Wage puzzle as jobless rate falls but rises remain static
Things are happening that never used to happen before, to misquote John B Keane. In the wake of the Great Recession, some of the laws of economics, as well as of politics, appear to be breaking down.
Doubtless you are familiar with Okun's Law, which basically links changes in unemployment to increases and falls in output. It seems like common sense but it is proving difficult to apply it to Irish output.
After that comes the one which relates wages to unemployment. The lower the jobless rate, the higher the wages employers will be willing to pay to attract and retain scarce workers, it says. But as unemployment tumbles, that does not seem to be happening.
It may seem strange to make such a claim amid threats of a rail strike but there are few workers with the industrial muscle of train drivers. Taking the country as a whole, the recovery in output and employment does not seem to have been matched by wage growth.
It is a pattern repeated in other places, including Britain. Each individual country is different and it seems unlikely that the same factors are at play in all of them.
In the UK and USA, there was no collapse in employment as happened in Ireland and the changing nature of jobs is often given as the main cause of the "wage puzzle" with casualisation, globalisation and technology all taking a share of the blame.
The Irish situation is especially curious because of the extraordinary rebound in employment after the collapse. The numbers in work have just returned to the peak recorded in 2007. This means all of the 300,000 jobs lost after that year have been replaced - but not the same jobs, and not at the same wages.
A new research paper in the Central Bank's recent quarterly notes that there has been little or no growth in hourly wage rates across the economy as a whole, even as unemployment was falling like a stone. The increase in national personal income is mainly due to the fact that more people are working, as well as working longer hours.
Those raw figures certainly help explain the lack of a political feelgood factor, at least until very recently. They also indicate one possible reason why income tax revenues are below Department of Finance forecasts. The research looked deeper into what is going on.
One key question is how close is the economy to full employment? A rule of thumb based on past experience suggests that an unemployment rate below 6pc is the answer, in the sense that anything lower than that will spark inflationary wage rises which damage competitiveness.
It is only a rule of thumb. The paper applies wider criteria, including job vacancies, hours worked and labour force participation (the percentage who are at work or looking for work). Participation rates are much studied by economists but rarely make it into mainstream discussion. Just using unemployment, though, can disguise the success or failure of an economy in providing jobs. Despite all the successes, Ireland's participation rate is low by EU standards.
Even an actual number like that 300,000 new jobs does not tell the whole story. The percentage of men working or seeking work, having fallen rapidly after the crash, has not fully recovered. The decline in female jobs was smaller and employment is back to 2007 levels.
The reason for this gender gap is plain enough. Construction is not anywhere near peak levels and may never be again. The increase in non-construction jobs has been rapid, but it has to be to offset the loss in male employment in the building industry.
It is a bit more of a surprise to find that the proportion of non-Irish workers is back to peak levels of around 16pc, given that so many of the pre-crash jobs for non-nationals were in construction, but immigrants go where the jobs are.
Most sectors display similar levels of vacancies and wage growth but there are some notable exceptions. Professional positions and finance have vacancy rates of more than 2pc and wage increases greater than 3pc. The health sector looks unusual, with a higher-than-average vacancy rate but lower-than-average wage rises.
From a policy point of view, the most important finding may be the one on how wages are affected by unemployment when the distortions in GDP caused by multinational activities are excluded by using some of the new measurements of the economy devised by the CSO.
These show that, while very low unemployment and very high unemployment have the expected effect of accelerating and depressing wages, there is a wide range in the middle where changes in the jobless rate seem not to affect the behaviour of wages.
The range in question is where unemployment varies between 5pc and 10pc. Obviously, it would never be policy to push unemployment above 10pc, but should policy try to prevent it falling below 5pc to stay in this 'Goldilocks' zone, which is neither too hot nor too cold?
That is now a pressing question. The research suggests there is still some slack in the labour market if one goes beyond the quite restrictive official definitions. On that basis, what might be called the "shortage of work" rate is closer to 10pc than the 6pc unemployment recorded in the statistics, but even that is approaching pre-crash lows of 8pc.
All of which suggests we may be getting close to full employment. It is not possible to say exactly where that point may be, because one cannot be sure that the future will be like the past. Too many things have changed.
Back then, national wage agreements were often replicated by firms which had not officially signed up to them, so that they did indeed create a national system. That has gone. Now, researchers find evidence of new kinds of wage setting, such as "pattern bargaining", where a leading firm does a deal which is followed by others in the sector.
If that changes the old pattern where public sector deals set the trend, it could make quite a difference to the overall picture. Another obvious difference from the last decade is the near-zero inflation, where smaller cash increases are needed to provide the desired gain in real wages.
The crash had its own impact. A startling piece of evidence, mentioned in a previous column, was that the wages of those hired during the recession and early recovery were 20pc lower than in the same jobs before the crash.
Lost jobs bore the brunt of the adjustment rather than the tremendously difficult alternative of cutting actual wages. The end result may be much the same but it does seem a pity that this appears to be the way we prefer to go about it.
The Central Bank researchers reckon their analysis allows them to make some short-term forecasts, which say unemployment will fall below 6pc next year and average hourly earnings will grow by 2.8pc; up from 2.2pc this year and giving an overall increase of 3.2pc when extra hours are added in.
That is still pretty restrained, but if Government medium-term forecasts are right, the economy looks like heading for full employment from 2019, with all the risks that might bring. On the other hand, as we know all too well, no model can forecast what will happen after April 2019. A hot economy may cool rapidly for reasons which have nothing to do with us.