Yesterday, I posted about a new working paper on the projected demographic impact of COVID-19 in Australia. They estimated an impact on the Australian population that would see the population smaller by 1.4 million people (about 4.2 percent) by 2040. I mentioned that I had been doing some work for local councils here in the Waikato region, so I want to talk a little bit about that now (ahead of presenting on this work to Hamilton City Council next week).
As I mentioned yesterday, projecting the population using the cohort-component method involves projecting three components: (1) fertility; (2) migration; and (3) mortality. Let's talk a little bit about the impact of COVID-19 on each of those components.
Although the scariest aspect of COVID-19 is undoubtedly its impact on mortality rates, New Zealand has thankfully been spared. And provided we are sensible about managing public health at the borders, it is possible we could stay that way. The impact so far on mortality has therefore been minimal, and so my projections assumed no change in mortality (note: this was the same assumption that the Australian paper mentioned yesterday adopted, even though deaths in Australia have been rising rapidly over the last couple of weeks, in Victoria especially).
In terms of fertility, as I noted yesterday:
Despite media articles suggesting that there could be a 'COVID baby boom', that seems unlikely. In fact, fertility tends to decrease in times of economic recession.
My projections assumed no impact on fertility. Given that we are undoubtedly in a recession, that might tend to overstate future births, and therefore over-estimate the population. However, in the absence of good data on changing fertility intentions, it is difficult to pin down the impact on fertility. The Australian study assumed a small reduction in fertility, which would persist for some time. So far, unemployment rates have not spiked as much as expected in New Zealand (probably thanks to the wage subsidy), so it is possible that there will be no change in fertility. We won't know for another six months or more.
Finally, the current model that I have been using combines net international migration and net internal migration into a single net migration rate for each territorial authority. However, based on past inter-Censal population movements, we know what proportion of migration flows arise from international rather than internal migration (and I am also using that data in updating my projections model to separately account for internal and international migration - more on that in a future post). I assumed that net international migration would fall to zero for one year (from March 2020 to March 2021), before bouncing back to normal. At the time I made the projections for Hamilton City (in early May this year), that seemed like a sensible assumption, based on what we knew at the time. Compared with the Australian study, my assumptions are a mixture of their severe scenario (zero net international migration) and light scenario (bouncing back to normal after one year). Probably now, we might expect lower net international migration to persist for longer than one year, and possibly to take some time to return to normal. My projections might therefore overstate the future population as a result.
What do the results look like? The diagram below shows two projections for Hamilton, based on different assumptions for fertility, mortality, and net migration: (1) a medium (baseline) scenario; and (2) a low scenario, that assumes lower fertility and net migration, and higher mortality. Both scenarios start in 2013, and are calibrated to replicate the estimated Hamilton population in June 2019, before diverging. Can you see any impact of COVID-19? Barely. If you tilt your head just right, you might see that the projection gets slightly flatter from 2019 to 2020. However, the takeaway message is that, even though net international migration is projected to be zero for one year, that impact is swamped by the long-run upward trend in the total population.
The impact is easier to see if we break down projected annual population change in Hamilton into two components: (1) natural increase (births minus deaths); and (2) net migration. This is shown in the diagram below. Here, you can clearly see that there is a sizeable impact on migration for the 2020 and 2021 years, after which things get back to normal fairly quickly. However, that impact is almost imperceptible when we look at population change in the longer term, as in the previous diagram.
You might doubt that the impact of COVID-19 would be so small. After all, it is the most salient thing to happen to population change in New Zealand since the baby boom. However, consider the diagram below, which tracks the total population of New Zealand from 1900 to 1935. In the middle of that time series is the combined impact of World War I and the Spanish Flu. Both events resulted in large increases in mortality and decreases in net international migration, relative to 'normal'. And, because mortality was concentrated in prime age adults, there was also a negative impact on fertility. And despite all of those compounding negative impacts on the population, there is only a flattening out of the population before it returned to its previous trajectory. Given that COVID-19 is a much smaller event in terms of population impact that the combined World War I-Spanish Flu, it should be no surprise that it is difficult to see any change in the projected total population.
Compared with the Australian research I discussed yesterday, my work is showing a much smaller (not zero, but small) impact on the future population of Hamilton City. I've done similar work for Waipā District, which shows an even smaller impact (because population change is drive more by net international migration in Hamilton than in Waipā). I expect to also see something similar in some forthcoming work for Waikato District. However, as I noted above, I may be overestimating future births by not adjusting fertility downwards in response to the recession, and I may be overestimating net international migration if border closures persist for some time. As in the case of the Australian research, that could be addressed by projecting alternative scenarios.
It is easy to see the world falling apart around us, and jump to doomsday scenarios for population change. However, it would take something much larger and longer-lasting than what we are observing in New Zealand right now, before we would see a substantial change in the trajectory of future population.
Read more:
No comments:
Post a Comment