Over at the UNISON website there’s a fascinating new report called Public Sector Pay Premium: Fact or Fiction? that I’d really recommend you read. It’s a model of union-commissioned research – commissioned from Incomes Data Services – and it doesn’t dodge difficult findings in other reports, which makes its findings all the more persuasive.
Every few weeks there’s a newspaper story claiming public sector workers are higher paid than private sector. Of course, Coalition politicians, eager to support the government’s public sector pay freeze, are quick to use these stories. Many of these stories are based on the Office for National Statistics’ Average Weekly Earnings data, published every month.The average earnings figures are a little higher for the public sector. However, as the report notes, the composition of the public and private sector are different – there aren’t many wholesale, retail, hotel and restaurant staff in the public sector but this low paid sector (average total weekly pay £309, compared with overall average of £472) accounts for 23% of all employment.
The report shows that private companies have comparatively large proportions of low and high paid workers compared to the public sector, and this difference has got worse in recent years by all the outsourcing of low-paid jobs in the public sector. These figures do not show that public sector workers are getting paid more than they would get for doing the same sort of jobs for a private firm. One approach that attempts to get round this problem is known as ‘regression analysis‘ and a number of reputable organisations – like the Institute for Fiscal Studies - have used it to study whether there is a public sector pay premium. Regression analysis allows investigators to allow for factors like the generally higher levels of qualifications and experience of public sector workers; IFS calculated that, after making these allowances, public sector workers have a pay premium of 8.3%.
One issue with regression analysis is that it’s very sensitive to the number of factors you take into account. Other studies, taking into account size of organisation and usual hours worked, have found that pay was higher in the private sector. It is difficult to work out what the data will and won’t allow and what factors should and shouldn’t be taken into account. In particular, studies have not taken workers’ level of responsibility into account even though HR professionals and organisations like IDS regard this as essential for benchmarking pay. Using IDS’s pay database and job evaluation methodology, the authors find that public sector pay tends to be higher at lower levels of responsibility, but that at middle and higher levels private sector pay opens up a growing gap.
As I noted at the outset, this is a model for union research – and partly because it doesn’t make an easy case for the answers we favour. It treats the complexity of the arguments seriously and after reading it you can’t just dismiss the studies that have found a public sector pay premium. But it makes a very strong case that these studies are not the final word and trades unionists should not run scared from this issue. My reading is that I’d be uncomfortable now talking about “a” public or private sector pay premium – workers with low levels of qualifications in jobs with lower levels of responsibility may be better off in the public sector than in private sector services, but other workers are lower-paid in the public sector than similarly skilled workers at similar levels of responsibility. At a time when growing inequality and in-work poverty are widely recognised as important economic problems I wouldn’t say that it’s the public sector that is the problem here.
Three of the authors of this report – Alastair Hatchett, Rupert Griffin and Jon Taylor – will be speakers at tomorrow’s TUC-IDS Pay Forum. One of the sessions is on “the use and abuse of pay data”; I’m looking forward to it.