Wage developments during and after the high inflation period

Prepared by Colm Bates, Katalin Bodnár, Peter Healy and Marc Roca I Llevadot

1 Introduction

Wages are key to the assessment of inflation and thus for monetary policy. Inflation both shapes and is shaped by wage developments. Wages represent an input cost to firms. In turn, productivity-adjusted wages affect firms’ price-setting decisions and thus inflationary pressures, while workers base wage claims on current and past inflation and on inflation expectations. Moreover, wages represent a significant part of income for households and therefore affect consumption and saving decisions.

The past decade has seen significant changes in the conditions for wage growth, in terms of both price pressures and labour market conditions. The period from 2013 to 2017 was characterised by high albeit declining labour market slack, low inflation and subdued productivity growth which, together with some structural factors, kept a lid on wage growth. While inflation began to gradually pick up from 2018 and labour market tightness was increasing, wage growth remained low.^[1] During the pandemic, headline inflation was muted, as were underlying wage dynamics, while both the unemployment rate and wages were affected by government measures to cushion the economic impact of the pandemic shock.^[2] Following the post-pandemic reopening of the economy and Russia’s unjustified invasion of Ukraine, inflation in the euro area increased to historically high levels which, combined with a tight labour market, resulted in historically high euro area wage growth. Headline inflation began declining considerably in 2023, while there have also been signs of weakening labour demand and wage growth has gradually been easing from a high level.^[3]

Such changes in the macroeconomic environment create challenges for assessing the relative importance of the drivers of wages. While lessons can be learned from the low inflation period, the changing economic landscape and data distortions during the pandemic require a reassessment of the standard tools and an extension of data sources used to analyse wage growth. Against this background, this article examines the drivers of wage growth in the extraordinary post-pandemic period (2022-24) through an augmented wage Phillips curve and through the analysis of new, granular data on wage agreements. It also illustrates the link between wage growth and inflation by applying the Bernanke-Blanchard model to the euro area.^[4]

2 Wage developments during the rise and decline of euro area inflation

The events influencing labour market developments after the pandemic had a heterogenous impact on wage indicators (Chart 1). A key indicator in the assessment of wage growth in the euro area is the annual growth rate of compensation per employee (CPE). This reflects the labour costs payable by employers – including wages, salaries and employers’ social contributions – expressed as an average per employee. The ECB monitors various other wage indicators for a more complete assessment of wage pressures, including compensation per hour (CPH) and negotiated wage growth. While CPE growth declined considerably during 2020, indicators of wage growth per hour worked, such as CPH, rose. These developments were driven by statistical factors linked to the pandemic and the use of job retention schemes, which distorted the information content of most wage indicators during this period in various ways. Accordingly, these indicators continued to be volatile in 2021 owing to base effects. By contrast, the ECB’s indicator of negotiated wages, which captures the outcome of collective bargaining processes, was not affected by statistical distortions.^[5] It remained relatively stable at a low level during 2020 and 2021.

Chart 1

Euro area labour cost indicators and HICP inflation

(annual percentage changes)

Following the surge in inflation, wage growth increased to levels not seen since before the advent of the monetary union, but different indicators accelerated at different speeds. CPE growth was the first of the wage indicators to follow the rise in euro area inflation, and in 2022 it was already above 4%. At that time CPH growth was about 1 percentage point lower. The difference reflected the recovery of average hours worked after the pandemic, which contributed to CPE growth on top of the more slowly rising hourly compensation. As the growth in average hours worked slowed down, the two measures began to grow at similar rates of above 5% in 2023. In contrast, negotiated wage growth strengthened more gradually, from just above 1% in 2021 to close to 3% in 2022 and above 4% in 2023. This more gradual adjustment reflected the fact that wage negotiations took some time to restart after the pandemic and were lengthy. Much of the difference between CPE and negotiated wages is captured by the wage drift, which made a much stronger than normal contribution to actual wage growth in the early stages of the high inflation period.^[6] This high wage drift reflected not only the recovery in average hours worked but also ad hoc inflation compensation at firm level, partly incentivised by preferential tax treatments. As formal wage negotiations began to directly reflect inflation compensation, the wage drift component started to decline again, which was also reflected in a greater similarity between CPE and negotiated wage growth.

Recently, wage growth has been moderating from a high level, but again with different indicators moving at different speeds. As in the acceleration phase, negotiated wage growth also seems to be the most persistent and the last to adjust during the period of wage disinflation. While CPE and CPH growth both peaked in the second quarter of 2023, two quarters after the peak in HICP inflation, negotiated wage growth has remained high and volatile, reflecting the fact that wage contracts are of different lengths and workers that were locked into multi-year contracts suffered delays in their wages responding to the surge in inflation. The volatility of the negotiated wage series largely reflects the greater importance of one-off payments in recent years, but base effects due to these payments will have a downward impact in 2025. Information from the ECB wage tracker (see Section 4) and the ECB’s Corporate Telephone Survey also suggests a gradual easing of negotiated wage growth ahead.^[7]

3 What can the Phillips curve tell us about post-pandemic wage developments?

Shocks in recent years have triggered changes in the drivers of wages. The cyclical position of labour markets, inflation developments and productivity growth are key drivers of wage growth, which economists frequently assess through the lens of a wage Phillips curve. Chart 2 shows how these drivers have developed. To correct for the considerable volatility related to job retention schemes in 2020, we interpolate the wage level and the productivity level – the variables most affected by job retention schemes – over the first and second quarters of 2020. This makes it easier to look through data distortions and understand how recent developments in most of the wage drivers have deviated from historical regularities beyond the distortions induced by the pandemic. Price inflation, captured in Chart 2 through HICP inflation, increased very quickly to historically high levels in 2022 but has returned to close to the ECB’s 2% target more recently. At the same time, the unemployment rate (inverted in the chart) has declined to its lowest level since the introduction of the euro. Finally, labour productivity growth has been below its long-term average for a prolonged period. The ECB’s augmented wage Phillips curve (Nickel, C. et al., op. cit.) captures the impact of these factors – (either past or expected) inflation, labour market state, and productivity – on actual wage growth while also accounting for lags. We will use this tool as a theoretical guide and an empirical device to assess wage developments using a variety of specifications.^[8]

Chart 2

Euro area CPE growth and key macroeconomic drivers

(annual percentage changes; percentages of the labour force)

Catching up with past inflation has been an important driver of recent wage growth. The quick, sizeable and unexpected surge in consumer prices resulted in a decline in real wages across euro area countries and sectors. On average, real wages declined by about 5% between the start of the inflation surge in mid-2021 and the peak of headline inflation (Chart 3). Workers will seek to recoup their real wage losses, meaning that wage demands will reflect the difference between actual real wage levels and real wage expectations, conditional upon labour market conditions.^[9] Thus real wage catch-up has been an important driver of recent wage growth. Real wage levels are not included directly in the basic wage Phillips curve, but these can be incorporated indirectly by including past inflation and lagged wage growth in alternative specifications. The assessment of this factor will be enriched by looking at granular wage negotiation data in the next section.

Chart 3

Nominal and real CPE and the price level in the euro area

(index: Q2 2021 = 100)

It is a priori unclear to what degree wage dynamics reflect past inflation as opposed to forward-looking inflation expectations. This may differ strongly among workers and may also depend on the actual level of inflation. To address this uncertainty, in the augmented wage Phillips curve we capture the impact of inflation on wages by including a measure of either backward or forward-looking inflation, selected from a wide range of possible indicators.^[10] All these indicators increased relative to their past values. However, they have deviated strongly since the beginning of the high inflation period (much more so than previously), with backward-looking indicators rising earlier and to a higher level than forward-looking indicators, and longer-term expectations remaining relatively stable (Chart 4).^[11]

Chart 4

Measures of past inflation and inflation expectations over different horizons

(annual percentage changes)

Wage Phillips curve estimations confirm that the reaction of wages to past inflation has been the major driver of recent wage inflation. Although the prevalence of wage indexation is relatively low in the euro area, before the pandemic inflation compensation in wage formation was influenced primarily by backward-looking inflation.^[12] Such wage formation may increase the persistence of nominal variables and could amplify second-round effects. Inflation may play a greater role in wage formation in a period of high inflation because firms and workers are more attentive to it than when inflation is low. During the high inflation period and the subsequent disinflation, past inflation and shorter-term inflation expectations explain a larger part of wage developments. Models that include past inflation and short-term inflation expectations can explain a large part of the upward phase of wage growth, while in the wage disinflation phase past inflation appears to have made a somewhat larger contribution. In contrast, models that include long-term inflation expectations point overall to a lower estimated impact of inflation on wage formation (Chart 5). Recently, as HICP inflation has eased, the contribution of inflation to wage growth has also been easing in most specifications with backward-looking or short-term inflation measures, but it nevertheless remains high. Overall, this finding confirms that there is a strong backward-looking element in wage formation in the euro area, which in the recent period is also related to the strong surprise element of the inflationary shock. This is in line with the application of the Bernanke-Blanchard model to the euro area.^[13]

Chart 5

Euro area CPE growth and the contribution of inflation in different wage Phillips curve specifications

(annual percentage changes and percentage point contributions)

Various indicators on the state of the labour market have been pointing to tightness, although to different degrees. The period since 2022 has been characterised by the lowest unemployment rate since the introduction of the euro in 1999. The headline unemployment rate has also been lower than estimates of its non-inflationary rate (for example, the European Commission’s non-accelerating wage rate of unemployment – NAWRU), confirming the signals for labour market tightness.^[14] Other measures of the labour market state that are used as alternative indicators in the augmented wage Phillips curve specifications (e.g. the vacancy-to-unemployment ratio and the European Commission’s measure of labour as a factor limiting production) have been pointing to an even stronger increase in labour market tightness than the unemployment rate. One possible reason for this is that these two indicators include labour demand more directly, while the unemployment rate reflects the balance between labour demand and labour supply, where adjustments in the labour force may satisfy labour demand. Recently, indicators of labour demand have been easing from a high level, while the unemployment rate has been more stable (Chart 6).

Chart 6

Measures of the cyclical position of the labour markets

(percentages; net balance; percentages of the labour force)

Labour market tightness has contributed to recent wage growth. The unemployment rate points to greater upward pressure on wage growth than the historical average. However, since 2022 wage growth has been more closely aligned with the dynamics of the vacancy-to-unemployment ratio and the European Commission’s indicator on labour limiting production. Wage Phillips curve specifications that include these indicators point to a stronger impact of labour market tightness, suggesting they are more informative about wage growth in the recent period (Chart 7). At the same time, the dynamics of recent wage growth more closely follows that of the contribution from (past) inflation, while labour market tightness has likely acted as a supporting factor for real wage catch-up.

Chart 7

Contribution of labour market tightness to CPE growth across wage Phillips curve specifications

(annual percentage changes and percentage point contributions)

A wage decomposition based on a range of augmented wage Phillips curve specifications attributes the recent high wage inflation primarily to high price inflation but, as the role of price inflation is declining, the relative impact of the labour market is increasing. In the decomposition, we consider an average of various specifications with different labour market and inflation variables. Comparing the main wage drivers, inflation seems to have been the main driver of wage growth in recent years (Chart 8). The more recent easing of CPE growth is also driven by a lower contribution from inflation. In contrast, the contribution of labour market developments is smaller but is not estimated to have declined.^[15] Labour productivity has been weak recently, and this may have dampened overall wage growth. The estimated correlation between productivity and wage growth is very low at the business cycle frequency, and the estimated contribution of productivity to actual CPE growth is very small and primarily negative.^[16] Finally, there has been a positive residual, reflecting the presence of some factors which are not captured by the augmented wage Phillips curve models. These factors could be the interplay of labour market tightness and high inflation, but the unexplained part may also reflect the fact that the wage Phillips curve is limited in its ability to capture real wage catch-up.^[17] A wage decomposition based on a subset of these augmented wage Phillips curve specifications where the inflation component is only backward-looking increases the relevance of the inflation component and reduces the residual in the post-pandemic era. Such a finding underscores the role the recent high inflation has played in shaping wage growth.

Chart 8

Wage Phillips curve decomposition of CPE growth

(annual percentage changes and percentage point contributions)

Box 1
Time variation in the slope of the wage Phillips curve

Prepared by Colm Bates, Katalin Bodnár and Peter Healy

For monetary policy, it is important to understand changes in the trade-off between the real and nominal sides of the economy and the link between inflation and wage growth embedded in the wage Phillips curve. As we move further away from the recent large shocks, it is becoming possible to assess whether these specific shocks caused a temporary break or a more permanent change in the slope of the euro area wage Phillips curve. In general, the literature is inconclusive about the time variation of the slope of the wage Phillips curve, which may in part reflect the variety of methodologies and indicators used. Nevertheless, there is some evidence of a flattening, in particular after the global financial crisis.^[18] There is, however, little evidence of a change in the wage Phillips curve after the pandemic.^[19] This box focuses on potential changes in the euro area since the pandemic.

With the unemployment rate at record low levels and wage inflation reacting to the inflationary surprise, the pure correlation between the two would suggest a steep Phillips curve in the past few years. However, the shift is largely due to the impact of past inflationary shocks on wage growth. Indeed, estimates of the wage Phillips curve based on a rolling sample do not point to a steepening of the slope of the curve (Chart A). There is some increase in the estimated parameter of the vacancy-to-unemployment ratio and the labour limiting production indicator, following earlier stronger volatility for both, but these changes are not statistically significant for most specifications. The estimated parameters for the unemployment rate and unemployment gap are fairly stable, as they have been historically. There is also no significant change in the estimated inflation parameter, although this might reflect that it is difficult to account for large shocks (like the recent inflationary shock) in this linear framework. Rolling window estimates point to some volatility in the inflation parameter, mainly for long-term inflation expectations.

Chart A

Parameter estimates of cyclical measures of labour market tightness in the euro area wage Phillips curve

(estimated parameters)

4 What can the ECB wage tracker tell us about the role of inflation compensation in wage negotiations?

The ECB wage tracker provides a timely assessment of negotiated wages. Wage indicators are released with a considerable time lag, and recently, as shown in the previous section, assessing these data has been challenging. In the high inflation period there was a particularly strong need for more timely indicators and approaches. One significant development has been the introduction of the ECB wage tracker, which uses granular data on existing collective bargaining agreements.^[20] The ECB wage tracker provides both backward-looking information on negotiated wage growth pressures and very timely forward-looking signals on expected future negotiated wage growth during the coming months.^[21] This indicator is one of several new data sources, including survey-based wage indicators, developed at the ECB since the onset of the pandemic.^[22]

Granular data on wage negotiations can help explain the role of inflation compensation as a driver of such negotiations. The ECB wage tracker made it easier to anticipate the gradual increase in euro area negotiated wages starting in 2022 and to understand the considerable country heterogeneity (Chart 9). In this article, we use this tool to understand the features of wage negotiations using granular data at the agreement level. In particular, we assess the role of real wage catch-up demands in the gradual increase in negotiated wage growth and in the composition of negotiated wages. We differentiate between structural increases – i.e. changes that affect base wages – and one-off payments, and we assess the role of institutional factors in the pace of real wage catch-up.^[23]

Chart 9

ECB wage tracker and contributions by country

(annual percentage changes and percentage point contributions)

One important factor behind the heterogenous speed of adjustment of negotiated wages to inflation is the length of previous collective bargaining agreements. Institutional factors, such as the frequency of collective bargaining negotiations and the length of agreements, can influence the dynamics of underlying wage pressures. Understanding how quickly social partners can react to economic shocks and renegotiate collective bargaining agreements is therefore key to assessing the pass-through of inflation to wages. Long contract durations limit the scope for employers to adjust to macroeconomic shocks but, at the same time, provide a safeguard for employees regarding their future compensation, as it is less dependent on the business cycle. However, in a high inflation period, long contract durations can lead to substantial real wage losses, especially when the strong inflationary shocks were not accounted for in the previous wage negotiation. Long contract duration can also increase the volatility of negotiated wage series when inflation is unexpectedly high, since in this case there is a large accumulated real wage gap to be closed or narrowed in future wage negotiations. For example, retail and wholesale sector workers in Germany recently negotiated a new wage agreement for the first time in three years. To compensate for the cumulated real wage loss, they agreed large one-off payments as well as significant structural wage increases (i.e. growth in regular payments, excluding one-offs) in the third quarter of 2024, resulting in record negotiated wage growth (as shown in Chart 1).

ECB wage tracker data suggest that recorded contract durations differ greatly within the euro area. In France and Austria, collective bargaining agreements last on average around 12 months. By contrast, in Germany, Italy, Spain and the Netherlands, agreements are usually valid for two years or more. In Italy and Spain, the distribution of workers by contract duration exhibits several peaks, some at significant intervals such as five or six years, which is uncommon in other countries (Chart 10, panel a). Aggregating the ECB wage tracker data at the euro area level reveals that 13% of workers are covered by contracts lasting one year or less and which have been renegotiated at least three times since 2021. As the contract duration lengthens, the renegotiation frequency naturally decreases. One-third of workers have contracts lasting one to two years, and there have been a large number of agreements in this segment since 2021, reflecting the sometimes staggered and slow nature of collective bargaining. Contracts with a duration longer than two years, which cover 54% of workers, have been renegotiated only once or twice since 2021 (Chart 10, panel b).

Chart 10

Duration and number of collective bargaining agreements, by share of workers since 2021

(panel a: x-axis – contract duration in years; y-axis – density; panel b: percentages of workers)

Shorter contracts enable wage growth to respond more quickly to inflation. Structural wage growth among contracts renegotiated every year increased from 2.0% in 2021 to 5.1% in 2022 and 6.0% in 2023, before easing to 3.5% in 2024. While these contracts still reflect staggered wage dynamics, they suggest that negotiated wage pressures will continue to ease as long as inflation does not suddenly increase again. Contracts with longer durations display a more gradual structural wage adjustment, with lower recorded wage growth in 2022 and higher wage growth in 2023 and 2024. Thus, contracts with longer durations displayed a weaker immediate response to the inflation surge in 2022 and stickier and higher wage growth in the subsequent years. Still, wage pressures for longer duration contracts are also easing, as already seen in 2024 in the wage growth dynamics of contracts with an average duration of two years (Chart 11).

Chart 11

Structural wage growth, by year in which current and previous contracts were signed

(annual percentage changes)

As most of the workers covered by the ECB wage tracker have renegotiated their wages since the start of the inflation surge, real wage catch-up is easing as a factor, as confirmed by the forward-looking tracker. A significant share of collective bargaining agreements expired in 2024, affecting more than 30% of the workers covered by the ECB wage tracker database. Moreover, a further 15% will see their agreements expire in the first half of 2025. Of these, a substantial majority were agreed in either 2023 or 2024, suggesting that most workers have already received wage increases or one-off payments compensating for inflation. Very few of the contracts that are expected to be renegotiated by the second quarter of 2025 have not already been renewed at least once since 2023. Accordingly, the vast majority of contract renegotiations are anticipated to result in lower wage growth than was agreed in 2023 and 2024, driven partly by the more flexible, or less staggered, wage adjustment of contracts with lower durations (Chart 12). This, together with the lower wage drift, will bring overall wage pressures down.

Chart 12

Share of workers under expiring agreements, by signing year

(percentages of workers)

Box 2
Wages and the pass-through of shocks to inflation

Prepared by Antoine Kornprobst and Carlos Montes-Galdón

This box illustrates the drivers of wages and inflation, as well as their interconnection in the recent episode of high inflation, through the lens of the semi-structural model of Bernanke and Blanchard.^[24] The model is used to identify the sources of wage growth and inflation, taking the dynamic relationships between prices, wages and expectations fully into account. These include the effects of the exogenous shocks that have hit the euro area economy since the first quarter of 2020 as well as the effects of the initial conditions in the first quarter of 2020, capturing the dynamic effects of pre-pandemic macroeconomic conditions. Importantly, the model can illustrate how inflation and wage growth interact, and it unveils the crucial role of wages in the propagation of shocks to price inflation in the euro area.

In the model, negotiated wages are determined not only by long-run productivity growth and the degree of slack or tightness in the labour market – as measured by the ratio of job vacancies to unemployment – but also by price shocks, as wage agreements reflect compensation for both unexpected and expected inflation in order to catch up with and prevent further real wage losses after inflation surges. In turn, wages enter the cost structure of firms via the price equation in the model and are gradually passed through to consumers. In the wake of the inflation surge, negotiated wage growth increased significantly: post-pandemic shocks accounted for around 3 percentage points of negotiated wage growth, in particular owing to higher energy and food prices, global supply chain pressures and, to some extent, the tightening of the labour market (Chart A).

Chart A

Sources of euro area annual wage growth and price inflation

(annual percentage changes and percentage point contributions)

Dynamic decompositions of euro area wage growth and (underlying) price inflation reveal that the response of wages to post-pandemic shocks has increasingly been feeding into prices since the first quarter of 2022 (Chart B). Our results underscore that there is a delay in the response of wage growth to inflation, and subsequently in the response of inflation to wage growth. This reflects the fact that wages take time to catch up with increases in prices owing to labour market rigidities, but also that firms have sticky prices and do not pass on changes in labour costs to consumers immediately. These results signal some remaining upside pressure to inflation from wage dynamics: past price shocks, while fading, were still being passed on to wages until the third quarter of 2024.

Chart B

Direct and indirect impacts of price shocks on inflation and second-round effects via wages

(annual percentage changes and percentage point contributions)

The wage channel has been playing an increasingly important role in explaining inflation dynamics since inflationary price shocks began to subside. Monitoring wage growth is therefore crucial to assess the risks to medium-term price stability. In the absence of significant price shocks in the future, the normalisation of wage growth will support the return of inflation to the ECB’s 2% target.

5 Conclusions

As the unusual shocks of recent years are slowly dissipating, wage growth is gradually easing. Wage growth is an important factor influencing the inflation outlook, and understanding its drivers is crucial. The augmented wage Phillips curve approach points to real wage catch-up as the main driver of recent high wage growth, which was supported by tight labour markets. In contrast, productivity growth has played a negligible role. The estimated inflationary impact reflects developments in past inflation and captures – although imperfectly – the real wage catch-up process. This factor is slowly dissipating, as is also confirmed by granular data in the ECB wage tracker, allowing wage growth to edge down, which is further supported by the easing of labour demand. At the same time, the wage channel is playing a key role in explaining inflation dynamics, so it remains important to monitor wage developments.