![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Making Demography Slightly More Exciting
Sep. 27th, 2008 11:17 amSo, I did a course on demography the other day. Nothing too taxing, but interesting to learn about the various techniques we use for population estimates etc. At one point we compared three different projections for Australia's future population, based on three different assumptions about Net Overseas Migration.
Only, it was acronymised, so the first column of the table read: "NOM NOM NOM".
Then we looked at Indigenous population estimates between Census years. Problem here is that while we get reasonably good estimates for the Census years, the birth/death records aren't reliable enough that we can just tally up the difference - while most Indigenous births/deaths do get recorded, a lot of them aren't identified as Indigenous. However, we do have moderately good estimates of Indigenous birth rates and death rates* for each age group. So, if we want to estimate the Indigenous population for (say) 2004, we get better data by taking the Census estimates from 2006, and then applying the birth/death rates in reverse.
Or, as the lecturer put it, "we go back in time and bring people back to life!"
*which, in case you're wondering, are pretty depressing - about three times the national average for most age groups, up to six times for the 35-44s. Detail p. 73 here.
Only, it was acronymised, so the first column of the table read: "NOM NOM NOM".
Then we looked at Indigenous population estimates between Census years. Problem here is that while we get reasonably good estimates for the Census years, the birth/death records aren't reliable enough that we can just tally up the difference - while most Indigenous births/deaths do get recorded, a lot of them aren't identified as Indigenous. However, we do have moderately good estimates of Indigenous birth rates and death rates* for each age group. So, if we want to estimate the Indigenous population for (say) 2004, we get better data by taking the Census estimates from 2006, and then applying the birth/death rates in reverse.
Or, as the lecturer put it, "we go back in time and bring people back to life!"
*which, in case you're wondering, are pretty depressing - about three times the national average for most age groups, up to six times for the 35-44s. Detail p. 73 here.
