New labour market data were out earlier today, covering all the headline estimates for Scotland and the UK.
These showed that the Scottish unemployment rate hit 3.7% in the three months to October 2019, down -0.3% points on the three months before and down -0.1% on a year before.
Meanwhile the employment rate hit 74.5%, down -0.4% points on the three months before and down -0.5% points on a year ago.
In this blog we pull out a few important points from today’s data.
Firstly, it is important to note that with the UK unemployment rate sitting at 3.8%, the unemployment rate in Scotland is now lower than in UK as a whole.
However, it is important to understand that this very small differences are not statistically significant. Like all survey based measures, there is a degree of uncertainty surrounding these numbers and it’s important not to lose sight of this. This is done by reporting 95% confidence intervals for each central statistic.
The chart below illustrates the sampling interval for the Scottish and UK unemployment and employment rates.
For the unemployment rate for the UK the sampling variability implies that the ONS have 95% confidence that the ‘true’ UK unemployment rate likes within 0.2% points of the central estimates – so +/- 0.2% points.
For Scotland given the smaller sample size, the ONS has 95% confidence that the ‘true’ figure lies within 0.7%-points of the central estimate.
For the employment rate a similar pattern is clear, with the headline UK employment rate at 76.2% with a confidence interval of +/- 0.4% points, and a central estimate for Scotland’s employment rate of 74.5% with a confidence interval of +/- 1.4% points.
Secondly, it is worth noting that there has been a rise in economic inactivity in Scotland over the past year. There are now 20,000 more people in Scotland who are economically inactive compared to a year ago. This explains why employment can be down 16,000 over the past year while unemployment is also be down by 2,000 over the same period. Although we again reiterate the sampling variability around these numbers.
There are many reasons why a person can be economically inactive, from the positive reasons like people being full time students, through to the more concerning reasons like sickness.
With more data, and in time, we will be able to say more about the reasons underpinning this recent rise in economic inactivity.
In the most recent data (which cover the year to the end of June 2019) the number of people economically inactive due to long term sickness had risen by 15,000 while the number of those economically inactive due to being students was down 25,000. The next set of data will cover the year to October 2019 and may provide additional insight.
Finally, a few words on different labour market data sources.
The data which were updated today on overall employment, unemployment and inactivity come from the Labour Force Survey (LFS) and not the boosted sample Annual Population Survey (APS) data.
The LFS data provide the aggregate headline measures and these are National Statistics.
But it is the APS data which are the best for the most robust sub-group analysis within the Scottish labour market . Of course, this LFS data for sub-groups are still useful from time-to-time but it’s important to be clear about caveats on data variability and to note that they are not National Statistics.
The ONS will update the APS data with the next release of regional labour market data next month.
However, the Scottish Government have announced a welcome new quarterly publication which will focus in on one use of these data to provide an estimate of youth (16 – 24 year old) unemployment.
This will presumably be a new consistent publication which will sit alongside their new look Labour Market Trends update which has been published today.
These are very welcome initiatives and will help guide users to the most appropriate statistical estimates.
The Scottish Government should be commended for making these data for Scotland available to users earlier, but it is disappointing that ONS are not able to make these headline estimates from the APS available directly with coverage of all the regions and nations of the UK, which would be useful for comparison.
Authors
The Fraser of Allander Institute (FAI) is a leading economy research institute based in the Department of Economics at the University of Strathclyde, Glasgow.