Academic journal article Journal of Economics and Economic Education Research

Tiebout Forces, Public School Competition and Educational Quality

Academic journal article Journal of Economics and Economic Education Research

Tiebout Forces, Public School Competition and Educational Quality

Article excerpt

(ProQuest: ... denotes formulae omitted.)

INTRODUCTION

Households in the US can choose the specific jurisdiction and school district by choice of i residence, which is the Tiebout sorting in the modern education system. Theoretical model of local 2 educational markets argue that, increasing choices among schools can improve the social efficiency of allocation of resources for two main reasons. First, students are sorted into learning envi-ronments and peers that conform to their own ability and preferences, which might facilitate learning; second, choice intensifies the competitive mechanism that rewards the most productive school, at the same time decreasing the principal-agency problem between households and school board. Therefore, schools have an incentive to increase efficiency within a more competitive market.

Policy reforms on improving choices in the local educational market has generated a substantial body of empirical literature on the productivity effect of competition. However, according to a review by Belfield and Levin (2002), the generality and scale of the effect of competition on the quality of school is still unclear. A number of previous studies (e.g., Hoxby (2000a), Belfield and Levin (2002), Hanushek and Rivkin (2003), and Bayer and McMillan (20i0)) discussed the challenges in estimating the relationship of competition and academic outcome. All of them agreed that the most substantial problem is simultaneity: the question of how much more productive schools are when marginally increasing household's school choices is a question on the supply side of schooling with demand side held constant. However, only the equilibrium of demand and supply is observed. Therefore, simply linking the measure of competition to educational outcomes in a hedonic regression function is suffered from endogeneity bias. This is mostly due to the measures of competition reflect both the initial administrative structure of school districts and the variation of school qualities within an MSA itself is endogenously determined by the location choices of families.

Therefore, an important econometric challenge is to find instrumental variables that implicitly reflect the variations in the initial supply of school districts, isolating the variations caused by the demand side shifts. In an in fluential paper by Hoxby (2000a), the author proposed 3 geographic "streams" which reflects the number of natural school district boundaries as instrument, and found large competition effects. This instrument is soon challenged by Rothstein (2007), which proves that the estimated effect is extremely sensitive to the way that the "streams" are coded, and if an alternative, more accurate measure is used, the difference between Hoxby's IV estimation and OLS is minimized.In this paper, I assess the problem of endogeneity of competition in two steps. First, I rank order all the MSAs in the US from those least likely affected by demand side Tiebout forces to those most likely affected by this force, conditional on a constant 4 "initial" supply of local jurisdictions in 1990. This is done by utilizing a nationwide panel school profile with geographic details obtained from Common Core of Data (National Center of Educational Studies) dated from 1987 to 2005; I then examine whether the estimated choice effect through both value-added and non-value-added function with time fixed effects is different for each subgroup of MSAs. I find significantly positive choice effects for the MSA group that are least affected by demand side forces. In the second step, I re-estimate the choice effects for all the MSAs using an instrument approach, which is a Herfindahl Index that is based on the share of total geographic distance of each school district to the center of MSA in year 1990. This instrumental variable measures the initial supply of school district within an education market, thus is unlikely to be correlated with the test score outcome. This step confirms the choice effect estimation obtained from the first step. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.