Yesterday's unemployment figures show a rise in Scotland of 7,000, or just over 3 per cent. In the UK, in contrast, unemployment fell by 50,000, a fall of just under 2 per cent.
What accounts for the difference?
There has been speculation in the media that this is due to an 'Olympic bounce' favouring London and the South East but not Scotland. Others cite the different movements in part-time and full-time employment between Scotland and the UK. Some mentioned weaker Scottish growth. Measurement error was also hinted at. Lags too!
All of these explanations may hold some truth. And, I certainly can't give a definitive explanation either. But what I can do, and what I think the discussion requires, is provide a framework within which to place the factors that generally determine changes in unemployment. I have tried to do this in the following chart:
What the chart suggests is that changes in unemployment reflect movements in both labour demand and labour supply. Taking demand first, we can see that when GDP changes - let's say falls - then jobs will fall proportionately if productivity remains unchanged. The fall in jobs will then translate into an equal rise in unemployment providing there are no changes on the supply side: the numbers active/inactive and new entrants, including individuals deciding to take on more than one job.
Therefore, holding productivity and supply factors constant, a comparable fall in GDP in Scotland and the UK should result in a comparable fall in jobs and a comparable rise in unemployment. There may be lags as employment adjusts to output change and unemployment to job change but for simplicity I'll abstract from those.
So, in principle, yesterday's rise in Scottish unemployment should be explained by differential movements between Scotland and the UK in one or more of the following:
- productivity, and
- labour supply.
Can we go further?
Yes, I think we can. But just a little!
The jobs and unemployment data published yesterday refer to the period June - August. Unfortunately, we don't have GDP data for this period. The latest Scottish GDP data, also published yesterday, cover the second quarter, that is April - June. There is only an overlap of one month. We must await the third quarter data to see if there is evidence of an 'Olympic bounce' because, if it exists, it will be picked up in the GDP figures. So, the strong jobs growth of 0.72 percent in UK employment in the quarter, compared to a fall of 0.06 per cent in Scottish jobs, might be largely the result of a strong output differential due to the Olympics. But we just don't know at the moment.
There is of course the possibility that Scottish productivity rose faster, or fell by less, than in the UK during the quarter, probably because of a bigger shift to part-time employment in the UK than in Scotland. The Scottish TUC, and Stephen Boyd in particular, highlight this factor. And they may be correct.
But we still have to get from jobs to unemployment. And for that we need to allow for changes in labour supply.
The jobs, or employment, rate fell in Scotland by 0.1 per cent while rising by 0.5 per cent in the UK. The activity rate rose by 0.3 per cent as more people offered themselves for work in the UK but it also rose very slightly in Scotland. In the UK, the rise in the activity rate was less than the rise in employment, so unemployment fell. But in Scotland, the rising activity rate meant that a weak fall in employment translated into a bigger rise in unemployment.
We can therefore conclude, tentatively, the following as an explanation of the large rise in unemployment in the most recent quarter in Scotland compared to the UK:
- Output growth was probably weaker here.
- Productivity probably rose faster, or fell by less, here due to a greater shift to part-time employment - with less hours worked - in the UK. And because of this:
- Jobs growth was disproportionately weaker here.
- The supply of labour rose disproportionately in Scotland.
The third quarter GDP data, when it arrives, should throw more light on this important issue.