سرمایه انسانی و سوئیچ از کشاورزی به صنعت

Endogenous technological change explains the transition from agriculture to industry. Initial production is agricultural. Human capital accumulation causes the economy to switch from agriculture to industry. The model explains slow growth in population and income before a switch to a balanced growth path of higher population growth and rapid income growth. Introducing a human capital externality in the total factor productivity of agriculture and industry allows for the reproduction of the previous outcomes and generates better per capita incomes in 200 and 3000 BC, the onset and duration of an agricultural revolution, time varying market size and modern income differentials.

From 200 BC to 1850 AD, the rate of population growth in Europe averaged barely one tenth of one percent per year.1 The rate of growth of income was, perhaps, even lower.2 After the onset of industrialization, both population growth and income growth accelerated. Between 1850 and 1996, the rate of population growth in Europe averaged 0.8 percent per year—more than seven times its prior rate. The rate of growth of per capita income from 1800 to 1996 increased by an even larger amount, from essentially no growth to 1.8 percent per year. This paper models the transition from agriculture to industry endogenously. Researchers have taken a variety of approaches to modeling these events (Laitner, 2000; Galor and Weil, 2000; Hansen and Prescott, 1999; Jones, 1999; Tamura, 2001). However, virtually all models to date rely on exogenous forces to generate the transition. Such features are understandable, for endogenous growth models quickly lose analytic tractability when they are extended to complex environments. Such models are, however, solvable using numerical methods. In this paper, I develop a model of economic and population growth that generates a transition between agriculture and industry endogenously. The model is similar to that of Lucas (1998), and embodies elements from the models of a number of other authors, including Kremer 1993a and Kremer 1993b, Rosen (1982) and Tamura 1992 and Tamura 1995. The model assumes the existence of two highly stylized methods of producing a single final output. As in Lucas (1998), output is produced in autarky using a fixed factor, called land, and labor, and is labeled the agricultural method. As in Rosen (1982) and Tamura 1992 and Tamura 1995, in the second, output is produced using labor along with intermediate services that are acquired through trade, and is labeled the industrial method. Trade, however, incurs a coordination cost that is inversely related to the level of human capital. In the model, individuals start out having low levels of human capital and specializing in agricultural production in autarky. In Lucas (1998) the precipitating event moving an economy away from a classical steady state in income and population is an exogenous increase in the rate of return to human capital investment. Here, however, human capital accumulates because individuals care about the incomes of their children, and human capital is productive even in agriculture. Eventually the coordination costs of trade are sufficiently low so as to make the switch from agriculture to industry profitable. In addition, the accumulation of human capital has implications for fertility. When specialized in agriculture, rates of both population growth and income growth are low. Both growth rates increase after the transition to industry.3 The resulting model is too complex to be solved analytically. The model is therefore solved numerically so as to be consistent with the history of Europe, Canada and the United States combined (henceforth called the West or Western). Laitner (2000) examines a model with two goods, agricultural and manufacturing. When capital is low the productivity of the economy is low. In the low income state, the economy is predominately agricultural. With capital accumulation and exogenous technological change, incomes rise, and because agricultural consumption has an income elasticity less than 1, there is rising demand for manufactured products. Thus an economy slowly evolves toward a modern economy with a larger manufacturing sector.4 In Galor and Weil (2000) there is a subsistence level of consumption. For low levels of capital, fertility is constrained and consumption is constant at the subsistence level. However in their model if population increases the rate of return to human capital accumulation, or if there exists learning by doing or if there is exogenous technological progress, the productivity of households rises. This initially leads to an increase in fertility and low levels of consumption. With continuous technological progress, eventually the rate of return to human capital accumulation is sufficiently high that individuals reduce their fertility and begin accumulating capital. Hence there is a demographic transition to modern economic growth. Hansen and Prescott (1999) present a model with two methods of producing a single consumption good. For low levels of capital, technology A is more productive than technology B. However A and B have different rates of technological progress. Eventually, technology B crosses technology from below, that is to say there exists a capital level that makes producing using technology B more productive than utilizing technology A. When the first crossing occurs, there is a strong incentive to specialize capital in the hands of some individuals to operate technology B and produce more output than if these same individuals produced using technology A. The model explains periods of income equality and rapid development of efficient inequality. Eventually capital accumulation is sufficiently large in the economy to allow all individuals to operate the more productive technology B. Jones (1999) and Tamura (2001) introduce mortality into the model. In these models fertility is positively related to the level of mortality. With falling mortality, due to rising consumption in Jones (1999) and rising average human capital in Tamura (2001), fertility falls. Jones achieves acceleration of economic growth from rising population, as in Kremer (1993b) and from an exogenous shock to the productivity of idea creation, which he identifies with increased property right protection. Tamura (2001) induces greater economic growth by relying on the rising level of human capital investment coincident with falling fertility. The model of this paper produces an endogenous switch from agricultural methods of production to specialized industrial production. Capital accumulation alone can produce a switch in productivity between methods of production. For low levels of capital, output is more efficiently produced using the agricultural method. With accumulation, eventually the specialized industrial production method becomes more productive. Thus capital accumulation induces a change in the rate of growth of capital and population, consistent with evidence on industrialization and demographic transitions. Finally the model is extended by introducing endogenous technological change in order to fit the history of the western economies from 3000 BC to the present. The paper is organized as follows. Section 2 introduces the baseline model. Section 3 analyzes the equilibrium problem with perpetual human capital accumulation, which produces endogenous economic development. The predictions of the model are compared with historical evidence from Europe and Asia in Section 4. Section 5 presents a numerical example designed to match as closely as possible the experience of the combined region of Europe, Canada and the United States from 200 BC to 1996. This exercise is reasonably successful in matching population in 200 BC, 1850 AD, 1950 and 1996, and US per capita income in 1850 and 1996. However, the implied incomes prior to 1850 are implausibly low. Section 6 therefore modifies the baseline model by introducing human capital spillovers. In addition to generating more plausible estimated incomes prior to the Industrial Revolution, these spillovers predict an agricultural revolution prior to industrialization.5 The model also produces a time series on human capital that seems consistent with interpretation of evidence by a number of economic historians, including Braudel (1979), Cipolla (1978) and Easterlin (1981), all of whom identify human capital accumulation as a key feature of industrialization. Section 7 presents a sensitivity analysis of the numerical solution. The conclusion follows.

This paper developed a model in which the transition from one mode of production to another, from agriculture to industry, occurs endogenously. The model necessarily abstracted from a number of details, including some addressed explicitly in other frameworks. For example, I ignored the issue of subsistence analyzed in Galor and Weil (2000), nor did I find food production play a special role as it does in Laitner (2000). The paper abstracts from the institutional environment and protection of private property, which Baumol (1990), Bethell (1998), North 1981 and North 1990, Rosenberg and Birdzell (1986), and Parente and Prescott (1999) have argued persuasively plays an important role in development. Finally the paper ignores the role of changing mortality risk examined in Jones (1999) and Tamura (2001). Despite the model's stylized nature, the transitional dynamics were too complex to analyze analytically, necessitating the use of numerical methods. The numerical solutions were reasonably consistent with the population growth experience of the West since 3000 BC, and with the levels of per capita income in 1850 and 1996. A number of additional features arose when the model was extended to include human capital spillovers. In this case, both human capital accumulation and population growth accelerated prior to the departure out of agriculture. 46 The rise in agricultural productivity is an agricultural revolution, corresponding to the transition from hunter-gatherer society to farming. The extended model also better fits per capita income levels in agriculture. A number of extensions are possible. One would be to account for the diffusion of the industrial revolution, as in Lucas (2000). This could be accomplished by incorporating conditional human capital spillovers in the accumulation technology as in Tamura (1996). There would occur just single Industrial Revolution, which could spread through a combination of integration of world markets and imitation.