In 1994 the majority of local school operational funds in Michigan shifted from a local property tax system to a state-level per-pupil allowance, and since 1996 districts have been allowed to enroll non-resident students. Together, these policies have created a supply and demand for student enrollment, and this market was further fueled by stagnant per-pupil funding allowances and a declining K-12 student population across the state over the past 10 years. Today, over 80% of Michigan districts allow open enrollment, in part to maintain or increase state funding for their local schools. Although only approximately 5% of students participate statewide, in 2008-09 there were over 25 districts whose non-resident enrollment was 15% or greater, including 11 districts with 30% or greater non-resident enrollment (largely, but not exclusively, in urban areas) (Addonizio & Kearney, 2012).

Although operational funding was largely centralized at the state level, funding for capital improvements and in some circumstances additional operating funds, are still generated through local millages. Therefore, local property taxes continue to provide an important stream of financing for local school systems. In this study, we will identify the impact that non-resident student enrollment has on local voting behavior with regards to school funding, under the assumption that residents are less likely to vote to tax themselves to educate non-resident students.  Specifically, we will test the following hypothesis: As the percentage of non-resident students increases in a district, the less likely it is that voters of that district will pass a school millage. To test this hypothesis, we will utilize publicly available data to estimate a logistic regression:

ln{P[M_t = 1] / 1- P[M_t = 1]} = θ₀ + θ₁R_t + θ₂M_t-5 + θ₃F_t + θ'₄P_t + θ'₅P_t-10 + θ'₆S_t + θ'₇S_t-10 + θ₈E_t + θ₉E_t-10 + e          (1).

This model predicts the odds that a school millage vote passed in a district between 2005-06 and 2010-11 (M_t) as a function of the proportion of students in the districts who were non-resident (R_t), controlling for whether a school millage had been passed in the past five years (M_t-5), state per-pupil allowance (F_t), current and past population characteristics of district residents (e.g., property wealth and percent school aged) (P_t and P_t-10), student characteristics (e.g., race and poverty status) (S_t and S_t-10), and total student enrollment (E_t and E_t-10) (error term assumed to follow a Bernoulli distribution).  

As state and federal policies continue to push “choice” as a model for improving educational outcomes for students, it is important to fully investigate the impact that these policies have on communities’ commitment to funding public education. Although more of school financing is being centralized at the state level, a significant amount of budgetary discretion continues to lie in the hands of local voters and property tax payers. Therefore, this work has important implications for evaluating the long term effects of school choice policies and the adequacy of community engagement and funding of public school systems.