Yesterday morning I was skimming through the online database for the OECD’s recently-released latest Economic Outlook when I noticed that their table for the labour force participation rate was for ages 15-74. Often tables and charts for the “working age population” are done for ages 15-64 but as a growing proportion of those 65 and over have continued working the 15-64 numbers have become increasingly unrepresentative. So I downloaded the data and had a quick look, resulting in this tweet. (2019 because last year’s numbers were thrown around for every country and aren’t necessarily that representative.)
I was a little surprised – not that we were towards the right-hand end of the chart, but that New Zealand was now highest.
But there I might have left it, except that Stuff’s Susan Edmunds got in touch. We had a bit of a chat, and she produced this article. The headline suggests that high house prices might be the explanation (but as I’d noted to the journalist a variety of factors will have affected a range of different demographics).
It is worth stressing here that a high participation rate is neither good nor bad in its own right. It depends. I’m not in the labour force these days, and since I’m fine with that and I’m not a burden on the taxpayer it isn’t really a matter for anyone other than me and my family. My mother was in the labour force for perhaps only two or three years of the 60 years between her wedding and her death. On the other hand, if people are deterred from working by very high effective marginal tax rates or by “overly-generous” welfare provision or highly inflexible labour markets that might be more of an issue. And in wars, states sometimes compel people into the labour force – there might be reasons of state for that, but it is hardly a first-best state of affairs.
But I was still a bit puzzled so dug about a bit more data. First, this chart showing (since 1986 when our HLFS began) the 15-74 participation rate for New Zealand relative to (a) the median OECD country, and (b) the highest participation rate of any country in each year.
(Technical note: it is only since 2002 that there is complete data for all today’s OECD countries. Using a fixed sample of the countries for which there is data for the entire period takes out much of the dip in the median after 1990, but converges back to the orange line in the last half dozen years.)
So we now have the highest 15-74 participation rate of any OECD country, but that is quite a new thing. Over the entire period our participation rate has been at or above the median (which was probably my impression), but the closing of the gap is striking. Note that back in 1986, the universal age pension (NZS) was paid from age 60, so from that change alone it is not surprising that our participation rate is now materially higher than it was 35 years ago (in 1986 32 per cent of 60-64 year olds were in the labour force, while about 74 per cent are now).
But where does this highest participation rate show up once we dig a bit deeper, by age and by sex?
It turns out that it is nowhere in particular. For not one of the age/sex breakdowns I looked at did New Zealand score highest in 2019.
Here are the young people, aged 15-24
Well above the medians but also quite a bit below the highest participation rates, whether for men or women.
This chart, for what is often referred to as “prime [working] age” people (25-54) took me a bit by surprise.
Not only are the New Zealand numbers not that near the maximum but for women of this age New Zealand was the median country in 2019. So much – it would appear – for the high house price story (at least as it is often applied, with both parents “having” to work even though they might prefer one to take more time out). Of course, house prices are high in some other OECD countries as well, but hardly any have quite the price/income ratios of New Zealand.
The next age group is the last before NZS eligibility now kicks in.
Here both male and female participation rates are well in excess of OECD medians, and not far off the highest rates (Japan for men, Sweden for women).
What about the 65-69 age group?
Again, well above the OECD medians for both men and women. but with a few countries higher than us especially for men.
And finally, the entire 65+ age group.
With much the same picture.
So standing back (and assuming the OECD did the 15-74 calculations correctly) we end up with the highest overall labour force participation rate for this working age band not by having spectacularly high participation rates for any particular age or sex sub-category, but from being well above the median in almost all categories, and at the median in a single sub-category (women 25-54).
(Note too that these are participation rates rather than employment rates – simply because that was the OECD table I started with – but note that New Zealand unemployment rates are typically below average among OECD countries.)
What should we take from all this? There are no doubt people who have dug into some of this stuff in more depth than I have here, but a few things strike me:
- There is no single cause, factor or policy that explains New Zealand’s current participation rate being the highest in the OECD,
- The high house price story seems unlikely to be that compelling, at least yet. For example, our working age participation rates are not particularly high (especially not for women), and today’s 65+ generation were already at least in their mid-30s 30 years ago (which is when real house prices really began to take off),
- Our NZS regime is relatively supportive to labour force participation (since there is no income-testing or abatement), but as other countries raise their state pension ages, the income effect will tend to dampen labour force participation here among the 65+ age group, relative to some other OECD countries,
- Much as I am not a fan of our high (absolute) rate of people receiving working age welfare benefits, or of the fact that our minimum wage is now high relative to median wages in New Zealand, it is only fair to note that our labour force participation rates are at or (mostly) above those in most OECD countries for every demographic shown here.
In the original tweet (above) I contrasted the high rate of labour force participation in New Zealand with the dreadful record on average labour productivity. In 2019 it would have taken a 68 per cent lift in average New Zealand real GDP per hour worked to have matched the leading bunch in the OECD (the US and some northern European countries). But by putting in more labour inputs (some mix of being employed and hours per job) we keep rather closer to the first world on average per capita incomes. The comparisons with the UK – not one of the highest productivity countries – is illustrative. UK average labour productivity is about 40 per cent higher than that of New Zealand, but UK real GDP per capita is only about 10 per cent higher than that here (both on OECD estimates). The UK does not have a low participation rate, but their participation rate (see first chart) is far lower than New Zealand’s (as are average hours worked per capita).
Sometimes the story is portrayed as one of employed New Zealanders working long hours. There is something to that story – working hours per worker in New Zealand tend to be in the upper quartile for advanced countries (only Lithuania was higher in 2019 among OECD countries – although Singapore, Korea and Taiwan are all materially higher). Some of that, in turn, doesn’t reflect long hours each week, but shorter annual leave entitlements than in many European countries. But equally important is long hours over a lifetime – our young people and our old people are simply more likely to be in the labour force than in the rest of the OECD, for a mix of reasons that can’t simply be dismissed as “good” or “bad”.
To the extent those reasons are “good” just think of the per capita incomes that could be sustained here if policies were ever adopted that once again delivered average productivity getting even close to the leading bunch in the OECD.