On April 2, the Milwaukee Public Schools will ask Milwaukee voters to allow the district to exceed state revenue limits by up to $252 million over four years to support its schools and programs. Here, we give our impartial take on the referendum, based on more than a decade of annual reviews of MPS’ budget and finances. Our intent is to frame for Milwaukee voters how they might consider this critical matter and to better inform citizens throughout Wisconsin given the increased frequency of school referenda throughout the state.
Four years after approving a 2020 referendum that allowed the Milwaukee Public Schools (MPS) to exceed state-imposed revenue limits by $87 million, Milwaukee voters again will be asked to support increasing property taxes for the state’s largest school district this spring. The combined increase in property tax levy and state aid resulting from the referendum would be almost triple the previous request.
MPS leaders are not alone in turning to referenda. As the Forum has documented in several reports, districts across Wisconsin have asked voters for permission to exceed state revenue limits with increasing frequency over the past decade. The increase in referenda has occurred as state-imposed revenue limits, which cap what districts can receive on a per pupil basis from the combination of their main form of state aid and property taxes, have failed to keep pace with inflation.
In April 2023, 83 referenda to increase school spending on capital projects or operations appeared on ballots across the state; this April there will be 91 (on top of 10 in February). It is also important to note that MPS could use a referendum (as it did in 2020) to net additional state aid in future years, offsetting some of the cost for Milwaukee taxpayers.
Still, the amount of money requested by the district, the referendum’s occurrence four years after the 2020 referendum and in the wake of MPS’ receipt of nearly $800 million in one-time federal pandemic relief aid, and the fact that Milwaukee property taxpayers as a whole are among the poorest in the state and have just begun paying a new 2% city sales tax, raise several questions for voters. Here, we address the central fiscal issues as they weigh this important decision.
On What basis can the district argue that it needs more money?
Overall Funding
Using data from the state Department of Public Instruction (DPI), the Forum has documented annually how growth in the dollars received by MPS under the revenue limit and from a related form of state per pupil aid has lagged inflation, leaving the district with fewer overall dollars available to educate its students. As shown in Figure 1, in 2004 MPS received $793.5 million in funding subject to revenue limits (that is, state general aid and the applicable part of its local levy), or $1.31 billion when adjusted for inflation. That was 45.5% more than the $903.2 million projected for MPS in the current 2024 school year even after including a state per pupil aid allocation that falls outside the revenue limit.
The revenue limit dollars comprise more than 80% of MPS’ roughly $1 billion budget for school operations, which means the decline of these dollars in real terms has meant fewer resources for schools. When phased in over four years, the extra $252 million from the April 2024 referendum (representing approximately a third of the district’s revenue limit) would reverse the decline, but it would still leave the district short of its 2004 purchasing power.
The drop in MPS’ revenue limit dollars is attributed in part to decisions by state lawmakers to reduce the revenue limit for all districts in 2012 and freeze or allow only modest annual increases in some of the following years. Yet the single biggest reason for MPS’ revenue decline has been the district’s loss of enrollment due to demographic shifts, competition from independent charter and state-funded voucher schools, and other factors. As shown in Figure 2, MPS first began to suffer sizable enrollment declines soon after the turn of the century. In 2004, MPS reported 98,663 students under the full-time equivalent enrollment count used to calculate its revenue limit, which includes students at the district’s charter schools. In 2024, this enrollment count for MPS stands at 67,577, a reduction of 31.5%.
The decline in inflation-adjusted dollars shown in Figure 1 would have been far higher if not for the decision by voters in 2020 to allow the district to exceed the state-imposed cap by $87 million by 2024 and every year thereafter. Also, after state officials opted not to increase revenue limits at all in the 2021-23 state budget, they raised them by $325 per pupil in both 2024 and 2025 as part of the current two-year state budget. Per the state’s Legislative Fiscal Bureau (LFB), these increases are still below what would have been necessary to match the rate of inflation over the four-year period.
Funding Per Student
Another way to consider MPS’ revenue picture is to analyze the numbers on a per pupil basis, and doing so reveals a less alarming picture. The passage of the referendum in 2020 has brought core per pupil funding in the district to $13,366 in 2024, which is slightly higher than the inflation-adjusted $13,319 it received in 2004 (see Figure 3). These totals include revenue limit funds as well as the form of state categorical aid known as per pupil state aid. MPS receives $742 per student in that form of aid, as do all other districts in the state.
Core per pupil funding also includes MPS charter students in schools operated by outside entities, who count toward districts’ revenue limits. MPS then sends to its charter schools a state-set per pupil payment. That payment has historically been less than the per pupil revenue limit amount, although the current state budget increased the payment by $2,121 per full-time student in 2024, adding to the district’s costs. A new referendum, if passed, would not directly affect the per pupil payments for charter schools or voucher programs.
Additional context can be gleaned from considering how MPS’ per pupil funding compares to other Wisconsin districts. MPS’ per pupil revenues under state caps are currently third highest among the state’s 10 largest districts and in the middle of the pack when compared to 10 large districts in the metro Milwaukee region, with whom MPS competes for students and staff.
As shown in Figure 4, adoption of the 2024 referendum would place MPS at or near the top of the two current comparison groups after it is fully implemented in four years. We also show how successful referenda would impact other districts already pursuing them, but the figure does not take into account any referenda that have yet to be announced. For example, the Madison Metropolitan School District has been actively considering a referendum for November 2024.
It should be recognized that student need is an important consideration when comparing funding levels, since it costs more to serve students entering school with greater needs. Among the 20 peers listed in Figure 4, MPS serves the highest percentage of students from economically disadvantaged backgrounds – 82.5% in 2023, more than 20 percentage points higher than the next closest district displayed. Other measures of need are less stark but still notable: MPS serves the highest percentage of students with disabilities (19.1% in 2023) out of the 20 districts and the highest percentage of English Learners (14.5%) out of the 10 identified metro area peers. MPS’ English Learner proportion ranks fourth highest compared to the largest districts in Wisconsin.
Other Special Aid
Additional “categorical aids” outside of core funding are available from the federal and state government to address some of these needs. MPS receives relatively large amounts of federal aid because of the high poverty rates in Milwaukee as well as rates of students with disabilities. The district received $2,476 in federal aid per student in 2020 (not including pandemic relief aid) for a total of $184.9 million, according to the National Center for Education Statistics. In line with Milwaukee’s comparatively high rate of poverty, that was nearly double the average of $1,337 per student among the 120 largest districts nationally and was fifth highest among those districts. A portion of these dollars serve students in private schools outside MPS.
MPS also receives categorical aids from the state for special education. Districts are required by state and federal law to meet the unique needs of students with disabilities and in doing so incur added costs. Yet despite some progress in recent years, the state is covering a smaller share of those eligible excess costs today than it was a generation ago.
In 1994, the state’s main form of special education aid covered 44.6% of the additional costs incurred by MPS and other districts and charter schools. Since then, the state has nearly doubled its funding for special education aid, yet statewide special education costs have nearly tripled over the period, driving down the reimbursement rate for MPS to 31.5% in 2024, according to data from DPI and the LFB.
The diminished reimbursement rates typically require districts to turn to unrestricted operating dollars to fund the required special education services. In 2024, MPS has estimated unreimbursed special education costs of $124.2 million – a sizable amount though down from an inflation-adjusted $188.5 million in 2014 and $151.0 million in 2004.
Overall, this analysis shows how both sides in the MPS referendum debate can make valid points about the district’s fiscal capacity. Referendum supporters can argue that the total dollars available to educate the district’s students have lagged inflation by a wide margin despite the successful 2020 referendum. This reality is particularly burdensome given recent high rates of inflation and the district’s unique challenges tied to the poverty and special needs of its students.
On the other hand, opponents can cite per pupil figures to show that MPS’ revenues in recent years match inflation and that its revenue amounts are comparable to other large state and area districts. They may argue further that per pupil figures are more relevant than total revenues given that they account for MPS’ falling enrollment, which should have allowed the district to accommodate its decline in total revenues at least somewhat by reducing expenditures on facilities, overhead, and staff.
what has MPS done to decrease its workforce and footprint in line with its reduced student population?
It is reasonable to ask whether MPS has taken appropriate steps to reduce spending by “right sizing” its workforce and infrastructure in light of declining enrollment. Any effort to objectively answer that question, however, must acknowledge that right sizing is a subjective term. There is no question that a school district that has lost 31.5% of its students since 2004 should have had an ability to reduce its workforce and close some school buildings. However, it is also critical to point out that for any district – and for MPS in particular – certain fixed costs for administration and other centralized services cannot be reduced proportionally to a loss in students, as the demand for such supports is not reduced in an equivalent fashion.
Moreover, there are various difficult considerations involved with school closures – not the least of which are the impacts on students and families – and reductions to the teacher workforce must be carefully calibrated against objectives related to class size and the district’s ability to maintain appropriate offerings in areas like art, music, physical education, and career and technical education (CTE).
With regard to the district’s physical footprint, MPS grew its number of school buildings to a high of 178 in 2008 and then gradually reduced them to 150 as of its 2022 financial report. That was only a 3.2% decrease from three decades ago despite a much larger drop in enrollment over that period. Furthermore, as we noted in our annual analysis of MPS’ budget last May, the district has not updated its long-range facilities plan since 2018. It has since released a request for proposals for a consultant to help conduct such an update – a welcome move but arguably one that should have occurred before the upcoming referendum so that any savings yielded could be factored into the district’s fiscal forecast. The district also completed a space utilization study in 2023.
With regard to its workforce, the district has made a reduction in its number of instructional staff – a designation that includes teachers, principals, guidance and other counselors, librarians, and supervisors. According to the district’s annual financial reports, its number of instructional staff fell from 7,266 in 2004 to 4,594 in 2022 (the most recent year available), a decrease of 36.8%, which is more than the enrollment decline in the same time period.
The district’s number of budgeted positions, however, has not been falling as quickly in recent years for roles such as teachers, which is one reason why the state of MPS’ workforce is difficult to assess. Per data received from MPS, the district had 639 vacant positions as of the end of February, including 292 teacher vacancies, and the 2024 budget contains a savings of nearly $70 million from assuming the bulk of those positions will remain vacant. This comes after the 2023 budget contained a vacancy savings of more than $50 million.
As we pointed out in our May budget brief, large vacancy adjustments are “somewhat understandable given the nature of today’s labor market, which makes it exceedingly difficult to fill vacant positions on a timely basis.” However, we also noted that “for MPS leaders to convincingly argue to state leaders that they need more funding despite declining enrollment, it will be imperative to show how they are adjusting permanent staffing levels to respond to their reduced student body.” That same statement arguably could be made today with respect to MPS’ arguments to city voters about the referendum.
Finally, it is worth noting that one positive outcome related to MPS’ declining enrollment is an improved pupil-teacher ratio, at least when it comes to budgeted positions. The district’s code of “teacher” applies to more positions than solely classroom teachers – for example, speech pathologists and library media specialists – meaning that a declining ratio may not immediate translate into smaller class sizes even if all positions were filled. Still, as shown in Figure 5, that ratio had fallen to below 14:1 by 2020, a low enough figure to suggest there may indeed be room to permanently eliminate some vacant positions without sacrificing the critical goal of maintaining appropriate numbers of teachers.
What about the extra money from the 2020 referendum – was that supposed to solve the revenue problem?
MPS officials certainly hoped that the extra $87 million they would receive annually once the 2020 referendum was fully phased in would mostly address structural budget gaps caused by years of stagnant revenues, while also allowing them to fulfill several educational objectives they cited to voters in advancing the referendum. Those included, most prominently, steps to enhance MPS’ ability to attract and retain qualified educators, as well as increases in the number of library, art, music, physical education, early childhood education, and CTE teachers.
What district leaders could not have predicted was the surge in the rate of inflation, which both reduced the purchasing power of the $87 million and led them to provide matching cost-of-living salary adjustments for staff (4.7% in 2023 and 8.0% in 2024). As a result, referendum monies were at least indirectly used in part just to try to keep district compensation in sync with the labor market. They also could not have predicted that revenue limits would be frozen by the state for both the 2022 and 2023 school years, thus further reducing the financial benefits obtained from the referendum, nor that teacher turnover would spike, further exacerbating their labor challenges.
On the other hand, while district leaders have budgeted a sizable portion of referendum monies for hundreds of new staff positions in areas that are consistent with their promises to voters, they have faced hiring challenges for these positions as they have for others. According to MPS officials, 265 of the 358 (74%) referendum-related positions included in the 2024 budget in areas like art, music, physical education (collectively referred to as AMP), library, early childhood, and professional support staff have been filled.
Even more relevant to this year’s referendum is that these 93 vacant positions – which contribute to MPS’ overall huge number of vacancies – have helped to produce sizable operating budget surpluses in each of the past three years. It is likely that a substantial portion of the 2021 surplus was linked to pandemic-related cost savings in areas like transportation as many schools continued to operate virtually; 2022 and 2023 surpluses were more closely correlated with vacancies.
According to information prepared by MPS’ Office of Accountability and Efficiency, the historically large surpluses in 2021 and 2022 allowed MPS to use a combined $74 million over budgeted amounts to pay down retiree health care liabilities and transfer $60 million to its long-term capital improvement trust fund. While the books are not yet closed on 2023, the district will have sufficient unspent funds from that year to add $60 million more to its long-term capital fund.
These are fiscally responsible moves that enhance MPS’ capacity to meet its challenges in future years. They also suggest the district has not yet spent a substantial amount of the dollars generated by the 2020 referendum in each of the past two years, thus raising the question of why it is requesting such a sizable additional inflow beginning in 2025. We will address that question below, but the impact of federal pandemic relief aid must first be addressed before we can do so.
On top of the 2020 referendum, MPS received nearly $800 million in federal pandemic relief aid – how was that spent and how does that factor into the referendum?
The federal pandemic relief aid packages for schools – primarily known as Elementary and Secondary School Emergency Relief (ESSER) – could be used for a wide variety of district needs. The majority of relief funds flowed through the federal Title I formula, which is based on students’ family income. As the most populous district in Wisconsin and a district serving a high proportion of students from low-income households, MPS received $797 million to spend or obligate by September 30, 2024.
MPS has reported its planned and actual spending of these funds via both line items and broad categories. Figure 6 shows allocations across these categories for the two largest funding packages, ESSER II and ESSER III, which together comprise 92% of the district’s total federal pandemic relief aid.
Facilities spending tops the list of allocations, accounting for 32.8% of MPS’ ESSER II and III funds. The largest line items in this category include $86.9 million for construction at various schools, $25.1 million for capital projects related to air quality, and $20.9 million for testing and adjusting air handling systems. Accelerating student learning is the second-largest category at 24.1% of the combined ESSER II and III allocation. The largest items in this area are $15.3 million for instructional technology; $15.1 million for science curriculum adoption and support materials; and $15.0 million on AMP teachers and library media specialists.
This range of spending illustrates MPS’ overall approach to pandemic relief aid funding: commendably funneling a sizable portion of the one-time money toward one-time expenditures like facilities upgrades, pandemic-specific needs, and time-limited instructional investments. At the same time, especially in later budget revisions, MPS drew upon the temporary federal dollars to fund some positions that the district hoped to make permanent, thus setting up a budget hole after the pandemic aid runs out. Other school districts have reported taking similar steps, especially in light of frozen revenue limits in 2022 and 2023.
And indeed, MPS has recommended to its board that $88.1 million of items funded through ESSER in 2024 be added to the district’s normal operating budget for 2025 (see Figure 6). In addition to the $15.0 million for AMP teachers, the largest proposed allocations are $14.4 million for textbook adoptions and $11.3 million for school support teachers previously funded by federal Title I dollars. MPS officials have emphasized the value of the latter positions, which are housed in schools and designed to provide instructional coaching, assessment, and other professional development for teaching staff, but presumably must weigh them against the district’s high levels of unfilled classroom teacher positions.
what’s the bottom line for MPS’ financial future and the students and families it serves?
MPS officials say they face a $200 million budget shortfall in 2025. It is not surprising to us that a sizable deficit is projected. In our budget brief last May, while acknowledging that many vacant positions may not need to be filled and the uncertainty of future state decisions on revenue limits, we warned of “the potential for significant spending cuts in future budgets” in light of the expiration of ESSER dollars, the impacts of inflation, and the continued negative effects of MPS’ declining enrollment on its major revenue streams.
The $200 million shortfall now being projected by MPS is much higher than we expected, however. Figure 7 shows the district’s tabulation of that projection. The largest component – at $45.1 million – is a reduction in the amount of savings MPS expects from leaving vacant positions unfilled. Other major contributors are increases in the cost of employee benefits, the need to replace federal pandemic grants currently being used for various positions with other funds, plans to invest in new textbooks, and a projected 4.12% cost-of-living adjustment for all staff based on the previous year’s rate of inflation.
At this early point in the budget process, it is justifiable to make conservative estimates on next year’s gap. However, some context is needed for key assumptions behind the projection, which is central to the dollar amount being requested in the referendum.
First, the $45.1 million reduction in the vacancy adjustment is ambitious. A reduction of that magnitude means the district would need to fill about 500 of its over 600 vacant positions (assuming an average of $90,000 in salaries and benefits for each position). That not only would be a very tall task in light of the challenges of today’s labor market, but it also raises the question of whether the district really needs to fill that many of its vacant positions given that it has mostly lived without them for several successive years and that its student population is lower than it was pre-pandemic.
The $32.2 million to replace the federal ESSER dollars funding staff positions in the 2024 budget also could be better explained. As mentioned above, the district’s ESSER reporting suggests that $11.3 million of that amount is to maintain funding for school support teachers, and another $15.0 million would support art, music, and health and physical education teachers.
These two items account for 39% of the projected $200 million gap. It is also likely that a sizable portion of the $38.8 million increase in fringe benefits is linked to added health care and other benefit costs associated with filling vacant positions or sustaining ESSER-funded positions. Filling and sustaining these positions may be defensible in many respects, but it is difficult to assess their need without the benefit of an overall staffing plan for the district that takes into account its reduced enrollment and the realities of today’s labor market.
These unknowns also complicate efforts to assess the impact on MPS staff, students, and families if the referendum does not pass. It seems clear that the referendum dollars would make a difference for Milwaukee students’ experience. Less clear is whether a failed referendum would mean $200 million of cuts, forgoing $200 million of hoped-for service expansions, the need for immediate steps to right size positions and facilities, or a combination of all of the above.
Finally, what would this mean for city of Milwaukee property taxpayers?
Although the referendum allows MPS to exceed revenue limits by up to $252 million, taxpayers would not be responsible for the full amount. Figures provided by MPS per its financial advisor (Baird) show that the projected impact on property taxpayers would be greatest in year one but then would stabilize. After rising $97.4 million in 2025, MPS’ overall levy would increase by an estimated $10 million each year in 2026, 2027, and 2028. MPS says the added levy in 2025 would produce about a $432 increase in property taxes for the owner of a $200,000 property. The higher levy amount in 2025 would be impacted not only by the referendum, but also by the full phase-out of a longstanding provision that in 2024 provided more than $32 million in additional state general school aids solely to MPS.
These numbers assume that a substantial portion of the referendum increase would be offset by 1) a loss in enrollment that would reduce the MPS revenue limit in the absence of a referendum; and 2) an increase in state aid in future years resulting from the referendum’s boost in MPS’ levy and overall spending. Some decrease in enrollment seems all but certain for MPS going forward, although it is not simple to predict.
The projected increase in state aid is also reasonable. For every $10 million of additional spending by MPS in 2023 due to its 2020 referendum, we estimate it received an additional $3.7 million in state aid in 2024. This has helped to offset the impact of the additional spending on local taxpayers and would do so again if the 2024 referendum succeeds (though, as referenda are approved by MPS and other districts, state aid amounts for other communities also could be affected).
It is important to recognize that, on an inflation-adjusted basis, the gross MPS property tax levy has fallen over the past generation because of the impacts of enrollment loss on revenue limits, the state shifting more of the costs of voucher schools off the MPS levy, state aid increases, and a variety of other state actions to hold down K-12 property taxes across Wisconsin. As shown in Figure 8, the MPS levy on December 1993 tax bills totaled $507 million after adjusting for inflation. The December 2023 levy for MPS totaled $320.2 million, a decrease of 36.9%. If the referendum is approved, then the levy would increase to its highest inflation-adjusted level since 2015 but would still be far below the amounts levied in the early-1990s.
MPS’ levy did increase in 2020 following the successful referendum, but the additional spending boosted state aid that – along with inflation and declining enrollment – helped to lower the levy again in the following years. That same dynamic would be in play this time. In fact, MPS officials say that, with the $252 million in extra revenue authority that would be in place by 2028, the levy would still increase by less than $130 million because of an increase in state aid and other factors.
One additional property tax-related factor that has received little public attention is the impact of the district’s community service fund (also known as Fund 80). This is a special fund for non-classroom activities that serve not just MPS students, but the entire community – like adult education, recreation, and day care services. These community service programs are run by school districts, which can levy property taxes to support them outside of the regular district levies that are restricted by revenue limits.
Last fall, while considering the district’s final budget for the 2024 school year, the MPS Board of School Directors voted to increase the district’s Fund 80 levy by $77.7 million – from $34.7 million to $112.4 million. This increase by MPS was enough on its own to raise the combined Fund 80 levy by all school districts in Wisconsin by 52.6%.
The action by the MPS board emanated, at least in part, from the unexpected discovery in the fall that the levy available under revenue limits to support school operations would be far lower than originally anticipated in light of state budget actions related to the caps and school aids. Some or all of that reduction in the available levy for schools could have been used to lower the amount charged to taxpayers, but MPS officials instead saw this as an opportunity to reinvest in community recreational facilities.
Consequently, they increased the levy for Fund 80 by $77.7 million to set the district’s total levy at $320 million, almost exactly matching the previous year’s amount. According to MPS officials, over $40 million of the Fund 80 increase will be spent on a new recreation community center and aquatic facility at the former Browning School and Browning Playfield location that has been in the planning stages since 2018. The district had previously considered issuing debt for the project but now will pay cash to avoid interest costs. The remaining amount will be used to address a backlog of repair needs for other MPS-owned recreational facilities.
We note this decision to point out its relationship to the referendum. Because MPS’ calculation of the referendum’s property tax levy impacts assumes that Fund 80 levy amounts will return to the typical $30 to $35 million range in future years, the $97 million increase in the total MPS levy for 2025 is a net amount that assumes a $75 million to $80 million reduction in the Fund 80 levy next year.
Conclusion
As we have discussed throughout this report, compelling arguments can be made both to justify and question the April referendum. On the pro side, it is understandable why MPS is asking voters for more money in light of its unique challenges, its workforce pressures and needs, and the fact that the amount of money otherwise generated under state revenue limits would continue to fall if the referendum is not approved.
At the same time, legitimate questions can be raised about how well district leaders have set the stage for such a big ask from voters. In particular, those leaders may wish to justify the decision to seek such a large sum by sharing a formal plan for continuing to adjust facilities and staffing levels in accordance with reduced enrollment.
Indeed, a huge unanswered question is whether the referendum amount is predicated on plans to continue expanding educational offerings to students according to the promises of the 2020 referendum; whether it is the estimated amount needed to maintain the status quo; or whether it already accounts for right sizing plans. Voters may decide the request is justified for any of those scenarios, but local residents and businesses should know whether the extra amount they are being asked to pay might be a pathway toward improved offerings and outcomes, as opposed to simply another effort to stave off programmatic reductions and financial instability.
In the end, voters have a lot to consider. We hope this report helps better frame the decision they must make and perhaps invites MPS leaders to answer critical questions before April 2.