| Abstract [eng] |
To identify the impact of automation on the wage-productivity gap, the Master's thesis starts by examining the theoretical aspects of the topic and analyses the literature collected. The theoretical part of the thesis presents the most important concepts – the wage-productivity gap and automation, which affects it – as well as other factors affecting the gap and their corresponding indicators. The most commonly used factors to explain changes in the gap are: automation, wage share, unemployment rate, trade union bargaining power and the terms of trade. The literature analysis shows that automation processes, often reflected in the ICT sector's share of GDP, increase the wage-productivity gap. Still, there are different views on the direction of the factor's impact. While no consensus has emerged on the factors highlighted, the most common view is that the increasing share of wages in national income, the improving bargaining power of workers and more favourable terms of trade tend to narrow the gap. Meanwhile, the unemployment rate tends to contribute to widening the wage-productivity gap by affecting the wage side. The methodological part of the Master's thesis justifies the selected research sample, the period and the indicators reflecting the chosen macroeconomic factors, and presents their calculations. An empirical model of the study is also created, and the limitations that may affect the study results are identified. The methodological part also describes the technical steps to ensure the robustness of the results: Gauss-Markov assumption testing, panel specification tests, and ways of solving potential problems. The empirical part of the Master's thesis starts with a dynamic analysis of the wage-productivity gap and its determinants to identify the changes in the indicators over the period analysed, comparing them with the European Union average. It is noted that, contrary to economists' fears, the gap fell by 0,8 percentage points in the EU countries over the analysed period; however, some countries, such as Romania and Ireland, have shown exceptional changes. Taking other factors into account, it is found that automation tends to increase in the EU countries over the period analysed, while workers' bargaining power and terms of trade also improve. Unemployment rates have fallen on average among EU Member States over the analysed period. However, there is also a downward trend, albeit slight, in the wage share. Based on the selected indicators and their specifics, an initial Ordinary Least Squares Method (OLS) model is constructed, with the necessary adjustments made to take into account the results of the tests for non-linearity, heteroskedasticity, autocorrelation, cross-sectional dependence and panel specification. It was found that automation does not significantly impact the gap in a fixed-effects (FE) model, reflecting short-run change analysis. Therefore, the hypothesis of the Master's thesis is rejected in this case. No significant effect is also present when examining wage shares and workers' bargaining power. Meanwhile, the unemployment rate is found to increase the gap, while the terms of trade are found to decrease it. To determine whether the automation process’s impact is, indeed, insignificant, alternative modeling methods are used: lagged effects analysis, removal of exclusions, differenced variables, and non-overlapping and overlapping-period models capturing the effects in the long run. The analysis of a five-year non overlapping model shows that automation significantly increases the wage-productivity gap when the long run is considered. This suggests that the market needs more than one year to react to the introduction of new machinery and equipment in enterprises. In the long run, it is found that the wage share reduces the gap significantly. |
