بازارهای کار معلم و خطرات استفاده از خویش شناسی برای برآورد اختلافات جبران در بخش دولتی
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|8756||2010||17 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Economics of Education Review, Volume 29, Issue 1, February 2010, Pages 1–17
Some scholars and policymakers who are concerned about the inequitable distribution of quality teachers suggest offering financial incentives for working in hard-to-staff schools. Previous studies have estimated compensating differentials using hedonic modeling, an approach potentially undermined by district-wide salary schedules and the lack of labor market competitiveness. To address this problem, we build hedonic wage models for both public and private schools using data from the 1999–2000 Schools and Staffing Survey and the 2000 Census. Empirical estimates suggest that both public and private schools compensate teachers for some working conditions, but there also appear to be differences between public and private schools in the magnitude of the compensating differentials, particularly for teaching low-income students.
A significant body of empirical research shows that compensation (including salary and benefits) and working conditions (such as neighborhood crime, transportation, job-related danger, or cleanliness of the workspace) both play important roles in individuals’ labor market decisions. This proposition holds true for the teacher labor market: compensation and working conditions explain much about who opts to teach, where they look to teach, and how long they remain in the profession. Among working conditions, teachers appear to care particularly about the type of students in the classroom, and evidence suggests that teachers prefer working with white, academically successful, and/or more affluent students. On average, teachers in high-need schools (e.g. those with low student achievement or high poverty levels) have lower levels of education, have attended less selective certifying institutions, and perform worse on teacher tests, characteristics thought by many to be associated with low teacher quality1 (Lankford, Loeb, & Wyckoff, 2002). When they have the opportunity, teachers who work at schools with lower performing, poorer, or minority, students tend to migrate to schools with higher achieving, more affluent, and fewer minority students (Hanushek et al., 2004, Hanushek and Rivkin, 2004 and Lankford et al., 2002). Of course, student characteristics may serve as a proxy for other job factors (such as safety, the quality of school leadership, or collegiality among teachers at the school), but nonetheless empirical evidence suggests that student characteristics are a key predictor of the type of teachers in a school and their propensity to remain there (Guarino et al., 2006 and Hanushek et al., 2004). In recognition of this phenomenon, there appears growing policy interest in using financial incentives to try to offset these patterns of teacher sorting (Goldhaber, 2006, Hoff, 2005, Jacobson, 2006 and Prince, 2003), so that students in high-need schools have a better chance of getting an experienced, credentialed teacher. For example, states such as California, Florida, and North Carolina have all recently offered salary bonuses to qualified teachers working in low-income schools.2 Some states, including Maryland and Texas, adjust funding formulas to reflect inter-district differences in working conditions beyond localities’ control that make it more or less difficult to attract teachers. For years, state formulas have adjusted district-level funding to reflect differential operating costs, such as the costs for real estate and transportation (CLCS, 2002, FDE, 2002, Rothstein and Smith, 1997, Thompson and Silvernail, 2001 and WDEA, 1999); adjusting for the different prices districts must pay to attract teachers represents the next logical step in creating an equitable playing field. But how much should be offered to entice a teacher to teach in a disadvantaged school? This question would not typically arise when assessing compensation in the private sector because private sector compensation is thought to adjust smoothly to reflect differences in employee skills and working conditions. All else equal, employees with highly sought-after skills earn higher wages, while, at the same time, these wages reflect non-pecuniary aspects of the job, such as location or safety. Basically, an employee serving under less desirable conditions receives a compensation premium, or a compensating differential, over otherwise identical employees. These wage adjustments happen through the competitive labor market as employees sort across different types of jobs and employers seek to hire employees that best fit their needs at the lowest possible rate. One way to calculate compensating salary differentials is to use hedonic modeling.3 Unfortunately, using a hedonic technique to estimate the magnitude of the differentials necessary to compensate a teacher for working in a less attractive position (e.g. in a crumbling school building) is not a straightforward proposition. This is particularly true in the case of public schooling because teacher salaries are set in the public domain where market forces are often distorted—the set of competitive pressures that influence compensation in the private sector may not exist. Thus, teachers may not be paid their precise ‘value’ in the public school system, and yet just such precision is assumed when utilizing a hedonic approach. The public school salary differentials calculated using this method may not accurately reflect compensation necessary for otherwise identical teachers in different job assignments and thus cannot precisely measure what would be required to attract a teacher from a more desirable to a less desirable teaching position. This paper explores the issue of compensating differentials in teaching by estimating hedonic models in public and private schools and then comparing these estimates to determine whether the setting of teacher salaries in the public sector may bias hedonic estimates of the relationship between working condition factors and teacher salaries.4 The paper is structured as follows: first, we examine the theory behind hedonics, introduce a conceptual model and provide a detailed discussion of why hedonic modeling may not work well in public education; then, we present our data and econometric models followed by our empirical results; finally, we examine the policy implications of our findings and offer some concluding thoughts.
نتیجه گیری انگلیسی
Both the No Child Left Behind Act (NCLB) and recent research highlight the importance of teacher quality and the teacher qualification gap that exists between rich and poor public schools. Some have suggested that this gap results from the salary structure used in the overwhelming majority of school districts. Specifically, within districts, schools with less-desirable working conditions have no means to adjust their compensation, and, as a result, the most qualified teachers choose the more affluent schools (Hanushek et al., 2004, Hanushek and Rivkin, 2004, Hill, 2006, Hill et al., 1997 and Lankford et al., 2002). While states and localities may wish to change this structure and offer teachers greater compensation to teach in disadvantaged schools, it is an open question as to the magnitude of the pay differences that will entice teachers to take more difficult positions. Our comparison of the estimated teacher-salary returns to individual teacher, school, and community characteristics in public and private sectors suggests that these variables are associated with different returns between the types of schools. In every model we tested, private schools appear to pay larger compensating differentials than did public schools; the differences range from $3770 (in the latent variable analysis) to $8355 (comparing high-poverty and low-poverty schools in the basic disaggregated model). Those differences are mitigated, but not eliminated, when we focus on the student characteristics at the least desirable school in the district. Of course, there is more than one explanation for why the returns might vary. Private schools are certainly not a perfect comparison group for public schools, and it is possible that we have not adequately accounted for the working conditions or teacher quality in either sector. It is also possible that private school teachers have preferences that are quite distinct from public school teachers. We argue that a more plausible explanation for the observed differences is that private schools’ teacher salaries are determined in a more competitive labor market than in public schools. Unfortunately, this suggests that one of the approaches to deriving such estimates – hedonic teacher salary models – does not work well for public school labor markets. So where does this leave policymakers wishing to get reasonable estimates of the ‘right’ salary differential? There are at least four options. The first is to more thoroughly pursue the strategy that we have employed here, and use estimates of compensating differentials from private schools as a measure of the pay differentials that might be used in the public sector. In practice, however, this would likely require new datasets, as the Schools and Staffing Survey has a relatively small sample of private schools in many states. Second, more efforts could be put into estimating salary differentials using teacher mobility as in Hanushek et al. (2004). The difficulty here is that if school systems continue to employ the single salary schedule, then estimated compensating differentials would be based solely on teachers that move from one district to another since there would be little or no within-district salary variation. A third possibility is to explore alternative estimation techniques, such as the one employed by Boyd et al. (2003). Their approach is novel, and very data- and computer-intensive, but will likely become better understood and more feasible in the future. More teacher and school data will be collected and statisticians will have a chance to better evaluate the methodology. Finally, the option likely to generate the best estimates of the values teachers place on various school attributes is to evaluate the direct effects of departing from a strict adherence to the single salary schedule. This would require varying salaries within and between schools and then observing the effects on the teacher labor market.