The myth of over-paid public sector workers

If Mitt Romney had won the 2012 presidential election his problems would really have started. An incoming president gets to make several thousand political appointments. The individuals have to be identified and security cleared and many of them have to be approved by the Senate. This all takes time and represents a major advantage for an incumbent president doing a second term, who already has many people in place.

The last three presidents – Messrs Clinton, Bush and Obama – have all served two terms. Bill Clinton had lots of problems in his first year in office as it took so long to get his appointments in place during his attempt to pass healthcare reform. In part this is the reason why there is a gap of a couple of months between the election in November and the inauguration in January.

The current system in the UK means that the transitions between governments are much more seamless and best left alone. Civil servants perform many of the tasks in the UK that political appointees perform in the US and the system appears to have worked well for a generation. If it isn’t broken don’t fix it.

The recession has given the Coalition a handy excuse for reducing the size of the state. David Cameron, for example, made it clear he thinks civil servants are “the enemies of enterprise”.

It is unsurprising then to see continued attacks on public sector workers who have been fired at an even faster rate than planned. Public-sector employment is down by 628,000 since June 2010 and is expected to be 1.1m lower by 2018 if Chancellor George Osborne has his way.

The attacks on the public sector are broad based. Francis Maude, Cabinet Office minister, apparently wants a bigger ministerial role in order to increase the accountability of civil servants. For example, Mr Maude has been pressing for secretaries of state to be allowed to choose their most senior civil servants from a shortlist of approved candidates. This apparently hasn’t gone down well with Sir Humphrey who has booted the proposal into touch. Mr Maude, pictured above, of course, was the clot who caused panic earlier this year by advising people to fill up their vehicles and to store fuel in their garages in jerrycans because of a non-existent fuel shortage.

In the Autumn Statement the Chancellor announced that he is imposing an extra two years of real pay cuts for millions of public sector workers. Plus, in another sop to the right-wing of the Tory party that thinks the UK’s labour market is inflexible even though it isn’t, Mr Osborne announced he plans to end national pay bargaining for teachers within two years. Teacher unions are not going to accept this without a fight. Half of all workers in education are union members.

But how does the pay of public-sector workers compare with those in the private sector? Public-sector pay acts as an automatic stabiliser as the economy slows, declining less in recessions than private-sector pay and is especially important in high-unemployment areas.

The evidence from the Office for National Statistics (ONS), Average Weekly Earnings (AWE) survey shows the public/private gap in weekly earnings is small and has changed little recently – in September 2010 it was 4 per cent, in September 2011 it was 4.1 per cent and it was 4.5 per cent in September 2012. AWE in the public sector in September 2012 was £488, up by 1.7 per cent this year compared with £476 (+1.9 per cent) in the private sector, excluding financial services.

In a paper published last week, the ONS examined the public/private hourly earnings pay gap in 2011*. Some of the pay differences between the two sectors they found were attributable to factors such as location, tenure and size of workplace or organisation. Big employers pay more, there is a London premium and workers are paid more the longer their tenure.

Using a so-called regression model, the ONS attempted to “control” for differences in worker characteristics between the sectors. It used data obtained from employers from the Annual Survey of Hours and Earnings (ASHE) and controlled for gender, age, occupation, the region that the job is located in, full time/part time, permanent/temporary, job tenure and size of organisation. It found public-sector workers earned 14.9 per cent per hour more than private-sector workers, but once they controlled for the above factors the differential dropped to only 2.2 per cent more per hour (excluding overtime).

It turns out such estimates are sensitive to not including relevant variables that are likely to influence pay, especially qualifications which are not available in the ASHE survey. The educated get a wage premium. This matters because the average public-sector worker has a higher level of educational qualifications than the private-sector worker. It turns out that 40 per cent of public-sector workers have a degree or equivalent compared with 25 per cent in the private sector.

A better source of data is the ONS’s Labour Force Survey (LFS) which gives a rather different answer. The LFS has all of the same data as ASHE but also includes highest educational qualifications, which turn out to matter.

I took earnings data on approximately 70,000 employees from January 2011 to September 2012 and also estimated a series of regressions. The results are reported in the table above. The first row reports the raw difference of 22 per cent without any controls. In the second row I report results which “control” statistically for age, gender, job tenure, region of work, hours of work and workplace size and the differential halves to 11 per cent.

In the third row I add controls for highest qualification. When I do that the differential becomes negative: public-sector pay is below private-sector pay, not above it. I then took this specification and estimated results separately by region: public-sector pay gaps are negative in London (-10 per cent) and the rest of the South (-5 per cent), are 0 per cent in the Midlands but positive in the North (+3 per cent) and Wales, Northern Ireland and Scotland (2 per cent).

Contrary to what some commentators have suggested, pay in the public sector is lower than in the private once differences in characteristics are accounted for. Public-sector workers do not appear to be overpaid.