Thursday, 22 July 2010

Having a go in the dark

Michael Taft: That the Sunday Business Post has another go at public sector workers should hardly be news. Every time they mention public sector workers, whatever the context or story, they are apt to attach the rider ‘highest paid in the Europe’. Now they are doing one better with the headline ‘highest paid in the world’. Their front page story is based on data leaked from a National Competitiveness Council report (a convenient leak as discussed over at Cedar Lounge Revolution) to be published soon but, in truth, the data they use is pretty old hat. No doubt, any attempt to deconstruct the data will be ignored – it has in the past. But let’s take up the cudgels one more time and look specifically at the health sector. For any fact-based analysis shows the newspaper headlines to be wide of the mark; indeed, it shows that they are missing a real story that should be pre-occupying us.

The story is probably close to the mark when comparing high-end professionals – the consultant doctors etc.; there is some data to show they are paid above European norms. Of course, this is part of a general story – in both the private and public sector – of phenomenally high-pay at the upper ends which distorts averages. This is why Ireland compares so badly to other EU countries in wage equality statistics. So there’s not a whole lot new there.

The story goes on to quote health statistics from the OECD Health at a Glance reports which, in the words of the OECD itself, should be ‘interpreted with caution’. And for good reason. Let’s look at hospital nurses pay in the 2009 publication. First, this does not measure hourly labour costs – which is the true determiner of cost to the employer; in the case of public sector employees, the Government. Second, it doesn’t compare like with like. For instance, data sources vary from country to country (some include private sector, some include part-time, some include non-hospital staff, some omit certain grades, etc.) – reflecting a real problem in making comparisons that can stick.

Third – and this is a problem throughout OECD wage statistics – most data is compiled on the basis of taking the total amount of wages and dividing it by the total number of employees to get, as they put it, ‘the average gross annual wage’. However, this can be highly misleading. Take this comparison from the EU Klems database (which does measure hourly labour costs).

• In Ireland, annual employee compensation (including non-wage costs) was €44,800 in the whole economy in 2007. In the Netherlands, it was €37,750. On that basis you’d say – wow, we are really high paid compared to the Dutch.

• However, when hourly labour compensation is examined we find the situation reversed. In Ireland, hourly labour costs were €24.82; in Netherlands it was €28.08. What accounts for this discrepancy?

Simple: the Dutch work less hours per employee. In 2007, each Dutch employee worked, on average, 1,344 hours per year; the Irish worked 1,805. Not only is the average working week lower in the Netherlands, there is a higher level of part-time workers.

So if you compare average annual incomes you’re likely to fall into this mistake. So when the OECD puts nurses’ average annual income in 5th place among countries surveyed (including low income Greece, Mexico and the Slovak Republic), it only tells us so much.

Where the OECD methodology is more helpful is when it compares one set of workers with another set in the same economy, since the measurement is internally consistent When this is done, it shows that Irish nurses’ making the same as the Irish average wage – which puts them 15th out of 19 in terms of intra-national comparisons. In other words, in 14 countries nurses make more in relation to the average wage in their own country. On this comparison, Irish hospital nurses are not raking it in.

So is there somewhere we can go to find a robust international comparison of labour costs in the public health sector? Unfortunately, not yet; though the OECD is trying to establish an internationally agreed base-line. However, we can turn to the EU Klemsto find a story that the Sunday Business Post might wish to investigate; though I suspect if they dug too deeply it might lead them to a politically unpalatable conclusion (unpalatable for them).

Hourly labour costs in the Irish health sector were €34.57 per hour in 2007; in the Netherlands it was €25.84. This, again, might lead us to conclude that Irish public health sector workers are, indeed, costly.

But there’s an odd trend here. In 2000, Irish labour costs in this sector were €19.74; in Netherlands it was €20.51. What explains this extra-ordinary growth and turnaround? Of course, we can always put it down to greedy public sector unions, benchmarking, milking the taxpayer, etc. However, this simplistic explanation doesn’t add up.

Given that all public sector workers were covered under the same wage agreements, the same benchmarking deals, we should – according to the greedy-public-sector-worker thesis – expect to find similar increased costs in the Public Administration sector. Except in that sector (which employs a third of all pubic sector workers), nominal hourly labour costs increased by only €6.29 per hour compared to health sector costs which increased by €14.65 per hour.

How could this be? The problem with measuring labour costs in the health sector is that 45 percent of labour is in the private sector. Unfortunately, we don’t have a breakdown between public and private health sectors. But what we may be seeing is substantially increasing labour costs arising from the costly, socially perverse and economically inefficient interpenetration of the public and private in what should be a free and public good. In other words, if public sector health workers’ labour costs have increased in the same manner as other public sector workers, the issue doesn’t lie in the public realm.

It’s hard to know; and that’s the problem. Labour costs in the Public Administration sector is lower than average Eurozone costs; ditto for the Education sector. But the health sector – which is split between public and private – significantly exceeds Eurozone averages.

We do know that of those private sector health enterprises which granted wage increases in 2009, the average increase was 15 percent; for public sector health enterprises it was 7 percent. In the heart of the recession, incomes in the private health sector were rising faster than those in the public sector (though we don’t know that nurses pay was rising this fast in the private sector).

And we do know the EU/CSO shows that between 2003 and 2008 the top 10 percent households took 73 percent of the total gross PAYE income increase in the State. Not only is Ireland suffering from wage inequality, it is getting worse – and I suspect that, public or private, not too many nurses feature in the top 10 percent.

Here is a real story – costs are increasing in public services where the private sector is playing a significant role; and income is rising faster in households which have a disproportionate number of high-end professionals. The Sunday Business Post might want to investigate this further. But I will give them this warning.

They might have to conclude that we need a truly public health sector where goods are delivered, not on the ability to pay, but on need; and we might have to do something about those high incomes.

In the meantime, for want of any analysis that goes beyond dubious headlines, repeat after me: ‘Surely, gosh, we have the highest paid public workers in the whole wide world.’


Anonymous said...

Michael, there are a couple of bad errors/omissions in your fact-based analysis:

1. You compare average hours worked per year in Ireland and the Netherlands, but you must know that there is there is a much wider gap between public and private sector hours worked in this country than there would be in Holland. So in a sense you attempt to justify high public sector wages by appealing to the long hours worked mainly in the private sector.

2. You attribute the higher rate of increase in hourly health sector wages than those in public admin by appealing to supposed higher increases enjoyed by private heath workers. Yet you don't even consider the obvious impact of the massive changes in core hours that occurred in the health sector in the past decade (nurses's working week reduced to 37.5 hours, other workers impacted by the European working time directive).

Michael Taft said...

Thanks, Anonymous, for your comments. On the first point, I didn't use the average hours worked comparison between Ireland and the Netherlands as an argument re: public sector labour costs. I merely used it as a cautiounary example against using 'average annual incomes' - regardless of the sector. I should have made that point clearer in the post.

As to the second point - you raise a good point. No, I hadn't factored that in. Obviously, if work hours are reduced but overall wage levels remain the same, then the employment of extra labour to take up the extra hours will result in an incrase in labour costs. However, that only holds if the employer does that; to the extent that they cut back on hours or squeeze producivity, that reduces the basic calculation. But the overall point is correct (I wonder if other countries' health systems were affected to the same degree as Ireland by the implementation of the working time directive).

I would point out, however, that I was not conclusive - 'hard to know' is the phrase. I'm only suggesting that there might be alternative and more complex factors involved, than merely 'public sector workers are over-paid'. And, when using NACE codes for health, we have to accept that we looking at a sector which is about 50/50 between public and private.