by Robert Kleine, Vice President and Senior Economist

After 25 years of futility and 12 ballot proposals, in 1994 Michigan finally revamped the way it funds public K–12 education. This Advisor presents an early appraisal of the new school finance system.


In March 1994 Michigan voters said yes to a 2-cent sales tax increase that replaces about two-thirds of the local school property tax (school property taxes were reduced by about 50 percent after accounting for the 6-mill state education tax). There were two keys to success. First, the previous July, the legislature had repealed all school property taxes without approving any replacement revenue. Second, the legislature and the governor agreed to give the voters no chance to vote down school finance reform.

The legislature approved the package in December 1993 and gave the voters only one choice: whether to replace the property tax with the income tax or the sales tax. In every previous vote, voters had been able to chose between a new system or the current one, and in every case they had opted for the old, however much they disliked it. Exhibit 1 presents a comparison of the old and the new school finance systems. The two key provisions are replacement of most school property taxes with the sales and other taxes and adoption of a minimum “foundation” grant of $5,000 for all school districts—that is, the state gives all districts at least $5,000 for each pupil.

Goals of Reform

The new system has five major objectives.

  • Reduce property taxes, which were about 35 percent above the national average, and, after slowing in the early and mid-1980s, had increased at an annual rate of 7.2 percent from 1986 to 1993 (inflation averaged 3.7 percent).  Proposal A reduced total property taxes by about 33 percent (26 percent after accounting for loss of state property tax credits).  For homeowners, the cut is 42 percent (32 percent after the loss of state tax credits); for the roughly 30 percent of taxpayers who itemize for federal tax purposes, the reduction is smaller.  For businesses, the property tax cut is about 13 percent.
  • Reduce reliance on local millage votes to provide school funding.
  • Increase the state share of total K–12 revenue, which in 1993 was only 32.1 percent, ranking Michigan 48th among the 50 states (see Exhibit 2).  This meant that two-thirds of K–12 revenue had to be generated locally, and this was a major reason for the heavy property tax burden and the unequal distribution of resources among districts.  Under the new system, the state picks up about 80 percent of the total cost, moving Michigan up to 2nd among the states.
  • Assure all local districts a minimum level of per pupil revenue, in part to permit them to institute education reforms, including establishing a core academic curriculum and achieving school accreditation.
  • Reduce interdistrict disparities in per pupil revenue.  In 1994 the highest-spending district was disbursing $10,356 per pupil, while the figure for the lowest-spending was only $3,277 per pupil; the ratio was 3.16 to 1.  The highest-spending 20 percent of districts (roughly 105 districts) disbursed an average of $6,542 and the lowest-spending 20 percent $3,980 per pupil; this ratio was 1.64 to 1.

Problems with the New System

Proposal A has been successful in achieving these objectives but not without some problems, some anticipated, some not.

  • The new state School Aid Fund was not expected to grow as fast as did the pool of money generated formerly by local property taxes, and, indeed, it has not.   From FY 1994–95 to FY 1997–98, the foundation grant increased from $5,000 to $5,462, an annual rate of 3 percent, only slightly above the estimated inflation rate (2.8 percent); local property tax assessments—but for being capped by Proposal A謡ould have increased at an annual rate (1997–98 is estimated) of 7 percent.
  • During an economic turndown, the new funding system is expected to grow more slowly and be less stable than the old system.  During the 20 years from 1972 to 1992, property tax assessments grew at an annual rate of 7.1 percent, while the replacement-revenue sources grew at a rate of 6.6 percent.  During the most recent downturn ( from 1989 to 1992), however, property taxes increased at an annual rate of 8 percent, but the replacement revenues increased only 3.6 percent.  For schools, the slower-growth problem will be ameliorated somewhat by the increased earmarking of the state income tax to the School Aid Fund (increased from 14.4 percent to 23 percent in 1996), but the fact remains that the new revenue sources, particularly the property transfer tax and cigarette tax, are not as stable as the property tax (see Exhibit 3).
  • The new system restricts local revenue-raising ability.  The provision to allow intermediate school districts (ISDs) to ask voters to approve an ISD-wide millage of up to 3 mills may not be practical because such approval will be too seldom achieved.
  • The new state formula depends heavily on enrollment, and no provision is made to help districts where enrollment is declining.  The old system didn’t either, but under it, the district had the option of asking voters for additional millage to offset revenue lost due to an enrollment drop.
  • One of the key provisions of Proposal A is to roll many of the categorical funds—e.g., transportation, retirement, FICA—into the foundation grant.   This has created two problems for the districts.  First, it increased legislative scrutiny of the categorical grants, as evidenced by the recent cuts in appropriations for adult education.  Second, eliminating the categoricals for teacher retirement and FICA is increasing the financial pressure on most districts, because these expenditures are increasing faster than revenue.  Some observers, however, view this provision of the reform as positive, because now districts must bear the full cost of the wage contracts they negotiate and cannot pass fringe-benefit costs to the state.
  • The new plan fails to address two longstanding shortcomings of the old system.  First, the formula does not take into account the regional cost differences of providing education:  $5,000 per pupil buys a lot more education in the Upper Peninsula than in southeast Michigan (for example, in 1994 salaries in the Oakland County ISD were 44 percent higher than in the eastern U.P. ISD).  Second, there is no provision for equalizing debt millage, which places low-property-value districts at a great disadvantage (for example, Birmingham can raise about $270 per mill per pupil, while Detroit can raise only $16).

Results to Date

In many ways, Proposal A has been a major success.

  • First, it has largely settled an issue—school finance reform—that had festered for 25 years and consumed the time and energy of legislators, state officials, school officials, association representatives, and many others.  Resolving this issue allows more attention to be focused on such other state problems as how to deliver high-quality K–12 education.
  • Second, reducing the property tax and capping assessment increases has significantly defused taxpayer anger about property taxes.  A side benefit is that taxpayers are less dissatisfied with state and local government.  The two-cent increase in sales tax has generated almost no negative reaction:  Most taxpayers barely notice the increase in the tax, except when buying large ticket items such as motor vehicles.  Few taxpayers know how much sales tax they pay annually, but almost everyone knows the amount of his/her property tax and the reduction in the latter is not going unnoticed.
  • Third, Proposal A has reduced the gap between the highest- and lowest-spending districts; the ratio of highest to lowest is estimated at roughly 2 to 1 for FY 1997–98 (narrowed from more than 3 to 1 in 1993–94, the last year under the old system).  The highest-spending districts include Bloomfield Hills, Birmingham, Grosse Pointe, Troy, Ann Arbor, and Midland; the lowest generally comprise rural areas and small towns.  The discussion that follows will be easier to understand if the reader keeps in mind that under Proposal A the wealthier—highest-spending—districts are receiving the smallest increases in state funding, and the poorer—lowest-spending—districts are receiving the largest.

As shown in Exhibit 4, Proposal A has significantly boosted resources for the lowest-spending districts. The first column shows the average cumulative growth rate since FY 1993–94, and the fifth quintile (20 percent of all districts—the roughly 105 lowest-spending) has received an almost 16 percent hike; their per pupil amount, from all income sources, still is considerably less than in wealthier districts, but the gap is narrowing. There is a correlation between the increase and enrollment: The average current enrollment for the fifth quintile is only 1,484, while for the first quintile (the wealthiest districts), it is 4,375. There also is a high correlation between the increase in revenue and property value (SEV) per pupil: The fifth quintile’s SEV per pupil for homestead property is $38,765; the first’s is $97,510 (the state average is $62,678).

For nonhomestead property, the fifth quintile’s SEV per pupil is only $35,787, but the first’s is $96,434 (the state average is $53,040). In the 45 highest-spending districts—now subject to the slowest growth rate in resources—there is a heavy concentration of wealth: The average SEV per pupil in these districts is $108,283 for homestead property and $142, 498 for nonhomestead property.

These data raise some interesting questions. First, since 1993–94, 89 school districts have received average annual revenue increases exceeding 5 percent, which is well above the average inflation rate (2.8 percent) for the period. In 1993–94, the poorest districts were spending less than $3,000 per pupil, but now they are spending a minimum of $4,816 per pupil (foundation grant only). What have these districts done with this money and have they been able to spend it efficiently.1 Second, 80 districts (the highest-spending) have received annual revenue increases of 2 percent or less. Their 1996–97 average per pupil revenue (foundation grant plus the categorical for at-risk pupils) is $7,243, 26.5 percent above the state average of $5,726. However, if their resources continue to increase at a rate below that of inflation, a time will come when they will have to reduce the quality of the education they offer. In many of these districts voters likely would approve additional local millage if asked. Why should their voters be denied this option?

The critical moment will come for the new finance system when the Michigan economy experiences another recession. In the past 30 years there have been five national recessions, and, on average, Michigan has been hit harder than the nation each time. There is some reason to believe that the state’s economy is less vulnerable to economic downturns than in the past and also that these downturns will be less frequent. However, it is unrealistic to expect that recessions are a thing of the past, and we can be almost certain that economic downturns will be harder on the revenue sources for the new School Aid Fund than they had been on local property tax revenue.

Needed Reforms

We do not believe that the new school finance system needs major change, but some tinkering could improve its fairness and flexibility.

The most obvious weakness of the new school finance system is that local districts have no ability to raise operating revenue on their own. The ISD-wide millage option should be eliminated and each district allowed to ask its voters for up to 2 or 3 mills. One reason for the current ISD millage option is to provide equalization of resources among member districts, and this still could be accomplished if the amount an individual district receives from a local millage is based on the average ISD property value per pupil. Revenue in excess of this amount, generated by districts with above-average property values, could go into a state or ISD equalization fund to be used to supplement millages levied by districts with below-average property value.

Proposal A provides adequate resources for districts with increasing enrollment, but where enrollment is declining, districts will be hard pressed to keep up with rising costs—and under Proposal A they are prohibited from asking local voters for additional millage. An enrollment dip can be caused by demographic or economic factors as well as by losing students to charter schools or, as allowed under schools of choice, to other districts. Probably, there should be little sympathy for a district that is losing pupils because of student/parent dissatisfaction with the education offered by the district, but there should be a “safety net” for districts that lose students because of local economic conditions or to specialized charter schools that can offer programs not offered in the local district.

Proposal A has reduced the disparities among districts for operating funds, but it has done nothing to equalize capital spending. A district having property value of $200,000 per pupil can raise $200 per pupil per mill, but a district with property value of only $50,000 per pupil can raise only $50 per pupil. The capital-spending inequality is critical because of districts’ need to upgrade current buildings or construct new ones, to meet technology and/or enrollment demands and also to compete with charter schools and traditional schools trying to take advantage of the schools-of-choice plan. One solution is to use for capital funds the same approach we suggest for operating mills—that is, base the amount each district receives from levying debt mills on the average ISD property value per pupil.

We also believe there is need for some adjustment for the differential cost of education across the state. As mentioned above, it is much more expensive to educate kids in southeast Michigan than in the Upper Peninsula. The failure to make such an adjustment erodes the equalization of resources that Proposal A was designed to achieve. The easiest approach would be to develop a cost index for each ISD (or for larger regions), using teacher salaries as a proxy for the difference in costs. (Other goods and services, such as utilities and supplies could be included in the index, but this would complicate the process without much improving the outcome.) The index then would be used to adjust the foundation grant and possibly the categorical grants. For example, if the average state grant were $5,000, a district with costs 10 percent below the state average would receive $4,500 per pupil and a district with costs 10 percent above would receive $5,500. Although it is possible that such an index could encourage districts to be less heedful of holding teacher salaries down, the size of the state School Aid Fund is going to be restricted by available revenue, and in lean years even an indexed grant could be insufficient to cover a district’s costs if they have been allowed to rise too high.


Many school districts, because their experience to date is that the foundation grant has increased at about the rate of inflation, are concerned that the new school finance system will not provide them with sufficient resources to keep up with the increased cost of providing a K–12 education. The problem is exacerbated for districts receiving increases at a rate less than that of inflation. As shown in Exhibit 4, under Proposal A, the average annual increase for the quintile comprising the highest-spending (per pupil) districts has been only 1.7 percent. The increase for the quintile comprising the lowest-spenders has been 5 percent, and this, of course, results in increased equalization of resources, which is one of the main purposes of Proposal A. Unfortunately, districts with revenue now growing more slowly than inflation have only very limited means to raise more revenue. Under the old system they could have asked the voters for more millage, and in many districts—particularly the more affluent—likely would have been successful, but that alternative is closed to them now. Their only option under the new system is to get ISD-wide approval for an increase of up to 3 mills. In most ISDs, such approval is unlikely.

Does this mean that many Michigan school districts face a permanent financial squeeze? Probably not. First, part of the reason that the increases to date in the foundation grant have been only modest is higher-than-projected enrollment increases, but we expect enrollment to level off soon and then begin to decline. As shown in Exhibit 5, the Michigan school-age population is expected to increase only 0.5 percent from 2000 to 2005 and then decline 2.5 percent from 2005 to 2010; the drop will be due largely to a reduction in the number of women in the childbearing age group.

Second, although the Michigan economy is strong, revenue growth is slowing as the recovery matures. The current recovery has reached 78 months, the third-longest since World War II. If Proposal A had been enacted two years earlier, the initial increases in the foundation grant would have been well above the rate of inflation. Exhibit 6 presents our projections for School Aid Fund revenue through FY 2004–05. Because of slower enrollment growth after 2000, we expect the increase in the foundation grant to average 4.7 percent from 2000 to 2005; the average increase from 1994–95 to 1997–98 was 3 percent.1 There will, of course, be periodic slowdowns in the economy, when state revenue will grow more slowly than currently. It appears, however, because of structural (permanent) changes in the economy, slowdowns will be less frequent and less severe than in the past.


So far, we have to give Proposal A pretty good marks. It needs some refinement, but that was to be expected. Final judgment cannot be made until we have several more years experience, including a downtown in the economy. In the meantime, we must keep sight of its objectives—to reduce property taxes, lessen reliance on local millage as a means to fund schools, increase state funding of K–12 education, guarantee every district a minimum level of state support, and diminish the revenue gap between rich and poor districts—and judge the new system on how well it meets them.

1There is some evidence that some of this money has been used to support larger pay increases for teachers, but the numbers are not dramatic. For the fifth quintile (averaging a 15.7 percent hike in their resources), the average teaching salary has increased 7.7 percent (adjusted for a small number of aberrations) from 1993–94 to 1995–96 (1996–97 data are not yet available), compared with only a 4.9-percent average salary increase for the first quintile (averaging a revenue increase of 5.3 percent).

090597-1.gif (32997 bytes)

090597-2.gif (18940 bytes)

090597-3.gif (9874 bytes)

090597-4.gif (25559 bytes)

090597-5.gif (12035 bytes)

090597-6.gif (16275 bytes)

Copyright © 1997