Saturday, 30 January 2010

Memo to IBEC

Michael Taft: Yesterday evening I was on Matt Cooper’s The Last Word with a representative from IBEC discussing wage levels. I quoted the numbers from the US Bureau of Labor Statistics, Eurostat and Destatis (German Statistical Board) to show that Irish labour costs are not high; indeed, they’re rather low by comparison with our EU partners.

The IBEC spokesperson insisted, however, that Irish wage levels are high – 15 percent higher than the EU-15 average. He quoted from the EU Commission’s AMECO database. I had no wish to get into an argument over this database or that; or get into a detailed deconstruction of the AMECO numbers. I just said I would put up the sources on this blog and let people decided for themselves.

Below I present the data and links. Then I look at the AMECO database. For it is the only one IBEC spokespersons use – and in doing so they are knowingly misleading the debate over our wage competitiveness.

Destatis: The German statistical board, using Eurostat data, presents the most recent numbers from 4th quarter 2008. They show, using hourly labour costs, that:

• Irish private sector wages are 1 percent below the EU-15 average (including lowly Portugal and Greece) and 14 percent below the average of our peer group – the other top 10 economies.

• Irish manufacturing wages are 2 percent below the EU-15 average and 16 percent below our peer group’s average

US Bureau of Labor Statistics: this database – based on hourly manufacturing compensation costs (including employers’ social security contributions) - is up-dated to 2007. This shows that:

• Irish manufacturing labour costs (including management salaries) are 3 percent below the EU-15 average, excluding Luxembourg and 16 percent below the average of our peer group.

• Labour costs for industrial workers (excluding management and clerical) are 3 percent below EU-15 average and 18 percent below our peer group.

These two databases are based on the actual cost of labour to employers on any hourly basis. This is the better type of measurement of costs in an economy.

OECD Benefit and Wages: this database, which measures private sector wages (NACE C – K) has a number of defects. First, it is not a measurement of labour costs but rather an attempt to identify annual wages. However, it acknowledges that some of the countries data may not include managerial and supervisorial wages, therefore under-stating some numbers. It also acknowledges that some countries data do not separate full-time and part-time wages (which we will see below can distort numbers). For Ireland, the figure is the average wage for production workers – not all private sector workers.

Especially curious are database figures for Irish wage across the years - showing inexplicable jumps:

2003: €33,939
2004: €27,781
2005: €39,206.

In 2007, the OECD shows Irish wages barely changing – up less than €300. In fact, in some previous editions of the 2007 database, average Irish production worker wages were much lower, below €33,000 – which is consistent with CSO data.

With these caveats, the database shows that average annual private sector wages in Ireland are 12 percent above the EU-15 average but 2 percent below our peer group average.

However, given the statistical inconsistency and methodological shortcomings, one should be extremely cautious about citing these numbers.

EU Commission AMECO: this database measures income but not hourly labour costs. Rather, take an ‘aggregate GDP’ approach. Essentially, they take the total amount of wages, salaries, bonuses, social security contributions and divide them by the number of workers. There are two problems with this:

First, it does not distinguish between full-time and part-time for most countries. Ireland has one of the lowest proportions of part-time workers. This can skewer the wage data (Germany, for instance has nearly twice as many part-time workers as a proportion of their workforce as Ireland). Let’s say that an employer needs 1000 hours worked with a total wage bill of €20,000. If that is divided up among 50 part-time workers, they will 20 hours a week with an average pay of €400. However, if that work is divided up among full-time workers, the average pay will be €800 per week. Whatever about the pay of different workers, there is no cost difference to the employer and no difference to wage competiveness.

Second, it does not distinguish between hours worked. Ireland has one of the highest levels of hours worked in the EU-15. If an employer needs 1000 hours worked with a total wage bill of €20,000 and divides it up among workers on a 40 hour working week, there will be 25 workers earning €800 per week. However, if that same employer divides up the working time a 35 hour working week, there will be 28.5 workers earning €700 per week. So, while there is a difference in wage, there is no difference in cost to the employer and no difference to wage competitiveness.

An aggregate approach rarely makes these distinctions. It can sometimes use a ‘full-time equivalent’ measurement – but AMECO acknowledges it cannot do this for all countries.

This is IBEC’s database of choice. As the IBEC spokesperson on the Last Word said – Irish wages would appear to be 15 percent higher than the EU-15 average. And this is what the AMECO database produces.

But does this figure measure hourly labour costs? No. Does it measure wage per hour worked? No. Does it measure the total amount of wages in the economy per total working hours? No. This database tells us what it tells us – and it tells us very little in terms of labour cost competitiveness. If the data compensated for hours worked and part-time employment, the figure would approximate the data from Destatis, Eurostat and the US Bureau of Labor Statistics.

No wonder that when our labour costs are examined with the proper measurements, other observers come to the same conclusion. The National Competitiveness Council stated:

‘Irish pay and income levels are moderate when compared to other developed high income economies . . .‘

Forfas’s report on the retail sector showed that average Irish wages are low in comparison with the Dutch retail sector.

This is more than just an argument over numbers and methodology. This is about identifying what exactly is wrong with the Irish economy and, from that, constructing policies to address the defects.

But IBEC is not interested in that. It is intentionally distorting the debate in accordance with its own agenda. Their use of wage statistics is deliberately misrepresentative.

Very simply, IBEC should stop it.


kirneh98 said...

What time were you on Matt Cooper yesterday? I'd like to listen to the interview.

Michael Taft said...

The 5:00 slot on Friday's Last Word. The link should you lead you there.

Rory O'Farrell said...

There are two issues:

1) Only labour costs per hour or per unit are worthwhile comparing.

2) Why bother with AMECO at all, all their data is taken from Eurostat and national sources anyway.

But then why let the truth get in the way of a good story?