Skip to main content

CEE: rapid wage growth to fuel above-target inflation

  • Central and Eastern Europe is one of the regions of the world where we think that the risk of sustained higher inflation in the next few years is greatest. The Phillips curve is alive and we think the combination of a cyclical recovery in demand for labour alongside structural labour shortages will feed into stronger wage growth and keep inflation above central banks’ targets. This is not fully appreciated by most and we think interest rates will ultimately settle at a higher level than most expect in two-to-three years’ time.
  • The Phillips curve has been a key part of macroeconomics for decades and shows the relationship between the unemployment rate and wage growth. Whereas the Phillips curve has flattened in advanced economies in recent years, there has been a strong and negative relationship between labour market slack and wage growth in Central and Eastern Europe (CEE) since 2005. A 1%-point fall in the unemployment rate pushed up wage growth in the G7 economies by 0.1%-points after the financial crisis, but by 0.85%-points in CEE.
  • It is certainly the case that there are structural factors that have reduced the bargaining power of labour in the G7 economies in recent decades (such as globalisation, technological change and structural changes in the labour market) and which are not present in CEE.
  • But we think the most important reason why wages are more responsive to labour market conditions in CEE is because falls in working age populations and strong GDP growth pushed unemployment rates far below their natural rates during the last two business cycles and led to widespread labour shortages. In Poland, Czechia and Hungary this acted as a tipping point for accelerations in wage growth. Such a sharp tightening of labour market conditions has not been experienced in DMs.
  • The combinations of a strong economic recovery, a continued decline in the working age population and a small rise in the employment rate as the recovery broadens out should push unemployment rates in CEE down to or below pre-pandemic rates in 2023. Positive output gaps are likely to re-emerge in twelve months’ time and we expect wage growth to rise from 6% y/y in 2020 to around 8% y/y in 2023.
  • Such strong wage growth is likely to outstrip productivity and firms are likely to have little incentive to absorb these costs in their profit margins if demand continues to recover strongly. As a result, we think wage growth will put upwards pressure on labour-intensive services inflation, leaving inflation rates firmly above central banks’ targets. Once pandemic distortions wash out next year, we expect inflation to rise to near 3.8% in Poland and Hungary and near 2.8% in Czechia by end-2023 and to pick up further in 2024.
  • This view does not seem to be fully appreciated by most analysts, which expect inflation to settle at central banks’ targets from 2023. While we agree with the amount of monetary tightening investors expect over the next twelve months or so, we think higher inflation will ultimately prove more persistent and pave the way for interest rates to settle at a higher level than most anticipate in a few years’ time.
  • We’re confident that the Czech central bank will deliver the monetary tightening necessary to stabilise inflation around its target, but the risk is that policymakers in Poland and Hungary tolerate a prolonged period of high inflation, tighten monetary policy too slowly and that wage and price expectations become unanchored. We think the direction for government bond yields will be higher in the coming years.

CEE: rapid wage growth to fuel above-target inflation

Central and Eastern Europe is one of the regions of the world where we think that the risk of sustained higher inflation over the coming years is greatest. A key reason for this is because we think that labour market conditions will tighten significantly as the recovery matures and that this will feed through into higher wage growth and inflation. In this Focus, we explain why the Phillips curve is alive in CEE and how this feeds into our view for inflation to remain above central banks’ targets for longer than most expect over the coming years.

Phillips curve flattening across the G7…

The Phillips curve has been a central part of macro-economics since William Phillips first popularised the relationship between UK unemployment and wage growth in 1958. The Phillips curve suggests that policymakers face a trade-off: they can lower unemployment at the cost of higher wage growth.

There has been an intense debate about the changing shape of the Phillips curve in the G7 economies as wage growth has become less responsive to labour market slack. (See Chart 1.) This flattening of the Phillips curve first started in the mid-1980s and the general view is that this reflects structural changes in the labour market, globalisation and technological progress that have reduced the bargaining power of labour. (See our Global Economics Focus.)

Chart 1: G7 Phillips Curve (1985 – 2019)


Sources: Refinitiv, Capital Economics

… but remains alive and kicking in CEE

In contrast to advanced economies, wage growth has been very responsive to labour market conditions in Central and Eastern Europe. We have created a CEE Phillips curve between 2000 and 2019 for Bulgaria, Czechia, Hungary, Poland, Slovakia and Romania.

Chart 2: CEE Phillips Curve (2000 – 2019)


Sources: Refinitiv, Capital Economics

There are two key observations from the curves in Chart 2. First, there has been a negative relationship between the unemployment rate and wage growth in CEE since 2005. The slopes of the Phillips curves show that between 2005 and 2008 a 1%-point fall in the unemployment rate pushed up nominal wage growth by 1%-point. Since the financial crisis, wages became slightly less responsive to labour market slack as a 1%-point fall in the unemployment pushed up wage growth by 0.85%-points. But wage growth remained very unresponsive to labour market slack in the G7 as the same fall in the unemployment rate pushed up wage growth by only 0.1%-points.

Second, there has been a leftward shift in the CEE Phillips curve since the financial crisis, similar to the G7. In other words, unemployment has needed to fall to lower levels than before the financial crisis to have the same effect on wage growth. The shift reflects a decline in productivity growth, low and better anchored inflation expectations due to policy credibility, and falls in natural unemployment rates.

Natural rates of unemployment are those that are consistent with stable inflation. And Phillips curves that take account of these provide a fuller depiction of the wage-unemployment trade-off at the country level. Natural rates are unobservable, difficult to estimate and subject to large revisions. Chart 3 shows two OECD estimates of natural unemployment rates for Czechia and Poland: one estimate published in the OECD’s May 2017 outlook, and the other in its June 2019 outlook. There has been a significant fall in estimates of natural unemployment rates since the 2000s, with Poland’s falling from 13% in 2004 to 5% in 2020. (See Chart 3.)

Chart 3: Natural Unemployment Rates (OECD, %)


Source: OECD

When comparing wage growth with the deviation of unemployment from its natural rate (hereafter referred to as the “unemployment gap”), Phillips curves have a strong non-linear shape in Poland, Czechia and Hungary. (See Chart 4.) In other words, wage growth is generally unresponsive to labour market conditions when there is a lot of slack, but picks up more sharply once the unemployment rate falls below its (estimated) natural rate.

Chart 4: CEE Phillips Curves (2000/02 – 2019)
Panel A: Poland

Panel B: Czechia

Panel C: Hungary


Sources: Refinitiv, OECD, Capital Economics

There is a debate about whether the “unemployment gap” is the most appropriate measure of spare capacity in a Phillips curve specification. We have also modelled the relationship with alternative measures and our findings show that these results are robust as wage growth has a non-linear relationship with estimates of the output gap and the vacancy-unemployment ratio in all three countries.

The idea is that during recessions or periods of weak demand firms can meet demand by producing more with existing capacity so there is little incentive for firms to raise wages above productivity. But when there is limited slack in the economy, firms produce at capacity and strong demand may force them to raise wages more quickly. The underlying cause of this non-linearity is the stage of the business cycle.

All of this raises a key question: why have wages in Central and Eastern Europe been more responsive to changes in labour market conditions than those in advanced economies?

Workers’ bargaining power: CEE vs. the G7

There are certainly some structural factors that have reduced the bargaining power of labour in the G7 and which are not present in CEE. As we explained in a Global Economics Focus in 2017, developments in recent decades such as (i) increased competition with workers in China and other emerging economies as supply chains have globalised, (ii) the substitution of robots and other forms of technology for labour and (iii) the rise of part-time, temporary and less secure forms of work, have acted to reduce demand for unskilled labour and workers’ bargaining power in the G7. But in many ways, the economies of CEE do not conform to these trends.

For one, the globalisation of supply chains and EU accession in 2004 supported wages in CEE. Cheaper labour costs have attracted international firms and increased the demand for labour. The share of employment in the industrial sector (32%) is higher in CEE than in advanced economies (23%) and has been stable since the late 1990s. (See Chart 5.)

Data from the International Federation of Robotics show that robot density in manufacturing in 2016 was much lower in CEE than in major economies. Low labour costs reduce the need to automate in certain sectors and this may mean that workers have not had their bargaining power undermined by the shift in production towards capital.

Chart 5: Industry (% of Total Employment)


Sources: World Bank, Capital Economics

Meanwhile, less secure forms of employment are also less prevalent in CEE, which has meant that a larger share of workers is in full-time, permanent employment and with greater bargaining power. It has been suggested that labour market slack should take into account the wider use of temporary and part-time contracts as well as the underutilisation of the labour force. But these are likely to have had little impact on wage growth in CEE. The share of people on temporary contracts is only 5%, versus 20% in major economies. (See Chart 6.) The figures for part-time employment tell a similar story. This partly reflects the smaller size of services sectors in CEE.

Chart 6: Workers on Temporary Contracts
(2017-19, % of Total Employment)


Sources: Eurostat, OECD, World Bank, Capital Economics

While these factors may imply that there has not been the reduction in bargaining power in CEE like there has in advanced economies, they do not tell the whole story. The wage negotiation process is complex and one key similarity between labour market structures is the extent of trade union membership, which has fallen more sharply in CEE since the 1990s than in major economies. (See Chart 7.) This mainly reflects the transition to a free market in the 1990s and large ownership changes related to the presence of international corporations.

Chart 7: Trade Union Density in the G7, Euro-zone & CEE Economies (2000 and 2018/19, %)


Sources: OECD, Capital Economics

Of course, trade union membership is no guarantee that workers will have the power to drive wages up. The threat of offshoring jobs convinced German trade unions to accept wage restraint in the 1990s. This restraint is still practiced today and collective bargaining in the euro-zone tends to prioritise job security over above-inflation wage increases.

Wage bargaining in CEE is extremely decentralised and typically held at the company level. Analysis by the ECB suggests that the bargaining structure in CEE may result in less nominal wage rigidity than in the rest of Europe, so that wages are more responsive to shocks (both up and down). (This may also help to explain the large variation in wage growth during periods of slack in Czechia and Hungary in Chart 4.)

This lack of rigidity may be partly related to inflation. While the anchoring of inflation at a low level may have helped to flatten the Phillips curve in the G7 in recent decades, it is not clear that this has had the same effect in CEE – analysis by the IMF shows that inflation and inflation expectations do not have a major impact on wage growth in Eastern Europe.

Labour shortages and the business cycle

All of the aforementioned factors may have played a part, but we think the key reason why wages are more responsive to labour market conditions in CEE is because falls in working age populations and strong GDP growth pushed unemployment rates far below natural rates during the last two business cycles. Such a tightening of the labour market has not been experienced in advanced economies but has pushed up wage growth up significantly in CEE.

Since 2010, working age populations aged 15-64 have fallen by 7% in Poland, Czechia and Hungary and 8% in Romania. Working age populations have increased by 3-7% in the UK, US and Canada and broadly stagnated in Germany over this time. Only Japan has experienced a similar demographic profile to CEE. (See Chart 8.) The decline in working age populations in CEE reflects the combination of falling birth rates in the 1990s and emigration to Western Europe following EU membership in the 2000s.

Chart 8: Population Aged 15-64 (2010 = 100)


Sources: Eurostat, OECD, Capital Economics

Unlike in Japan where the decline in the population started in the 1990s and its impact on overall labour supply has been partly offset by rising participation of women and retirees, demographic shifts in CEE have generally happened more abruptly and labour supply has been slow to respond. Birth rates slumped in the late 1990s and only in the last 5-10 years or so have governments adopted policies to encourage higher fertility and retirees to remain in work. The share of people over the age of 65 in or available to work stands at just 4-7% across Central Europe, compared to 25% in Japan. (See Chart 9.)

Chart 9: Labour Force Aged 65+ (% of Pop. 65+)


Sources: Eurostat, OECD, Capital Economics

By any measure, CEE economies were running hot during 2006-08 and 2017-19. Industrial capacity utilisation rates had risen, and large positive output gaps opened up as GDP ran above potential. Labour markets tightened, with the number of vacancies relative to the number of people looking for work rising and businesses reporting widespread labour shortages. (See Chart 10.)

Chart 10: Labour a Constraint on Activity (Avg. Of Services, Industry & Construction, % of Firms)


Sources: European Commission, Capital Economics

What’s more, there was very little hidden slack in the region’s labour markets. The number of people either (i) underemployed working part-time, (ii) seeking work but not available to work or (iii) available to work but not seeking work remained a small share of the labour force. Against this backdrop, unemployment rates fell far below their natural rates and by more than in advanced economies – unemployment rates fell 2%-points or so below natural rates in Czechia and Poland and more than 4%-points in Hungary in 2017-19. (See Chart 11.) This acted as a tipping point for an acceleration in wage growth across the region. (See Chart 4 again.)

Chart 11: Difference between Actual Unemployment and Estimated Natural Rate (2019, %-point)


Sources: Refinitiv, OECD, European Commission, Capital Economics

Labour market in the next stage of the pandemic

To be clear, labour market conditions have softened during the pandemic as unemployment rates have risen by 1%-point since Q4 2019. But the economic recovery is now in full swing, and the key question is how quickly this slack will be used up.

Surveys of labour shortages (as shown in Chart 10) have already started to turn and central banks are now increasingly factoring in a tightening of labour market conditions in their forecasts. Poland’s central bank expects a fall in the unemployment rate to a multi-decade low of 2.3% in 2023 to push wage growth up to 8% y/y; Hungary’s central bank has warned about a possible wage-price spiral in which stronger wage growth pushes up prices and this feeds into expectations for even higher wage negotiations; the Czech central bank has said that the labour market has already started to overheat.

We think unemployment rates will fall to or below pre-pandemic levels in 2023 due to a combination of a strong economic recovery driven by household spending and investment, poor demographic trends and a rise in the employment rate as the recovery broadens out. We think GDP growth will average 4.0-5.0% over 2021-23, which would leave the size of GDP in the CEE region not far off its pre-pandemic trend in 2023 and output above potential (a return to a positive output gap). Based on our estimated CEE Phillips curves in Chart 4, this would push CEE wage growth up from 6% y/y in 2020 to around 8% y/y in 2023. (See Chart 12.)

Chart 12: CEE* Nom. Wage Growth & Unemp. Rate


Sources: Refinitiv, OECD, Capital Economics

Persistent wage pressures to keep inflation high

History provides a guide as to the extent of the wage pass through to inflation. The rise in wages between 2015 and 2019 outpaced productivity and pushed up unit labour costs (ULCs) by 15-20% in Czechia and Hungary. The rise in ULCs was smaller in Poland due to strong productivity growth, but average ULCs in CEE still rose at their fastest sustained pace in over two decades and significantly outpaced that in Germany. (See Chart 13.)

Chart 13: Nominal Unit Labour Costs (% y/y)


Sources: Refinitiv, Capital Economics

Firms initially absorbed these higher costs in their profit margins as seen by the fall in gross operating surplus between 2015 and 2019. (See Chart 14.) But they eventually passed them on via higher prices.

Chart 14: CEE* Nat. Accounts (4-Qtr Avg., % of GDP)


Sources: Refinitiv, Capital Economics

Domestically generated wage pressures do not show up in all parts of the consumer basket. Durable goods prices are typical determined by external factors as a high share of goods are imported. But wage pressures do show up in cyclically sensitive areas such as services. Our in-house measure of labour-intensive services inflation that includes education, hotels, restaurants, housing and hairdressing surged across the region in 2017-19. (See Chart 15.)

Chart 15: Labour-Intensive Services Inflation (% y/y)


Sources: Refinitiv, Capital Economics

We’re likely to see a repeat of these events in the coming quarters. For one reason, we think that nominal wage growth of 8% y/y will outstrip productivity growth. Although there is uncertainty about how productivity will fare in the post-pandemic world, we think it is unlikely to be higher than 4-5% y/y on a sustained basis. And two, those services-sector firms hit hardest by the pandemic and which experienced the largest loss of revenues will have less incentive to absorb higher labour costs in their margins as aggregate demand continues to expand.

In Poland, there has usually been a tight relationship between wage growth (advanced by one year) and our measure of labour-intensive services inflation. Inflation rose sharply to almost 7% y/y in Q2 last year due to pandemic-related distortions. (See our Focus and Chart 16.) Distortions to the wage figures mean that wage growth will be difficult to interpret for a while. But our forecast for wage growth to average 7-8% y/y in 2023 would be consistent on past form with labour-intensive services inflation settling at 4% or so, above the central bank’s 2.5% target.

Chart 16: Poland Nominal Wages & Labour-Intensive Market Services Prices (% y/y)


Sources: Refinitiv, Eurostat, Capital Economics

Prolonged period of above-target inflation ahead

The strength of the Phillips curve relationship and the diminished amount of slack means that there is likely to be more scope in CEE than in other parts of the world for the recovery in economic activity to feed through into higher wage growth and inflation.

Inflation this year has been stronger than we had expected due to more persistent goods shortages and strong “re-opening inflation” in services sectors. These factors are likely to fade into next year which, alongside the easing of energy inflation, should push inflation down sharply. But we think that inflation will pick up again and rise further in 2023 as demand outstrips supply and capacity constraints bite harder.

We expect inflation to rise to 3.8-4.0% in Poland and Hungary and 2.5-2.8% in Czechia by end-2023. This would leave it 1.0-1.5%-pts above central banks’ targets in Poland and Hungary and 0.5-0.8%-pts in Czechia. This is higher than in 2017-19, due to the length of the pass-through and higher starting point and would mark the longest sustained period of above-target inflation in a decade. (See Chart 17.)

Chart 17: Difference Between CPI Inflation
& Central Banks’ Targets (%-points)


Sources: Refinitiv, Capital Economics

This view does not seem to be fully appreciated by most. Forecasts for nominal wage growth in 2023 and beyond compiled by FocusEconomics settle at rates that would generally be consistent with unemployment rates at their natural rate. Similarly, the consensus forecast is for headline inflation to return to central banks’ targets in 2023 or 2024.

We agree with the amount of monetary tightening investors expect over the next 12-18 months or so. Our forecasts are for interest rates to rise by 75bp in Hungary, 110bp in Poland and 175bp in Czechia by Q1 2023. This compares with investors’ expectations of 110bp, 150bp and 125bp, respectively. While we do not expect inflation to take off in 2024, we think interest rates will ultimately settle at a higher level than most expect in a few years’ time.

The Czech central bank is the most prudent and forward-looking in the region and we are confident that it will deliver the monetary tightening required to stabilise inflation and inflation expectations close to its target. We expect the Czech policy rate to settle above the estimated neutral rate (of 2.5-3.0%) in the next few years. But central banks in Hungary and Poland have a higher tolerance for inflation – we think they would be willing to let inflation settle above their targets in the coming years and tighten monetary policy only gradually, with short-term interest rates settling slightly below their neutral rates (of 3.0% or so) in 2024. It is for this reason why we think inflation risks are greatest in these two countries. Our policy rate forecasts and those of the consensus are shown in Chart 18.

Chart 18: Policy Interest Rates Forecasts (Capital Economics & Consensus, End-2020 to End-2024, %)

Sources: Bloomberg, FocusEconomics, Capital Economics

Exploring the market implications of higher inflation

Higher inflation might normally be expected to put upwards pressure on government bond yields, either because central banks react by raising short-term rates or because investors discount higher inflation in the future. But this hasn’t happened in CEE – inflation has risen since 2015 while a simple average of 10-year bond yields in Poland, Czechia and Hungary has fallen by 100bp. The result has been a decline in the spread over inflation to -1%. Nominal bond yields in CEE have been depressed by global developments, particularly falls in “safe” government bonds as the spread of 10-year government bond yields in CEE over the 10-year German bund has remained flat at 200-250bp or so. (See Chart 19.)

Chart 19: Difference Between CEE* 10-Year LCU Gov’t Bond Yield, CPI Inflation & German Bund


Sources: Refinitiv, Capital Economics

Of course, it is more appropriate to compare 10-year bond yields with expected, rather than past, inflation but this tells the same story that deeply negative real long-term interest rates in CEE are generally unique, bar the key exceptions of advanced economies like the US, UK and Germany. (See Chart 20.) This may signal an environment of high inflation and low growth (“stagflation”), but we do not think this is appropriate in CEE as growth prospects are strong. It may also suggest that investors are tolerant of high inflation in the medium term, so there may not be much of a re-adjustment in yields at some point in the future. But we think there will need to be a tightening of monetary conditions.

Chart 20: 10-Year LCU Government Bond Yields Less 10-Year Consenus Inflation Forecasts (%-points)


Sources: Refinitiv, Capital Economics

There’s likely to be upward pressure on CEE bond yields from higher “safe” government bonds as the global recovery matures and inflation in the US proves more persistent than most expect. We expect the 10-year US Treasury yield to rise by 100bp and the German bund by 50bp by end-2023. At the same time, CEE central banks will respond to above-target inflation by raising short-term interest rates. We think investors will discount higher interest rates down the line, particularly in the 3–5-year horizon as high inflation pressures prove persistent. This will put upwards pressure on CEE bond yields. Our forecasts for 10-year local currency bond are in Table 1.

Table 1: 10-Year Local Currency Government Bond Yield Forecasts (%)





















Sources: Refinitiv, Capital Economics

Liam Peach, Emerging Europe Economist, +44 (0)20 7808 4990,