After the housing bubble collapse of 2007-08, I thought of forming Economists Anonymous, a self-help group of forecasting addicts. I couldn’t get anyone to join me.
Forecasts and projections are in demand. Someone will do them, and they are best if done by people who are informed, trained and cautious. Caution is important because statements about the future are taken seriously; lives and fortunes may depend on them.
Forecasts and projections are compared to what we believe we know about the present. Annual population estimates by the U.S. Bureau of the Census are as close as we can get to how many people live where. These estimates, however, are not perfect. Measuring projections against the estimates involves using two imperfect sets of numbers.
Matt Kinghorn at the Indiana Business Research Center does population projections. He meets all the criteria for a good forecaster and he doesn’t give up. Whether his results are on or off target, he keeps going, testing, improving. He knows people in business and government need guidelines for their investments.
The projections he produced in 2012, based on history up to 2010, indicated Indiana would have a 2015 population of 6.68 million persons. The 2015 estimates from the Census Bureau came in at 6.62 million or 58,000 persons (0.9 percent) shy of the projections.
In the world of state population projections, that is on target; not a bulls-eye, but very close.
The projections exceeded the estimates in 74 of Indiana’s 92 counties; they fell short of the estimates in only18 counties.
Which is better: to exceed or fall short? If projections exceed our best guess (the estimates) of reality, we may have over-built, invested too much. But are we better off having more growth than expected and then finding we have not invested enough, ending up short on essentials?
Over a five-year span, when the differences in absolute numbers are small and the percentage differences are slight, it probably does not matter much. We should be pleased that the projections were within one-half of one percent (0.5 percent) of the estimates in 18 counties and within two percent of the estimates in another 42 counties.
Yet users of Indiana population projections will be asking questions about the remaining third of the counties. And they will ask: Why did the projections come in 14,400 and 8,800 too high for Lake and Hamilton counties respectively? Why, in Marion County, were the projections 11,000 too low. Does the Great Recession explain everything?
According to the projections, Indiana will have a population of 6.85 million in 2020. But, if we continue on the growth path of the first five years of this decade, we will fall short of that figure by 93,700.
Matt Kinghorn is probably working on a new set of Indiana population projections. As the world continues to change, changes need to be incorporated into our thinking and planning for Indiana’s future.
Morton Marcus is an economist, formerly with the Indiana University Kelley School of Business. Send comments to firstname.lastname@example.org.