Spatial Explanations for Deferred Teacher Compensation: Unions and Competition for Teachers

• In Event: Strategic Staffing and Incentives

Abstract

Perspectives and Frameworks
To remain competitive in the labor market, school districts may need to strategically compensate teachers. Yet despite evidence supporting the effectiveness of frontloading compensation for recruiting and retaining teachers (e.g., Hendricks, 2014, 2015), districts often defer substantial compensation until late in teachers’ careers (Grissom & Strunk, 2012; Vigdor, 2008).

One explanation for this is that deferred compensation reflects teachers’ unions using their monopoly power in the labor market to extract rents from districts (e.g., Anzia & Moe, 2015; Lankford & Wyckoff, 1997). A second theory holds that deferred compensation persists where competition for teachers is limited (e.g., Rose & Sengupta, 2007).

Objectives
Studies testing these theories’ predictions are few, have mixed results, and are often hampered by data and methodological limitations. This motivates my two research questions. Do districts defer compensation due to (1) teachers’ union influence or (2) lack of competition for teachers?

Methods
I estimate spatial autoregressive models used in previous work to estimate how compensation in a district is associated with that of nearby districts (Greenbaum, 2002; Millimet & Rangaprasad, 2007; Wagner & Porter, 2000; Winters, 2011). I extend prior work in several ways. First, I use more detailed data on compensation. Second, I observe districts longitudinally, and can thus estimate models that account for district and year fixed effects. Third, I disentangle two potential causes of spillovers between districts (Goldhaber et al., 2014), estimating models that include separate spatial lag terms for compensation in (1) physically-nearby districts and (2) other districts belonging to the same union service center of the California Teachers Association.

Data
I take school district boundary shapefiles from the National Historical Geographic Information System (Manson et al., 2018). Other data come from the California Department of Education (CDE), including annual district reports on benefit packages and the salary offered at each step of each lane of the salary schedule. Other CDE data are used to control for various time-varying factors (e.g., revenue and student demographics). I restrict my sample to a strongly balanced panel of 496 school districts observed every year from 2008-2009 through 2018-2019.

Results
Results suggest that districts compete for teachers based on salary, though not by frontloading salaries. A 10% increase in mean salaries in nearby districts predicts 2% higher salaries at the same experience level in a district, and this is similar across all experience steps. I find stronger evidence that union influence results in deferred compensation. A 10% increase in salaries among other districts in the same union service center predicts a district’s salary at that step will be 2.5% higher for a first-year teacher, but up to 5.4% higher for salaries for teachers in year 30. Results for health and welfare or retirement benefits are mixed.

Significance
These results contribute to our understanding of teacher labor markets and administrative dynamics in districts. Policymakers may also gain insights about why school systems have not adopted proposed compensation reforms (e.g., Kellough & Lu, 1993; Springer, 2023) or how districts might respond to staff shortages.

Author

• Paul Bruno, University of Illinois at Urbana-Champaign