Lies, damned lies - and extracting simple messages from complex statistics

Beware the over simplification of statistics, especially when carried by the right wing press in the middle of TUC conference. The Telegraph is one of many which carries the result of Office of National Statistics research into public and private sector pay. ”Myth of the underpaid public sector worker” it screams following up with a report entitled “The trouble with the public sector is bone-idle staff”. Nice.
According to the research described as “the first comprehensive analysis of the pay divide by Britain's national statistician”, the average weekly salary for public sector workers in April last year was £539, compared with £465 in the private sector and their pensions are better too.

Yet reference to ONS figures collected over many years for public and private sector pay rises show that neither has significantly outstripped the other over a time frame or ten, twenty or even thirty years.
How can that be so? Unless public sector pay has always been better (and no-one seems to be arguing that) how can this disparity be explained?

Consider the simplistic example below

You work in a small office next door to another small office. You are paid £30,000 (significantly more than the public sector average). The other person in your office is paid £20,000 (approaching half of public sector workers earn this figure or less). The office next door also has two people undertaking the same tasks for the same pay. The average pay in each office is £25,000.

One day it is decided that your co-worker should move next door – even though they will continue to work with you in the same way. Despite a few problems, work continues unabated until you are summoned by management to be told that you are uneconomic because average pay in your office has risen to £30,000 (from £25,000) whilst average pay in the next door office has fallen from £25,000 to £23,333. Your pay hasn’t increased but your office average has at the same time as that of the next door office has reduced.
A more complex version of this is at play when changes in relative public/private earnings are considered. Over the past few decades a whole range of catering, care, cleaning and other work has been contracted out. And as the example above illustrates, as soon as you transfer low paid staff from public to private sector you increase the relative pay averages in the former.

Even though they are hardly over-paid, many professionals in public service receive more than the average wage. So the immediate effect of increasing the proportion of teachers, nurses and police in the public sector workforce is to increase public sector pay relative to private sector pay.

Of course there are still some care workers, cleaners and catering staff directly employed in the public sector. And on average their pay is better than in the private sector (but very often below the Living Wage). Some also benefit by making contributions along with the employer to modest pension schemes. (The benefit is of course shared as these are precisely the schemes which will in the future limit the payout from the public purse for elderly benefits.) Paying above poverty wages for these workers does increase the public sector pay bill – but would we have it any other way?

And there’s a final factor. Although the introduction of the minimum wage should have seen a bit significant jump in private sector earnings it has been at least offset by the clustering of minimum wage pay in certain professions – retail and hospitality being the largest. The consequence, along with the weakening of the manufacturing sector is a growing army of low paid private sector workers anchoring private sector pay and significantly increasing wage inequality.

It would be so much more straight-forward of course if we could compare like with like. But which private sector profession would we choose to compare to the pay of a teacher or a nurse such that we could elicit meaningful results?

No-one has come up with an answer to that, because it is not that simple.

Dave Moxham - STUC