INVESTIGATING THE WAGE–PRODUCTIVITY NEXUS: A PANEL DATA APPROACH

Abstract

This study examines the determinants of labour productivity in ten selected European Union countries over a decade (2014-2023). Using a fixed effects panel data approach, the analysis aims to examine the relationship between labour productivity and four key macroeconomic indicators: GDP per capita, average wages, unemployment rate and investment in research and development (R&D). The main objective of the study is to assess whether these indicators can provide significant insights into variations in labour productivity both over time and between countries. Labour productivity, an important measure of economic efficiency, is used as the dependent variable. It is assumed that GDP per capita and average wages, which reflect economic performance and income distribution, have a positive influence on productivity. The unemployment rate, on the other hand, is likely to have a negative effect and serve as an indicator of inefficiencies in the labour market. Investment in R&D is seen as an important driver of technological progress and innovation, with an expected positive effect on productivity. The analysis uses a fixed-effects regression model to control for unobserved heterogeneity between countries and to ensure that country-specific factors, such as institutional framework conditions or cultural influences, do not distort the results. The data set was compiled from publicly available sources (Eurostat) and covers ten EU member states. All variables are log-transformed to standardise the units and to facilitate the interpretation of the coefficients as elasticities. The study contributes to the growing literature on productivity determinants by providing empirical evidence specific to the European Union context. It offers practical implications for policy makers and emphasises the need for targeted strategies to promote investment in R&D, ensure equitable wage growth and reduce unemployment in order to foster sustainable economic growth. In addition, the study emphasises the importance of country-specific factors and suggests that policies need to be tailored to the unique economic and institutional environment of each Member State.

Headnote

ABSTRACT

This study examines the determinants of labour productivity in ten selected European Union countries over a decade (2014-2023). Using a fixed effects panel data approach, the analysis aims to examine the relationship between labour productivity and four key macroeconomic indicators: GDP per capita, average wages, unemployment rate and investment in research and development (R&D). The main objective of the study is to assess whether these indicators can provide significant insights into variations in labour productivity both over time and between countries. Labour productivity, an important measure of economic efficiency, is used as the dependent variable. It is assumed that GDP per capita and average wages, which reflect economic performance and income distribution, have a positive influence on productivity. The unemployment rate, on the other hand, is likely to have a negative effect and serve as an indicator of inefficiencies in the labour market. Investment in R&D is seen as an important driver of technological progress and innovation, with an expected positive effect on productivity. The analysis uses a fixed-effects regression model to control for unobserved heterogeneity between countries and to ensure that country-specific factors, such as institutional framework conditions or cultural influences, do not distort the results. The data set was compiled from publicly available sources (Eurostat) and covers ten EU member states. All variables are log-transformed to standardise the units and to facilitate the interpretation of the coefficients as elasticities. The study contributes to the growing literature on productivity determinants by providing empirical evidence specific to the European Union context. It offers practical implications for policy makers and emphasises the need for targeted strategies to promote investment in R&D, ensure equitable wage growth and reduce unemployment in order to foster sustainable economic growth. In addition, the study emphasises the importance of country-specific factors and suggests that policies need to be tailored to the unique economic and institutional environment of each Member State.

Keywords: GDP, GDP per capita, linear regression model, productivity, wages

1. INTRODUCTION

Labour productivity is a cornerstone of economic performance and a key determinant of longterm growth and competitiveness. Defined as the amount of output produced per unit of labour, productivity reflects the efficiency of resource use and serves as a key indicator of a nation's economic health. In the context of the European Union (EU), it is particularly important to understand the drivers of labour productivity given the different economic structures, income levels and labour market dynamics in the Member States. This study attempts to contribute to this understanding by examining the relationship between labour productivity and several macroeconomic indicators: GDP per capita, average wages, unemployment rate and investment in research and development (R&D). Economic theory assumes that a higher GDP per capita and higher average wages are positively correlated with labour productivity. GDP per capita reflects total economic output in relation to the population and is often seen as an indicator of economic prosperity and living standards. Similarly, average wages reflect labour market conditions and income distribution, both of which influence employee motivation and investment in human capital. Conversely, high unemployment rates are generally associated with underutilised labour resources, inefficiencies and lower economic performance, and thus have a negative impact on productivity. Investment in R&D, on the other hand, is expected to promote technological innovation, improve production processes and increase competitiveness, all of which contribute to higher levels of productivity. The European Union provides a compelling context for analysing these relationships. With its mix of developed and developing economies, the EU provides a unique opportunity to analyse how these macroeconomic factors influence productivity in different institutional, cultural and economic settings. Moreover, understanding these dynamics is crucial for policy makers seeking to address the persistent differences in productivity and economic performance between member states. In this study, a fixed-effects panel data approach is used to analyse the evolution of labour productivity in ten selected EU countries over a decade (2014-2023). By controlling for country-specific characteristics, this method ensures that the analysis isolates the effects of the main variables of interest while accounting for unobservable heterogeneity. The dataset comes from reliable sources (Eurostat) and all variables are log-transformed to facilitate interpretation based on elasticities. It is expected that the results of this study will provide valuable insights into the determinants of labour productivity in the EU and highlight the role of economic and political factors in shaping productivity outcomes. By considering both theoretical and practical dimensions, this paper aims to contribute to the broader discourse on productivity growth and economic policy in the European Union.

2. LITERATURE REVIEW

2.1. Productivity and Wage Correlation

The relationship between wages and labour productivity is often the subject of research in the economic literature, as it is a key aspect for understanding labour market dynamics and economic efficiency. According to previous research, the relationship between these two variables is generally positive and is characterised by various economic and institutional conditions. The theoretical basis for the expectation of a positive relationship between wages (pay for work) and labour productivity is found in "human capital theory" (Becker, 1964), as wages reflect the productivity of workers and employers have an incentive to pay more for workers with higher levels of knowledge, skills and experience. In this book, Becker defines "human capital theory" as activities that increase an organisation's opportunities by developing its most valuable resource, its employees. Germany and the Scandinavian countries, for example, often show a clear parallel growth in productivity and wages related to their investments in innovation, technology and education. Studies in developed countries, such as the EU and the US, show that there is a strong positive correlation between productivity growth and wage growth. When productivity increases, companies also have the opportunity to increase employee wages. Historical data shows that total compensation per hour has increased at a similar rate to productivity growth, suggesting a stable relationship between wages and productivity over time (Bauducco and Janiak, 2018). The relationship between wage increases and GDP per capita is positive. Empirical studies show that higher wages can lead to higher economic output, especially in different regional contexts (Anghelache and Anghel, 2018). The stability of labour compensation as a share of national income indicates that while productivity can increase, income distribution remains relatively constant, which has an impact on overall economic prosperity (Bauducco and Janiak, 2018). At this point it should be noted that there is also evidence of the inverse relationship between labour productivity and wages. In some cases, productivity growth does not follow proportional wage growth. This can be observed in many countries where income inequality is a major problem. The reasons for this relationship may be: the weakening of workers' bargaining power (there are no trade unions) and a greater focus on capital and returns to shareholders rather than wage increases. In less developed countries, productivity may grow faster than wages due to the greater labour supply and lower cost of living, but also due to the slow development of institutions regulating the labour market.

In our study, we use data on the average wage in a country in a given year as a variable to define wages. The influence of the state-set minimum wage has a large impact on the level of the average wage, which leads to the conclusion that the level of the minimum wage also influences labour productivity. There is evidence that moderate increases in the minimum wage can stimulate capital accumulation and production, suggesting a nuanced relationship between wage policy and economic performance (Bauducco and Janiak, 2018). The introduction of higher minimum wages can lead to a compression of the wage distribution, benefiting lowwage workers, although this can also lead to a reduction in hours worked and employment for these workers (Neumark et al., 2004). Based on the literature, we can see that the relationship between wages and labour productivity is influenced by several factors, including "wage inequality", "organisational context" and "age structure of employees". These topics illustrate how wage inequality can affect overall productivity, the role of workplace practises in shaping earnings, and the impact of worker demographics on productivity levels. Wage inequality can lead to lower effort among workers with lower wages, which in turn reduces overall labour productivity due to lower overall effort levels. Wage inequality can lead to lower effort among workers with lower wages, which in turn reduces overall labour productivity due to lower overall effort levels (Policardo et al, 2019). The variance in earnings for high-wage occupations across industries is largely explained by the prevalence of pay-for-performance practises. The influence of organisational context on wage determination underlines the importance of workplace dynamics in understanding wage-productivity relationships (Hanley, 2011). Research suggests a negative correlation between the proportion of older workers and labour productivity, although this effect may vary depending on the analytical model used (Garnero et al., 2016). Some studies have found a positive correlation between the proportion of older workers and productivity, suggesting that older workers can contribute positively to organisational performance (Mahlberg et al., 2013).

2.2 Productivity and unemployment rate

The relationship between labour productivity and the unemployment rate is complex and depends on the economic context, labour market conditions and the time frame under consideration. There is a relationship between productivity and the unemployment rate, but it is nuanced. Unemployment can negatively affect productivity through labour underutilisation, but it can also lead to short-term productivity gains through labour market adjustments and automation. Economic theory and empirical studies point to different dynamics between these two variables. In the short term, productivity growth can lead to an increase in unemployment, as shown by empirical analyses that demonstrate a temporary increase in unemployment following productivity increases (Chen and Semmler, 2018). The relationship is scaledependent, suggesting that the impact of productivity on unemployment may vary over different time horizons (Gallegati et al., 2016). In the long run, productivity growth and unemployment tend to have a negative co-variance, suggesting that unemployment decreases as productivity increases (Chen and Semmler, 2018). The long-term analysis suggests that productivity growth ultimately creates employment opportunities and offsets the short-term negative effects (Gallegati et al., 2016). Economic policy has a huge impact on the labour market. Economic policies, such as the introduction of minimum wages, can have complex effects on employment that can lead to negative overall employment outcomes despite the benefits for some low-wage workers (Holtemóller and Pohle, 2020). The global economic crisis of 2007-2009 has shown significant changes in labour markets, with different effects on employment and unemployment in different countries, influenced by factors such as production volumes and productivity (Kwiatkowski, 2011). To summarise, the relationship between productivity and unemployment rates is complex. Short-term increases in productivity can lead to higher unemployment, while long-term trends suggest a negative correlation between the two.

In addition, economic policy plays a crucial role in shaping labour market outcomes, which further complicates this relationship.

2.3 Productivity and R&D

The relationship between productivity and investment in research and development (R&D) is generally positive and well documented in the economic literature. R&D plays a crucial role in promoting innovation, technological progress and efficiency gains, all of which contribute to labour productivity growth. The relationship between productivity and R&D is explored through various channels, including "technological change", "efficiency of R&D investment" and "complementarities between innovations". These topics highlight how R&D contributes to productivity growth and the importance of efficient R&D practises in different industrial contexts. R&D and ICT have been shown to account for almost 95% of productivity growth in OECD countries, indicating a strong link between technological investment and productivity performance (Pieri et al., 2018). In high-tech industries, technological progress is identified as the primary driver of growth, especially in the national and eastern regions of China (Chen et al., 2022). In summary, the relationship between productivity and R&D is characterised by the significant role of technological change, the efficiency of R&D investment and the complementarity between different forms of innovation. Together, these factors contribute to increasing productivity in different industries and emphasise the importance of strategic R&D management and investment.

3. METHODOLOGY

In the empirical part of the paper, we investigate the relationship between labour productivity and the wages that workers receive as compensation for their work. Proving the existence of a relationship between these two variables is the main objective of this paper. By reviewing the literature and previous research in this area, a research problem was defined, which is reflected in the hypotheses put forward:

* H1: GDP per capita has a positive and significant impact on labour productivity (β1>0).

* H2: The average salary has a positive and significant influence on labour productivity (β2>0).

* H3: The unemployment rate has a negative influence on labour productivity (β3<0).

* H4: Investments in research and development have a positive effect on labour productivity (β4>0).

To successfully confirm the defined hypotheses, we used the multiple linear regression method applied to panel data to prove the relationship between the variables. The data were collected from Eurostat (all data are publicly available). The study was conducted in 10 countries of the European Union, mainly in the countries of Central and South-Eastern Europe (Austria, Bulgaria, Czech Republic, Greece, Croatia, Hungary, Poland, Romania, Slovenia, Slovakia) over a ten-year period (from 2014 to 2023). The multiple linear regression model was created as follows:

...(ProQuest: ... denotes formula omitted.)

where is:

GDPit gross domestic product for country i in year t

emit employment for country i in year t

PRLit - labor productivity for country i in year t

GDPpcit - gross domestic product per capita for country i in year t

AWit - average wage for country i in year t

UERit - unemployment rate for country i in year t

GERDit - investments in research and development for country i in year t

Logarithmising the dependent variable and the most important independent variables reduces heteroscedasticity and allows the coefficients to be interpreted as percentage changes. Control variables were used in the model to eliminate confounding, improve the accuracy of the assessment and increase the reliability of the conclusion. For the control variables, we opted for the unemployment rate (UER) and investment in research and development (GERD). The unemployment rate was chosen as a control variable because it affects the labour force and productivity. A high unemployment rate may indicate an inefficient use of labour, which has a negative impact on the overall productivity of the economy. On the other hand, in some situations, unemployment can influence labour selection, with employers retaining or hiring only the most productive workers, which increases productivity per worker. Including the unemployment rate provides a better understanding of how the labour market shapes the relationships between productivity and independent variables such as wages. Investment in research and development as a control variable is a key factor in economic modelling as it influences innovation, technological progress and productivity. In the context of our research, this variable serves to eliminate possible biases and provide a more accurate insight into the relationship between the main variables (e.g. GDP per capita, average wages) and labour productivity. It should be emphasised at this point that we have taken the values of the variable with a one-year lag, as investment in research and development cannot have an immediate impact and influence productivity. Only in the period in which the investments are used can they have an impact on labour productivity. After we had set up the model and the variables, it was necessary to resolve the doubt with which method to calculate the regression coefficients (linear regression of fixed effects or linear regression of variable effects). To remove the doubt, we performed the Hausman test. This test is one of the most important instruments for determining which panel model is the best for us, i.e. which panel model will give us a more efficient result. The null hypothesis of the Hausman test assumes that there is no significant difference in the estimated coefficients, i.e. in this case the random effects model should be used. The Hausman test was calculated using the STATA 18 software package and the results are shown below (Figure 1.).

The result of the Hausman test for the panel data used in this study confirmed our expectations and showed that the fixed effects model was more appropriate. The previously defined regression model was calculated using the weighted least squares method with fixed effects. The weighted least squares (WLS) method is used with panel data when heteroscedasticity or an irregular distribution of variance within or between different panel units is present. The aim of WLS is to adjust the model so that it provides reliable coefficient estimates by reducing the influence of units with higher variance. As in our data, some residual variances differ substantially between panels (e.g. countries with different economic size such as Austria). In this case, WLS is an appropriate method as it gives more weight to the units with lower variance and thus reduces the bias of the model. The results and the analysis of the regression coefficients are presented in the following title.

4. EMPIRICAL RESULTS

Applying multiple linear regression to panel data covering ten different countries over a tenyear period, we obtained the following results. We used the STATA 18 software package to calculate the model (Figure 2.). We can accept the first hypothesis that defines the relationship between labor productivity and GDP per capita for the reason that there is a positive relationship between the mentioned variables, β1 = 0.9054, while the p-value is equal to 0.000, which confirms the statistical significance of the calculated coefficient. So, when GDP per capita increases by one percent, profitability will increase by 0.90 percent. With the second hypothesis, we defined the relationship between labor productivity and average salary. The value of the coefficient is positive, β2 = 0.1072, and the p-value is 0.000 indicates that the calculated value is statistically significant. As we aspected, we can confirm the second hypothesis. With the third hypothesis, we defined the relationship between labor productivity and the unemployment rate, where we expected a negative relationship. The value of the calculated coefficient is positive, β3 = 0.0679, and it is statistically significant because the pvalue is equal to 0.000, s we must recject third hypothesis. Although a positive relationship between labour productivity and the unemployment rate is counterintuitive, it can occur under certain economic conditions. During periods of high unemployment, firms often retain only their most productive workers, resulting in higher average productivity, a phenomenon known as labour hoarding or selective retention. In addition, unemployment can prompt companies to introduce technological solutions and automate processes, which increases productivity per worker. In times of economic uncertainty, workers may also work harder to secure their jobs, further increasing productivity. Structural changes in the economy, such as the transition from low-productivity industries to high-tech sectors, may also explain this relationship. Finally, the observed relationship may also result from model limitations, such as the lack of a time lag between unemployment and its effect on productivity. Future research should investigate this relationship in more detail and include a time lag. The fourth hypothesis defines the relationship between labour productivity and investment in research and development. In this hypothesis, a positive relationship between the variables mentioned was expected. The value of the calculated coefficient is β4 = -0,0416, while the p-value is equal to 0.000. The value of the calculated coefficient is statistically significant, which forces us to conclude that the relationship between these variables is opposite and we must recject fourth hypothesis. The author had already come to this conclusion in his earlier studies. The negative relationship between labour productivity and investment in research and development (R&D) may seem unexpected given the widely accepted view that investment in R&D promotes innovation and increases productivity. However, there are possible reasons and explanations as to why this result might occur. Firstly, we mention the insufficient time lag, as investments often take a long time to deliver results. Innovations need time to be researched, developed, tested and implemented before they contribute to an increase in productivity. The next reason we cite is that a lack of skilled labour can limit the effectiveness of investment in research and development. In countries or regions with an inadequately educated or low-skilled labour force, investment in research and development is not used effectively, which can reduce productivity.

5. DISCUSSION

The results of this study shed light on the determinants of labour productivity in ten European Union countries in the period 20142023, using a fixed-effects panel regression model to identify complex relationships between labour productivity and key macroeconomic indicators: GDP per capita, average wages, the unemployment rate and investment in research and development. The positive relationship between GDP per capita and labour productivity is in line with theoretical expectations and underlines the role of economic prosperity and efficiency in increasing productivity. Similarly, the statistically significant positive effect of average wages on productivity emphasises the importance of fair pay in motivating employees and promoting the development of human capital. These results suggest that policies that promote fair wage growth and improve economic performance are crucial for increasing productivity in the EU. Unexpectedly, a positive relationship was found between the unemployment rate and labour productivity. While counterintuitive, this result is consistent with scenarios in which companies retain only their most productive employees in times of high unemployment or invest in automation and technological solutions to address labour shortages. In addition, economic uncertainty could lead employees to work harder. These dynamics should be further analysed, especially with regard to the long-term effects and possible differences between sectors. The most surprising result was the negative relationship between investments in research and development and labor productivity, which contradicts established theories suggesting that R&D drives innovation and efficiency. Possible explanations include insufficient time lags, inefficiencies in R&D allocation, or workforce skill mismatches limiting the practical application of innovations. This finding highlights the need for better alignment between R&D strategies and labor market dynamics to ensure productive outcomes. This study contributes to the understanding of the determinants of labour productivity in the EU, but also raises questions for future research. In particular, the inclusion of time lags for variables such as R&D and the analysis of sector-specific effects could provide deeper insights. Furthermore, analysing the role of institutional factors and regional differences could refine the understanding of these complex relationships.

6. CONCLUSION

This study examined the relationship between labour productivity and four key macroeconomic indicators in ten EU countries from 2014 to 2023. Using a fixed-effects panel data approach, the analysis yielded significant, albeit nuanced, results. GDP per capita and average wages were found to have a positive and significant impact on labour productivity, confirming their role as key drivers of economic efficiency. These findings emphasise the importance of policies to promote economic growth and ensure fair wage practises to support sustainable productivity gains. Policy makers should focus on promoting equitable wage growth by supporting wagesetting mechanisms that reflect the contribution of workers while maintaining competitiveness. Moderate increases in minimum wages, accompanied by measures to support small and medium-sized enterprises (SMEs), can increase worker motivation and productivity without unduly burdening businesses. While the positive relationship between the unemployment rate and productivity is unexpected, it reflects underlying dynamics such as selective retention, technological adaptation and increased effort by workers in times of economic uncertainty. This finding suggests that labour market policies should strike a careful balance between reducing unemployment and productivity-enhancing measures. Policy makers should aim to reduce unemployment while promoting productivity-enhancing measures. This includes incentives for companies to invest in automation and digital transformation in times of economic downturn, combined with active labour market policies to retrain and reintegrate laid-off workers. Contrary to expectations, R&D investments showed a negative relationship to productivity. This result points to possible inefficiencies in the implementation of R&D, insufficient time horizons for the realisation of returns or structural problems in the labour markets. Policy makers should prioritise targeted R&D strategies that are aligned with the skills of the workforce and sectoral needs in order to unlock the full potential of innovation. Policymakers should ensure that R&D investments are strategically targeted to sectors with high innovation potential and that resources are allocated effectively. Improving co-operation between universities, private companies and governments can also improve the efficiency of R&D initiatives. In conclusion, this study highlights the complex interplay between macroeconomic factors and labour productivity. It provides a framework for understanding these dynamics, but also highlights areas that require further investigation, including time lags, sectoral heterogeneity and institutional influences. The results provide actionable insights for policy makers seeking to improve productivity and competitiveness in the European Union.

References

LITERATURE:

1. Anghelache, C. and Anghel, ?. ?. (2018). Analysis Models and Methods of the Life Quality in Romania. Romanian Statistical Review, (2) pp. 59-86.

2. Bauducco, S. and Janiak, A. (2018). The macroeconomic consequences of raising the minimum wage: Capital accumulation, employment and the wage distribution. European Economic Review, 101 pp. 57-76.

3. Becker, G. (1964). Human Capital: A Theoretical and Empirical Analysis, with Special Reference to Education. New York: Columbia University Press.

4. Chen, P. and Semmler, W. (2018). Short and Long Effects of Productivity on Unemployment. Open Economies Review, 29(4) pp. 853-878.

5. Chen, X. Q., Liu, X. W. and Zhu, ?. Y. (2022). Comparative analysis of total factor productivity in China's high-tech industries. Technological Forecasting and Social Change, 175.

6. Claveria, O. (2019). A new consensus-based unemployment indicator. Applied Economics Letters, 26(10) pp. 812-817.

7. Gallegati, M., Gallegati, M., Ramsey, J. ?. and Semmler, W. (2016). Productivity and unemployment: a scale-by-scale panel data analysis for the G7 countries. Studies in Nonlinear Dynamics and Econometrics, 20(4) pp. 477-493.

8. Garnero, A., Giuliano, R., Mahy, B. and Rycx, F. (2016). Productivity, wages and profits among Belgian firms: do fixed-term contracts matter?. International Journal of Manpower, 37(2) pp. 303-322.

9. Hanley, C. (2011). Investigating the organizational sources of high-wage earnings growth and rising inequality. Social Science Research, 40(3) pp. 902-916.

10. Holtemóller, O. and Pohle, F. (2020). Employment effects of introducing a minimum wage: The case of Germany. Economic Modelling, 89 pp. 108-121.

11. Hu, J. L., Yang, C. H. and Chen, C. P. (2014). R&D Efficiency and the National Innovation System: An International Comparison using the Distance Function Approach. Bulletin of Economic Research, 66(1) pp. 55-71.

12. Kwiatkowski, E. (2011). Global Crisis and Labour Market in Poland and other Countries of the Vyshehrad Group. Ekonomista, (1) pp. 37-53.

13. Mahlberg, B., Freund, 1. and Prskawetz, A. (2013). Ageing, productivity and wages in Austria: sector level evidence. Empirica, 40(4) pp. 561-584.

14. Mohnen, P. and Hall, ?. H. (2013). Innovation and Productivity: an Update. Eurasian Business Review, 3(1) pp. 47-65.

15. Neumark, D., Schweitzer, M. and Wascher, W. (2004). Minimum wage effects throughout the wage distribution. Journal of human Resources, 39(2) pp. 425-450.

16. Pieri, F., Vecchi, M. and Venturini, ?. (2018). Modelling the joint impact of R&D and ICT on productivity: A frontier analysis approach. Research Policy, 47(9) pp. 1842-1852.

17. Policardo, L., Punzo, L. F. and Carrera, E. J. S. (2019). On the wage-productivity causal relationship. Empirical Economics, 57(1) pp. 329-343.