Share

A boring blog about a hot topic

Posted on December 11, 2016

Tom Healy, Director NERI
Tom Healy, Director NERI

With the possible exceptions of GAA sports, soccer and rugby there is no topic more likely to cause intense passion and heat (but very little light) than that of public sector pay in the Republic of Ireland. What follows, in this Blog, is a most dull and dis-passionate consideration of the statistics concerning public sector pay.  Three questions arise:

  1. How much is spent on public sector pay each year?
  2. How does public sector pay, here, compare with that in other jurisdictions?
  3. How does public sector pay compare with that in the private sector?

To begin to do any justice to these questions would require three separate blogs.  Here, I am going to attempt to reduce it into one blog and, hopefully, return to the three individual questions if and when more timely and comprehensive data are made available by the relevant statistical or reporting bodies. I begin with a confession:

There are two ways of measuring public sector pay at this time:

  • The Westminster Exchequer-based ‘cash revenue in’ and ‘cash spending out’ in the current year
  • The Eurostat-CSO approach to government finance statistics and national accounts based on international norms of statistical reporting.

With the greatest of respect and fondness for the practitioners of the Westminster exchequer system of governance accounting who have served us well over at least two centuries I do not propose to use that way of counting and reporting. Rather, I am going to stick with the Eurostat-CSO approach because:

It is internationally standardised. It directly feeds into EU government finance statistics and, thereby, forms part of the reporting procedures for the purpose of compliance with EU fiscal rules which have been incorporated into domestic law, here. 

In saying this I fully acknowledge that a transition from 19 th century Westminster accounting to 21 st century EU accounting could prove as disruptive and traumatic to all concerned as a switch from driving on the left hand side of the road to driving on the right. I am not without sympathy for those encased in the Westminster approach even if it is of very limited use and relevance in today’s world.

How much is spent on public sector pay?

The answer that Eurostat/CSO give is ‘Compensation of employees’ in the General Government sector.

Chart 1 tracks this measure from 1995 up to 2015 and forward to 2018 (using published Department of Finance estimates for 2016-2018).  It should be noted that total employee compensation includes an estimated real value of employer contributions to pensions of current employees. So, for example, total compensation in 2013 came to €18.6 billion (D1) of which an estimated €15.0 billion was for wages and salaries of employees in General Government in that year (referred to as D11 in national accounts). The difference, €3.6 billion, is referred to in the Eurostat rules as “Social contributions paid by government as employer” (code D12). This latter figure is made of two components:

Actual social contributions of government as employer (D121) coming to an estimated €1.6 billion and imputed social contributions of government as employer (D122) coming to €1.4 billion.  In summary, the ‘mark-up’ used by the statistical agencies in respect of the current value of future pension liabilities for current employees is probably somewhere in the region of 19% of the total wage bill (€3 billion over €15.6 billion in the case of 2013 data).  It is my understanding that the matter is under review as part of a long-term exercise to arrive at a good estimate of employer cost of future pension liabilities. The matter is complicated by the fact that there are many different groups of public sector workers including a significant and growing number of ‘pension career-average’ public servants for whom the pension contribution by the employer will be considerably less than that of ‘pension end-of-career’ public servants. The difference is to do with the salary on which the final pension is calculated (years of service multiplied by final salary in the case of ‘end-of-career’ or ‘career-average’ salary). 

 The total for employee compensation grew rapidly throughout the period up to 2008. This growth reflected very rapid rates of growth in GDP and, in association, total government revenue. It is, therefore, reasonable, to compare the growth in total compensation with (i) total GDP (iii) total government spending and (iii) total government revenue. As stated already, I am adhering to Eurostat-CSO ways of measuring revenue and spending which will differ from exchequer-based approaches.

Chart 2 shows the size of total public sector pay relative to these three key aggregates over a period of fast growth up to 2008 followed by contraction from 2009 to 2014 and, then, recovery. Clearly, the recovery in total employee compensation is strongly related to growing public service numbers with the lifting or relaxation of the employment embargo in 2015. The trends, over time, point to changes in the aggregates against which total employee compensation is measured. Total pay (compensation) was remarkably stable as a percentage of GDP between 1995 and 2007 – belying the traditional narrative that public sector pay was ‘out of control’ in the period immediately prior to the Crash of 2008. Pay cuts and employment embargoes notwithstanding, total public sector as a percentage of GDP rose between 2008 and 2009 to reach a maximum of 12.2% in the latter year. This reflects on the sharp contraction in GDP that occurred during these two years. The share of public sector pay has fallen, year by year since 2010 and reached an estimated 7.4% in 2015 (and is projected to stay close to that level out to 2018). Clearly, there have been problems with GDP as a measure over time and especially in 2015 following the Great Leap Forward which had more to do with the curious activities of a handful of large multinational enterprises combined with the application of new international national accounts rules than anything approaching ‘real’ economic activity. That said, GDP is GDP according to the rules of international statisticians and the total value of GDP is, potentially , revenue and expenditure for the Irish Government depending on tax arrangements.

The peculiar behaviour of revenue in 2009 and expenditure in 2010 was due, respectively, to a recessionary collapse in revenues and a one-off transfer of capital to the most infamous dead bank in Irish history.  Leaving aside recessions and the Great Leap Forward, it does appear that the sizes of the Irish public sector pay bill has been, and will continue to be close to around one quarter of total government spending or revenue.

  How does public sector pay, here, compare with that in other jurisdictions?

There are many ways of measuring public sector pay across countries. I will start with the most aggregate measure – namely total staff compensation as a percentage of total government revenue. This approach has the merit of side-stepping the Irish GDP conundrum.

Chart 3 shows the percentages of total government revenue, expenditure and GDP, in 2014, accounted for general government employee compensation. Care is need in comparing across countries due to different proportions of government spending and revenue as well as the way in which governments ‘compensate’ its citizens including government employees. It is possible that many European countries ‘compensate’ their own employees along with other citizens through relatively higher levels of access to ‘public goods’ such as health, housing and education. The price of such access is lower cash wages or cash compensation. A striking example is given by health insurance where it is typical for many average to above average earning employees to spend a thousand or many thousands of Euro a year on private health insurance (depending on household size and characteristics). Recourse to private health insurance is dramatically lower in most other European countries including the UK. These differences matter in comparing pay as do other areas such as housing, transportation, early childhood care.

An important further consideration that a comparison based on total Government Revenue or total Government Expenditure, only, would fail to account for the unique structure of Irish economic activity. Comparisons anchored on GDP do matter.  GDP is an unreliable measure of actual economic activity due to the tax-avoiding behaviour of large multinational corporations. Yet, all value-added and corporate income (whether repatriated or not) is taxable by Irish authorities.  On this measure, total public sector pay is not evidently lower than in other countries. Put another way, public sector pay accounts for a larger proportion of a relatively smaller pool of government revenue due to a regime of lower than average international corporate taxation (including employer social contributions by way of pay-related social insurance).

 An important further consideration that a comparison based on total Government Revenue or total Government Expenditure, only, would fail to account for the unique structure of Irish economic activity. Comparisons anchored on GDP do matter.  GDP is an unreliable measure of actual economic activity due to the tax-avoiding behaviour of large multinational corporations. Yet, all value-added and corporate income (whether repatriated or not) is taxable by Irish authorities.  On this measure, total public sector pay is not evidently lower than in other countries. Put another way, public sector pay accounts for a larger proportion of a relatively smaller pool of government revenue due to a regime of lower than average international corporate taxation (including employer social contributions by way of pay-related social insurance).

An alternative approach is to compare individual occupations using data published by the OECD in Government at a Glance (the latest edition being in 2015 but with international data on public sector employee compensate in the 2013 edition for which Ireland is missing). Unfortunately, the last available comparative data for the Republic of Ireland is 2009 as new data have not been provided since then. Just as public sector pay got really interesting the data cells for Ireland switched to empty! The 2009 data indicate that Ireland does not necessarily have particular high staff compensation costs. Two particular grades in the civil service (the OECD survey was confined to civil service occupations and grades) are shown, here, in Charts 4 and 5 for, respectively, clerical officers (or equivalent) and principal officers. An interesting result is that lower-paid occupations in the Irish civil service appear to be less well remunerated than higher-paid grades. This is in contrast the within-Ireland comparison of public and private pay which I will come to now.

 

How does public sector pay compare with that in the private sector?

This is the mother of all statistical questions.  Much ink and sweat has been poured by analysts in reviewing the data. The Tabloid headlines on average pay levels in the public and private sector are all too familiar.  Such reporting adds little light to the discussion because no account is taken of the composition and structure of the workforce in the public and private sector. We know that the public sector is composed of workers who, on average, have higher levels of education, have served longer in their current job and are older than their peers in the private sector. Complex econometrical analysis has been undertaken by researchers at the Economic and Social Research Institute as well as by the Central Statistics Office. The latter published an updated analysis of 2009/2010 earnings data on in April 2015. The Report was entitled ‘Specific Analysis of the Public/Private Sector Pay Differential for National Employment Survey 2009 & 2010 data’ provided an analysis of pay taking into account the Public Sector Pension Levy introduced in 2009. Previous analysis by the CSO (published in 2012 under the title ‘National Employment Survey 2009 and 2010: Supplementary Analysis’) had not taken account of the deduction levy in a reckoning of public-private sector comparisons.  The CSO took a broader analytical approach to that taken by the ESRI. Whereas, the ESRI arrived at a series of single-point estimates of the ‘wage premium’ for public sector workers compared to the private sector (having taken account of age, educational and other factors), the CSO published a range of estimate including the results of the different models. 

The estimation of differences between the public sector and private sector is conditional on which explanatory factors are taken into account. One of these concerns the size of the organisation that employees work for. This factor can be significant as an explanation of part of the differences in wages in both sectors. There is no universal agreement among economists or statisticians about the appropriateness of including or excluding organisational size in the analysis. The CSO show the results with and without organisational size included as an explanatory factor.

The following are some key points to note in the CSO analysis:

  • There was a pay gap between public and private sectors in 2010. However, this is linked to many factors including education, experience and the structure of the workforce in each sector.
  • The data are six years of out date and do not reflect pay trends since 2010.
  • The most recent analysis takes into account the impact of the pension deduction levy which was, in effect, a pay cut for public servants.
  • The difference between public and private sector is considerably narrowed as a result of taking the pension levy into account. For some categories of workers – especially men above the average for pay there is likely to be no gap in favour of the public sector. Depending on estimates made by the CSO some male workers above the average earn more in the case of the private sector.
  • Senior managers may be less well paid than their counter-parts in the private sector in Ireland adjusting for different characteristics (this contrasts with the international findings which shows Irish public sector senior managers relatively well paid in 2009).
  • Cutting across all of the analysis, is a marked gap in gender pay: female workers in the public sector are relatively better paid than their counterparts in the private sector while the relative gap for male workers tend to be smaller.
  • It is evident that a proliferation of low pay occupations and sectors in the private sector (notably retail and hospitality) in conjunction with a high concentration of female employment in these sectors impacts on the comparison of the broad public and private sectors.
  • As noted by the CSO, many occupations in the public and private sectors have very limited or no direct comparators in the other sector (e.g. workers in food, accommodation and retail, prison security, police, army and teachers).
  • The analysis refers to 2010.
  • It is likely that the inclusion or exclusion of particular occupations and sectors (e.g. army, Garda, education, health, etc.) impacts on the results.

While every effort is made by CSO to provide robust and comprehensive data on earnings it is possible that some degree of under-reporting of earnings arises in the private sector especially at the higher end of the pay scale due to a range of pecuniary and near-pecuniary benefits to highly paid private sector employees.

On the other hand the monetary benefit of most public sector pensions vis-à-vis comparable occupational pensions in the private sector is difficult to factor into this analysis. Pensions are a significant part of 'lifetime' income. They are a form of 'deferred income'.

As with many other areas, analysis of public sector pay at international, national and other levels is a work in progress.

In conclusion it can be said that the measurement of pay in the Irish public sector is a complex area fraught by an increasingly shrill and loaded public debate in which selective data insights are used to support particular points of view.

Digital Revolutionaries