLS-B, NELS88, ELS2002, HSLS2009, and the National Household Education Surveys (NHES). Our methodology follows that which Sean Reardon (2011) used in estimating trends in inequality of educational outcomes. First, we estimate the relationship between family income percentile and probability of enrolling children in a private school of a particular type in each dataset.  We then examine trends over time in the size of income-based gaps in enrollment in private schools. 

 Our initial work findings indicate that the gap in private elementary school enrollment between children from the 50^th and 95^th family income percentile increased by six percentage points between 1968 and 2013, as the percentage of middle-class families that sent their children to private schools declined. In the most recent two decades, the greatest increase in the gap in private school use is between students from the 95^th percentile of the family income distribution and students from the 90^th or lower percentile of the distribution.  The increase in the gap is concentrated in enrollment in non-Catholic religious private schools.  The paper we would like to present at the proposed APPAM panel will include more detailed evidence on income-based trends in enrollment in different types of private schools, and evidence of heterogeneity of trends by region, race/ethnicity, and city size.