Graduate underemployment

I was on holiday last week, hence no reaction, until now, to the Association of Accounting Technicians report, conducted by the CEBR last week.

The report casts doubt on the value of a degree based on their interpretation of DLHE data and some ONS data.

Now, the Work Foundation have already expressed scepticism in this blog post by Laurence Hopkins, and I’m going to expand on some of Laurence’s points.

The CEBR claim that in 2009/10 that 52% of new graduates are either out of work or underemployed. They seem to reach this conclusion (it’s not quite clear) by using DLHE data for employment outcomes, but then tacking on the unemployment data gained, not from DLHE, but from the ONS research we dissected here. This is an interesting decision as it takes a measure of unemployment immediately on graduation and combines it with an employment outcome measured six months post-graduation even though a measure of unemployment six months post-graduation already exists in the same dataset. 

The way they calculate what jobs don’t require degrees is interesting as well. The authors state that ‘we regard the regard the following occupational classifications as not requiring degrees’, and then list, essentially, all SOCs from 4000 down. Now, that’s ok if you want a crude approximation, but it does ignore the fact that there is already a well-known measure of graduate level occupations produced by the IER and available from HESA as part of DLHE.

It might be a bit creaky, but it’s evidenced by some rather splendid research. I am not sure why the CEBR did not use this research to determine a rather more nuanced view of graduate level employment.

Using their measures, the CEBR show that in 2006/7,  32% of graduates were ‘underemployed’, rising to around 34% in 2007/8 and rising further to 37% in 2008/9.

Let’s do some analysis. 


In 2006/7, 147,315 UK-domiciled first degree graduates were known to be in UK employment six months after graduation.


Of those, using what appear to be the author’s measures (I’d prefer a full SOC methodology), 46,730 were working in occupations coded with SOC 4000 or above – 31.6% of the employed cohort. 


In fact, if you use the IER graduate job categories, 33.5% of employed graduates were in non-graduate jobs six months after graduation. Unemployment was 5.5%. Now, of course, you can’t add 33.5% to 5.5% and say ‘hey presto, 39% of graduates were underemployed’ because that 33.5% are only 33.5% of the employed cohort.

The whole UK-domiciled cohort was 209,115 graduates. In all, 60,705 of 2006/7 graduates were in non-graduate employment or unemployed six months after graduation in the UK, and that comes to 29% of the whole cohort. The CEBR report it as about 40%, presumably using a much higher unemployment figure from ONS data taken at a different time.

Doing the same calculations gives us 69,000 graduates from 2007/8 in non-graduate employment or unemployed six months after graduation in the UK, or 31.3% (the recession had started)). In 2008/9, at the depths of the recession, we have 75,385 graduates from 2008/9 in non-graduate employment or unemployed six months after graduation in the UK, or 33.5% of the whole graduating cohort. That is using the respective DLHE datasets, the commonly acknowledged measure of non-graduate employment and a measure of unemployment taken at the same time and using the same dataset.

Finally, the CEBR reaches the interesting conclusion that as the proportion of graduates in non-graduate employment grew between 2006/7 and 2008/9, it will therefore continue growing in 2009/10 and 2010/11, ‘given underlying trends and weaknesses in the UK labour market’. This seems to contradict reports of more optimistic employment prospects for graduates this year and I believe they’re wide of the mark. The economy is in recovery, albeit a sluggish and choppy recovery, and graduate numbers actually fell last year.

If you look at the data at subject level, the methodological information provided is not sufficient to reproduce the data in the report, but the subject groupings used suggest that the LFS data mentioned previously is being used to derive unemployment figures even though a more appropriate measure exists in the DLHE data used to derive employment data.  If you do use DLHE throughout, you get a rather different subject pattern.

So, have employment prospects for graduates deteriorated as a result of the recession?

Unequivocally, yes?

Are they as bad as this report makes out?

No, I rather think not. 

Anyway, we'll find out who is right about 2009/10 graduates in about six weeks when the new DLHE is out, and I'll tell you then. Keep an eye on what we have to say.