Muokkaa hakua
Pääsivu Media Oheistietoa Tutkimus ja julkaisut Tilastot Rahapolitiikka Euro Maksut ja markkinat Ura EKP:sssä
Hakuehdotuksia
Järjestä tulokset
Agostino Consolo
Senior Team Lead - Economist · Economics, Supply Side, Labour and Surveillance
Claudia Foroni
Senior Economist · Economics, Supply Side, Labour and Surveillance
Ei saatavilla suomeksi

Drivers of employment growth in the euro area after the pandemic – a model-based perspective

Prepared by Agostino Consolo and Claudia Foroni

Published as part of the ECB Economic Bulletin, Issue 4/2024.

The adjustment in real wages has been key in supporting post-pandemic employment growth in the euro area. This box looks at developments in employment growth after the end of the pandemic, focusing on the period since early 2022. The strong growth of employment compared with economic activity has been remarkable in an economic environment characterised by very high energy prices. The adjustment mechanism was different from that seen following the 1970s energy crisis, when real wages increased substantially and outpaced productivity growth. By contrast, since early 2022, real wages have fallen more than productivity, providing firms with leeway in their margins and thus helping to sustain job creation amid a scarcity of skilled labour. The following empirical analysis sheds light on the key drivers behind recent employment growth.

Since the end of the pandemic, employment dynamics in the euro area have been significantly stronger than economic activity (Chart A, panel a).[1] Employment growth since the first quarter of 2022 has been exceptionally strong from a historical perspective. One way to examine employment dynamics is through the lens of the “Okun’s law” relation between employment growth and economic growth.[2] According to empirical estimates of this relation for the euro area, employment would be expected to grow at around half the rate of real GDP in both the short term and the long term[3] (simple estimates of Okun elasticities range between 0.2 and 0.5). Instead, employment growth has undergone a strong adjustment since the first quarter of 2022, recently overtaking real GDP growth. This cannot be solely attributed to the pandemic recovery, as the realignment had already happened towards the end of 2021.

Chart A

Okun’s law, productivity and real wages

a) Output and employment

b) Productivity and real wages

(index: Q4 2019 = 100)

(index: Q4 2019 = 100 and percentage points)

Sources: Eurostat and ECB calculations.
Notes: Panel b) shows real wages deflated by the private consumption deflator. When using the GDP deflator, the gap between productivity and real wages remains positive but is quantitatively smaller and closes in the fourth quarter of 2023. Productivity is measured as output per worker. The red bars show the deviations (in percentage points) from Okun’s law. The latest observations are for the fourth quarter of 2023.

Job creation has been supported by real wages falling by more than productivity. In contrast to developments during the 1970s, real wages underwent a stronger adjustment than labour productivity following the 2022 energy crisis, thus supporting job creation (Chart A, panel b). The opening of a gap between real wage growth and productivity growth pushed employment growth above the rate consistent with its historical relation with GDP, in line with what a large positive Okun’s law gap would predict. Not only does such an adjustment in real wages provide a rationale for job creation, it can also limit job destruction by making firms’ existing relationships with workers more valuable (hence encouraging labour hoarding) in times of buoyant profitability for firms and scarcity of labour market skills.[4]

Looking through the lens of an empirical model of Okun’s law, we find that demand fluctuations and real wage adjustments following the energy crisis have helped decouple employment growth from output dynamics. We use a structural Bayesian vector autoregressive (BVAR) model comprising an aggregate demand shock and three aggregate supply shocks (neutral technology, factor substitution, and labour market-specific) to explain the key drivers of employment fluctuations.[5] The neutral technology shock can be seen as a fall in total factor productivity generating a proportional contraction in the use of capital and labour, typically also consisting of other disturbances, such as global supply-chain bottlenecks.[6] The factor substitution shock, meanwhile, captures the direct substitution between labour and other production inputs, not necessarily only capital. In the face of a large increase in energy and intermediate input prices, it is natural to assume that reallocation and substitution effects are also at play.[7] The quantitative relevance of each driver is measured by applying the model to the data. Chart B shows the historical breakdown of how the four shocks have contributed to output and employment growth. As of early 2022, the contribution from neutral technology shocks (red bar) has been largely negative for both output and employment, primarily reflecting global supply bottlenecks. The factor substitution shock (green bar) has been important in driving output down and employment up, highlighting the possible significance of substitution among production inputs. This reflects a relative price change of input factors as labour became cheaper compared with energy and capital.

Chart B

Historical decomposition of employment and output

(percentage changes and percentage point contributions)

Source: Consolo, A. and Foroni, C., “The euro area labour market: and yet, it moves!”, mimeo.
Notes: The lines depict the year-on-year growth rates of employment (left panel) and output (right panel) in terms of deviation from their respective deterministic component. The bars show the percentage point contribution of each shock. A description of the economic meaning of each shock is provided in the main text. The latest observations are for the fourth quarter of 2023.

Recovering demand supported employment and output growth until the beginning of 2023. Since then, weaker demand has led to a stronger slowdown in economic activity than in employment growth, contributing to the greater deviation from Okun’s law. Demand-side factors – including monetary and fiscal policies – have been conducive to employment growth, as captured by the yellow bars in Chart B. Examples of fiscal policies supporting employment have been the increase in public employment as well as indirect effects from apprenticeship schemes and tax credits in some euro area countries. Together with the aggregate demand effects of targeted energy-related policies, this has sustained demand and jobs, with a stronger impact on employment than on output.[8] However, the effects of monetary policy tightening have led to a considerable dampening of demand, causing employment growth to slow and economic activity to weaken. The model suggests that demand depressed activity more quickly than it did employment. In 2022 the model attributes only limited negative effects to labour market-specific drivers (blue bars) stemming from lower labour force participation and the greater bargaining power of workers.[9] These effects have partly reverted to previous levels in 2023, possibly reflecting the lack of automatic wage indexation schemes and the fact that inflation expectations were well anchored in 2022. As opposed to previous energy crises (such as that experienced during the 1970s), these factors have led to the adjustment of the labour market through real wages rather than through a higher unemployment rate.

According to the model, a large share of the recent fall in productivity has been driven by cyclical factors. After a positive rebound in productivity following the pandemic, the energy crisis weighed substantially on productivity. The cyclicality of euro area productivity was even more amplified by the exceptional strength of the labour market. The model shows that the resilience of employment, triggered by the factor substitution shock, has amplified the fall in labour productivity (Chart C). The same shock has also made real wages less costly and, in so doing, amplified the positive effect on employment and negative effect on productivity. Moreover, the adjustment in productivity is also related to demand-side factors (yellow bars), the contribution of which has become more negative recently but is expected to fade in the longer run. The factor substitution shock is dissipating, also reflecting the changes in energy prices and real interest rates, but other supply-side factors such as innovation and digitalisation, population ageing, and other structural trends may have more persistent effects on long-term productivity dynamics in the euro area.

Chart C

Historical decomposition of labour productivity

(percentage change and percentage point contributions)

Source: Consolo, A. and Foroni, C., “The euro area labour market: and yet, it moves!”, mimeo.
Notes: Productivity is measured as output per worker. The line depicts year-on-year productivity growth in terms of deviation from the deterministic component. The bars show the percentage point contribution of each shock. The latest observations are for the fourth quarter of 2023.

Looking ahead, several factors currently weighing on productivity growth are expected to subside. The effects of fiscal support measures that have sustained employment are expected to fade in the coming quarters. Energy prices are normalising, reducing the factor substitution effect between labour, energy and capital. And, as the euro area inflation rate declines towards target, the restrictive effect of monetary policy on demand is likely to normalise. All these drivers of the slump in productivity are expected to fade, thereby helping to close the gap between real wage growth and productivity growth. With fewer incentives for firms to hire workers or hoard labour, the relation of employment to GDP is likely to return to normal, with a consequent cyclical adjustment in productivity growth.[10]

  1. For a thorough overview of the euro area labour market that complements the results presented in this box, see the box entitled “The euro area labour force: recent developments and drivers”, Economic Bulletin, Issue 6, ECB, 2023; Arce, O., Consolo, A., Dias da Silva, A. and Mohr, M., “More jobs but fewer working hours”, ECB Blog, 7 June 2023; and the box entitled “The role of public employment during the COVID-19 crisis”, Economic Bulletin, Issue 6, ECB, 2022.

  2. For a review, see Ball, L., Leigh, D. and Loungani, P., “Okun’s Law: Fit at 50?”, Journal of Money, Credit and Banking, Vol. 49, Issue 7, 2017, pp. 1413-1441.

  3. See Anderton, R., Arank, T., Bonthuis, B. and Jarvis, V., “Disaggregating Okun’s law”, Working Paper Series, No 1747, ECB, Frankfurt am Main, December 2014.

  4. As regards the role of profitability, see the box entitled “Higher profit margins helped firms to hoard labour” in this issue of the Economic Bulletin.

  5. The BVAR model disentangles the different potential driving forces behind output and employment growth (see Consolo, A. and Foroni, C., “The euro area labour market: and yet, it moves!”, mimeo). The model is estimated using quarterly data on employment, real GDP, Harmonised Index of Consumer Prices (HICP) inflation, real wages (deflated by the GDP deflator) and nominal interest rates from 1970 to 2023. The identification strategy is implemented via sign restrictions derived from theoretical models of the labour market (see Abbritti, M. and Consolo, A., “Labour market skills, endogenous productivity and business cycles”, Working Paper Series, No 2651, ECB, February 2022, and Foroni, C, Furlanetto, F. and Le Petit, A., “Labor supply factors and economic fluctuations”, International Economic Review, 2018, Vol. 59, pp. 1491-1510.

  6. See Blanchard, O. and Galí, J., “The Macroeconomic Effects of Oil Price Shocks: Why Are the 2000s so Different from the 1970s?”, Working Papers Series, No 15467, NBER, 2010.

  7. The literature is increasingly focusing on shocks which create a similar negative co-movement between output and employment. These include what are sometimes defined as automation shocks (see Foroni, C. and Furlanetto, F., “Explaining deviations from Okun’s law”, Working Paper Series, No 2699, ECB, August 2022) and sectoral shocks that alter capital deepening across sectors in the spirit of Acemoglu, D. and Guerrieri, V., “Capital Deepening and Nonbalanced Economic Growth”, Journal of Political Economy, 2008, Vol. 116, pp. 467-498.

  8. See the box entitled “Update on euro area fiscal policy responses to the energy crisis and high inflation”, Economic Bulletin, Issue 2, ECB, 2023.

  9. A labour market-specific shock covers the degree of bargaining power of workers or drivers of labour supply. These shocks originating in the labour market have been shown to be quantitatively relevant to explain employment and wages (see Consolo, A., Foroni, C. and Martínez Hernández, C., “A Mixed Frequency BVAR for the Euro Area Labour Market”, Oxford Bulletin of Economics and Statistics, 2023, Vol. 85, pp. 1048-1082).

  10. See Arce, O. and Sondermann, D., “Low for long? Reasons for the recent decline in productivity”, ECB Blog, May 2024.