Every May and November some Indiana school districts ask their voters to approve property tax increases.
There are construction project referenda to pay for school buildings and tax levy referenda for general operating costs. This May, schools proposed 12 referenda. All 12 passed.
It was the first sweep since referenda became common in November 2008. Up to November 2011, only 42 percent of referenda passed. Since then, 76 percent have passed. Over the last four elections, 89 percent have passed — that’s 34 wins and 4 losses.
I can think of three possibilities. We’ve had a slow but steady expansion since the Great Recession. Perhaps expansion improves people’s ability and willingness to pay taxes for schools, so they’re more likely to vote yes.
School district leadership has more experience with referendum campaigns. Before 2008, there was a referendum every couple of years or so. Since then, there have been 176. Maybe practice makes perfect.
Then there’s a possible reason called self-selection. School districts that win once almost always win again. Maybe school districts try again if they think they can win. If there’s a greater share of past winners in the mix (and fewer “rookies”), then the number of winning referenda would increase.
Let’s look at some evidence. County unemployment rates are a measure of recession and expansion. We can match school districts with their counties and look at the unemployment rates when each referendum took place. Sure enough, since November 2008, in counties with unemployment rates 6 percent or greater, only 44 percent of referenda passed. In counties with unemployment rates less than 6 percent, 81 percent passed.
The problem with this test, though, is that unemployment rates were all greater than 6 percent from 2009 through May 2011, and they’ve all been below 6 percent since May 2016. Unemployment would appear to matter even if the trend toward passage had another cause.
So, let’s look at only those years when some unemployment rates were high, and others were low — from November 2011 to November 2016. Some counties recovered more quickly than others after the recession. Let’s also look at just those school districts that proposed their first referendum, to take out some of the effects of experience.
There were 45 referenda proposed by rookies during that period. Where the unemployment rate was greater than 6 percent, 59 percent passed. Where the unemployment rate was less than 6 percent, 65 percent. That’s not much of a difference. Maybe the economy doesn’t matter as much as we might think.
Let’s test experience. Districts that proposed their first referendum passed 51 percent. Districts that had tried before passed 84 percent. Maybe experience counts for a lot.
But again, there’s that trend. Naturally, in the first few years after November 2008, most of the referenda were first tries. Only later did we see many second attempts. Second tries would appear to be more successful as time passed, even if that wasn’t the cause.
So, let’s tighten the test. Let’s look at school districts that tried again after losing. And let’s pick only the second tries that were two years or less after the first, to lessen the effect of changing conditions in the district. There have been 16 such referenda, and 9 passed. That’s a lot of winning after starting out 0 for 16. Looks like the odds are a little better than 50-50 when a district tries again. Experience seems to count.
Then there’s self-selection. Since May 2016, 23 of 48 referenda have been proposed by school districts that have won referenda before — 22 of 23 won. But 20 of the 25 districts that had never tried before also won. There are more past winners in the mix, but rookies are winning more, too.
What can we conclude? There’s evidence that school referenda are winning more often because of school district experience, and also because of self-selection. Evidence for the improving economy is weak.
I’m an economist. I think the economy must matter. We won’t really get to test that possibility, though, until the next recession. Then, if passage rates go down, despite experience and self-selection, we could say that the economy matters. Let’s hope we don’t get to make that test for a long, long time.
Larry DeBoer is professor of agricultural economics at Purdue University. Send comments to [email protected].