[여기에입력]
Paper to be presented at Development Economics Seminar at USC on January 13, 2015
Labor Productivity and a Test of Kaldor-Verdoorn Law in East Asia
Hak K. Pyo
Professor of Economics Emeritus
Seoul National University
Seoul, 151-742, Korea
Abstract
While mainstream explanations point out technological changes as the main determinant of income inequality and admit that globalization has had negative effects on the wage share in advanced economies, the proponents of wage-led growth argue that the declining share of wage income and the increasing share of capital income has been the major cause of stagnation in the global economy. In advancing their proposition, proponents of wage-led policy rely on the Kaldor-Verdoorn law in Kaldor (1957) which claims that there is a positive relation between the growth rates of GDP and the growth rate of labor productivity. It suggests that demand-led growth will have an impact on the supply components of growth. McCombie (2002;106) reports that a one percentage point addition to the growth rate of output will generate a 0.3 to 0.6 percentage point increase (the Verdoorn coefficient) in the growth rate of labor productivity. It is consistent with the estimates by Storm and Naastepad (2008) and Hein and Tarassow (2010) which find a similar range around 0.30 for European countries (1960-2007 data) but a lower range for the UK and the US between 0.1 and 0.25. The purpose of this paper is to investigate the causality between aggregate demand and labor productivity and examine the implication of the Kaldor-Verdoorn law in East Asia using three East Asian Countries’ (China, Japan and Korea) data set. We find that the Kaldor-Verdoorn Law is partly accepted from the data set of China and Korea and also the causality of real wages on labor productivity growth from the dataset of China and Korea not from the data set of Japan.
*An earlier version of this paper was presented at the third Asia-KLEMS Meetings
On August 13-14, 2015 in Taipei, the Republic of China.
1. Introduction
There have been two stylized facts in the global economy after the global financial crisis in 2007-2008. The first one is the decline of labor income share in both advanced and emerging nations and the second one is the continuation of slow recovery and stagnation despite massive monetary and fiscal expansion measures by the United States, Euro area and Japan. Onaran and Galanis (2012) reports that there is a clear secular decline in the wage share in both developed and emerging developing countries (Turkey, Mexico, South Korea (henceforth Korea), Argentina, China, India and South Africa) from late 1970s or early 1980s onwards. In the Euro area and in Japan, the decline in the unadjusted wage share (not adjusted by the wage income by self-employed) exceeded 15 percent points and 20 points respectively in the index value. The fall is lower, but still strong, in the US and UK with a decline of 8.9 percent and 11.1 percent respectively. They note that a correction of the wage share by excluding the high managerial wages, which have increased very sharply in these developed countries would have provided a more realistic picture about the loss in labor’s income share. They also note that in the developing world, Turkey and Mexico have experienced the strongest decline in the wage share (- 31.8 % and - 37.9 % respectively) followed by South Africa (- 17 %), Argentina (- 12 %), India (17.6 % from the base year of 1980. In China the improvement in the wage share in the 1980s was reversed in 1990 culminating in a cumulative decline of 12.8 percent in the index value. In Korea, the increase in the wage share from mid-1980s onwards was reversed by the crisis in 1997.Pyo (2015) estimates the unadjusted wage share (year) in Korea as: 0.45 (1970), 0.55 (1997) and 0.5 (2012) and Cho, Kim and Schreyer (2014) estimates the adjusted wage share (adjusted by assuming the wage level of self-employed and unpaid workers as 80 percent of wage earners’ average wage rate) as: 0.82 (1970), 0.78 (1997) and 0.67 (2012). Onaran and Galanis (2012) reports the mean values of the wage share from the sample as: Euro area-12 (0.693), UK (0.727), US (0.684), Japan (0.721), China (0.581) and Korea (0.845: adjusted wage share assuming the average wage level of self-employed would be equal to 100 percent of average wage level of workers in the aggregate economy).
The empirical finding of the declining wage share in both developed and major developing economies is equivalent to and consistent with the empirical finding of the increasing capital share as reported in Piketty (2014). Piketty (2014: PP. 211-212) notes that capital income share was around 15-25 percent of national income in rich countries in 1970 but reached 25-30 percent level in 2000-2010. He further notes that the upward trend in capital’s share of income is consistent not only with an elasticity of substitution of capital for labor greater than one but also with an increase in capital’s bargaining power vis-à-vis labor over the past few decades, which have seen increased mobility of capital and heightened competition between states eager to attract investments. He argues that over a very long period of time, the elasticity of substitution between capital and labor seems to have been greater than one: an increase in the capital/income ratio seems to have led to a slight increase in capital’s share of national income, and vice versa. While he does not estimate capital’s share of income in developing countries, he notes that the upper centile’s share of national income in poor and emerging economies is roughly the same as in the rich economies: the top centile’s share of national income moved from around 20 percent in four countries – India, South Africa, Indonesia and Argentina - then fell to 6 -12 percent level during 1950-1980 but rebounded in the 1980s to reach 15 per cent of national income.
While there is secular decline in the labor share of national income and correspondingly increase in the capital share, two stylized facts seem to have emerged: polarization of personal income distribution and weaker growth performance in most countries except China and India. The polarization of personal income and its rising inequality has been observed by Atkinson et. al. (2009) and Piketty (2014). Piketty (2014) relies on the historical trend of capital/income ratio (ß) and has observed its U-shaped curve implying that the income inequality in recent years is getting worse going back toward the level observed at the end of 19th century and early 20th century in both developed countries and emerging developing countries. He estimates the capital/income ratio in Italy and Japan reaching above 6, France around 5.8 and US and Germany slightly less than 5. He also observes a similar U-shaped pattern from the historical data of India, South Africa, Indonesia, Argentina, China and Colombia and concludes that the share of income held by the richest 10 percent income group in these emerging economies is as large as the level observed in developed countries. But he notes that the income inequality measured by household survey in developing economies tends to underestimate the degree of inequality. For example, in case of Colombia and Argentina, the richest 10 percent income group is estimated to hold more than 20 percent of national income according to tax records while it is estimated to hold 5 percent of national income according to household survey. Kim and Kim (2013) estimates the income concentration ratio by top 1 percent income group from tax records in Korea and observes that the ratio increased from 7 percent in 1996 to 12 percent in 2010. They also observe that the top 1 percent’s income share in the US, Japan and Korea in mid-1970s was similar at the level of around 5 percent but it reached in the US to the level of 12 percent in 2008 and in Japan to the level of 5.5 percent.
While the income inequality is rising, the growth performance in both developed economies and most of developing economies has been stagnant after the global financial crisis in 2007. Onaran and Galanis (2012) reports the average growth rates of GDP between two sub-periods (1970-79 and 2000 - 07) in Euro area-12 (3.78 % and 2.13 %), US (3.32 % and 2.61 %), Japan (5.21 % and 1.73 %), Korea (10.27 % and 5.20 %), China (6.11 % and 10.51 %), and India (2.68 % and 7.26 %).
With these stylized facts observed from the recent economic trend in the global economy, there has been debate on wage-led growth vs. profit-led growth. The wage-led growth is often called as income-led growth or demand-led growth while profit-led growth is also called as supply-led growth. The purpose of this paper is to reexamine theoretical background of two contesting economic policy platforms and derive implications for East Asian countries of China, Japan and Korea who have a diverse historical background and economic regimes.
The paper is organized as follows. Section 2 reviews theoretical backgrounds and empirical studies. In section 3, we examine the result of the Granger causality tests from the three countries’ dataset. The last section concludes the paper.
2. Wage-led Growth and Profit-led Growth: Two Contesting Economic Regimes
2.1 Theoretical Background
(1) Kaldor-Verdoorn Law and Wage-led Growth
According to Bleaney (1976) and Lavoie and Stockhammer (2012), a wage-led economic strategy has a long history and was advocated in the form of ‘underconsumption theory’ in the 19 th century by classical economists including Malthus, Sismondi and Hobson. It was Keynes (1936) who endorsed it in his theory of effective demand arguing that excessive savings rate relative to deficient investment rates, were at the core of depressed economies. Under-consumption theories have two groups of followers. One group is Marx (1979) and subsequent Marxist theorists such as Baran and Sweezy (1966) who related under-consumption theories to the principle of infinite accumulation and the problems of the realization of the profit and the other group is Kalecki (1971) and post-Kaleckian authors such as Steindl (1952) and Bhaduri (!986) who have brought together the theory of effective demand and the problem of realization of profit..
As reviewed and revisited recently by Piketty (2014:227-230), Marx (1979) predicted that capitalists would accumulate ever increasing quantities of capital, which ultimately would lead inexorably to a falling rate of profit (i.e., return on capital) and eventually to their own downfall. The under-consumption theory is related to the problem of the realization of profit. It was Kaleckian and post-Keynesian authors such as Rawthorn (1981), Taylor (1983) and Dutt (1987) who have revisited the relationship between the under-consumption theories and the problem of realization of profit. In the context of the present paper, we should pay attention to Taylor (1988) who showed early on that when emerging countries had enough capacity to adjust, a wage-led growth strategy is preferred to profit-led growth strategy.
Kaldor (1957) had defined technical progress function and had initiated a debate on what he called the Verdoorn Law in Kaldor (1967) which claims that there is a positive causal relationship going from the demand-led growth of GDP to the growth rate of labor productivity which is the supply components of growth. It was later called the Kaldor-Verdoorn Law by Kaldorians such as Boyer (1988), Setterfield and Cornwall (2002) and Naastepad and Storm (2010) who argued for a long time that supply-side growth is endogenous which predates neoclassical theory of endogenous growth. The Kaldor-Verdoorn Law has now a series of empirical evidence such as McCombie and Thirwall (1994), McCombie (2002), Leon-Ledesma and Thirwall (2002) and Dray and Thirwall (2011).
Mainstream neoclassical economists such as Hicks (1932:124-5) and Samuelson (1965:354) have argued that rising real wages would induce firms to invest in more capital-intensive methods which would lead to higher labor productivity. Kaldor (1961) called the constancy of the wage share in total national income a stylized fact of economic growth implying that there a long-run relationship between real wages and labor productivity. Dumenil and Levy (1995) introduced a stochastic model of induced technical change where the selection of new technologies is based on the profitability criterion, with only techniques yielding a profit rate higher than the present being adopted. The model suggests that if the labor share is larger than the capital share, then the savings I labor will tend to be larger than in capital and that a rise in real wages also increases the labor share and the probability of the selected new technology being labor-saving and capital-using. It also suggests that real wages affect the trajectory of technical change through the profit rate such that an increase in real wages reduces profitability, driving profit seeking capitalists to implement labor-saving technologies in order to reduce labor costs. Marquetti (2004) has carried out Granger (1988) causality tests on the causal relationship between real wages and labor productivity for the United States over the period 1869-1999 and supported a unidirectional Granger causality from real wages to labor productivity. Piketty (2014:220-221) argues that over a very long period of time Kaldor’s constancy of the wage share has been rejected by the rising capital’s share of national income which is consistent with an elasticity of substitution between labor and capital greater than one confirming the inadequacy of the Cobb-Douglas model for studying evolutions over the very long run.
In order to highlight the effects of an increase in the wage share on profitability and aggregate demand, we may revisit the post-Kaleckian model as illustrated in Figure 1 following Lovoie and Stockhammer (2012:9-12). We start with a closed economy without government of which short-run equilibrium level of GDP is established by the intersection of the saving and investment functions. Following standard Keynesian and Kaleckian models, we assume that saving (S) and investment (I) are positive functions of income (Y). We further assume both investment multiplier effect and acceleration effect which is reflected in the steeper savings function than the investment slope.