The idea of consolidating school districts is rising to the top of the list of preferred ways to minimize the costs of Michigan's public education system. At first glance, it appears to be a commonsense reform that will create economies of scale and lower the state's education expenditures. Reflecting on this proposal, however, reveals it's not quick or easy, and holds little hope of reducing spending.

It's true that Michigan has more school districts — 551 — than most states, but has this contributed to higher per-pupil costs? If it did, states with fewer school districts would have lower per-pupil costs, but that isn't the case. Data from the National Center for Education Statistics shows that there is no correlation between the number of school districts in a state and its per-pupil expenditures. Many factors contribute to the costs of public education, but the number of school districts in a state is not one of them.

A 2007 study by Andrew Coulson, an adjunct scholar with the Mackinac Center, solidifies this point. Controlling for a number of inputs that do impact per-pupil spending, Coulson found little potential for savings through district consolidation. There is an ideal district size in terms of cost effectiveness, but creating more similarly sized districts is problematic and costly. In fact, the study suggests that the state could create more savings by breaking up larger districts than by consolidating smaller ones — exactly the opposite vision of consolidation optimists.

We don't have to rely on just Coulson's model. History tells a similar story. The very same arguments in support of consolidation heard today were used in the 1960s to back the creation of Intermediate School Districts. ISDs have done nothing to help control spending and are no longer necessary. Since their creation, inflation-adjusted per-pupil spending in Michigan has almost quadrupled. Which raises the question: why doesn't consolidation in the public sector work like it does in the private sector?

Incentives drive human action and the public sector operates under a different set of incentives than does the private sector. Public choice theory explains how incentives in the public sector encourage individuals to maximize expenditures. Public employees in charge of budgets work to multiply their expenditures, because they can then more easily win the favor of those who depend on these budgets for their livelihood and enhance their reputation for "getting things done." They are, essentially, merely acting in their own self-interest. The conventional public schools system creates the same incentives for school leaders.

In spite of this, there are real ways schools can cut costs. School districts could reduce spending by sharing some services, but they must privatize those services and allow competition to drive down costs to see any significant benefit. Hundreds of school districts already privatize noninstructional services and save millions of dollars.

Since about 80 percent of a school district's costs are tied up in labor, restructuring school employee compensation holds the greatest chance of reducing costs. Health insurance and pension contributions are the primary catalysts for these rising costs. NCES reports that these expenses increased 15 percent from 2003 to 2008. Schools should start reducing these costs through Health Savings Accounts or defined-contribution retirement plans.

Also, expanding the number of charter schools in the state would drive down costs since they operate at an overall lower expense than conventional districts. Data from the state's Center for Educational Performance and Information shows that charter schools' operating expenses were almost $1,500 less per-pupil than conventional school districts in 2007-08.

These are ideas school districts and the state should consider when finally dealing with the handwriting that's been on the wall for years — schools must control their spending.

#####

Mike Van Beek is director of education policy at the Mackinac Center for Public Policy, a research and educational institute headquartered in Midland, Mich. Permission to reprint in whole or in part is hereby granted, provided that the author and the Center are properly cited.