Are government wages interlinked with private sector wages?

Abstract

The ongoing fiscal and financial crisis with significant macroeconomic imbalances in the euro area raises the question in how far public–private wage interaction and even wage spirals contribute to fiscal and competitiveness problems. In this vain we analyse empirically whether the evolution of public sector wages is decoupled from that of the private sector. Using data for number of OECD countries, we find: (i) a strong and extremely robust positive annual contemporaneous correlation of public and private sector wages over the business cycle; (ii) evidence of co-movement of these variables over the medium and long run. From a policy perspective, the findings of this study suggest public wage restraint and private wage flexibility coupled—where needed—with institutional reform to prevent or break public–private wage spirals.

Introduction

In this paper we analyse, from an empirical standpoint, how public and private wages interact and whether public sector wages are decoupled from that of the private sector. This issue is relevant from an analytical and a policy perspective as public wages and employment comprise almost one quarter of the total dependent work force and total compensation of employees in the OECD. Moreover, the financial crisis that erupted in 2007 and that turned into a fiscal crisis in 2009/10 has shown that public finance problems go hand in hand with competitiveness loss and macroeconomic imbalances in a number of euro area countries where also public wage dynamics have been very strong (Holm-Hadulla, Kamath, Lamo, Pérez, & Schuknecht, 2010).

From a policy perspective, the main questions are (1) what drives public and private wages, and (2) whether the interaction of public and private wages is benign or whether it could give rise to public–private wage spirals which undermine both fiscal stability and regional economic competitiveness in the euro area. Wage negotiations in the government sector and their potential spillover effects on other sectors of the economy may have a marked impact on unit labour costs and inflationary pressures (including via changes in administered prices). If the latter happens within a monetary union such as the euro area (or in a fixed-exchange regime for an individual country), public–private sector wage spirals may impinge on the cost and price competitiveness of individual countries, as well as on the union as a whole. Beyond competitiveness, the potential contribution of the government wage bill to fiscal consolidation packages and its potential moderating effect on private wage developments make it worth the analysis of the linkages between public and private sector wages.¹

Conventional wisdom would hold that a decoupling of public and private wages is most problematic as, for example, higher public wage increases would undermine countries fiscal positions. But the co-movement of public and private wages is only unproblematic if wage developments in both sectors are in line with domestic sectoral or intra-euro area competitiveness requirements. For example, if in a country due to a buoyant economy, government revenue is very strong, the government may be tempted to pay higher wages which in turn could further increase pressure on all private wages (through market forces or institutional linkages) even though the tradable sector cannot “afford” it. The result would be close correlation of wages in the two sectors which, however, may not be desirable from a normative/efficiency perspective.

What seems most problematic from this perspective is, hence, if we find both a co-movement of public and private sector wages and private wage increases well above what is “affordable” in monetary union (or in a fixed-exchange regime). While the “affordability” will only be dealt with casually by presenting long term public–private wage developments relative to the euro area average, the main focus of the study is on inter-linkages and co-movements.

From a theoretical point of view it is not clear what sign and size one should expect for the correlation between public and private sector wages. On the one hand, according to standard theory, (real) public wages are assumed to be exogenous or to follow the same determination patterns as private wages (Ardagna, 2007, Calmfors & Horn, 1986, Demekas & Kontolemis, 2000, Holmlund, 1993, Holmlund & Ohlson, 1992, Quadrini and Trigari, 2007). In most cases, free labour mobility implies that wages should be equalized across sectors for the same type of labour. Either because labour in the public and the private sector is paid its marginal product, because of competition by trade unions or due to other political economy arguments (as in Fernández-de-Córdoba et al., 2012a, Fernández-de-Córdoba et al., 2012b) this literature predicts a close positive correlation between public and private sector wages.

On the other hand, though, there are theories that predict that the determination of public and private sector wages is not related. Indeed, a political economy view with self-interested politicians and bureaucrats would suggest that public wages are not necessarily determined by market forces to the same extent as wages in the private sector at least in the short to medium run (Borjas, 1984, Matschke, 2003, Mueller, 2003). The wage setting behaviour of the public sector is likely to differ from that of the private sector due to a number of factors such as a higher degree of unionisation, political objectives, the different status that civil servants enjoy and that might make public wages less reactive to the business cycle, or the separate agenda of public employees (rent-seeking behaviour). These arguments would imply a weak correlation (or co-movement) between the evolution of public and private sector wages than standard theory. Election-oriented wage policies, for example, would suggest deviations at least in the short term. More problematic from a normative perspective, would be a situation where public and private wages decouple in the medium to long run and bureaucrats can ration entry and extract rents via above-equilibrium wages on a more durable basis.

Our study focuses on the assessment in how far the observed evolution of public and private wages over the period 1960–2006 has been consistent with theories that predict a positive and high correlation/co-movement between wages in both sectors, or rather with theories that prescribe that the correlation should be negative or weakly positive. In this paper we estimate the sign of the correlation between public and private wages in many industrialised countries: the euro area aggregate, ten euro area countries (Germany, France, Italy, Spain, Netherlands Austria, Belgium, Greece, Ireland, Portugal, Finland), Sweden, Denmark, Norway, United States, United Kingdom, Canada and Japan. The use of annual data is imposed by the absence of quarterly data for the variables in this study for most of the European countries in the sample. For the sake of robustness, we look at the results for the sample 1960–2006 but also 1980–2006. Along this line, the study aims to obtain the most robust results possible, by examining nominal wages, and also nominal wages deflated with two alternative price deflators (private consumption deflator and GDP deflator), by using a large number of detrending techniques, and by applying a host of approaches to correlation and co-movement over different time horizons.

The paper provides, firstly, robust empirical evidence on the correlation of public and private wages over the business cycle. We look at the unconditional correlations between detrended series (at the standard business cycle frequencies) using eleven methods. The study finds that in the euro area and for most of the countries of our sample private wages are positively and strongly correlated with public wages over the business cycle in a mostly contemporaneous manner. This is a very robust result across countries, in spite of very different institutional settings and different inflation regimes witnessed in the different decades covered by our study. Secondly, the paper provides an analysis of short-, medium-, and long-run co-movements between public and private sector wages using the methodology of den Haan (2000). For the short-run correlations the results of the previous (robust) analysis are confirmed. For the co-movements at longer frequencies than the standard business cycle frequencies our results show a strong correlation between public and private sector wages.

Putting these findings in perspective with the fact that a number of countries in the euro area experienced very large public wage increases over the first decade of EMU, this suggests that the co-movement of public and private wages in the euro area may not be fully benign. Under certain institutional settings (e.g. wage indexation, public wage leadership, wage inflexibility between the trade and non-tradable sectors) such co-movements may reflect and give rise to public–private sector wage spirals that would be detrimental to fiscal soundness and competitiveness.²

From this, the policy lessons would be two-fold: first, when fiscal and competitiveness problems exist, public wage restraint could help correcting both fiscal imbalances and –through the interlinkage with private wages– competitiveness problems. In other words, public private wage spirals can be reversed starting with the public sector. Second, institutional features of wage setting are key for this adjustment to work or for such spirals to be prevented in the first place. For example, the elimination of wage indexation or of downward inflexibilities may be needed, depending on country circumstances.

The paper is organised as follows. Some stylised facts on the developments of public wages for the period 1960–2006, on public private wage dynamics in the euro are for up to 2008 and a description of the data as presented in Section 2. Section 3 looks at the co-movements of public and private wages over the business cycles (an Appendix A details the methods used), while Section 4 analyses medium- and long-term co-movements. Section 5 concludes and provides detailed policy implications.