How to count training, and explain the results

There’ a rather curious thing about our attitudes to training.  On the one hand, an endless mantra about a high-skill economy, and how we need to push up skills training.  On the other hand we seem not to care very much about how we should be measuring training so that we understand better what we do and don’t know about it.  We tend to use very crude measures which tell us about the proportion of people participating over a given period, and not much else.

How there’s a very welcome contribution to the training debate, from Francis Green and colleagues at LLAKES,  which goes well beyond this.  It looks at the overall volume of training, capturing the duration of training as well as participation rates.  It traces the patterns over a full decade or so, so we get a good sense of trajectory.  And it offers competing hypotheses to explain the results, rather than just  ‘more is better, less is worse’ .

The basic message is quite dramatic.  Once you look at how much people in the UK do in the way of training, rather than just how many of them take part, it’s clear that there has been a steep decline over the last decade – starting way before the recession.  The analysis draws on several datasets, but the overall conclusion is clear:

“average training hours per week per employed person – the closest indicator of training volume – fell substantially between 1997 and 2006, then continued to fall until 2009. The best estimate of the decline in volume over the 1997-2009 interval is from 1.24 to 0.69 hours per employed person, a startling cut of 44%. ”   This is not quite the picture of a country forging its way towards a C21 knowledge-based economy.

Along with this goes an estimated real-terms cut of 14.5% in the annual investment made by employers in training.

It might not all be gloom, though.  Green and his colleagues don’t leap to the conclusion that everything is going pear-shaped.  We know, for example, that more employers are integrating learning more closely into the job, so that ‘training’ as such might be less distinct from daily work and so less reported.  The  training may just be becoming more efficient.  That said, it’s a fairly challenging overall trend.

What has the decline meant for different groups?     Overall it’s the usual picture:  higher shares of training going to those already well qualified.  There’s a particularly sharp drop amongst young people, by 49% for those under 25 compared with 22% for those aged 35 and over.

Now for gender;  are, as the PP suggests, women still increasing their human capital at a greater rate than men, by engaging more in training?  The answer is definitely yes.   If we first take the crude participation rate, women were nearly two points ahead in 1995 (13.8% to men’s 12%) and this had increased to nearly 3.5 by 2012 (14.9% to 11.5%).   Now the decline in average duration has affected women as well as men, so in spite of their increased participation rate the total volume of training has dropped for women, by a very substantial 37% over the period.  But for men the decline has been even greater – 46%.  So the competence gap goes on widening.

Green and colleagues conclude with a plea for better indicators which give us a less crude idea of what is happening with training, beyond crude participation rates.  Amen to that – and of course we need to know much more about what effect this learning actually has on careers….