The inexact science of projecting growth

Nathaniel Clayville has an idea of how much school enrollment will grow next year.

So does Tim Hill.

And, says Hill, both guesses will probably be wrong.

It will be up to Idaho legislators to decide which estimate is less wrong. If this sounds like a mere academic exercise, it isn’t. This represents a key difference between Gov. Butch Otter’s K-12 budget proposal and state superintendent Sherri Ybarra’s request. And it puts several millions dollars of K-12 funding on the line.

Breaking down the numbers

As of Wednesday, there were enough students in Idaho public schools to fill 14,677.5 “support units.” A support unit isn’t exactly a classroom, but the two are reasonably close.

When the 2014 Legislature set this year’s K-12 budget, lawmakers paid enough to cover 14,627 support units. That’s not too far off — it’s a variance of only .3 percent — until you realize that the state allocates $86,000 per support unit. Then it comes to a $4.3 million gap, one that was awaiting the 2015 Legislature when it convened last month.

Since a small discrepancy quickly adds up to real numbers, lawmakers are taking a close look at the competing growth projections.

Clayville, an economist with Otter’s Division of Financial Management, is projecting 14,693 support units for next fall.

Hill, a State Department of Education veteran who works on school finance issues for Ybarra, is going with a projection of 14,781.

Using that $86,000 multiplier, the difference comes down to $7.6 million.

How did they get there?

On Wednesday, Hill and Clayville walked through the math with Senate Education Committee.

Hill bases his estimates on past numbers, a three-year rolling average of support unit statistics.

Clayville wanted to come up with a more far-reaching approach. Instead of using the three-year rolling average, which has been off by more than 100 support units three times since 2004-05, he wanted to incorporate other variables into the equation. The new model also tracks school-age population by three subsets: 5- to 9-year-olds, 10- to 14-year-olds and 15- to 19-year-olds.

And according to his projections, the 5- to 9-year-old student cohort is likely to drop through 2017-18.

Uncertainties

By his own account, Clayville is still working through his new model. In 2012-13, the number of support units grew faster than his modeling would have projected, and he’s still not sure why.

For that reason, he has added 65 support units into his 2015-16 model — which, when left alone, projects only modest growth.

Hill says his projection strategy is simple: aim for the middle, and minimize the margin of error. And that makes him nervous about ditching an old model for a new one. “I’m just nervous about being handed a parachute that hasn’t been tested,” he told the Joint Finance-Appropriations Committee in late January.

Nonetheless, Hill doesn’t see this as a competition. Hill told Senate Education that he has a friendly wager with Clayville: a lunch is riding on the actual 2015-16 figure.

For lawmakers, tasked with setting the 2015-16 budget based in part on growth projections, the stakes are a little bit higher.

Republish this article on your website