Für neue Autoren:
kostenlos, einfach und schnell
Für bereits registrierte Autoren
87 Seiten, Note: 1,5
List of Abbreviations
List of Figures
List of Tables
2 Theoretical framework of European secular stagnation
2.1 Importance of growth
2.3 The Importance of productivity and TFP
2.4 Definition of secular stagnation
3 Secular stagnation, productivity and their causes in Europe
3.1 Literature concerning secular stagnation and productivity
3.2 Determinants of productivity in the EU-15 area
3.2.1 European values
3.2.2 Education and human capital
3.2.3 Energy prices and use of energy per capita
3.2.5 European cooperation, economies of scale, resource allocation
4.2 Considered variables
4.5 The TFP-slump environment
5 Regression Analysis
5.1 Logit panel data analysis
5.2.2 Education and human capital
5.2.3 European values
5.2.4 Europe-wide cooperation, economies of scale, resource allocation
6.1 Monetary policy
6.3 Structural reforms
7 Conclusions and Implications
List of Abbreviations
I would like to thank Professor Dr. Florentine Schwark for her very helpful guidance throughout every step of the development of this thesis in such a difficult time. Furthermore, I would like to thank the Humboldt University of Berlin for providing the material and references relevant for my research.
The aim of this master thesis was to identify the background and determinants of the stagnating economic growth in the EU-15 countries and to evaluate whether the region has entered a secular stagnation. To this end, the influence of a structural total factor productivity (TFP) slowdown on European economic growth was empirically examined. In order to obtain the results for this study, a logit panel regression was carried out with the EU-15 countries over the period from 1950 to 2017. It became particularly clear that the TFP slowdowns increased over the years while TFP upswings decreased. This tendency can mainly be attributed to a lack of human capital, too little intra-European cooperation, too many regulatory hurdles and too much employment in the services sector. These highly structural factors and their influence on TFP growth and, ultimately, economic growth, demonstrate that the EU-15 countries have fallen into an economic stagnation dynamic that is strongly reminiscent of the model of secular stagnation. Only the way in which policymakers deal with the current pandemic will show how strong and protracted the current stagnation of economic growth will be.
Abbildung in dieser Leseprobe nicht enthalten
1 Level GDP per capita (in current USD)Euroastat (2020)
2 Annual TFP growth estimates (left) Productivity Waves (right) Bergeaud et al. (2014)
3 Downward Revision in Potential GDP, Eurozone Summers (2014a)
4 Structural break
5 Histogram per income (3Y, 1.5% criterium)
7 GDP per capita (current USD) - comparison largest / smallest
8 GDP per capita (current USD) - comparison largest total econimies
9 TFP EURO + GB & NOR (1970 - 2012) Bergeaud et al. (2014)
10 TFP EURO + GB & NOR (1890 - 2012) Bergeaud et al. (2014)
13 Accelerations - frequency table
14 Slumps - per country observations (3Y, criterium)
15 Slumps - per country observations (2Y, criterium)
16 Accelerations - per country observations (all criteria)
17 Accelerations - per country observations (3Y, criterium)
18 Acceleration - per country observations (2Y, criterium)
19 Slumps - per year observations; Slump equal to first Year of period (all criteria)
20 Slumps - per year observations; Slump equal to highest growth rate in period (all criteria)
21 Slumps - per year observations (3Y, criterium)
22 Slumps - per year observations (2Y, criterium)
23 Slumps - per year observations (5Y, criterium)
24 Accelerations - per year observations Slump equal to first Year of period (all criteria)
25 Accelerations - per year observations Slump equal to highest growth rate in period (allcriteria)
26 Accelerations - per year observations (2Y, criterium)
27 Accelerations - per year observations (3Y, criterium)
28 Decreasing wage standard deviation
29 Regression wage volatility on mean TFP growth
31 Regression cross-county correlations (Harroika Puzzle) on mean TFP growth
1 T-Test 3 Years - 1.5% criterion
2 T-Test 2 Years - 1.5% criterion
3 T-Test 2 Years - 1.5% criterion from 1971
4 T-Test 3 Years - 1.5% criterion from 1971
5 List of variables
6 TFP Slump Regression from 1971 (3Y 1.5% criterium - important observations)
7 TFP Slump Regression (3Y 1.5% criterium)
8 TFP Slump Regression from 1971 (3Y 1.5% criterium)
9 TFP Slump Regression (2Y 1.5% criterium)
10 TFP Slump Regression from 1971 (2Y 1.5% criterium)
11 TFP Slump Regression Joint consideration ARegQ, GovE)
12 TFP Slump Regression Joint consideration HC, years of schooling, yearstertiary
13 TFP Accelerates Regression (3Y 1.5% criterium)
14 TFP Accelerates Regression from 1971 (3Y 1.5% criterium)
15 TFP Accelerates Regression (2Y 1.5% criterium)
16 TFP Accelerates Regression from 1971 (2Y 1.5% criterium)
European economic growth has been on the decline for decades. Through the reinterpretation of Summers (2015b) and, shortly afterwards, Eichengreen (2014), the concept of secular stagnation has gained new popularity. The term secular stagnation was originally developed by Hansen (1939) who, at the time of his work, claimed that the economy of United States had fallen into a phase of unsatisfactory overall demand and low investment. More precisely, the argument follows the line that, by closing borders to immigration and, given a slowdown in population growth, investment opportunities had been significantly reduced. In light of a current phase of low economic growth and steadily falling interest rates, this model of secular stagnation was for the first time applied to the present day by Summers (2015b) at the 2013 IMF research conference.
The fact that developing economies have entered a phase of long-term, weak growth rates have triggered a major debate among economists—many of whom focus on stagnating technological progress and its significance for economic development. Eichengreen et al. (2015), for example, questions whether stagnating productivity growth will lead to long-term economic stagnation and the factors that may cause this. In general, however, the economists can largely be divided into two camps. The first of these are the so-called technology optimists, like Rogoff (2016), who argues that we currently find ourselves in a very extended economic cycle that will soon come to an end. On the other hand, there are technology pessimists, such as Gordon (2012), who predict that all the great and important inventions have already come to pass, meaning continuing with past productivity growth rates will be barely possible.
At the time of writing, 13 years after the beginning of the financial and euro crisis, the European Union (EU) is in a situation where growth is not expected to gain momentum despite negative key interest rates. The crisis, which began in the USA in 2007 when the real estate bubble burst and then steered into the “subprime crisis” (Shiller (2012)), resulted in a global financial crisis that hit the EU particularly hard (Dabla-Norris et al. (2015)). In the aftermath of the crisis, Summers (2015b) emphasizes that the development of GDP, employment and overall demand was conspicuous. Whereby he stresses that the figures should usually pick up again after such a crisis, since the lack of credit is over and claims can be made. This was not observed after the financial crisis, however and, in looking for a reason for this, Summers (2015a) points to the unhealthy growth of the last 30 years. Indeed, he concludes that it may not simply be the aftermath of the financial crisis of 2007-2008, but that we may have been in secular stagnation for almost half a century. Now, with the consequences of the corona pandemic, we may be facing another recession. If economies fail to rebound sufficiently during recovery phases, long-term growth will fall even further due to the “ratchet effect” (Wohltmann (2005)). Whether the EU-15 countries have entered a phase of structural stagnation is therefore more relevant than ever.
With this in mind, the present study sets out to empirically analyze the effects of a structural slowdown in total factor productivity on European economic growth. The thesis uses the paper, “The global productivity slump: Common and country-specific factors” by Eichengreen et al. (2015) as methodological basis and applies it to the EU.
Even if structural patterns are difficult to change by policy measures, especially in the short term, it is important to identify the causes of this stagnation. Since the end of the Malthusian world (Malthus (1798)), productivity has increased massively. Historically, there have always been periods with higher, and lower, rates of productivity. From the 1970s, rates of productivity have been declining almost unchecked, especially in the eurozone (Bergeaud et al. (2014)). Thus, if we assume that technology is the main driver of economic growth (Barro and I.Martin (2003)) and that technological progress has slowed since the 1970s (Summers (2015a)), then it is clear factors influencing the technology and its growth require analysis. These points raise the following questions:
Will European economic growth increase again in the coming years, or are we heading for secular stagnation? Is stagnating productivity growth responsible for low economic growth, and what are the factors influencing this low productivity growth? What can politics do to stimulate productivity and economic performance?
Following the introduction, Chapter 2 explains the theoretical principles. Here the importance of growth for an economy, the relevance of technology for economic growth, and the model of secular stagnation are defined. Subsequently, Chapter 3 summarizes the common literature on secular stagnation in connection with technology and works out the most important factors influencing technology for the European area. Based on this, the data and the methodology for the following regression analysis are presented in Chapter 4. Based on the introductory literature, it is then possible to interpret the regression analysis in Chapter 5, and to determine causal relationships. With the help of these correlations, in Chapter 6 political recommendations are formulated that can counteract these downward trends over the long term. Finally, Chapter 7 summarizes the findings in the conclusion, and offers an interpretation of their implications for the future.
In order to narrow down the analysis geographically, the following analysis limits itself to the EU-15 states that had joined the EU by April 2004. This is because these countries had already gained substantial experience cooperating within the EU. Of course, these states also generate most of the European economic output, making them very representative of the EU as a whole. If we look at the share of gross domestic product (GDP) of the EU-15 compared to the total EU-27, 80.4% (Eurostat (2017)) is generated by the former. Hence, it can be said that have made a significant contribution to the EU, both historically and economically, which makes the analysis of the economic development of this region all the more appealing.
To get a feeling for the importance of economic growth, it is advisable to look at the period before the Industrial Revolution, the time of the Malthusian trap. Prior to 1800, there was no growth in per capita income and thus no sustainable increase in living standards. If income changed over the short term, the mortality rate adjusted accordingly, so that in the long term the subsistence income, that which is just sufficient to survive, was always achieved (Malthus (1798)). 1820 onwards saw fundamental change, however, where for the first time economic output grew faster than the population (Maddison (2006)), hence the Malthusian trap was sprung open and per capita income grew steadily.
Looking at the current development of the GDP per capita of the EU-15 states between 1970 and 2018, a clear increase from (current) USD 2,535.38 to USD 50,277.26 can be seen (Euroas- tat (2020)). Western European GDP’s powerful growth of 4.8% in the Golden Age (1950-1973) later weakened to 2.1% in the neoliberal phase (1973-1998), mainly as a result of increased em- ployment—commonly associated with the post-war baby boomer generation—coming to an end and, with that, rising unemployment and declining productivity (Maddison (2006)). Although a similar development can be observed with the USA, it is striking that the aggregated EU-15 countries show a much higher volatility (Figure 6GDP per capita (current USD) - comparison world regionsfigure.caption.7). Between 1980 and 1985, GDP growth in the EU-15 stagnated, particularly due to the sharp rise in unemployment, which peaked in 1985 (Fitoussi and Phelps (1986)). Between 1995 and 2001, slightly declining GDP per capita growth could be observed, which ultimately turned into a strong upswing culminating in the global financial crisis of 2007-2008. Since the beginning of the crisis, no sustainable GDP per capita growth can be observed (Figure 1Level GDP per capita (in current USD)Euroastat (2020)figure.caption.2). This may be due to structural stagnation, but could equally point to temporary stagnation that will soon regain momentum.
Looking at the development of GDP within the EU-15 countries, it is clear they differ greatly.
For example, with the highest GDP per capita in the world, especially since 2001 Luxembourg has increased the gap between itself and the EU-15 countries, a gap which since 2007 stands at a GDP double the EU-15 average (Figure 7GDP per capita (current USD) - comparison largest / smallestfigure.caption.8). On the other hand, countries such as Portugal and Greece have fallen further behind the EU-15, with GDPs at less than half the EU-15 average since 2010. Turning to the three largest economies of the EU-15, namely, Germany, France and the United Kingdom (UK), which together generate 51.4% of the GDP (Eurostat (2017)) of the entire EU, one observes that they follow the EU-15 average relatively simultaneously (Figure 8GDP per capita (current USD) - comparison largest total econimiesfigure.caption.9).
In international comparison, the EU-15 countries are among the richest in the world (Barro and I.Martin (2003)). Historically, there have been periods of concentration and de-concentration in terms of world industrial leadership. China and India, for example, were superseded by Britain and later other western European economies during the Industrial Revolution and, with the onset of the first globalization, the USA became increasingly important economically. The question today, therefore, is how sustainable European economic growth continues to be, and whether it will be able to keep up with the emerging Asian economies in the future. Together with the other OECD countries and a few Asian countries, the EU-15 countries started at a very high level in the 1970s and were able to maintain their position due to medium to high growth rates. However, since even very small fluctuations in GDP growth rates can have a major effect on the future level of economic performance (Barro and I.Martin (2003)), it is important to look at the main drivers of economic growth and to work out their influence on the EU-15 countries, points which are addressed in Section 2.2 below.
This section aims to highlight the link between economic growth and total factor productivity (TFP) growth (Sickles (2005)). For this purpose, economic growth is broken down into its individual components using the neoclassical growth accounting model of Solow (1957). The following derivation of the Solow residual is based on the approach of Barro and I.Martin (2003).
The total economic output or GDP Y is calculated using a simple standard production function:
Abbildung in dieser Leseprobe nicht enthalten
This indicates the share of GDP growth that is not due to growth in the use of production factors, labor or capital. This remainder can thus be defined as technical progress, or growth in TFP (Barro and I.Martin (2003)).
Generally speaking, productivity is the ratio of output to input, distinguishing between labor productivity (output/labor hours) and capital productivity (output/capital) (BPB (2016)). In order to place the standard of living of different economies in relation to each other, at the very least a measure of output, inputs and productivity is necessary (Inklaar and Timmer (2013)). In this analysis, as in the Penn World Table 9.1 (PWT 9.1), GDP is used as the output measure. Due to the input and productivity efforts of Hall and Jones (1999) and Caselli (2005), the data set can be seen to have gained in quality in terms of cross-country comparability since PWT 8.0. In addition, the PWT 9.1 TFP dataset contains figures for the EU-15 countries over the period under consideration, namely from 1950 to 2017.
After defining total factor productivity as a clear component of output and technical progress using the Solow residual, the relevance of technical progress for an economy is determined below. This is performed using the simplified growth model of Solow (1956). The aggregated production function for each period t, can again be described as follows:
Abbildung in dieser Leseprobe nicht enthalten
Where L t represents labor input, K t represents capital stock and T represents Harrold-neutral total factor productivity (TFP). The growth rate of TFP is denoted by g, and labor input by n. The intensive form in efficiency units is obtained by division by TL t, and is:
Abbildung in dieser Leseprobe nicht enthalten
If y=(Y/TL) is now constant in the steady state, the output Y grows with a growth rate of g + n and the output per capita Y/L with a growth rate of g. Thus, long-term growth per capita in this model is possible only through the exogenous growth of technical progress g.
Empirical studies (Heston et al. (2006)) show that, due to its very simplified representation, the model only delivers meaningful results when developed countries such as the OECD countries are considered. Since the countries of the EU-15 can be described as very highly developed economies, this model is a very simple and clear method with which to demonstrate the relevance of technological progress for economic growth in this region.
If productivity is the main driver of economic growth and the aim is to analyze the causes of economic stagnation, it is important to examine whether technical progress has stagnated and what its possible causes might be. For this reason, the following looks at the long-term trends in total factor productivity.
In order to be able to consider as large a time horizon as possible, the TFP data is from Bergeaud et al. (2014) is used between 1890 and 2012. The data is available for the eurozone and selected OECD member states. Twelve of the 15 EU-15 countries are located and generate most of their economic output in the eurozone1. Hence, so as to make statements about the productivity development of the EU-15 countries, it is important to examine the eurozone and its TFP development in more detail below.
Looking at the absolute TFP figures in the period 1890 to 2012, it appears at first glance as if TFP has been growing unchecked since the Second World War (Figure 2Annual TFP growth estimates (left) Productivity Waves (right) Bergeaud et al. (2014)figure.caption.3). However, on taking a closer look at technological progress, a first slowdown in the mid-1970s can be observed, followed by a more drastic second slowdown at the beginning of the millennium, which seems to justify asking the question of whether technological progress in Europe has stagnated.
To illustrate the waves of TFP development and assign them to specific periods, Bergeaud et al. (2014) use average annual growth rates of TFP smoothed with Hodrick-Prescott filtration (HP) with a Lamda equal to 500. The first period, from 1890 to the First World War, is characterized by rather moderate TFP growth, driven by technological inventions such as the steam engine developed in the first Industrial Revolution (Gordon (2014c)). This growth reaches its nadir by the First World War, but is followed by a catch-up process driven by the inventions of the second industrial revolution (Gordon (2015)). These include the internal combustion engine and telecommunications, as well as a huge increase in educational standards (Bergeaud et al. (2014)). This productivity growth is then interrupted by the Second World War, although its impact was much stronger in Europe than in the USA. After the Second World War, the “Golden Age” (Maddison (2006)) began, characterized by a very strong growth in technical progress. Since the 1970s, the TFP growth rate has slowed immensely and is often referred to as the time of the global productivity slowdown (Eichengreen et al. (2015)). However, in the mid-1990s the TFP growth rate rose again, particularly in the USA—in large part due to information and communication technology (ICT) (Gordon (1999)). In Europe, this trend is barely visible, which is due in part to the significantly higher telecommunications and Internet costs (Felderer (2001)) and its negative influence on the implementation of ICT. Since the beginning of 2000, growth in TFP has flattened even further, reaching lower growth rates than in 1890, which raises the question of whether this decline in technological progress has a structural background. Given that technology and innovation are the driving forces for economic growth (Freeman and Soete (1997)), this stagnation in TFP growth is worrying and has led many economists to question whether the European economy, especially as a result of flattening productivity, is heading towards secular stagnation.
Looking at the European economic development and its main driver, TFP growth, a long-term slowdown of the growth trend can be observed. For this reason, discussions focus on whether the 7 phase of slow growth will soon come to an end, or whether this will become a new normal state (Storm (2017)) set to remain in place for an uncertain period of time, in other words entering a new era of so-called secular stagnation.
The term secular stagnation was first used by Alvin Hansen (1938). He argued that the US investment opportunities in the 1930s had been significantly reduced by closing borders and cutting off immigration, as well as a general slowdown in population growth. In a speech to the president, Hansen (1939) emphasized that without significant measures, the global economy could face the beginning of a new era of persistent unemployment and economic stagnation.
It is possible to apply these points to the present day, where similar scenarios can be observed in Japan, the USA and also in the EU-15 countries. Particularly due to the continuing phase of low economic growth and ever falling interest rates, the topic was first applied to the present time by economist Summers (2015b) at the IMF Research Conference 2013. Specifically, Summers observes that economic growth is not being stimulated despite the lowering of the key interest rate to the zero lower bound (ZLB), stating that, “The nature of macroeconomics has changed dramatically in the last seven years. Now, instead of being concerned with minor adjustments to stabilize about a given trend, concern is focused on avoiding secular stagnation” (Summers (2014b), p.1). Indeed, Summers holds that the crisis is far from over, and that nominal zero interest rates inhibit economic activity (Summers (2015b)). Thus the consideration is by no means inconsiderable that a new phase of secular stagnation could already be underway since the 1970s. Naturally, these drastic forecasts for the future initiated a discussion among renowned economists. While many share Summer’s opinion, some maintain that the economy will recover in the long term. In this context Eichengreen stated, that secular stagnation ”means different things to different people” (Eichengreen (2014), p.4).
Summers defines secular stagnation quite generally as a long-term and permanent weakening of overall economic demand. Eichengreen, on the other hand, gives a more precise definition of the term by naming concrete factors. For him, secular stagnation means: “A downward tendency of the real interest rate, reflecting an excess of desired saving over desired investment, and resulting in a persistent output gap and/or slow rate of economic growth” (Eichengreen (2015), p.1).
Although there is in fact no uniform definition, it is still thought important to summarize some of the common definitions here. Accordingly, an economy is exposed to secular stagnation if it records low or no economic growth despite high per capita income. Slower growth over several decades, low interest rates, low investment rates, high savings rates and declining demographic dynamics provide information about demand-side secular stagnation. Many economists cite the long-term slowdown in growth as the primary cause of demographic change. Low interest rates, low investment and high savings rates are seen as endogenous results of the primary cause. On the other hand, secular stagnation can be defined as a supply-side phenomenon, whereby, due to structural factors, long-term productivity stagnates and thus has a negative impact on economic growth.
The manifestations of the above-mentioned dynamics of secular stagnation are clearly recognizable within the EU-15 countries. This is why the present study seeks to determine what factors are responsible for economic stagnation and whether it is secular stagnation.
As described in the previous section, definitions and opinions on secular stagnation differ. The common literature is divided into supply-side and demand-side explanatory approaches.
The demand-side is particularly driven by the ageing population. If a population ages and the retirement age remains the same, the amount of savings required to keep consumption constant throughout retirement will increase (Weizsaecker (2015)). Of course, an increased need to save leads to a demand deficit (Summers (2015a)), which delays the velocity of income. In order to bring the savings of a certain region back into balance with investments, conventional monetary policy provides for a reduction in interest rates. However, this is no longer possible once the ZLB is reached since, otherwise, all private households would convert their deposits into cash.
Turning to supply-side effects and the causes of secular stagnation, Eggertsson et al. (2019) conclude that the decline in the TFP growth rate is the key factor for this enduring development.
Abbildung in dieser Leseprobe nicht enthalten
Figure 3: Downward Revision in Potential GDP, Eurozone Summers (2014a)
Caption.4 shows the down-ward revisions of potential GDP for the euro zone. It is obvious that the estimates of potential GDP are ad-justed downwards from year to year and in 2014 Gor-don saw these as being significantly lower than in 2008. Low potential growth means that growth is not within a business cycle due to a recession, but that higher growth is not expected in the long run, which clearly indicates secular stagnation. A “technology pessimist”, Gordon (2012) attributes the sharp rise in total factor productivity over two decades to three industrial revolutions.
- The first Industrial Revolution (1750-1830) is characterized by the invention of the steam engine and the associated reduction in transport costs and the substitution of wood by steel.
- The second Industrial Revolution (1870-1900) includes the discovery of electricity, the use of petrol motors, industrial chemicals, running water and telecommunications.
- The third Industrial Revolution, that is the current ICT revolution, is characterized by the Internet, wireless telecommunications and artificial intelligence (Gordon (2012)).
As illustrated above, an industrial revolution with new inventions and strong productivity growth is always followed by a period of slower TFP growth rates. However, it is important to consider the delay theory of David (1990), which argues that every new technology takes time before its effects become visible in economic figures. In response, Gordon (2014a) argues in 2014 that the third industrial revolution, which peaked in 2000 when the “dot.com” bubble burst, has a much smaller impact than the second. Thus, he predicted much lower real GDP growth of between 1.4% and 1.6% for the US for the years after 2014.
The causes of slow growth can be seen as four headwinds (Gordon (2012)), these being: first, demographics, namely the growth of the pension population relative to the working population; second, stagnation in the quantity and quality of human capital; third, rising inequality and uncertainty (Baldi and Harms (2017)), which ensures that less is invested to provide for the future and fourth: private and public debt, which means too much attention is devoted to debt relief (Gordon (2014c)). Gordon and Sayed (2019) also adds globalization, energy and environment to these four headwinds. Using a simple subtraction exercise, he forecasts that these headwinds will reduce the real GDP per capita growth of 2.0% per year that predominated between 1891 and 2007, to 0.9% per year in the period 2007-2032 (Gordon (2014c)).
It is therefore important to question whether growth since the 1970s has been too slow, or whether growth in the period between the 1930s and 1970s was simply too high. Similar growth rates can be observed both before and after this period. Clearly, the reconstruction that followed the First World War (Gordon (2014c)) and the aftermath of the Second World War resulted from policies that promoted growth at all costs. Since the 1970s, such a policy has changed significantly, with economic growth no longer being the main priority.
An indication of a possible slowdown of this ICT revolution could be the low productivity growth, combined with rising spending on research and development (Eurostat (2019)). This means that companies can only exploit technological potentials at increasing expense, which may lead to a decline in innovation and investment (Gordon (2015)).
While it cannot be denied that these headwinds have a strong structural impact on European economies and their growth, it is difficult to predict how current technologies will develop. However, despite the number of technology “pessimists”, there is also a large camp of “optimists”, who in the future posit a central role for the use of personal data in conjunction with artificial intelligence. These optimists suspect that TFP growth, and thus GDP growth, will increase again over time, allowing an escape from secular stagnation.
Of course, time will tell whether innovations in AI, robotics and the human genome will lead to another industrial revolution characterized by significant productivity gains, or whether we will never again experience the high productivity growth associated with the second industrial revolution (Gordon (2014c)). However, as has become apparent when looking at developments in TFP growth rates, as has been made clear above, historically there have always been periods of stronger and lower productivity growth. The next section examines the causes of a possible secular stagnation of the EU-15 area, identifying specific factors influencing the productivity of these states.
This section presents various factors that have historically been relevant to productivity growth in the EU-15 countries and summarizes the current literature on them. This review serves to identify the determinants of productivity growth that are required for the regression analysis in Chapter 5. Within the EU-15, it is possible to observe a very heterogeneous historical and economic development of the individual member states. The cross-country comparison of the individual countries must therefore agree on common factors.
For hundreds of years, Europe has been characterized by a great diversity of cultures and traditions living together in a relatively restricted space. Because of this very heterogeneous population density, historically Europe has very often experienced wars, conflicts and competition. Jones (2003) argues that it was precisely these wars, institutional competition and political fragmentation that were the key to Europe’s rise and increased living standards in the industrial revolution. After so many years of war and competition, Europeans finally decided to work together, to support each other economically and to protect each other in case of military threats. Thus a union of values was formed with the EU, which has created a cohesion which, due to its long history, goes far beyond the points laid down in the Maastricht Treaty (1992, Lenz and Borchardt (2010)).
As a result of the gaining of power by the workers’ societies in many parts of Europe, such as during the German revolution in 1848/49 (Beer (1971)), a social idea grew in Europe through the formation of trade unions. These were intended to protect the labor rights of the lower classes in particular. Even today, a higher social demand is still visible in Europe than, for example, in the USA and it is important to examine here whether the level of occupational health and safety, approximated by the Employment Protection Index, or Regulatory Quality has an impact on productivity in the EU-15.
Efficient implementation of employment protection is primarily concerned with the trade-off between stabilization of employment and the undesired loss of efficiency (O’Mahony (2013)). On the one hand, employment protection is highly valued by workers because it provides them with occupational and financial security, however it also generates costs for the company, resulting in a loss of flexibility and, ultimately, lower total factor productivity growth (O’Mahony (2013)). In the 1990s, Europe saw a liberalization of the labor market, characterized by the deregulation of temporary employment contracts. Pompei et al. (2011) holds this has had a negative impact on TFP growth in European countries, as it means that companies are less likely to invest in firm-specific human capital (Engellandt and Riphahn (2005)) and therefore have fewer permanent employees. Hence, it is thought useful to examine this variable and its influence on productivity.
Added to this is the regulatory quality. If the EU-15 manages to close its regulatory gaps and thus remove intra-European barriers, this will make an enormous contribution to stability and productivity (Brunet (2012)). It is therefore important to improve regulatory quality within Europe in order to increase the productivity of the entire economy (van der Marel et al. (2016)). The development of regulatory quality in Europe shows large fluctuations over time. For example, it can be seen that in the mid-1970s the differences in regulatory quality became more pronounced, then around the 2000s the figures were very good and from then on, they declined again. Given the high relevance for intra-European cooperation, it is interesting to look at the influence of regulatory quality on productivity.
Even if the production of goods and services is generally not produced by the government and TFP growth is thus mainly driven by the real economy, the government and its political construct play a very important role in the promotion of TFP. When taking another look at the beginnings of innovation at the time of the Industrial Revolution, it quickly becomes apparent that all this technical progress was only possible with secure and stable institutions and an appropriate infrastructure (Gerschenkron (1962)). Even though the law and infrastructure have developed immensely since then, the institutional framework still plays an important role in economic development. Thus, a responsible and fair government can have a positive impact on state accumulation of social infrastructure and productivity (Rio (2018)), which is why government efficiency is also be part of the empirical analysis.
Education and human capital have historically played a significant role in the development of productivity and the economy as a whole. Especially with the rise of new technologies, the relevance of education for economic development has increased dramatically, since technologies could only be adapted and developed with a certain amount of human capital. The aim was to use foreign technologies for domestic production, which they achieved through secure institutions, a certain degree of openness and a necessary level of education. (Abramovitz (1986)).
The evaluation of Education can take place on three levels, namely quantity, quality and inclusiveness (Canton et al. (2018)). Although human capital is difficult to assess, the OECD (2019) has developed a human capital measure that combines the number of years of schooling and returns to education. School years and the human capital index are analyzed to examine both the quantity and quality of education.
An increase in the level of education has a positive effect on the human capital of OECD countries (Goldin and Kutarna (2016)). Furthermore, the level of human capital in OECD countries has a positive effect on total factor productivity growth (Engelbrecht (1997)). Aghion et al. (2005) observes that the closer European countries are to the potential of the technological leader, the better they can exploit the positive potential of human capital. This means that with European convergence, human capital will become increasingly important.
The financing of European education is seen as problematic, however, with investments only becoming visible after a long delay. Due to very short legislative periods, these investments are often neglected by politicians. Although even very small increases in human capital can have a very strong impact on future economic growth through its influence on innovation and thus on technological progress (OECD (2010)).
Since the beginning of industrial production, the availability, use and cost of energy has become increasingly important. Thus, the invention of the steam engine was only possible in England through a combination of high wages, a large market and very low energy costs (Allen (2011)). Through micro-inventions, technologies become more efficient over time, use less energy and can eventually be used in other regions, which increases overall energy consumption.
Historically, Europe’s economic development has been characterized by up and down movements in energy consumption. For example, per capita energy consumption was stable or slightly declining until the Industrial Revolution, when it rose sharply in the 18th century, driven by new inventions such as the steam engine. In the second Industrial Revolution, there was a new acceleration in energy consumption driven by the use of oil and electricity, which finally stabilized in 1970. Now, the ICT revolution and the enormous importance of new data storage platforms are giving a new importance to the use of energy (Kander (2014)).
However, energy costs not only have a direct influence on revenue and profit due to the price increase caused by the higher costs, they also influence productivity. Jorgenson (1984) examines the influence of energy price volatility on capital and human capital in the US and finds that energy prices can influence TFP growth when technological progress depends on energy.
Especially because energy is so important for the European economy, the EU’s energy policy aims to provide affordable energy. However, this is not so easy as it has to be reconciled with the EU emissions trading scheme and the promotion of renewable energies (Fischer and Geden (2013)).
The past has shown that the sectoral composition of the industry has had a major influence on technological change, especially in Europe. For this reason, it is interesting to look at the development of the sectoral composition in order to draw conclusions about technological change.
The global economy can be roughly divided into three sectors: agriculture, manufacturing and services. Each sector favors different factor and human capital endowments. The degree of economic integration also plays a major role. Whether the effect of agriculture on economic growth is positive or negative depends largely on the openness of an economy (Matsuyama (1992)). However, the transition from mercantilism to free trade was only possible with advancing technical improvements and the resulting increasingly profitable trade (Acemoglu et al. (2005)). As a result of increased global trade, European grain prices, for example, were suddenly influenced by the US harvest. Consequently, the incentives to produce grain in Europe diminished, which ultimately led to a shift of labor to the manufacturing sector and thus to an increase in wages. (O’Rourke and Williamson (1999)). This shows that production in different sectors, and in different regions, has a different influence on technological growth.
This global trade phenomenon can be explained by the Heckscher-Ohlin model (Heckscher et al. (1991)). This states that, in economies that open their markets to free trade, commodity prices converge first, followed by factor prices. Thus, according to the Heckscher-Ohlin model, open economies have an incentive to specialize in the production of goods where they have a comparatively high factor endowment. Europe is considered to be more capital-intensive, whereas the USA is considered to be more land-intensive. This means that production shifts to the respective sectors. As a result, since the rise of the worldwide trade there has been a massive shift from agriculture to industry in Europe. Thus, the demand for labor in agriculture dwindled as technologies grew (de Vries (1994)), and this led to more workers leaving agriculture and moving into industry, where higher wages and productivity could be expected. This redistribution led to massive productivity gains across Europe, but these declined in the 1970s with the onset of deindustrialization (OECD (1975)). Since the 1970s there has been a strong shift from the manufacturing sector (Lanciotti (1971)) and to some extent also from agriculture (Herrendorf et al. (2013)), to the service sector. After Baumol’s cost disease (Baumol (1967)), the service sector finds it more difficult than the manufacturing industry to drive productivity forward. Consequently, the growth rates of productivity in service sectors are significantly lower than in manufacturing (Duarte and Restuccia (2010)). This is due to the fact that the services sector is generally characterized by a smaller scope for innovation and technological change. Although the service sector promotes lower productivity growth, the European services sector tends to employ a more skilled workforce than manufacturing (Messina (2004)). Whereby Samuelson (1964) finds that in economies with higher per capita income, employment in the service sector is significantly higher.
As discussed above, TFP growth is one of the key elements of economic growth, but this has been declining in recent years. This is particularly due to productivity growth in the services sector, which in Europe is generally below productivity growth in manufacturing industry (Uppenberg and Strauss (2010)). For the reasons mentioned above, it is therefore highly relevant to take a closer look at the distribution of the economy as a whole across the individual sectors. This should allow for conclusions on the development of productivity to be drawn.
The aim of the introduction of the European internal market was ultimately to promote the free movement of goods, services, capital and people, as well as to boost productivity in the member states (Notaro (2002)). If the EU succeeds in reducing the barriers and obstacles between member states, considerable economies of scale can be generated, along with lower costs and prices as well as welfare gains. This is why European integration plays a central role in the productivity of individual economies. Following the efficiency criterion, integration describes a number of economies that use their production patterns efficiently within the EU (Baldwin and Wyplosz (2004)). This state can be achieved if the factors of production (capital, labor) can and do move freely within the EU.
One way to measure capital mobility between economies is the Feldstein-Horioka puzzle (Feldstein and Horioka (1980)). The idea behind this puzzle is that if capital can move freely without obstacles, investors will always invest where the returns are highest. For this reason, the Feldstein-Horioka puzzle looks at the cross-country correlations of savings and investment rates. If the correlations fall, this indicates increasing capital mobility. Theory suggests that capital should flow from relatively rich to relatively poor countries and that this is effective because a higher marginal product of capital exists in those countries (Chen (1997)). Through the efficient use of factor endowments, this reallocation ultimately accelerates the convergence of per capita income and leads to integration (Abiad et al. (2014)). Finally, Edwards (2001) notes that economies with a certain degree of economic development achieve more productive results when they have a more open capital account.
A further aim of European integration is to promote the mobility of workers within the EU (Regulation 1612/68, Catana (2012)) so that labor potential can be used as efficiently as possible. Removing barriers to migration has naturally increased the migration of skilled workers between European countries, which has not only had a positive effect on migrants and the western “receiving countries” but has also on the workers remaining in eastern Europe through the potential increase in wages in the poor ”sender-region”. At the same time, however, this can also lead to the so-called ”brain drain” problem whereby many of the most talented people leave eastern Europe for western Europe, with negative consequences for the former’s economies. Nevertheless, it can be seen that, compared with the USA, shocks on the European labor market are hardly absorbed by migration and are thus mainly reflected in the employment rate (Decressin and Fatas (1995)). In order to measure labor mobility, it is important to take a look at wages and their convergence within the EU. If wages converge over time, this may be an indication that workers are becoming more mobile between countries.
The EU thus offers enormous potential, which must be used efficiently by setting the course for efficient trade and productive cooperation. If the EU succeeds in optimizing the efficient allocation of resources throughout Europe and removing the barriers to greater labor market flexibility, this would increase productivity enormously (Buch (1999)).
The basic idea of trade is that each individual is no longer responsible for the production of all essential ingredients, but rather specializes in certain manufacturing processes. Through the learning effect and the economies of scale, production can eventually become more efficient. In exchange for the goods produced, profits can be generated (Smith (1776)). Of course, when we talk about trade, there are always winners and losers. Thus, when a region specializes in manufacturing a product, it increases its competitive advantage and may even drive other firms or even economies out of the market. In addition, increasing trade increases competition, which means that the profits of individual firms become increasingly smaller as prices near marginal costs. In his paper, Eichengreen et al. (2015) emphasizes that TFP growth has been particularly positive in European countries where tradable production has experienced less of a decline. Piton (2017) confirms this thesis and explains that sectors in Europe where there is less trade have lower TFP growth and therefore a shift to these sectors has a negative impact on overall economic productivity. This means that the positive effects of trade in Europe outweigh the negative effects of a shift to these sectors and that an increase in the openness of the economies of the EU-15 countries has a positive impact on productivity. As already considered in the Solow model (Section 2.3), growth of TFP often results from capital accumulation. Underinvestment has a negative impact on growth because the capital stock becomes obsolete over time. Consequently, it can no longer be produced with the latest technology and productivity declines (Havik et al. (2008)). The accumulation of human and physical capital particularly drives productivity growth in European countries, which is why it also plays a significant role in the stagnation of TFP and, ultimately, economic growth. For this reason, it is useful to look at the share of investment in GDP and to examine how this affects the productivity of the individual economies of the EU-15 countries. However, especially for poorer economies, domestic capital accumulation is a very lengthy and costly way of driving the economy. Another way to promote investment or TFP growth can be through foreign direct investment (FDI) directly in the real economy. This promotes the capital intensity of the recipient region, with an increase in the production of capital-intensive goods (Baldwin and Wyplosz (2004)). The interest of foreign investors in European companies has grown steadily in recent decades due to increasing global networking and the growing importance of world trade. In the European context, the acquisition of company-specific knowledge-based assets, which is an important component of the dissemination of innovation and technology, plays a major role in this context (Barrell and Pain (2012)). Thus, in order to be able to investigate the impact on TFP growth within the EU-15 more closely, FDI is included in the analysis. Of course, many different factors influence TFP growth and many of these have existed for a long time and reveal themselves as structural influences. Indeed, other factors have become more relevant, especially due to the emergence of the EU and the importance of European integration. The long-term development of productivity is also very relevant in the debate on the continued existence of the EU. All that follows in the rest of the thesis, therefore, the development of productivity will be examined in more detail and finally the influence of the relevant factors on TFP will be analyzed.
The analysis in this paper goes beyond the paper of Eichengreen et al. (2015) and looks at all EU- 15 economies, i.e. only economies that are declared by the World Bank as high-income countries2 (Worldbank (2020)). Since a completely different population is now considered, on the one hand the factors influencing the TFP fluctuations must be adjusted to the region and on the other hand, due to the smaller fluctuations, an attempt must be made to approach shorter time periods (Eichengreen et al. (2015)) in order to achieve significant results. Therefore, an optimal balance of sufficient observation points and therefore information, as well as a trend that is nevertheless relatively free of disturbances is attempted. Due to the more homogeneous countries in relation to their income, it is easier to deal with the typical influencing factors for this region in the following regression analysis. Nevertheless, it must be stressed that not all country-typical characteristics can be taken into account, as it is necessary to find the highest common denominator of the EU-15 countries. Therefore, the variables considered in the analysis are presented below on the basis of the literary summary.
Most of the variables (Table 5List of variablestable.caption.38) were taken from the Penn World Table (Feenstra et al. (2015)), which is a compilation of national accounts data over a long period (1950-2017) and many economies (currently 182). All variables below are taken from PWT9.1 and are considered over the period 1950-2017 unless otherwise stated.
All EU-15 countries (including the UK) are considered, some of which have different currencies. For this reason, the 2011 real GDP (cGDPo) on the expenditure side is measured in USD. Thus, GDP is shown in a hypothetical currency—the International Dollar—which has the same purchasing power as the USD in 2011. The Purchasing Power Parity (PPP) is the ratio of a country’s GDP in national currency to its GDP in international prices, allowing a comparison of countries over time. The expenditure approach was chosen because it assesses the sum of all goods and services purchased in an economy. In order to ensure comparability between countries, it is divided by population, resulting in real GDP per capita. In addition, GDP is considered in levels and squared to take account of possible non-linearity.
The variable rTFPna describes real total factor productivity at constant national prices (2011=1), where TFP is the part of economic output that can not be explained by the quantity of inputs used in production (TFP). This variable thus allows for comparability over time and across regions, with measurement points starting from 1954 onwards.
The next historically important variable for the development of TFP in the EU is the Human Capital Index (hc). This is based, on the one hand, on the number of years of schooling and, on the return to education on the other—which means that it contains both a quantitative and a qualitative component. By considering the return to education, besides the school education component, this variable takes information on board that goes beyond the acquisition of school knowledge. This means that company-specific human capital and knowledge from continuing professional training measures are also accounted for.
In addition, the economic openness (openk) is considered, which indicates the openness at current prices (%). This is measured by the ratio of exports plus imports to GDP and thus indicates total trade as a percentage of GDP (openk).
The investment ratio (csh i) is the last variable considered from the Penn World Table data set. It indicates the share of real GDP that can be represented by investments (capital formation) at current purchase price parities (PPP). By using production-based real GDP, a comparison between countries over time becomes possible (csh i).
The next two educationally relevant variables are the average total number of years of schooling and the average number of years of tertiary schooling from the long-term time series of Lee and Lee (2016). It should be noted that the data are only available in five-year steps. Due to the very structural development of school and university years, especially in EU-15 countries, an approximate linear growth between the data points can be assumed. For this reason, the data points between the observation points were completed using linear growth. By choosing three education-related variables, the analysis in the present study attempts to break down human capital into its individual components and, where appropriate, to draw conclusions about the strength of its influence on productivity. The variable EPL (Employment Protection Legislation) describes the strength of employment protection against dismissal by means of eight data elements. High values therefore reflect high employment protection and low values reflect low protection. The dataset was provided by the OECD (2013) and covers the period 1985 to 2013, including all EU-15 economies except Luxembourg. To pin down government effectiveness and the regulatory quallity, GovE3 and RegQ4 are taken from the World Bank Governance Indicators (Kaufmann et al. (2010)). The dataset includes data from 1996 onwards and is considered up to 2017.
In addition, the measured value ElePC describes the energy consumption per capita in kg oil equivalent per capita, whereby energy consumption refers to the use of primary energy before conversion into other fuels for final consumption. The data set is taken from the IEA database (OECD and IEA (2014) and contains data from 1960 to 2015. FDI is the next variable considered. This measure is the net inflow of FDI for the acquisition of a permanent business interest5 that operates in an economy outside that of the investor (Worldbank (2018b)). It is expressed as a percentage of GDP and covers the period from 1970 to 2017 (Luxembourg from 2002).
The variables gross investment (gCapFor)6 and gross saving (gDomSav)7 are taken from World Bank national accounts data and OECD national accounts (Worldbank and OECD (2018) data files.
The various sectors of the economy contribute differently to technological change. For this reason, the analysis will take a closer look at the shares of the economy in manufacturing (industry), services (services) and agriculture (agriculture). They are also all derived from World Bank national accounts data and OECD national accounts data files (Worldbank and OECD (2018)) and provide data for the period 1995 to 2017. The manufacturing variable corresponds to ISIC divisions 10-45 and can therefore be defined as manufacturing. The services sector includes wholesale and retail trade, transport, government, financial, professional and personal services such as education, healthcare and real estate services (ISIC divisions 50-99). Agriculture is dominated by forestry, hunting and fishing, and crop growing and animal husbandry (ISIC divisions 1-5).
To examine labor mobility in Europe and its influence on technological change, the standard deviations of private hourly wages in manufacturing were used to determine wage adjustments. It is particularly interesting to look at wages in manufacturing, as these are, on the one hand, subject to technological change and, on the other, require a physical presence of workers in the enterprise compared to the service sector. Due to the different collection methods, no data are available for Greece and Portugal. The data set is taken from the OECD Labor Earnings Database (OECD (2020) and provides data for the remaining 13 EU-15 countries from 1981 to 2017.
The final variable considered is the share of all employees in relation to the total population (employ s). The data are taken from the World Bank Development Indicators and provide data (Worldbank (2018a) from 1990 to 2017.
Thus, the data origin for all variables could be documented and information about the definition of the data could be obtained. With all this information, the question now is how these data sets can best be analyzed so that significant results relevant to the question can be obtained.
In the regression analysis of the TFP figures for the EU-15 countries, the aim is to identify sustained slumps or upswings and to determine their causes. If one were simply to use the annual measurements, the data would show very large fluctuations, whereby trend-related interpretations could become insignificant. Thus, economic research (Eichengreen et al. (2012)) often uses annual averages, defining periods over several years in which an average growth rate is determined, which can then be compared with the other periods.
As already described, the next challenge is about finding the optimal balance between enough observation points and therefore information and a trend that is nevertheless relatively free of disturbances. Eichengreen et al. (2015) shows that in his case, the loosest criterion provided the most logical results. However, since this data set only considers “rich” countries (Figure 5Histogram per income (3Y, 1.5% criterium)figure.caption.6), whose fluctuations are significantly lower than the Eichengreen et al. (2015) data set due to financial security, for the purposes of the present study it was decided to develop even more relaxed criteria starting from the loosest criterion Eichengreen et al. (2015) so as to find the optimal criterion for the EU-15 countries.
1 The UK and Sweden are also included in the data. Derivation: Eurozone good approximation for EU- 15,(Appendix 10TFP EURO + GB & NOR (1890 - 2012) Bergeaud et al. (2014)figure.caption.11)
2 see methodology in more detail (Figure 5Histogram per income (3Y, 1.5% criterium)figure.caption.6)
3 GovE, reflects the perception of the quality of public services, their independence from political pressure, the quality of the formulation and implementation of policies as well as the confidence in the implementation of these political promises (Worldbank and OECD (2019)).
4 RegQ, relies more on a government’s ability to drive private sector development and thus reflects public perception of its ability to formulate and implement robust policies and regulations(Worldbank and OECD (2019)).
5 10 % or more of the voting shares
6 gCapFor, represents expenditure on investments in the fixed assets of the economy plus net changes in the level of inventories in % of GDP (Worldbank and OECD (2019))
7 gDomSav can be defined as GDP minus consumer spending in % of GDP (Worldbank and OECD (2019))
Der GRIN Verlag hat sich seit 1998 auf die Veröffentlichung akademischer eBooks und Bücher spezialisiert. Der GRIN Verlag steht damit als erstes Unternehmen für User Generated Quality Content. Die Verlagsseiten GRIN.com, Hausarbeiten.de und Diplomarbeiten24 bieten für Hochschullehrer, Absolventen und Studenten die ideale Plattform, wissenschaftliche Texte wie Hausarbeiten, Referate, Bachelorarbeiten, Masterarbeiten, Diplomarbeiten, Dissertationen und wissenschaftliche Aufsätze einem breiten Publikum zu präsentieren.
Kostenfreie Veröffentlichung: Hausarbeit, Bachelorarbeit, Diplomarbeit, Dissertation, Masterarbeit, Interpretation oder Referat jetzt veröffentlichen!