Karakaplan, Mustafa (2012-10). The Effects of Multi-Dimensional Competition on Education Market Outcomes. Doctoral Dissertation.
In this dissertation, I analyze the effects of competition in education markets. In my first essay, I analyze the effects of different concentration measures on school personnel salaries. I find evidence that principals have more bargaining power over their salaries than teachers in Washington that through rent-sharing, principals start getting positive returns from increasing concentration at lower levels of concentration than that of teachers. Moreover, I present that the pattern of teacher salaries versus concentration in Washington is similar to that in Texas, but the inflection point in Washington is at substantially lower levels of concentration-a finding which can be attributed to Washington's being a union state versus Texas's being a right-to-work-state. In my second essay, I examine the effects of various measures of competition on school district cost inefficiency in a stochastic frontier framework. My results show that cost frontier is U-shaped in Texas with large positive returns to the scale over a relatively big range and mild diseconomies of scale over an extended range. In addition, I find that school district cost inefficiency increases significantly when market concentration increases. Furthermore, I present the competitive effect/scale effect trade-off through a couple of simulation exercises. The findings from both of my studies show that the effects of competition are barely sensitive to measuring the competition with different sets of relevant competitors. On the other hand, sensitivity of the effects of competition to using different definitions of the education markets is significant. Yet, the range of these estimated effects is relatively small, and the sign and the significance of the effect of competition generally do not change when a meaningful definition of education markets is employed to measure concentration. Furthermore, I present that the concentration measures employed in my essays are endogenous. I control for the endogeneity with several instrumental variables including degrees of lagged educational outputs in the neighboring schools, lagged education market characteristics, and counts of streams. My results imply that the hypothesized effects of competition may be underestimated due to the endogeneity. While the plausibility of competitive effect's being underestimated bolsters the importance of the competitive effects I find, it also strengthens my criticism of using uni-dimensional concentration indices as indicators of competition in the education markets.