Monday, 18 November 2013

We overpay, too

Last week I posted about some new research from European Commission economists showing the existence of pay premiums in a range of EU countries in favour of public sector employees relative to private sector employees, which did not disappear when the compositional nature of the two workforces was corrected for (the public sector, for example, tends to have more-educated people on board). Like for like, the public sector paid more, sometimes substantially so.

Perhaps naively, my first guess about New Zealand (what I might call my "prior", if "prior" means knowing nothing at all about the local issue or its literature) was that we might not replicate the typical EU pattern, and I asked if anyone knew whether there was local research on the topic.
And Eric Crampton over at Offsetting Behaviour did - thanks, Eric.

Eric had blogged about the issue two years ago, referencing two papers by Prof John Gibson at Waikato, one based on 2005 data and the other looking at 2003-07.

The result, I'm afraid, is that we appear to have a public sector pay premium, too. The first paper found (p63) that "Taking account of a wide range of worker characteristics and attitudes, job attributes, and the effects that jobs have on workers and their family life, there appears to be a pay premium of 17-21%, which is not due to compensating differentials". And the second one found (quoting from the Abstract) that "Comparing with observably similar private sector workers shows that public sector workers have received a pay premium that has grown in each year, from almost zero in 2003 to 22 percent in 2007". You might well think that the expanding premium had something to do with the government of the day: I could not possibly comment.

I introduced Gibson's results by saying that we "appear" to have a public sector premium. The reason I'm a little cautious is that the "like for like" comparisons between public and private sector jobs aren't quite as comprehensive as they might be (due to data limitations, nothing to do with how Gibson went about it). The European Commission research was able to correct for different occupations and different levels of managerial responsibility, whereas the Gibson comparisons weren't (the closest was a 'years of education' criterion, which could be a rough proxy, but isn't the genuine article). I wonder if anyone's minded to have a go at an update, with newer data sources - Statistics New Zealand has now got very extensive linked employer-employee datasets that you'd think would allow a very fine comparison of public and private sector pay for comparable work.

Why does any of this matter?

For starters, efficiency reasons: there's no reason for us to be throwing taxpayers' dollars out the window for work that's being overpriced. Especially on this scale: the premium is a really big number. Just looking at the core public service alone (government departments and the like), a premium of 20% across the wage bill is in the region of $600 million (44,500 FTEs at an average base salary of $68,561 adds up to a payroll of $3.05 billion, going by the numbers in the latest Human Resource Capability Survey from the State Services Commission).

And then there are the fiscal stabilisation issues. Like a lot of other western governments, we're trying to get our fiscal affairs back into order. If a public sector pay premium exists on this scale, then reducing it could be a more preferable way of helping to balance the books than (say) outright cuts in services valued by the public. In fact, we may be headed in this direction de facto: quite a few outfits around the public sector are being held to baseline budgets for the next three years that are frozen in nominal dollar terms. That doesn't prevent further public sector pay increases, but it makes them harder to concede and facilitates eroding the premium.

Overgenerous public sector remuneration can also be on the of the ingredients to walk you into fiscal problems in the first place. I mentioned last week that the European Commission research found that the fiscally challenged PIIGS (Portugal, Ireland, Italy, Greece, Spain, and it's my PIIGS terminology, not the Commission's) all showed up badly for overpaying the public sector. Since then I've read a European Central Bank working paper, 'The public sector pay gap: in a selection of Euro area countries'. It's not quite as recent as the Commission research, and doesn't cover as large a sample (10 countries rather than 26), but it found exactly the same thing (p21) : "Notable differences emerged across countries, with Greece, Ireland, Italy, Portugal and Spain exhibiting higher public sector premia than other countries".

It's no comfort then to observe that even on a low-ball estimate of the public sector premia in the PIIGS (the ones in Table 9 of the ECB paper), they ranged from 10.9% (Portugal) to 17.2% (Spain). They're lower than the numbers Prof Gibson found for New Zealand.


  1. I would have my doubts about the Gibson results (speaking as an ex public-sector employer), and I am not going to study them too closely either since his results are nearly always so many years behind the times that, well, "who cares". The really large employers in the State (e.g for teachers) have no like-for-like comparison in this country. For the record private schools just build off state pay rates and add a small premium. Lawyers in the state get paid overall less than in private practice (though not always early on in their careers). And so many jobs in the State are just jobs like child welfare and caring for sick and vomiting hoons on a saturday night, that nonone in the private sector wants to do.night

  2. Thanks for the comment, I appreciate the feedback. As I said in my initial post, my own (limited) experience of hiring or helping to hire people in the public sector would also have led me to believe that there wasn't a public sector premium in NZ, you're right about lawyers, and I can add the same would be true of many public sector economics and regulator functions, who wouldn't be paid what they would get in a private sector consultancy. But then the international evidence seems to be that where there is a premium, it tends to be in the lower-level jobs, and I have no personal feel for that, so all I can do is go by whatever research there is (and try and put interesting academic research in plain English, which is kind of the point of much of my blog). I agree too that 2007 is getting a bit dated, and there have been sharper pencils in public sectors around the world since then. Your point about good comparators being hard to find for some activities, that sounds right, too. I suppose what strikes me as a bit odd is that there is this immense, very detailed database available that doesn't seem to have been mined at all to throw light on questions like these


Hi - sorry about the Captcha step for real people like yourself commenting, it's to baffle the bots