Share

We need to talk about wages

Posted on November 25, 2016 by Tom Healy

Tom Healy, Director NERI
Tom Healy, Director NERI

Did you know that wages – as measured by real average weekly earnings – have not increased since 2008 in the Republic of Ireland ?  Never mind the Great Leap forward in GDP last year or the rapid and very welcome growth in employment (continuing to run at an annualised rate of close to 3%).   A release of data on wages is not accompanied by a media one-day major focus such as happens in the case of employment numbers (including those released last week).  The only regular official source of information is in the little heard of and even less talked about EHECS – Earnings, Hours and Employment Costs survey conducted in each EU member state.  EHECS is a survey of enterprises about how much their employees earn (as well as how many hours they work and what employers pay in terms of social insurance). Predictably, the focus of any little public attention on the latest wage data is on the public sector and especially particular sub-sectors within it and comparisons with the private sector.

What do the latest data on earnings tell us?  The most meaningful measure of wage movement, in my view, is average weekly earnings (AWE).  Implicitly, it combines information on hourly pay rates and number of hours worked. In gross terms (before deduction of taxes) average weekly earnings came to €706.9 per week in the third quarter of this year (seasonally adjusted time series).  Chart 1 presents the trend in AWE since January 2008.  Nominal wages have not changed much since 2008 (when the figure was 702.3). However, we need to factor in price changes. In other words, we need to consider what a given wage average will buy.  Choice of price index does matter. For simplicity and ease of reference I use the Consumer Price Index (CPI).  Chart 1 plots this index since 2008 and, also, the value of ‘real’ average weekly earnings when the CPI is applied to the nominal figures.  What the chart shows is remarkably little change in the overall level of prices over the period as a whole with a temporary fall in prices in 2009 (triggering a rise in real AWE in that year) followed by a modest recovery in prices since then. Some goods and services have increased in price while others have dropped (for a light technical elaboration on price indices see postscript to this blog). 

It is surprising that wage trends receive such little attention in media discourse, here. Last week, following the Chancellor’s Autumn Statement in the UK, the direct of the Institute for Fiscal Studies, Paul Johnson commented that:

On these projections real wages will, remarkably, still be below their 2008 levels in 2021. One cannot stress enough how dreadful that is – more than a decade without real earnings growth. We have certainly not seen a period remotely like it in the last 70 years. [source here ]

That wages, in the UK, have been stagnating in real terms has been a major focus of public debate there. Yet, in the Republic of Ireland, the statistics of wage changes receive remarkably little attention apart from the predictable obsession with the public-private gap.  There is very little discussion of trends in average wages across the entire economy or of earnings dispersion within the public sector and within the private sector. There is even less discussion about trends in income received by households from capital or self-employment. This refers to income from propriety self-employment, rents, dividends and other investment sources.  In a previous Monday Blog I discussed trends in the ‘wage share of national income’ [ A worrying trend in wages ].  Measurement of wage share in national income (whether via GDP, GNP or GNI) is problematic – as we know only too well – due to the activities of tax-avoiding multinationals and associated activities.  Another approach to examining the share of income going to ‘labour’ and ‘capital’ is to focus, only, on ‘primary income’ received by households. Primary income is total household income before deduction of taxes on income and before addition of social transfers.

Combining data from various sources I have derived a time series for household income showing the shares of different components. I have used reported CSO statistics for 2000-2014 and estimated or projected forward to 2015 and 2016.

Household income other fhan from employee labour shows much more volatility – especially in the course of recent years – in response to the huge economic shock of 2008-2010 followed by a slow and painful recovery for most businesses as well as workers. However, the nature of the shock and its impact as well as the distribution of recovery gains has been uneven.  Yes, non-employee income took a major hit in 2009-2010. However, there has been a rapid growth in household non-employee incomes since 2011. This may be seen in Chart 2 which shows the growth in Household ‘Primary Income’ and its two components from 2000 up to 2016 (with 2015 and 2016 estimated).  For data up to and including 2014 I am using CSO data. For 2015 and for 2016 I am using a combination of sources including Department of Finance (for taxes), European Commission Ameco database (for estimated growth in employee compensation) and the ESRI Quarterly Economic Commentary (for growth in ‘Other non-agricultural household income). What is striking about the recovery period is the rapid growth in non-employee income – about which we hear next to little in public discourse.  The projected growth in total employee compensation (which includes employer social contributions) in 2015 is set at 5.7% and in 2016 is set at 5.1% using European Commission figures. These assumed or estimated growth rates will be probably higher than the final figures.  Since employment is growing at an annual rate of 3% and there is no evidence that average weekly earnings are rising up to now I reckon that total employee compensation is growing at less than the rate projected by the Commission in these two years.

However, going with the projected commission figures we can calculate the employee labour share of total primary household income. The share of employee compensation has risen and fallen over the last 16 years.  The share has been on a steady downward trajectory since 2009.  On closer examination, it emerges that total non-employee income of households will come to an estimated €27.3 billion in 2016. This compares with €21.3 billion in 2008 just at the beginning of the recent recession. By contrast, total employee compensation fell from €80.7 billion in 2008 to €78.4 billion this year. The fall in the absolute level of employee compensation is related to two trends:

  • A fall in employment over 2008-2016 (the 2016 level is still not quite back at the 2008 level yet)
  • Stagnation in wages (or employee compensation) over the period as a whole.

The contrast with other forms of household income is stark.  In 2015, the growth in household income from capital or self-employment was just over 15% compared to an estimated 5.7% for employee compensation (I estimated that the true increase was probably closer to 3.5%).

 

Why, it may be asked, are wages stagnant or even falling in some sectors while pay settlement increases of 2-3% are typical across many sectors and occupations? This is an important question and deserves more careful analysis than is possible in this short blog. Suffice it to say that, on the basis of employment trends and earnings data it seems that the average wage is being held down as a result of changes happening in the structure of the economy as well as the composition of the workforce in employment.  In particular, it is highly likely that younger less well paid workers are gradually replacing retiring better paid workers. This process of replacement seems to be outweighing the impact of pay progression in many employment (where, for example, workers may receive increments linked to service or performance). It is striking that, in the case of the public sector the recent EHECS data shows very modest increases, on average, in earnings in the 12 months ending September 2016. This must reflect significant structural and compositional changes within the public sector. It may also be noted that average weekly earnings are down by approximately 1% over the same period in both the civil service and defence sub-sectors (Table 8a of EHECS). On the other hand there has been relatively strong wage rate growth at around 5% in the 12 months ending September 2016 in construction and some professional and scientific sectors of the economy.

In conclusion, we are confronted with four challenging developments:

  1. Wages in the republic of Ireland are stagnating despite a rapid recovery in corporate profits (not discussed in this blog) as well as household income from capital.
  2. The recovery in wages reflects different pressures across the economy with some sectors still falling but most stagnant or increasing modestly.
  3. Where modest wage growth is happening recovery in the public sector still lags that in the private sector.
  4. Total household income - before taxes and social welfare - is up by €4 billion from €102 billion to €106 billion since 2008. At the same time wages are down in absolute terms.

Houston we have a problem.

  postscript

The Consumer Price Index (CPI) is different to the Harmonised Index of Consumer Prices (HICP). The HICP, which is used as a harminised measure of consumer price inflation across the EU, does not include the cost of mortgage interest payments by households whereas the CPI does. In very recent times, the CPI and HICP have moved closely together in the case of the Republic of Ireland.

Because the CPI is based on fixed weights for a given basket of goods and services it is not sensitive to changes in actual consumer behaviour in response to price changes (what economists call the impact of ‘substitution’ of cheaper goods for more expensive goods). As readers will not need to be reminded the cost of renting accommodation has risen sharply in recent times and this is not reflected in the weighting given to accommodation costs (as distinct from the actual measured increase in rental costs in the CPI). The CPI is based on historical weights which are derived from the Household Budget Survey.  In time, the CPI will align with trends in the actual price of consumer goods bought by consumers. This is why, for the purposes of medium-term projections forward use of the CSO national accounts personal consumer expenditure deflator is best for the purposes of assuming cost of living changes in the future. 

In the very short-run choice of price index does make a difference. In the medium term it makes little difference. In the long-run it makes no difference (at least in theory!).

Digital Revolutionaries