This thesis presents three papers on the determination of factor income shares and economic growth. The aim is to address a topical but often overlooked issue in economics, namely, the functional distribution of income, i.e., the division of national income amongst factors that concur to its creation (labor and capital). The relative stability of the labor share of income is a sound foundation in modern macroeconomic models. However, there is strong empirical evidence suggesting that the global labor share in advanced economies has remarkably declined since the early 80s, with a generalized fall taking place in most countries and sectors. Alongside this trend, recent research on the field stresses the macrodimension of income distribution but no theory has emerged yet, thus relegating the topic as a secondary one. In the first part of the work, we make some technical and statistical considerations on the measurement and composition of labor share - since not all types of income in the national accounts can be easily attributed to either capital or labor. Once the issues related to the measurement of selfemployment income have been considered, we will see that there are multiple forces which contribute to the observed fluctuations of the labor share, and among them there is a significant amount of selfreinforcement. However, what emerges is that, despite the number of insights one can gain from the empirical literature, these theories alone cannot provide a complete explanation of the current worsening of income distribution. It is precisely this reason that motivates our work. From a theoretical perspective - which is crucial to understand the cross-relationships between the economic variables involved - we further show that the cost of modifying existing models is in some cases low if compared to the benefits that can be obtained and is especially important in understanding the implications of this phenomenon on economic growth. In the first paper we show that the CES production function presents at least two criticalities which are not compatible with the necessity to set up a modern theory of distribution. Therefore, we present a novel theory on how factor shares are determined at the aggregate level. By using AMECO data for a set of 20 industrialized countries over a 58-year period we provide empirical evidence against the CES and develop micro foundations for an aggregate production function that better fits the data and has the property of a variable elasticity of substitution. Then we relax the assumption of perfect competition by proposing a time-series calculation of the aggregate price mark-up and provide estimates of the elasticity of substitution under such a framework of imperfect competition in product markets. Finally, we test the prediction of the model by means of a numerical simulation. We find that firms’ rising markups along with biased technological change can account for a significant part of the labor shares’ decline observed in the last 40 years. The results also suggest complementarity between labor and capital in most of the cases, with the elasticity of substitution σ that has been below unity on average, fluctuating around 0.7-1.16. In the second paper the focus is shifted to heterogeneous sectors and firms and we study the micro-Level determinants of factor shares. Existing empirical analysis typically rely on industry or aggregate macro data, thus strongly downplaying the role of firm-level variables in the determination of wages. We analyze micro panel data from AMADEUS and seek to understand the dynamics of labor share in 19 broad sectors of the EU28. More specifically we explore the role of technological change, product and labor market imperfections. We build on insights from Bentolila and Saint-Paul (2003), Blanchard and Giavazzi (2003) Azmat et al. (2012), Karabarbounis and Neiman (2014) and econometric techniques from De Loecker and Warzynski (2012), to propose a model in which firms can be heterogeneous in terms of capital employed in production, market power and productivity. The core argument of our theoretical framework is the relationship between the labor share and the capital-output ratio. We show that this relationship holds also for firms and acknowledge the possible existence of relevant non-linearities. Labor share’s movement turns out to be driven by a complex interplay of conditions for capital and labor, the nature of technological progress and imperfect market structures which can shift the curve but also - by creating a discrepancy between the marginal product of labor and the real wage - cause departures from it. Broadly confirming results from previous cross-country and industry-level studies, we find that the main factor decreasing labor shares are connected to capital deepening (a 1% increase in the capital-output reduce the labor share by -0.03 percentage points) in conjunction with capital-augmenting technical progress and labor substitution (-0.15). Although institutional factors play a significant role in some specific industries (like for instance Construction, Manufacturing and Transportation), they appear to be less important for the aggregate economy. Finally, in the third and last paper we introduce new elements into a standard macroeconomic growth model and develop a post-Keynesian model which allow us to acquire valuable insights on the macroeconomic effects of changes in the aggregate distribution of income. We show, analytically, that: (i) income distribution matters mostly in the medium run; (ii) real wage restraint policies along with labor market deregulation can depress capital accumulation and growth; (iii) a decline in EPL may reduce the equilibrium unemployment. Then we test the predictions of the model by estimating the impact of a change in the labor share on economic growth in a sample of 20 industrialized OECD countries. At the national level, a decrease in the labor share leads to lower growth in Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, Luxembourg, Netherlands, Norway, Portugal, Spain and Sweden; whereas it stimulates growth in Australia, Canada, Ireland, United Kingdom and the United States. However, a simultaneous decline in the labor share in all these countries lead to a decline in global growth. Finally, we turn our attention to the long-run dynamics of the model by means of a structural vector autoregression (VAR). We focus on the relationship between effective demand, income distribution, labor market regulation, capital accumulation, labor productivity and unemployment. The VAR model is estimated for France, Germany, Italy, Spain, the UK and the USA. We find that employment is demand-led, and that income distribution may influence either demand or employment. Technological progress affects income distribution as well as employment. The policy conclusions of the paper shed light on the limits of international competitiveness strategies based on wage competition in a highly integrated global economy.
Three Essays on Income Distribution and Economic Growth
Bellocchi Alessandro
2021
Abstract
This thesis presents three papers on the determination of factor income shares and economic growth. The aim is to address a topical but often overlooked issue in economics, namely, the functional distribution of income, i.e., the division of national income amongst factors that concur to its creation (labor and capital). The relative stability of the labor share of income is a sound foundation in modern macroeconomic models. However, there is strong empirical evidence suggesting that the global labor share in advanced economies has remarkably declined since the early 80s, with a generalized fall taking place in most countries and sectors. Alongside this trend, recent research on the field stresses the macrodimension of income distribution but no theory has emerged yet, thus relegating the topic as a secondary one. In the first part of the work, we make some technical and statistical considerations on the measurement and composition of labor share - since not all types of income in the national accounts can be easily attributed to either capital or labor. Once the issues related to the measurement of selfemployment income have been considered, we will see that there are multiple forces which contribute to the observed fluctuations of the labor share, and among them there is a significant amount of selfreinforcement. However, what emerges is that, despite the number of insights one can gain from the empirical literature, these theories alone cannot provide a complete explanation of the current worsening of income distribution. It is precisely this reason that motivates our work. From a theoretical perspective - which is crucial to understand the cross-relationships between the economic variables involved - we further show that the cost of modifying existing models is in some cases low if compared to the benefits that can be obtained and is especially important in understanding the implications of this phenomenon on economic growth. In the first paper we show that the CES production function presents at least two criticalities which are not compatible with the necessity to set up a modern theory of distribution. Therefore, we present a novel theory on how factor shares are determined at the aggregate level. By using AMECO data for a set of 20 industrialized countries over a 58-year period we provide empirical evidence against the CES and develop micro foundations for an aggregate production function that better fits the data and has the property of a variable elasticity of substitution. Then we relax the assumption of perfect competition by proposing a time-series calculation of the aggregate price mark-up and provide estimates of the elasticity of substitution under such a framework of imperfect competition in product markets. Finally, we test the prediction of the model by means of a numerical simulation. We find that firms’ rising markups along with biased technological change can account for a significant part of the labor shares’ decline observed in the last 40 years. The results also suggest complementarity between labor and capital in most of the cases, with the elasticity of substitution σ that has been below unity on average, fluctuating around 0.7-1.16. In the second paper the focus is shifted to heterogeneous sectors and firms and we study the micro-Level determinants of factor shares. Existing empirical analysis typically rely on industry or aggregate macro data, thus strongly downplaying the role of firm-level variables in the determination of wages. We analyze micro panel data from AMADEUS and seek to understand the dynamics of labor share in 19 broad sectors of the EU28. More specifically we explore the role of technological change, product and labor market imperfections. We build on insights from Bentolila and Saint-Paul (2003), Blanchard and Giavazzi (2003) Azmat et al. (2012), Karabarbounis and Neiman (2014) and econometric techniques from De Loecker and Warzynski (2012), to propose a model in which firms can be heterogeneous in terms of capital employed in production, market power and productivity. The core argument of our theoretical framework is the relationship between the labor share and the capital-output ratio. We show that this relationship holds also for firms and acknowledge the possible existence of relevant non-linearities. Labor share’s movement turns out to be driven by a complex interplay of conditions for capital and labor, the nature of technological progress and imperfect market structures which can shift the curve but also - by creating a discrepancy between the marginal product of labor and the real wage - cause departures from it. Broadly confirming results from previous cross-country and industry-level studies, we find that the main factor decreasing labor shares are connected to capital deepening (a 1% increase in the capital-output reduce the labor share by -0.03 percentage points) in conjunction with capital-augmenting technical progress and labor substitution (-0.15). Although institutional factors play a significant role in some specific industries (like for instance Construction, Manufacturing and Transportation), they appear to be less important for the aggregate economy. Finally, in the third and last paper we introduce new elements into a standard macroeconomic growth model and develop a post-Keynesian model which allow us to acquire valuable insights on the macroeconomic effects of changes in the aggregate distribution of income. We show, analytically, that: (i) income distribution matters mostly in the medium run; (ii) real wage restraint policies along with labor market deregulation can depress capital accumulation and growth; (iii) a decline in EPL may reduce the equilibrium unemployment. Then we test the predictions of the model by estimating the impact of a change in the labor share on economic growth in a sample of 20 industrialized OECD countries. At the national level, a decrease in the labor share leads to lower growth in Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, Luxembourg, Netherlands, Norway, Portugal, Spain and Sweden; whereas it stimulates growth in Australia, Canada, Ireland, United Kingdom and the United States. However, a simultaneous decline in the labor share in all these countries lead to a decline in global growth. Finally, we turn our attention to the long-run dynamics of the model by means of a structural vector autoregression (VAR). We focus on the relationship between effective demand, income distribution, labor market regulation, capital accumulation, labor productivity and unemployment. The VAR model is estimated for France, Germany, Italy, Spain, the UK and the USA. We find that employment is demand-led, and that income distribution may influence either demand or employment. Technological progress affects income distribution as well as employment. The policy conclusions of the paper shed light on the limits of international competitiveness strategies based on wage competition in a highly integrated global economy.File | Dimensione | Formato | |
---|---|---|---|
Phd_uniurb_286386.pdf
accesso aperto
Tipologia:
DT
Licenza:
Creative commons
Dimensione
13.64 MB
Formato
Adobe PDF
|
13.64 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.