Transportation, Distribution & Logistics

EYE ON THE PIE Morton Marcus: Stats born of self-interest not worth the trouble

September 18, 2006

When I arrived at Indiana University in 1970, we had many copies of a 1966 projection of Indiana county populations sitting in file cabinets. Most state and local agencies also had these projections sitting about. Some used them for doorstoppers. After all, they were thick with details for age and sex in each county going out about 30 years, to a distant point in the 1990s.

Once the 1970 census was released, Bob Calhoun of the State Board of Health and I (representing the Indiana Business Research Center) got together and developed a new series of population projections for each county. There was a need for a single, consistent, county series. State agencies used different projections leading to much confusion. Most regions in the state, and some counties individually, had produced their own projections by different methodologies with incompatible detail.

If you added up all those individual projections to the state level, you had a number twice as high as any reasonable expectation. Why? State, regional and local agencies were competing for federal funds. Often, population projections were required to support their applications. Consultants hired to produce such data came up with numbers that made the case for the applicants.

It's not difficult to use different methods and data sets to get different results. But different alone is not necessarily valid. If South Bend benefits from one set of assumptions, naturally that area will use what works best for its interests. The same applies to Evansville, Terre Haute, New Albany, Richmond, etc. A consultant with a reputation for "satisfying" clients, regardless of the consistency or quality of the work, will get more work. A report that fails to show the desired results will be blasted by local boosters as inadequate, blatantly wrong, and biased on the part of the researchers.

Bob and I sought a consistent, objective baseline. Over the years, we were able to get a single series of population projections accepted by the state and recommended for use by all state and local agencies. We traveled the state trying to educate anyone who would listen to our explanation of how the projections were developed and could be used.

These were projections, not estimates of the population. Estimates are numbers that relate to either now or a past point in time. Projections are about a future point in time.

We (the IBRC and the Board of Health) never produced estimates. Those were products of the U.S. Census Bureau. We developed county population projections. Our methods were simple. We used data on the long history of fundamental demographic trends-births, deaths, and, yes, migration into and out of our counties. From these factors we calculated the future numbers of residents in each county.

We did not use the aspirations of developers, civic officials and others who believed in the inevitable greatness of their counties. We did not project the population of Jay County if a casino were to open there. We did not project the population of Grant, Delaware, Howard or Madison counties if General Motors Corp. were to reorganize and downsize. Nor did we project the population if meteors struck Henry and Randolph counties.

A confused article in the Sept. 9 Gary Post-Tribune is one indication that perhaps the work Bob Calhoun and I did to educate reporters and others did not take. Estimates and projections are confused. Once again a local agency, the Northern Indiana Commuter Transportation District, presents projections that reinforce its interests.

Maybe the arguments about consistency and independence in population projections need to be made again. Otherwise, we will have overbuilding of schools, retail facilities, roadways and many other capital projects to the benefit and glory of a few but to the cost of the many.



Marcus taught economics more than 30 years at Indiana University and is the former director of IU's Business Research Center. His column appears weekly. To comment on this column, send e-mail to mortonjmarcus@yahoo.com.
Source: XMLAr01101.xml
ADVERTISEMENT
Comments powered by Disqus