پویایی بازار کار و توزیع درآمد در ارتباط با سرعت اشباع تقاضا

The purpose of this paper is to provide a coherent framework to explain the unusual phenomena of employment, real wage, and profit share observed in industrialised economies since the 1980s, in relation to the speed of demand saturation. To this end, we use a multisectoral model including capital goods and the stochastic emergence of new products. We show that faster demand saturation tends to accelerate the growth of employment but decelerate the growth of the real wage. Furthermore, we show that faster demand saturation tends to increase the profit share and the share eventually converges irrespective of the difference in the speed of demand saturation. Finally, it is argued that our model has a critical implication for the non-accelerating inflation rate of unemployment (NAIRU).

Since the 1980s, income distribution in industrialised economies has been a research subject given great attention. This is because unusual phenomena were observed: (1) the stagnant growth of real wage despite the recovery of the growth of labour productivity and employment, (2) as related to the above, price level stability despite strong economic growth, and (3) the tendency of profits and profit share to increase in comparison with the past trends. Such phenomena were typically observed in the U.S., whose growth process was unprecedented enough to be called a ‘new economy’ in the 1990s. Let us review the process briefly. The average growth rate of the real wage in 1947–1973 was 4.4%, while the average growth rate of labour productivity during the same period was 2.8%. The average unemployment and inflation rates during the period were 4.7% and 3.5%, respectively. After 1973, the U.S. economy stagnated, and then it subsequently recovered in the early 1990s; the average growth rate of labour productivity was 2.1% in 1990–2000 and 2.5% in 2000–2007. The average unemployment rates in these periods were 5.6% and 5%, respectively. The average inflation rates were 3% and 2.8%, respectively. However, the average growth rates of the real wage were 3.4% and 2.3%, respectively. From the above review, it is obvious that the growth rate of the real wage alone was far below the level of the period 1947–1973, although the average growth rates of labour productivity, inflation, and employment improved almost to the levels of that period.2 The real wage did increase in 1990–2007, but the linkage between the growth rate of labour productivity and that of the real wage became weaker. Furthermore, increases in profits and the profit share of national income were observed in many industrialised economies. The U.S. average of the profit share of the period 1980–1990 was 8.6%, and the averages for the periods 1990–2000 and 2000–2008 were 10% and 10.8%, respectively.3 We can observe the recent tendency of the share to increase. The above phenomena have been already scrutinised by many economists and policy-makers. With respect to the relation among unemployment, real wage, and stable price level, the most influential argument is based on wage aspirations built on the concept of a time-varying NAIRU.4 According to this concept, the long-run unemployment rate is independent of the level of effective demand, and the growth of labour productivity has no long-run effect on the level of the NAIRU (Friedman, 1968). This is because, in the long run, the real wage is adjusted in accordance with the growth of labour productivity. Although labour productivity increased in the 1990s, workers had grown accustomed to slow wage growth in the 1980s. Such a mismatch in the growth of labour productivity and wage aspirations caused a favourable shift in the Phillips curve. Therefore, low inflation and low unemployment were consistent, decelerating the growth rate of the real wage. The IMF (2006, chap. 4) closely examines the recent change in corporate profits of industrialised economies. It has been indicated that the increase in corporate profits was so strong that the corporate sectors of industrialised economies ran financial surpluses. Although such a strong increase includes the profits generated in the financial corporate sector, the profits generated in the nonfinancial corporate sector also increased enormously.5 According to the IMF, the increase in nonfinancial corporate profits is caused by low tax and interest rates, the decline in nominal capital spending due to the declining relative prices of capital goods, and the strong desire of the nonfinancial corporate sector to hold liquidity due to a more uncertain business environment. The above arguments concerning these unusual phenomena are logically reasonable, but they lack a formal demand-side analysis. It seems important to us to consider the demand-side factors because we can observe the tendency of the saturation speed of demand for recently emerged products to increase. Many studies observe such a tendency; it is particularly strong in products related to information technology (IT). For example, Stefik and Stefik (2006, p. 204) presented a figure indicating that the diffusion of products which fostered the economic growth of industrialised countries from WWII to the early 1970s (e.g. electricity, telephones, and automobiles) took a longer time than did the diffusion of products which fostered the economic growth of the 1990s and 2000s (e.g. PCs, cell phones, and the Internet); electricity, telephones, and automobiles required more than 40 years for the ownership level to reach 30%, while PCs, cell phones, and the Internet required less than 20 years for the ownership level to reach 30%. Van den Bulte (2000) analysed the diffusion of various consumer durables using U.S. data from 1923 to 1996 and revealed the tendency of later-emerged products to diffuse rapidly. Furthermore, some empirical studies reveal that the recent tendency is increasingly accelerated, especially in IT-related sectors (e.g. Dewan et al., 2010 and Stefik and Stefik, 2006). Why does the speed of demand saturation become faster? First, it has been indicated that the complementarities effect becomes stronger as products emerge later. The effect works for several reasons. The first is related to the infrastructure: Stefik and Stefik (2006, p. 204) pointed out that the later-emerged products can reuse the infrastructures pioneered by older ones. Similarly, Van den Bulte (2000) argued that demand for products requiring a large investment in complementary infrastructures is prone to be saturated rapidly. These studies imply that newer products are diffused through ‘multi-innovation interrelationships’; new products do not diffuse in a vacuum, and often there are interactions across overlapping innovations that affect the diffusion paths of the individual innovations (Dewan et al., 2010). This holds especially true for IT-related products. In fact, there are many empirical studies on the complementarities effect in IT-related products. For example, Dewan et al. (2010) investigated the complementarities effect between the PC and the Internet. Bucklin and Sengupta (1993) investigated the complementary effect between scanners and universal product code (UPC) symbols, and Bayus (1987) examined complementarities in hardware and software sales data. Second, it has been indicated that the diffusion of products for which multiple competing standards were available early is prone to be faster (Van den Bulte, 2000). The indication also holds true for IT-related products. According to Van den Bulte, the effect of standards incompatibility is not necessarily important. He used the example of the competition between VCRs and PCs to point out the possibility that consumers expect to enjoy their VCRs so much that they accept the risk of ending up with an obsolete machine 2 years after purchase. Furthermore, Besen and Johnson (1986) argued that when intrinsic utility is high, consumers might not delay their adoption and diffusion may be rapid very early on in spite of the presence of incompatible standards. Even innovative consumers might postpone the adoption of new products until the supporting infrastructures are sufficiently developed or the uncertainty over the dominant standard has been resolved. However, once the infrastructures are fully developed and the standards are set, the products tend to diffuse rapidly. In some cases, the government must be massively involved in the preparation of the infrastructures and the setting of standards. Of course, other factors also have an effect on the rapid diffusion, such as the rise in real income and demographic changes. The diffusion of new products is also affected by prices. However, it is difficult to precisely measure the effect of prices on diffusion since this is not generally significant. According to Van den Bulte (2000), this difficulty is due to the collinearity of prices with other regressors, such as real income. Furthermore, for some products, older products might be substituted due to the emergence of new ones. However, such a substitution is a gradual rather than a revolutionary process. As Norton and Bass (1987) pointed out, the emergence of new products does not necessarily mean the demand for older ones disappears, even though the substitution process occurs. Therefore, we do not focus on substitution in this paper. In fact, Kurose (2009) has already addressed the tendency of the demand to be saturated rapidly. He investigated the effect of the difference in the speed of demand saturation on the dynamics of the labour market (i.e. the evolution of unemployment and real wage) by using a pure labour model à la Pasinetti (1993), in which the stochastic emergence of new products is assumed. The reason that Pasinetti (1993) is used as the basic model is that it demonstrates that the change in demand-side conditions, combined with the disparity of sectoral growth rates of labour productivity, gives rise to the complex repercussion for a whole economy: structural dynamics. This is a relevant viewpoint of macroeconomic analysis, which tends to be overlooked by mainstream economics despite that it is the characteristic feature of capitalistic growth. The result obtained by Kurose (2009) is that the growth of employment tends to be accelerated, but that of the real wage tends to be decelerated, ceteris paribus, as the saturation speed of the demand for products that emerged stochastically becomes faster. Moreover, it has been indicated that the results are consistent with the price level stability. Kurose (2009) investigated just the relation between the dynamics of the labour market and the speed of demand saturation. The change in profit share could not be addressed since it is based on the pure labour model. In this paper, we therefore examine whether, independently of the institutional differences between economies, not only the dynamics of the labour market, which was scrutinised in Kurose (2009), but also the contemporary characteristic movement of profit share can be coherently explained in terms of the speed of demand saturation by introducing capital goods into a model based on Pasinetti (1981).6 This is the characteristic of our analysis. Moreover, the introduction of capital goods enables us to analyse how the direction of technical progress affects the evolution of the rate of unemployment, real wage, and profit share in relation to the speed of demand saturation. Furthermore, what was not indicated in Kurose (2009) as a theoretical implication was that our model presented in this paper has a critical implication on the relationship between the unemployment rate and the stability of price level: the relationship between the Phillips curve and the NAIRU. As is shown later, our model demonstrates that the unemployment rate consistent with zero inflation is not constant, even though the structural characteristic in the sense of Friedman (1968) is unchanged. This paper is organised as follows. In Section 2, we review and extend Pasinetti's (1981) model and subsequently show that the concept of the NAIRU is misleading. In Section 3, we present the model based in Section 2, assuming the stochastic emergence of new products and demand saturation. Furthermore, we analyse the effect of the speed of demand saturation on the dynamics of the labour market using a computer simulation. In Section 4, we examine the effect of the speed of demand saturation on profit share. Finally, in Section 5, we present concluding remarks.

In this paper, we provide a coherent framework to explain the unusual phenomena observed in industrialised economies in relation to the speed of demand saturation. Kurose (2009) has already investigated the effect of the speed of demand saturation on the dynamics of the labour market on the basis of Pasinetti (1993). We demonstrate in this paper that the same result concerning the dynamics of the labour market as Kurose (2009) is still tenable even though we take capital goods into consideration. However, we indicate that the result is affected by the factors which do not exist in the pure labour model. Moreover, we show here that the above result concerning the dynamics of the labour market is valid irrespective of the direction of technical progress. We also show that the faster demand saturation tends to increase the profit share, at least in the early period. Finally, we demonstrate that the concept of the NAIRU is misleading from a viewpoint of the model dealing with the evolving structure of demand and production.