The Census Bureau has released the complete demographic data files for Texas, which is allowing me to revisit some issues in more detail.

A few months ago, I wrote about demographics shifts in zip code 78704. (For readers outside of Austin, 78704 is the zip code directly south of downtown and a large chunk of Austin’s urban core.) The most striking shift was 78704’s loss of children:

If we disaggregate the population into children and adults, we see that 78704 lost 1,387 children and added 385 adults. Put differently, 78704 shed 17.9% of its population of children even as it saw a slight increase in its adult population. That's a huge decline, percentage-wise, for just one decade. That decline is also reflected in the change in the number of children per occupied housing unit. The average number of children per occupied housing unit shrank by 20.9% between 2000 and 2010. By contrast, the number of children per occupied housing unit in Austin as a whole actually increased slightly (from 0.56 to 0.58) between 2000 and 2010.

The Census Bureau has released the single-age populations. As I expected, the population of children is skewed toward very young children. There are almost twice as many babies as thirteen-year-olds, for example.

This is a simple story, really. South Austin attracts singles and young couples. Many families start here but move away when children hit a certain age. The data show particularly steep drops between ages 3 and 4 and between ages 6 and 7, although these could simply be random variation. But the larger trend is not random.

It’s worth emphasizing that this is not a new trend. The population of children also declined with age in 2000. There are simply fewer children today. In fact, 78704 had fewer children of *each age* in 2010 than 2000.

None of this should surprise anyone. When I first looked at the numbers, I actually was surprised that 78704 has as many teenagers as it does. But I was guilty of extrapolating from the population of single-family homeowners. 78704 has a fair amount of low-income housing. I haven’t tried to determine whether low-income households are more likely to have a teenager, but the low-income developments are clearly a large source of school-age children in 78704.

Take the Bouldin Creek neighborhood (Census Tract 13.05). The neighborhood had 5,488 residents in 2010 spread among 110 Census blocks. 1,112 of the residents were under age 18. But over half of the children lived in just two census blocks (Blocks 2008 and 3025) containing large affordable-housing developments. (In fact, these two blocks contained more than 10% of the children in the entire zip code.) As is obvious from this chart, school-age children are spread very thinly over the remaining 108 census blocks.