Capital stock and its impact on unemployment in Chile

Introduction

Since reversals of inflations shocks happened on seventies were not able to reduce the unemployment rate, which remained high even though the reversals reduced the oil price and inflation in general, the supported relationship between inflation and unemployment was questioned as a main explanation of unemployment. In addition, the continuously decrease in real salaries during 1990s in some European countries has not had the expected effect of arising the employment levels, therefore labor-wage elasticity theory has not worked at all. All of above has encouraged to labor economist to find alternative explanations on unemployment phenomena. On this path, labor markets regulations, international trade, productivity changes and investment have arisen as feasible causes of unemployment.

Even though labor market regulations have been highlighted as one of the most important variable affecting labor demand because the expected negative impact of market rigidities on employment, the economical research has not been able to reach conclusive evidence. Then, capital stock, between other variables, has been researched as affecting the unemployment rates, getting a significant amount of positive evidence. Its relevance as an explanatory variable rests on the Keynesian assumption regarding the influence of good markets over labor markets which would not be able to explain the unemployment just by itself. As a consequence, this study intends to describe how the capital stock affects the employment level by reviewing the concerned theory and by analyzing the empirical evidence. Finally, a conceptual analysis over the particular Chilean case is done.

Economic theory

According to the Keynesian point of view, it is not possible to solve the unemployment phenomena for using policies and tools operating only from the labor market, especially when we find ambiguous outcomes from the literature concerned to labor market institutions" impact on labor demand (Stockhammer, 2010; Arestis et.al. 2007). In difference to classical studies undertaken by Layard, Nickell and Jackman (1991) about unemployment in the UK which assumes that changes in labor market institutions would lead to changes in the supply-side equilibrium independently from any variations in the level of economic activity, Keynesianism assumes a NAIRU continually changing because investment is always causing the capital stock to vary. Thus, because the pace of investment is always being influenced by the level of economic activity, the variability of NAIRU will be influenced continually by the path of aggregate demand (Stockhammer and Klar, 2010; Arestis et.al. 2007; Karanassou and Sala, 2008).

How does capital stock affect unemployment? Even though the channels identified can differ among researchers, most of them can be classified into three main streams (Stockhammer and Klar, 2010). Firstly, following the Keynesian assumption telling that the level of output and employment as a whole depends on the amount of investment (Bellais, 2004), is argued that capital accumulation may play a role as a demand factor: investment is the most volatile of the macroeconomic aggregates and is considered the driving variable in business cycle theory as well as in growth theory (Stockhammer and Klar, 2010; Stockhmammer, 2011). According to Malley and Moutos (2001), nonetheless, we need to take into consideration that the strength of relationship capital stock – unemployment will decrease (almost horizontal in terms of correlations) if unemployment is nearly frictional (lower rates of unemployment).

Secondly, elasticity of substitution between capital and labor is presented as being lower than classical economics says (Rowthorn, 1999; Malley and Moutos, 2001; Stockhammer and Klar, 2010). Lower substituibility, which is assumed in a CES production function (figure 1), is quite relevant is order to explain why a Cobb-Douglas function fails to predict the labor market performance when it says that only wages varies nor employment level when capital stock becomes higher (Kapadia, 2005; Rowthorn, 1999). This lack of substituibility can be demonstrated by arguing that as effects of demand shocks on employment and investment are reversed unemployment may not fall to previous levels due to insufficient capital (Arestis et.al. 2007; Kapadia, 2005).

Figure 1

Comparison between Cobb-Douglas and CES function performances by introducing increases in capital

stock.

Source: Graphs taken from Kapadia (2005), pp. 5 and 6.

Third channel says that there would be a bargaining effect. Rowthorn (1995) argues that "unemployment reduces the ability of workers to push up wages, while excess capacity limits the ability of firms to raise prices" (p. 28). Thus, an insufficient capital stock will require a higher unemployment rate to equilibrate income claims of workers and employers. In this sense, the capability of capital stock to diminish the unemployment is an issue concerned to medium and long term because in short run changes in capital stock are not feasible (Arestis and Biefang-Frisancho, 2000).

As a result of all of these theoretical assumptions, Stockhammer (2011) claims that effective labor demand need not be downward sloping at all. Given that main variable affecting unemployment would be capital stock, increases in wage can rise the aggregate demand because wage earners are likely to have a higher consumption propensity than earners of profit income (upward sloping of labor demand hypothesis is supposed to work in profit maximization firms based markets).

Evidence

We can summarize the evidence found by classifying three types of findings supporting the theoretical mainstreams assumptions and having relevant methodological implications. Firstly, the hypothesized relationship between capital stock and unemployment has been strongly supported by empirical research. It has been demonstrated, for instance, that countries that experienced the largest slowdown in capital accumulation per labor hour faced the highest unemployment rates in the 1990s (Arestis et.al 2007). On the same path, Karanassou and Sala (2008) state that if capital stock growth had not increased as it did from 1993 to 2006, unemployment in Australia would have ended the period about 5 percentage points higher than it actually did. In a cross Nordic countries study, Karanassou et.al. (2008) found that capital stock explains around 30% of the increase in unemployment in Denmark whereas in Sweden, capital accumulation contributes to 50% of the unemployment during the 1990s. The same trend was found by Kee et.al. (2005) who aims that capital accumulation in the export sector explains most of the decline in Singapore"s unemployment rate.

Secondly, in terms of comparative research between capital accumulation and LMRs impacts on labor demand, Arestis et.al. (2007), by estimating cointegrating relationships, found that capital stock has the most consistent impact of all the variables included in the estimations of the unemployment relationship such as wage flexibility, long-term unemployment, militancy and benefit provision. Stockhmammer and Klar (2010) performed a panel data analysis for OECD countries controlling a set of OECD-LMRs and found strong capital accumulation effects, substantial effects of interest rates, but very small effect of LMRs. On the same path, Kapadia (2005) demonstrated that capital stock has an stronger impact on unemployment than labor institutions by comparatively testing labor market implications for both Cobb-Douglas and CES production functions.

Finally, in terms of research regarding capital stock and economic policy, it has been found that interest rates have empirically important effects on unemployment. For instance, Stockhammer (and Klar, 2010; 2011) states that "macroeconomic variables have a much greater impact and among these capital accumulation and the real interest rate are the most important ones" (pp. 455). Malley and Moutos (2001) conclude that policies encouraging a faster rate of capital accumulation should be a necessary component of any policy package against the unemployment. Kapadia (2005) argues that policy implications in order to reduce equilibrium unemployment can be significant since the reduction in unemployment in countries such as United Kingdom and Netherlands (which presented investment booms during the 90s) has been sustained long after investment rates have fallen.

In spite of this robust evidence confirming the positive impact of capital stock on employment, other research offer alternative explanations. Malley and Moutos (2001), for instance, say that impact of absolute growth rate of capital stock in OECD-European countries has been less determinant that relative evolution of capital stock to other countries because an increase in the domestic capital stock relative to the foreign capital stock would allow to domestic firms to compete more effectively by capturing market shares at the expense of foreign firms. However, there are not enough studies testing this assumption. By the other hand, Kapadia (2005) also confirms the negative impact of capital stock on unemployment but its study limits this impact over a certain range.

Methodological Issues

Methodologically, there would be a tendency to examine the influence of capital stock on unemployment by using time-series methods and usually these studies do not analyze together both capital stock and LMRs impacts because time-series data is only available for a small number of LMRs (Stockhammer and Klar, 2010). Another methodological characteristic of capital stock studies is they usually use single unemployment rate equations and proxy variables such as real interest rates, real balances or investment ratios (Karanassou and Sala, 2008). For instance, Stockhammer (2011) offer us the following equation where LMI, ACCU and MS represent labor market institutions, capital accumulation and macroeconomic shocks, respectively. C represents other control variables to be specified later. FEt and FEj are crosssection and period fixed effects, respectively.

However, according to Karanassou and Sala (2008) this trend for using single-equation estimation brings implicitly two problems: (1) the proxies could not capture the effects of capital accumulation of other influences due to often the influence of capital stock is hidden behind noncontroversial accounts of the unemployment upturns due to rises in interest rates or financial crises. (2) Since the unemployment rate is a nontrended variable, single equation models have to use exogenous variables that do not display a trend. This would not the case with multi-equation labor market models.

Capital stock and unemployment performance in Chile

Unfortunately there is not research analyzing the capital accumulation impact on labor demand in Chile. However, by simply analyzing the trends of unemployment and growth of capital stock, an inverse relationship between both of them variables can be argued (figure 2). As we can see, an opposite trend from 1987 to 2006 is seen by putting together the corresponding curves. This trend, however, is not kept regarding last three years when both unemployment and growth of capital stock follow the same increasing direction. A possible explanation can be provided by Kapadia´s assumption (2005) which aims increases in unemployment can be sustained long after reduced investment rates.

Figure 2

Evolution of unemployment rate and growth of capital stock in Chile (1987 – 2009)

Source: Statistical Database of Central Bank of Chile

The inverse relationship between capital stock and unemployment is confirmed by running a simple correlation analysis. As we can see in figure 3, a strong correlation coefficient of -0.48 is obtained which indicates a closely negative relationship between both of them variables. This correlation is on the path of Karanassou et.al.´ studies regarding Nordic countries (2008).

Figure 3

Correlation between unemployment rate and growth of capital stock (1987 – 2009)

Source: Own calculations using data from Central Bank of Chile

A possible explanation about this trend can be obtained by analyzing shares of capital growth by types of capital stock in Chile (table 1). As we can see, both housing and rest of construction (share of 74% together) show a steady trend through the time in difference with machinery and equipment, which shows significant variation among several periods. Following the theoretical assumptions previously reviewed (Rowthorn, 1999; Malley and Moutos, 2001; Stockhammer and Klar, 2010; Kapadia, 2005), lower substituibility could be determining a lower unemployment as a higher growth in capital stock is introduced in Chile because most steady and biggest shares of capital stock are related to assets (housing and rest of constructions) which cannot as easily replace labor as machinery and equipment.

Table 1

Growth of capital stock evolution by type of capital (1987 – 2008)

Source: Statistical Database of Central Bank of Chile

Conclusions

Even though the evidence seems to be quite robust in order to confirm a negative relationship between capital investment and unemployment, into economic theory still remain a considerable amount of open questions underlying this relationship. The lack of research does not still allows to define precisely which is the channel underlying the impacts on labor demand although the lower elasticity of substitution between capital and labor has been widely demonstrated by empirical evidence. Thus, the classical assumption arguing that a high elasticity of substitution determines the presence of a Cobb-Douglass production function, which gives more importance to labor market institutions as unemployment determinants, has been almost rejected.

The implications of these findings by economic policy also has been discussed, concluding that empirical evidence permits to argue a relevant impact from fiscal and monetary policy on unemployment through both interest rate and capital accumulation.

Finally, even though time-series studies analyzing the relationship between capital stock and unemployment are needed for the Chilean case, a simple correlation model can initially demonstrate a negative relationship which could be determined by a lower elasticity of substitution between capital and labor (such as is predicted by the theory) supported in a massive share of capital stock in non-renewable assets in the short term.

References

Arestis, Philip; Mariscal, Iris. 2000, "Unemployment and Wages in the UK and Germany", Scottish Journal of Political Economy, Vol. 47 (5), pp: 487-503.

Arestis, Phillip; Baddeley, Michelle and Sawyer, Malcolm. 2007. "The relationship between capital stock, unemployment and wages in nine EMU countries", Bulletin of Economic Research, Vol. 59 (2), 125–48.

Bellais, Renaud. 2004, "Post Keynesian Theory, Technology Policy, and Long-Term Growth", Journal of Post Keynesian Economics, Vol. 26 (3), pp: 419-40.

Kapadia, Sujit. 2003, "The Capital Stock and Equilibrium Unemployment: A New Theoretical Perspective", Economics Series, Department of Economics, University of Oxford, Working Papers N° 181.

Karanassou, Marika and Sala, Hector. 2010, "Labour Market Dynamics in Australia: What Drives Unemployment?", Economic Record, Vol. 86 (273), pp. 185-209.

Karanassou, Marika; Sala, Hector and Salvador, Pablo. 2008, "Capital Accumulation and Unemployment: New Insights on the Nordic Experience", Cambridge Journal of Economics, Vol. 32 (6), pp: 977-1001.

Kee, Hiau Looi and Hoon, Hian Teck. 2005, "Trade, Capital Accumulation and Structural Unemployment: An Empirical Study of the Singapore Economy", Journal of Development Economics, Vol. 77 (1), pp: 125-52.

Layard, R., Nickell, S. J. and Jackman, R. 1991, Unemployment: Macroeconomic Performance and the Labour Market, Oxford University Press, Oxford.

Malley, Jim and Moutos, Thomas. 2001, "Capital Accumulation and Unemployment: A Tale of Two 'Continents.'", Scandinavian Journal of Economics, Vol. 103 (1), pp: 79-99.

Rowthorn, R. E. 1995, "Capital formation and unemployment", Oxford Review of Economic Policy, Vol. 11(1), pp: 26–39.

Rowthorn, R. E. 1999, "Unemployment, wage bargaining and capital–labour substitution", Cambridge Journal of Economics, Vol. 23 (4), pp: 413–26.

Stockhammer, Engelbert and Klar, Erik. 2010, "Capital accumulation, labour market institutions and unemployment in the medium run", Cambridge Journal of Economics, Vol. 35, pp. 437–457.

Stockhammer, Engelbert. 2011, "Wage norms, capital accumulation and unemployment. A Post Keynesian view", forthcoming in Oxford Review of Economic Policy, Version 2.0 (School of Economics, Kingston University).

Autor:

RodrigoValdivia Lefort

Writing Economic Reports – MA Business Economics

Module Leader: José Sanchez Fung

London, May 2011