Check, correct pattern of census estimates

If the census estimates for North Dakota consistently are inaccurate, then the way the government makes the estimates should change.

That’s the bottom line from the latest estimates, which show Grand Forks to be losing population at a time of brisk home sales, stepped-up tax collections and low hotel and apartment vacancy rates.

“Just drive around, and you’ll see that this is a city that’s growing,” said Hal Gershman, president of the Grand Forks City Council

The president of The Chamber of Grand Forks and East Grand Forks agreed. The official estimates “just don’t track with reality,” said Barry Wilfahrt.

The two officials could be dismissed as “boosters wearing blinders” if it weren’t for two things. First, census estimates matter: Among other functions, they’re used by businesses in making expansion and relocation decisions. That means they threaten to become self-fulfilling prophecies, if businesses steer clear of an apparently “shrinking” area and so bring about a decline.

Second and much more important, the estimates have a decades-long history of undercounting Grand Forks’ growth.

And not just Grand Forks: As late as July 2005, the estimates had Minot’s population down a full 4 percent from 2000. Likewise, Williston, N.D., supposedly shrank 2.5 percent over that time.

Meanwhile, “oil prices are at record highs, and the contractors can’t build new homes fast enough,” the Minot Daily News noted at the time.

“Williams County (N.D.) sales tax collections are up nearly 24 percent between 2000 and 2005, and Williston employment has grown from 10,360 jobs to 11,568 jobs.”

As a Herald editorial about the episode concluded, “Some decline.”

In fact, the underestimates are part of a pattern that stretches back not years but decades, said Earl Haugen, executive director of the Grand Forks-East Grand Forks Metro Planning Organization.

“Right after the Census, they start being very bearish about growth in Grand Forks,” Haugen said. “During the 1980s, they were very bearish on our growth, but when the 1990 Census came in there was a huge growth from their 1989 estimate to the actual count. The same thing occurred in the 1990s.”

And the 2000s saw it happen as well, so that “there was a huge gap between their 2009 estimate and the 2010 actual count,” Haugen said.

City officials have complained, and the Herald and other newspapers have editorialized about the low estimates before. Enough: North Dakota state government should take an official look. Are the complaints justified? Has there been, in fact, a pattern of underestimation that stretches back 30 years?

The claims can be easily checked. If they’re valid, then the model being used to make estimates should be changed, plain and simple.

The National Weather Service routinely updates its forecasting models, especially when those models get it wrong. And if their population forecasts are proving consistently inaccurate, then North Dakota’s census takers should do the same.

Tags: opinion, editorials, census