The Contribution of Human Capital towards Economic Growth in China

John Joshua

• Publisher: Palgrave Macmillan
• Year: 2015
• Print ISBN: 9781137529350
• eBook ISBN: 9781137529367
• Topic: Betriebswirtschaft
• Subtopic: Internationale Wirtschaft

This book develops a human capital model to explain transformational growth within different stages of economic development, which will induce technological changes and consequently will require a change in human capital. China is a case study in transition and can provide useful lessons to other emerging economies.

1

Neoclassical and Endogenous Growth Models

Introduction

Human capital in this thesis is discussed in terms of a factor of production which may increase output and productivity as the aggregate stock of skills, training, education, knowledge and ability of its workforce increases. Different levels of human capital may contribute to different degrees towards economic growth, depending on the stage of economic development and the intertwined relationship with technological change.

As China is moving through different stages of development, new types of human capital are required, so that we may speak of a human capital transformation curve, which is a direct consequence of the shift in demand for different qualities of labour. This chapter will discuss how human capital is considered in neoclassical and endogenous growth models.

This chapter is divided in three sections. The first section provides a short comparative analysis of neoclassical and endogenous growth theories in order to put the thesis in a proper theoretical context. This theoretical discussion is a prelude to the thesis. Further analyses of such theoretical concepts will be made throughout the thesis when appropriate in regard to Chinese transformational growth and human capital accumulation. The second section introduces human capital and technological change as endogenous factors of production. The third section highlights the importance of endogenous and exogenous factors as facilitators of economic growth.

A comparative analysis of neoclassical and endogenous growth theories

Theories of economic growth evolve over time. Hence, it may be useful to commence with the Harrod-Domar model of economic growth, as it was the first model that tried to create a coherent economic growth model, and so puts the present theories in their historical perspective. The Harrod-Domar and the other earlier models typically neglected to analyse the role of technological advance and human capital in particular in economic growth.

1. The Harrod-Domar theory of economic growth

During the 1950s, economic growth theory and its applications were mostly influenced by Harrod-Domar models (see Harrod, 1939, 1948, 1963; Domar, 1946, 1957) which assumed a constant capital-output ratio and ignored technological change altogether. According to the Harrod-Domar growth model, there are three factors which determine growth: (1) the rate of savings; (2) the capital-output ratio; and (3) the rate of depreciation of assets. As depreciation increases, less savings will be available for new investment so that growth will be hindered. An increase of capital relative to output will also reduce growth because of a decrease in productivity; but a constant output implies diminishing capital efficiency. Hence, ceteris paribus, declining capital efficiency will hamper growth.

One of the questions that Harrod (1939, 1948) and Domar (1946, 1957) were concerned with was which particular circumstances within an economy may lead to a steady-state growth; whereby knowledge as embedded in human capital, output, and wages grow at a constant rate. Within their model, n is population and labour force which grow at a constant proportional rate (Solow, 2000, p. 8); s is net saving and investment are a fixed fraction of net output at any instant of time (s = the fraction); and v is the technology of the model economy which is fully described by two constant coefficients: One is the labour requirement per unit of output, and the other is the capital requirement per unit of output. If s is smaller than vn, the proportion of investment to output is less than that required to keep the capital stock growing as fast as the labour force. If the rate of unemployment is kept constant, the economy will eventually run out of capacity. A steady state exists when the saving rate, capital-output ratio, and the rate of labour force are constant. This will only occur when s = vn so that an increase in capacity is matched by the increase in the labour force to obtain a constant capital-output ratio. If however s, v, and n are independent constants, then s is unlikely to equal vn. However, the Harrod-Domar model did not allow for an increase in labour or an increase in aggregate output as explicit mechanisms, and thus was inadequate in explaining economic growth.

The Harrod model (1948), which may be regarded as a precursor to exogenous growth models, defined a static equilibrium as a situation in which: “work is steadily going forward ... but without increase or diminution” (p. 3); whereas “dynamics” implies “an economy in which the rates of output are changing” (p. 4). Hence, even a small steady rate of savings becomes dynamic because it provides a source for capital which enhances growth and thereby will affect other factors associated with growth. In this case, we can concur with Harrod (1948) who stated that “[i]n static economies we must assume that saving is zero” (p. 11). Harrod’s determinants of the natural rate of growth are primarily the technological conditions of the economy. Although he did not specifically mention R&D and the formation and accumulation of human capital, investment in R&D and education can ameliorate the technological conditions of a country.

Harrod (1939 and 1948) and Domar (1946 and 1957) regarded the rate of savings and the productivity of capital as determinants of economic growth. Like Keynes (1967 [1936]), they did not accept that there is an inherent reason for a balanced economic growth; rather that economic growth depends on deliberate economic policies to increase investment by increasing saving and technological change which enhances the productivity of capital.

Within a steady state, the rate of growth of gross investment equals the requirement to maintain the capital stock and the economy is growing at the warranted growth rate (see Harrod, 1939; and Robinson, 1962, p.44; 1965a, 1965b, 1969, 1970a, 1970b). Harrod (1948) saw technological progress as proceeding continuously at a steady rate; and a steady growth implies that knowledge as embedded in human capital, output and wages, grow at a constant rate. The vintage model of neoclassical theories allows for the fact that new capital formation involves new technology (Solow, 1960; Nelson, 1964) which is assumed to be exogenous. However, technological progress has been endogenized as “learning-by-doing” by Arrow (1962a) and Kaldor and Mirrlees (1962), whereby new technology becomes embodied in new vintages of capital goods which then is assumed to percolate into the next generation’s vintage. The improvement of technology is then assumed to be a continuing process.

2. The neoclassical model of economic growth

Solow (1957) argued that for any increase in GDP, an incremental increase in capital will result in a smaller increase in output. In this case, there will be less inclination to invest in capital so that the capital-output ratio will decline. In Solow’s view, diminishing returns to capital in the long term would be counteracted through improved technology in order to increase efficiency. As Solow (1987) in his Nobel lecture stated, the effect of diminishing returns is: “that the equilibrium rate of growth is not only not proportional to the saving (investment) rate, but is independent of the saving (investment) rate” (p. xii); so that a permanent rate of growth of output per unit of labour can only be accomplished through technological change. On the other hand, the Harrod-Domar model assumes a constant capital-output ratio.

Solow’s growth theory predicts that growth in productivity would decline to zero when capital per worker increases, unless there occurs some technological change which was assumed to be exogenous. A further assumption of Solow’s model is that all countries would eventually converge at the same level of productivity.

Even though Solow linked technological change to economic growth, traditional neoclassical growth theory had nothing to say about the origin of technological change, and knowledge itself was regarded as a residual that remained once inputs of capital and labour had been accounted for. Furthermore, Solow’s (1956 and 1957) model emphasizes technological progress, but does not give equal emphasis to capital accumulation; even though technological progress and capital accumulation usually appear jointly as new technological know-how, it is generally embodied in physical capital so that if one factor increases, so, too, will generally the other.

Solow’s (2000) concept of a steady state is defined by an ultimately constant employment to capital ratio. That is, “the capital stock is growing at the same rate as the supply of labour” (p. 30), so that “aggregate output grows at the same rate as employment, and output per head is constant” (pp. 30–31). Hence, Solow (1956, 2000) assumed a continuously differentiable production function which is linear homogeneous, whereby output per unit is a function of capital per capita.

According to the neoclassical growth theory (see Solow, 1956, 1962), the rate of economic growth is determined by the initial level of output and the level of savings. In equilibrium, savings equal investment. Hence, an increase in the rate of savings will increase capital. However, a shortage of savings does not necessarily imply that a country has a lack of physical investment because net international capital inflows may compensate for a lack of savings. Nevertheless, in the short run, an increase in the rate of savings will reduce consumption and consequently it will also reduce aggregate demand and output; but in the long run, output per capita will again be equal to the rate of technological change, which is assumed to be exogenous in the neoclassical model.

The neoclassical growth model, as developed by Solow (1956), also assumes that the long-term growth rate is determined by technological progress, but the rate of technological progress itself is left unexplained. On the other hand, endogenous growth theories as developed by Romer (1990a, 1994) and Aghion and Howitt (1998) argue that technological innovations may be induced by previous economic conditions. They also allow for the possibility of government policies to influence long-term economic growth. For example, research and development may be sponsored or subsidized by government policies and thereby may facilitate technological progress through an increase in inventions and innovations, which may also induce positive externalities and spill-over effects.

Solow (1956) in his study of economic growth used a standard neoclassical production function with decreasing returns to capital. The steady-state level of income per capita was determined by the rate of population increase and by the rates of saving; both of which he assumed to be exogenous. Mankiw, Romer and Weil (1992) pointed out that their own research confirmed Solow’s model in so far as it predicts the direction but not the magnitude of the effect on income through growth in savings and population. Nevertheless, in their empirical analysis they found that: “more than half of the cross-country variation in income per capita can be explained by these two variables alone” (p. 407).

As population growth will increase the available labour force, the steady-state values of capital and income will increase at the same rate as labour. Hence, long-term growth can be sustained through a growth in population so that there will be no constant levels of capital and output. However, what a population increase cannot do is to increase the capital or output per worker in the long run. As capital accumulation is subject to diminishing returns, economic growth, which is input-driven, is not sustainable (Krugman, 1994); so that long-run economic growth is only achievable through technological progress and the enhancement of human capital which may increase total factor productivity (TFP).

Within a neoclassical framework, economic growth must come to an end if there are no further technological improvements because of diminishing returns of capital. However, Barro (1997) pointed out that over the long run: “positive rates of per capita growth can persist over a century or more and ... these growth rates have no clear tendency to decline” (p. 3). Of course, this may be explained through technological progress which has been caused in an exogenous fashion as Barro (1999b) himself recognized (see also Barro and Lee, 1994). However, Solow (2000) argued that to have: “more R [research effort] forever will do nothing for the growth rate, although it will do quite a lot for growth” (pp. 99–100).

Barro argued that the convergence hypothesis (see Chapter 5 of this book) as proposed by neoclassical growth models and largely adopted by Solow appears to be inconsistent with cross-country evidence which does not show any correlation with a country’s starting point of its per capita product and its subsequent rate of economic growth. The central focus of Barro’s (1997) growth theory is his concept of convergence between poorer and richer economic regions. That is, regions with lower GDP per head are expected to grow faster than richer regions; but it is a slow process. Barro argued that the gap closes “at the iron-law rate of around 2 % per year” (p. 15), which according to Barro, has not been broken by any regions as yet. As there is a tendency towards convergence, countries with lower initial values of capital and GDP or national income eventually will catch up with the richer ones; the lower the initial values, the faster the growth rate will be during the transition period towards convergence.¹ As we proceed towards convergence, capital per worker will increase so that the growth rate declines. Consequently, the growth rate will be highest when capital per worker is lowest. Capital movements across provinces and industries may narrow the gap in the marginal productivity of capital.

Barro (1997, p. 12) found in his cross-country empirical study that there has been an increase in inequality as the standard deviation of the log of real per capita GDP has increased from 0.89 in 1960 to 1.14 in 1990; however, according to Barro, this does not in any way refute the neoclassical prediction of conversion, because “the poor tending to grow faster than the rich is not the same as a declining trend in inequality” (p. 12) and also because the neoclassical model only predicts a conditional convergence.

Romer (1986, 1989a) argued that the neoclassical growth model is incapable of explaining the variations in growth rates and per capita incomes in different countries, because it assumes decreasing returns to capital, exogenous technology and perfect competition. Romer noted that as the growth rates of the most technologically advanced countries have increased, rather than fallen, countries have not fulfilled the prediction of the neoclassical model of convergence towards a common level of per capita income if such countries have approximately the same saving ratio and technology. However, according to Mankiw, Romer and Weil (1992), data on international differences in per capita income and rates of growth do not contradict but rather confirm Solow’s model when human capital has been included and allowances made for differences in the saving ratios between countries.

If differences in savings and in the accumulation of human capital are introduced into regression models of growth accounting, there is a tendency towards conditional convergence (see Mankiw, Romer and Weil, 1992). Human capital in the form of educational attainment has been found to be most important in a country’s ability to catch up with countries that have higher productivities. Thus: “countries with similar educational levels were shown quite consistently to be converging ...

Table of contents

• Cover
• Title
• Introduction
• 1  Neoclassical and Endogenous Growth Models
• 2  The Accumulation of Human Capital as a Factor of Production
• 3  Pathways of Human Capital towards Economic Growth
• 4  Human Capital Transformation and Cycles of Innovations
• 5  Stages of Economic Development and Convergence and the Changing Demand for Human Capital
• 6  Human Capital, Income Distribution and Economic Growth
• 7  An Economic Profile of China
• 8  Transformational Growth and Human Capital Transformation in China
• 9  Convergence and Divergence between Chinas Provinces
• 10  Review of Empirical Studies on China
• 11  New Evidence on the Contribution of Human Capital to Chinese Provincial Economic Performance
• Conclusion
• Notes
• References
• Index