What does the school workforce census tell us about teachers?

Lately I’ve been trawling through the numbers from the school workforce census – the annual census of staff in English state-maintained schools, taken during a week in November.

It records information including teachers’ age, their degree and initial teacher training qualifications. These two pieces of information were both used in the creation of the graphs shown in this blogpost. Further published statistics from the Department for Education (DfE) can be found on the government website.

When classifying teachers’ qualification subject, I took the subject of their initial teacher training qualification, or, if they had no initial teacher training qualification recorded, I took their highest-level other qualification.

This is exploratory work, and isn’t necessarily absolutely robust. A significant amount of data cleaning was required to produce these statistics, and there is also an amount (that is unknown to me) of missing data in the school workforce census.

The data I have used are from 2010 and 2013, which isn’t ideal, but this is exploratory work using the data we currently have available. Acquiring new data has proven tricky over the last half-year or so – we’re still waiting for more to be made available by the DfE.

The data used for these graphs includes primary and secondary school teachers. We have also performed analyses looking specifically at the secondary school workforce, which we’ll be blogging about soon.

A look at the ages of teachers shows that physics-qualified teachers are an aging population – and that this is the complete opposite of computing-qualified teachers. Teachers with physics qualifications had the highest modal age (48), compared to engineering (44), maths (42) and computing (34). This is likely due to a 25 years of underrecruitment of physics teachers.

In 2013 I was also able to identify in the data more computing-qualified teachers than physics-qualified teachers (4,827 and 3,767, respectively). We’re not certain which subjects they’re teaching, though, and will have to look into this further.


Crunching the numbers on when teachers began their training, it seems that the physics teacher workforce in 2010 undertook their initial teacher training in a relatively equal spread over the past 40 or so years. In contrast, a great proportion of the teachers in the workforce with computing degrees had done their initial teacher training in recent years.

It’s important to note here that attrition from teaching means we only see when the survivors (people still in the workforce in 2010) undertook their initial training. Similarly, it’s a feature of the data – rather than necessarily what has occurred – that it appears that many fewer teachers trained in the most recent years. We can’t speculate using this data about very recent entry numbers or attrition from the workforce.

Finally, there is a curious dip in teachers undertaking ITT in the second half of the 1970s and first half of the 1980s. I don’t know why this is – can anyone speculate?


Looking at the years in which teachers completed their training, we can see that for computing, maths and physics, the most common age to finish initial teacher training was 22 or 23. Entrants to teaching with engineering degrees are older than those with qualifications in other subjects. This is partly due to four-year integrated masters degrees being common for engineers. Beyond these modal ages, the number of teachers with a physics qualification drops off more quickly than those with a computing qualification.


Related posts

2 thoughts on “What does the school workforce census tell us about teachers?

  1. I think that the dip in recruitment in the late seventies and early eighties may be due to the fact that teacher pay was a very big issue then and teachers were relatively poorly paid compared to other professions in the public sector. Physics and Maths graduates will have had lots of opportunities to find good jobs elsewhere.

Leave a Comment