Let’s get some of the basics out of the way.
Ninety-three school districts (approximately 22% of the state) collectively asked 103 referendum questions. Of those, 61 passed (59.2%). (By the way, for this analysis, we’re including results from both February and April as “Spring 2024” data.) Now to the good stuff…
Takeaway 1: This is the lowest percentage of referendums voters approved in over a decade.
The 2010-11 school year is a dividing line in the data (see graph below). In the years prior, voters approved about half of the referendums on ballots (50.2%). Between 2011-12 and 2022-23, voters approved three-quarters of the referendums on ballots (73.9%).
At 59.2%, this year is much closer to the norm in Wisconsin’s since revenue caps were put in place. In fact, over the past 30 years, voters approved an average of 59.7% of referendums. That makes this year almost spot-on normal relative to that timeframe.
What’s more, the trend is looking like it’s reversing downward (see black box below). Between 1993 and 2019, the percentage of referendums voters supported trended upward until a peak in the 2019-20 school year. Since then, the trend has ticked down.
Takeaway 2: This presidential primary was different than other presidential primaries.
The most apples-to-apples way to look at spring results is by what happens in November of the same year. In other words, we should compare this April to referendums held in presidential primaries in 2020, 2016, 2012, and so on. (This comparison also helps give us similar turnout measures).
You’d have to rewind to the 2008 presidential primary to see results as low as this year’s. Even that year, however, is a bit of an outlier (the Great Recession began in December 2007). You could argue that we should compare 2024 to 2000 (54%).
Takeaway 3: The lower percentage of approvals isn’t due to one kind of referendum. It’s across the board.
In April 2023, we could clearly tell that the overall decline in approvals was driven by non-recurring referendums. (Quick reminder here. Capital referendums allow a school district to take out debt, i.e., a bond, to pay for facility improvements. The other type of referendum is an operational referendum. A non-recurring operational referendum provides schools with more money but expires after a set number of years. Recurring referendums provide money that doesn’t expire.)
This time ‘round, we can’t point to one type. Voters approved the lowest percentage of facility referendums since 2014-15, and the rate was below what we’ve seen recently. The same can be said for non-recurring (NR) and recurring referendums (RR) as well. In addition, the trend reversal we talked about earlier is evident here too. Peak voter approvals of all these referendums occurred between five and ten years ago and have trended downward since.
Takeaway 4: Voters support fewer recurring referendums than non-recurring referendums.
Occasionally, we would hear from folks that recurring referendums (the “forever” kind of funding) pass at a higher rate. This is false.
A greater percentage of non-recurring (the “expiring” kind of funding) referendums have passed in 23 of the last 28 years.
Not only that, there’s nowhere near the same number of them. Over the last 30 years, there have been 985 non-recurring referendums and 594 recurring. There’s also a self-selection problem. Typically, districts don’t pursue a recurring referendum unless they feel confident voters will support it. Those that aren’t have already weeded themselves out.
Takeaway 5: Voters are back to favoring facility over operational referendums.
Zooming out to the entire revenue cap era, it’s a mixed bag. Some years, voters supported more operational referendums. Other years, it was facility referendums. Sometimes, the data swung back and forth even in consecutive years.
In 11 of the past 14 years, voters supported a greater proportion of operational referendums. Two of those three years in which facilities were preferred were actually the past two years. This is something that’s ripe for more analysis as more data comes in, but it definitely has our eye.
Takeaway 6: Referendum approvals were down regardless of locale.
Earlier, I showed that referendum approvals were down, regardless of type. That lesson holds if we slice and dice a different way: by locale (rural, suburban, and urban). Referendums are near or even above the average passage rate over the past 30 years. However, the rates are below (and trending down) relative to the last five to ten years across locale, just like they were for type of referendum.
A couple notes about this data point. First, the overwhelming majority of districts in Wisconsin are rural. Thus, the rural chart very closely mirrors the chart for all districts. Second, urban districts represent the other side of that spectrum. Wisconsin has very few urban school districts (especially after the NCES redid how they calculate locale about 20 years ago). In fact, only two urban districts held a referendum this spring (Milwaukee and Beloit). Thus, that data is harder to interpret because there are so few. (In 2021-22, 100% of urban referendums passed. Only one took place.)
Takeaway 7: The fails failed worse, and the approvals had more nailbiters.
For this takeaway, we have to shift our frame of reference for percentages. Looking only at failed referendums, the percentage of people who voted “Yes” is tied for the lowest it’s been over the last 10 years. During that time, failed referendums had an average margin of 43.2% “Yes” to 56.8% “No.” This spring, the failed referendums only managed to gather 42.1% voter support.
The same is true for approvals. Approved referendums typically gather about 60.5% support. This spring, it was down to 57.9%
While 1.1% and 2.6% declines respectively don’t sound like much at first blush, remember that having more than 100 referendums in any given school year is not uncommon. Thus, to move the data by that much, voter support really needed to decline.
These declines are largely true regardless of type. Approvals for capital, non-recurring, and recurring referendums all gathered a smaller majority than usual and were much closer to a 50-50 split. The “fails failing harder” was true for capital and non-recurring referendums. (Recurring wasn’t much different than expected.)
Takeaway 8: The amount of money requested was fairly normal.
This takeaway is news primarily because it’s not news. The amount of money districts requested across those 102 ballot questions this spring wasn’t much to write home about.
For a while, it felt like the amount of money school districts requested and the amount voters approved were setting new records all the time. This year, when totaling all three types of referendums, the money fell mostly in the middle of the 30-year pack – not the most, not the least (see the yellow and blue bars in the chart below).
(Note: Recurring referendums set a record this year, largely driven by Milwaukee voters’ approval. That one referendum accounted for 86% of all the recurring funding approved across the state.)
Yearning for more? Couldn’t possibly carry on with your day without knowing how April referendums in presidential election years compare to April referendums in gubernatorial years?
Football's done. Basketball's done. And I'm a terrible baseball fan. Let’s nerd out.
The School Perceptions Blog and Resource Center features the voices of our team members. This post was written by Rob DeMeuse, Research Director.