Today’s Labour Market Statistics – also a note on interpreting the data

Today saw the release of fresh data on the Scottish labour market from the ONS which (mostly!) cover data up to June 2017. Some additional data only cover the period up to March 2017.The main data can be accessed here.

In this blog, we pull out and highlight the key trends in these data.

Essentially, headline indicators of Scotland’s labour market remain good, but we need to bear in mind that, as always, the health of Scotland’s labour market is more complicated than headline numbers.

Summary

The headline numbers are contained in the following table.

Capture

The key things to note from these data are:

  • The employment rate has hit a record (data go back to 1992) high in Scotland of 75.2% as the chart below shows.

Capture

  • The unemployment rate among those 16+ and those 16-64 is down to 3.9%, with the youth (16-24 year old) unemployment rate down at 9.9% (based on data to March 2017) shown in the chart below.

Capture

While headline indicators are good (and they are good!) we need to bear in mind that, as always, the health of Scotland’s labour market is more complicated than headline numbers. In particular, we need to better understand the nature and quality of destinations that those leaving unemployment are going into.

We note that almost all of the recent rise in employment is among the self-employed, which may have implications for tax revenues and the number of hours and type of work undertaken. More work needs to be done to explore these wider implications.

Similarly, a slight caveat to the welcoming of the fall in youth unemployment is that we have also seen a rise in the economic inactivity rate among this age group over the past year, so not all those leaving unemployment are doing so to begin work. This is illustrated in the chart below.

Capture

We’ve noted before that inactivity category is a ‘catch-all’ for a wide variety of labour market outcomes from people being in full time education through to being long term sick, or simply not actively seeking work or in work. To the extent that any rise in economic inactivity is driven by an increase in the number of people in training or education that’s a more positive move, but if people are becoming detached from the labour market, that would be more of a concern.

A note on interpreting the labour market data

There was some social media discussion of a change that has been made to a BBC Scotland News report of these data. The initial report suggested that the unemployment rate had increased in the most recent data.

Without getting too geeky, the LFS data report the unemployment numbers for a rolling three month period. That means that – for instance – the data released in August’s publication covers the period April-June with the next release in September covering the period May to July (and so on).

In this way, each time the statistics are released the numbers will share two months of data in common with the previous release from the month before. This helps to smooth out ‘noise’ in the sample, but makes calculating changes in these data more complicated.

For example, it means that comparing the most recent estimate of the unemployment rate published today with the estimate from last month and calculating the change between these two estimates doesn’t produce a robust measure of the change in the unemployment rate.

Instead, one has to calculate the change in the unemployment rate in the latest data, relative to the previous – non overlapping – three month period.

In other words, data covering the period April – June should be compared with data covering January – March, not data covering March – May.

This is why, in the ONS Statistical Bulletin accompanying these data, they note:

“Comparing non-overlapping periods (April to June 2017 with January to March 2017) provides a more robust short-term comparison.”

It’s easy to get confused by this (!) and easily corrected. We always find the best table to quickly see the headline numbers is here.