Random Analytics: Peak Employment (Part II): UK
by Shane Granger
I recently published a blog looking at Australian Peak Employment. Given some very good feedback and responses from the UK especially I thought it might be useful to have a similar look at the situation there.
Although the Office of National Statistics longitudinal depth was not as good as the Australian Bureau of Statistics in breaking down full-time and part-time employment pre 1992 there was some very good data not readily available in legacy ABS data captures, especially in areas like the self-employed.
Although I could only get data as far back as 1992 the picture it tells is as interesting as the Australian story (with data going back as far as 1978).
Here are the analytics.
Figure 1: UK Full-Time and Part-Time Employment (Mar – May 1992 to Oct-Dec 2012). Source: ONS.
Like the Australian example, the first graph highlights the almost identical increase in contingent (part-time) and full-time employment over the past decade. Since early 1992 full-time employment has increased by 10.3% (2.021-million) while part-time employment increased by 34.5% (2.072-million).
Figure 2: Increases in UK Full-Time and Part-Time Employment (Mar-May 1992 to Oct-Dec 2012). Source: ONS.
Like the Australian example the increases in both full and part-time employment are identical. Unlike the Australian example which has not seen a downturn since 1991/92 the UK data shows a massive 1-million full-time jobs disappearing post GFC but over the past 12-months or so that has improved by around 400,000 new positions. Additionally, the start of this data series showed negative full-time employment through to early 1994 as the UK struggled out of the late 1980’s and early 1990’s downturns (oh, how I wish I had the data going back to 1987).
Figure 3: Overall UK Employment versus Working Age Labour Force (Mar-May 1992 to Oct-Dec 2012). Source: ONS.
Here’s a look at the differential between total employment creation and the UK defined, economically inactive numbers. The UK created an additional 4.093-million new jobs; from 25.635-million in early 1992 to a record 29.729-million jobs at the end of 2012 (the pre-GFC high was 29.572-million in Mar-May 2008). The UK has steadily increased its working age population since 1992 (due to a combination a slightly lower than replacement fertility rate plus a higher immigration to migration ratio) which has left it with 3,734,000-million less jobs than that required to employ the 7.828-million increase in the working age labour force.
A quick point on the above. Like my previous article on this subject (avoiding wonkish angst) it should be noted that my Working Age Labour Force number in this graphic has been worked out utilising the ONS economically active rate data rather than my preferred method of using actual population statistics (you can see an example of this in my 2012 Abbott’s Promise piece).
In conclusion looking at the UK ‘employment type’ data is further confirmation of a global trend toward greater reliance on part-time employment, which on one hand is increasing employment to record levels while at the same time decreasing the amount of work available.
Has the UK reached peak employment yet? I’m not convinced it has but the more I look at the global data the more I am convinced we are reaching that point in the next decade. As I stated in my initial Australian analysis:
With an increasing working age population and a growing gap between jobs available the future is looking anything but certain, especially with the rise of labour augmentation and robotics replacing jobs quicker than they can be created.
Random Analytica 
[…] then looked at the UK and drew similar conclusions, noting, as in Australia, an increase in the proportion of part-time […]
Hi Shane,
I come to your blog courtesy of Flip-chart Fairy Tales, whose take on the world of work I enjoy. Thanks for the thought-provoking posts.
Here are some thoughts about another country, which may or may not be relevant.
A few days ago I spent a bit of time on FRED, making charts of US employment trends in each of the major industries. Over the last few decades some of these have been growing and some declining as a percentage of total employment. I think it’s likely that, in the US at any rate, the shift to part-time employment is at least partly due to these shifts.
I’ll leave aside the biggest sector (professional and business services, 13%) for now. There are four industries just below that at around the 10.5% to 11% level: healthcare, government excluding education, retail trade, and leisure and hospitality.
Health and L&H are growing quite quickly, taking employment from many other sectors: manufacturing mostly, but also FIRE, wholesale trade, transport, construction, and the aforementioned retail and government sectors, as well as the rats and mice: utilities, agriculture, mining & forestry and the like.
It’s my understanding that Health and L&H have long-standing traditions of part-time and casualised employment (for health, jobs like home help, public health nurse, orderly and so on tend to be fairly casual), whereas the declining sectors don’t have so many part-timers. So if these sectors are growing as a proportion of the whole we should see part-time employment rising in relation to full-time.
Why are Health and L&H dominating new employment? That’s a question I don’t know the answer to. Possibly it’s because as people get richer they want to enjoy their wealth for longer (i.e., live longer) and experience more with it (so they spend on leisure “experiences”). Possibly it’s the reverse: many households are getting poorer in real terms so they switch preferences to cheaper goods–a $5000 holiday instead of a $25000 car. Possibly tastes haven’t changed, but due to ongoing productivity gains in some industries households need to spend less meeting those needs. Possibly it’s the effect of the Baby Boomers getting to retirement age. Possibly it’s a combination of rent-seeking by an unholy alliance of medical companies, drug companies, and health insurers in the case of health, and a cover for near-unemployment in the case of L&H.
Another factor that could explain a rise in part-time and casual employment as a fraction of the whole is that employers in other sectors (retail, professional and business services, education) are adopting the employment practices of L&H. That would be a case of unequal negotiating power–probably.
The question is, does it have to be this way? I think that the example of Sweden shows that it doesn’t. Sweden is subject to the same technological forces as the rest of the advanced world but its employment looks quite different–at least on the surface. At 80% participation, it’s not showing obvious signs of stress in employment yet.
Greg,
Great reply, but a couple of points to consider in your very detailed response.
First of all, if we are going to consider peak employment, we must all think of the US as the pure test case. It is the ‘tip of the spear’ in terms of analytics. In terms of data the BLS has much better data than the US or UK because they quantify the differentials between unemployment hours… More on that in a later post.
I won’t comment directly on your US analysis, given that I haven’t done any analysis work in the US since 2006 and that was company/sector specific.
Saying that I would like to make some suggestions on your analysis and your question.
In terms of Scandinavia, using Sweden as a good employment outcomes example is not so good, given that Sweden is the lesser compared to Norway and quite high compare to some other EU nations. If you are trying to compare Sweden vs. the USA I would suggest you go to my argument that they are very different economies, thus don’t really compare. I’d also suggest in terms of manufacturing outcomes that current US productivity is much better than Sweden due to its better concentration of automation outcomes over employment outcomes since the GFC (another subject). I wont go into the participation rate that the moment as it is very nuanced and I would bore you in the explanation.
It will come up in my quarterly #peakjobs article but the question of why health is on the rise will be because we are coming up to a generational peak in elderly due to the baby boomers spike. The problem will be that we are now creating infrastructure to compensate for this aged spike.
More to follow on this but many thanks for your comments (always appreciated).
[…] countries. He has analysed data on long-term labour market trends in Australia and the UK to see if we might be approaching peak jobs. Shane’s conclusion is stark: With an […]
[…] Shane Granger: Peak Employment (Part II): UK If Jon Evans’ post (above) introduced you to the topic of ‘peak jobs,’ might I also recommend this fascinating and sobering analysis of the topic from Shane Granger? This post focuses on this topic as it relates to UK. Shane argues that the UK – in common with many other countries worldwide – could reach peak employment “in the next decade.” Follow Shane on Twitter. […]