A Scottish government account has tweeted that Scotland’s youth unemployment rate “fell by 0.3 percentage points over the year”.
This is looking at a dataset which is less reliable than the usual statistics on youth unemployment, which actually show the reverse happening.
The UK Statistics Authority has today asked the Scottish government to make it clear that the data it used is not considered reliable.
The data that the Scottish government refers to comes from an experimental release from the Office for National Statistics (ONS) which does not have the status of a “National Statistic”. National Statistics are considered fully compliant with the Code of Practice for Statistics, whereas experimental statistics are still being assessed for their suitability and quality.
The experimental statistics say that youth unemployment (among those aged 16-24) in Scotland fell from 10% in July-September 2018, to 9.6% in the same period in 2019. That’s what the Scottish government reports on.
But the ONS publishes another release, based on the Annual Population Survey of the UK, which does have National Statistics status. It’s based on a larger sample size, which is a key reason why it’s considered more reliable.
This data shows that the rate of youth unemployment in Scotland increased from 8.9% to 9.8% in the year to July 2019, compared to the year before.
The Scottish government told us it used the experimental data for its tweet because “those figures were more timely as they cover the period up until September 2019 and are published on a monthly basis”, whereas the other figures “cover the year up to June 2019 and were last published last month as a quarterly publication.”