Friday, January 11, 2008

Producing People for Export: A look at demographics

This post was inspired by some comments by Bruce Kettele on, I think, the Gettysburg Kroger issue, about changes in the young adult age cohort. I've had this in the back of my head, then got into some data sets, census and others, to investigate demographic changes in Montgomery County.

I really don't know anything about demography; it sounds so dry. But, getting into it, one can see beyond the numbers to consider the economic and social implications of change in population groups.

For example, the young adult group, say between 18 to 39 (particularly those in their 20s and early 30s), would be the most likely to be patrons of bars, nightclubs, and restaurants, creating a young singles scene.

Young adults would also be the source of new households and young couples, potential market for urban housing, and then furniture and house wares. Ikea customers.

And, perhaps most important, the young adult cohort would be the one which would be reproducing the population via family formation, having babies and raising young children. If the young adult cohort declines, one can anticipate a decline in births and young children as there are fewer couples to have them (assuming the fertility rate is constant).

A wrinkle in this birthrate consideration is immigration. Immigrants have a higher fertility rate, meaning more babies per women, so if there is a substantial in-migration from outside the USA to a community (like Chicago, or California) one could still see a growing number of children. The Dayton area hasn't seen much immigration, so this issue is mostly moot here.

So are some numbers from the Census American Factfinder site, grouped into certain age cohorts or groups.

It's clear there are some drops, not big ones, in the Children & Teens (0-17) and Young Adult (18-39) age groups from 1990 to 2000.

Taking a closer look, it seems the big changes are in the 20-29 and 30-39 groups of young adults. The 18 & 19 group is pretty stable.

But then, looking at percentage changes between 1990 and 2000, it's interesting to see that the largest % drop are in the young adult and young children cohorts. One can infer a vicious cycle developing, as there are perhaps fewer young couples to have kids, leading to fewer kids.

Finally, taking a look at births and deaths, using projections from this website. As one can see, using a simple linear projection, deaths will equal births around 2020, though one can anticipate an increase in the death rate as the large baby boomer generation passes on into history.

One wonders if the birth rate decline remains linear or becomes or exponential at some point, if there is a continuous decline in the young adults (assuming minimal in-migration and a static fertility rate).

Once could also assume a long-term negative effect on the local economy, as employers are faced with a shrinking labor pool, and might be inclined to relocate to growing areas, or just shut down as the workforce and management age, and then retire. This would be another vicious cycle, as reduced employment opportunity forces young adults to leave the area (but one would assume in-migration would pick up to replenish the labor pool).

This has been interesting but I'm not satisfied with my approach to demographics. What are we really measuring here?

Note that these census numbers are just snapshots. The census counts people in the, say, 20-29 category in 1990, and then counts people in that category in 2000. We can only compare counts and say "there are fewer people in this category in 2000 than there was in 1990".

But what happened to those people in the 20-29 category in 1990? Where are they in 2000?

That would be more useful, to see if Dayton (Montgomery County) is producing people for export.

No comments: