The late great George Carlin had a routine about a weather forecaster.

“The forecast for tonight: dark. Continued dark throughout the night, with scattered light in the morning.” Some predictions are easy.

Predicting the base rate of farmland used to be easy too. The base rate is the starting point for setting the assessed value of farmland for property taxes. The state’s Department of Local Government Finance recalculates it every year with a capitalization formula. They divide measures of farm income by a rate of return. The base rate for taxes this year is $1,960 per acre.

Here’s how easy it was. The base rate for taxes in 2015 was calculated in 2014, averaging data from 2006 through 2011. There was a four-year lag between the most recent numbers and the base rate used for taxes.

I could take numbers that were already in the books, feed them through the DLGF’s set formula, and come up with a really accurate prediction. In January 2012, I predicted the base rate for 2014 at $1,760. Nailed it. In January 2013, I predicted $2,050 for 2015. Right again. It was no big deal, like “scattered light in the morning.”

Then, in 2016 the General Assembly changed the formula to eliminate the four-year lag. Now, the most recent numbers are used. The base rate for taxes in 2017 was calculated in 2016, with numbers up through 2015. That’s a two-year lag.

This is a good idea. The Indiana Supreme Court says that assessments for property taxes should be based on “objective measures of property wealth.” Surely the most recent “objective measures” should be used.

But in 2016 I wanted to predict the base rate for taxes in 2018. Now, I needed numbers on prices and yields for corn and soybeans for 2016. That year wasn’t over. So I needed predictions of those prices and yields. My co-author Tamara Ogle and I waited until late summer, then used the best projections of all those numbers that we could find. The predicted base rate for taxes in 2018 came out to $1,770.

In March, the DLGF announced the base rate for 2018. It was $1,850.

I was off by $80! It turned out that the actual prices for corn and soybeans were higher than the predictions I’d used, so the DLGF’s actual base rate calculation was higher. That produced an error of 4 percent — I’m tempted to say “just 4 percent” — but after years of hitting the number on the nose, it looks bad. You can find the DLGF’s base rate memo and documentation at in.gov/dlgf/7016.htm.

How about a prediction for 2019? Our plan is to wait until summer to get a better read on prices, then publish an update in the Purdue Agricultural Economics Report in the fall.

But let’s take a look at the numbers now. The base rate is a six-year rolling average, ignoring the highest number of the six. In 2019, the numbers for 2017 will be included, and the numbers for 2011 will be dropped.

The capitalized value — income divided by rate of return — for 2011 was $3,699. That’s when corn was $5.85 and soybeans were $12.04 per bushel. Prices were much lower in 2016, $3.69 for corn and $9.50 for beans. The capitalized value for 2016 was $2,416.

Prices in 2017 will be more like those from 2016 than those from 2011. That means $3,699 will be replaced with a number like $2,400.

That should drop the base rate calculation to something like $1,600, a 14 percent decline from $1,850 in 2018.

The average capitalized value in the base rate formula from 2012 through 2014 is more than $4,000. If commodity prices remain low, small numbers will replace big numbers in the base rate calculation for the next few years. The base rate will decline, and so will farmland property taxes.

The days of precise-to-the-dollar predictions of the base rate are over. There will be forecast errors. But we can say this, with near certainty. The forecast for the base rate: down. Continued down through the early 2020s, with scattered ups and downs after that.

*Larry DeBoer is professor of agricultural economics at Purdue University. Send comments to awoods@tribtown.com.*