مهاجرت از روستا به شهر، سرمایه انسانی، و تراکم

A new general-equilibrium model that links together rural-to-urban migration, the externality effect of the average level of human capital, and agglomeration economies shows that in developing countries, unrestricted rural-to-urban migration reduces the average income of both rural and urban dwellers in equilibrium. Various measures aimed at curtailing rural-to-urban migration by unskilled workers can lead to a Pareto improvement for both the urban and rural dwellers. In addition, the government can raise social welfare by reducing the migration of skilled workers to the city. Moreover, without a restriction on rural-to-urban migration, a government's efforts to increase educational expenditure and thereby the number of skilled workers may not increase wage rates in the rural or urban areas.

In this paper, a new model of rural-to-urban migration is developed, with an emphasis on the role of human capital in both urban and rural economic activities. Notable exceptions notwithstanding, a substantial literature on rural-to-urban migration has taken its cue from the dual economy model of Harris and Todaro (1970). Their model assumes that the urban sector produces manufactured goods using (homogeneous) labor and physical capital as factors of production, and that the rural sector produces agricultural goods using (homogeneous) labor and land as factors of production. The model has been widely used as a basic analytical framework for studying rural-to-urban migration in developing countries and as a platform for policy formation. However, since the Harris and Todaro model ignores human capital as a factor of production, it appears to have become increasingly less applicable to many developing countries in modern times. For example, due to continuing structural changes in recent decades, cities in the developing world have become more oriented toward service and (relatively) high-tech industries. Also, with a growing share of manufactured goods becoming standardized, these goods can be produced anywhere that offers basic infrastructure such as piped water and reliable electricity. Consequently, the production of manufactured goods increasingly takes place in the rural areas of developing countries.1 This realignment enhances the importance of human capital in the rural areas. For example, Taylor and Yunez-Naude (2000) find that in rural Mexico, the returns from schooling are high both in crop and noncrop activities; as schooling levels increase, the returns from schooling arise from activities other than crop production. In China, the township and village enterprises (TVEs) have played a significant role in the country's economic growth since the early eighties. In 2000, for example, TVEs accounted for 47 percent of the total industrial output in China (Fu and Balasubramanyam, 2003), and the output value of TVEs has been far greater than the output value of agriculture. Yang and An (2002), and Yang (2004) show that education not only increases productivity in the nonagricultural sector in rural China, but that it also facilitates and encourages the relocation of productive inputs from agricultural to nonagricultural pursuits. Jonasson (2007) finds that in rural Peru, nonagricultural rural employment is a prerequisite for positive returns to education, and that education is rewarded by rural-based nonagricultural work. A perception that the rural areas in developing countries are an exclusive domain of uneducated peasants who apply physical labor to eke out a living had better be discarded. In this paper we develop a new policy-yielding model of migration in which human capital is important in both the urban and rural areas. In line with considerable research in urban economics and economic growth, agglomeration economies in the cities are built into the model.2 In a simple general-equilibrium framework, our model interlinks three key factors: the process of migration from the rural area to the urban area, the externality effect of the average level of human capital, and agglomeration economies.3 We postulate that a city's productivity is determined by its average level of human capital and by the size of its labor force. The productivity of the rural area is determined by its average level of human capital. Right at the outset, the analysis yields a rather surprising implication: in developing countries, unrestricted rural-to-urban migration reduces the average income of both rural and urban dwellers in equilibrium. This result implies that although a city attracts all the skilled individuals and enjoys the benefit of agglomeration economies (which the rural areas do not), with free labor mobility, the city's productivity is still very low in equilibrium. The intuition underlying this result is quite simple: since the returns to skills are higher in the urban areas than in the rural areas, as is typically the case in developing countries, skilled workers are likely to concentrate in the cities. Consequently, the wage rate of the unskilled workers in the rural areas will be low, which in turn will induce a large number of unskilled rural workers to leave for the cities. With free labor mobility, the rural-to-urban migration process will come to a halt when the urban and rural wages for unskilled workers are equalized: the urban wage will decline continuously with the in-migration of unskilled workers which, in turn, will reduce the average level of human capital in the city. In other words, unrestricted rural-to-urban migration results in a lower wage for unskilled workers in both the urban and rural areas. Furthermore, since the wages of skilled and unskilled workers in the urban area are affected by common productivity factors, the wage for the skilled workers will be driven to a low level by the unrestricted rural-to-urban migration. Thus, our model explains the negative consequences of rural-to-urban migration in the developing world, as is amply highlighted, for example, in nearly every leading development economics textbook (cf. Gillis et al., 1996, Ray, 1998 and Todaro, 2000).4 In essence, our results arise from the difference between the private human capital and the social returns to human capital, a difference that implies that free labor mobility leads to an equilibrium that is not socially optimal. In all countries in general, and in developing countries in particular, there is an urban–rural wage gap for educated workers.5 Consequently, with free movement of labor, all skilled workers will cluster in the urban area, leaving the rural area with an average level of human capital that is below the social optimum. Our analysis yields several interesting policy insights. First, our model shows that various measures aimed at curtailing rural-to-urban migration by unskilled workers, which include subsidizing the rural sector and restricting rural-to-urban migration,6 can lead to a Pareto improvement for both urban and rural dwellers. Second, and somewhat surprisingly, our analysis shows that the government can raise social welfare by reducing the migration of skilled workers into the city and by subsidizing some skilled workers to move from the city to the rural areas or, for that matter, to stay in the rural areas. Third, our model shows that in developing countries, when there is nothing to deter rural-to-urban migration, a government's effort to increase educational expenditures and thereby the number of skilled workers may not increase the wage rates in rural or urban areas as long as there are still a large number of unskilled workers in the rural area in equilibrium. In fact, a rather perplexing result is that as the number of skilled workers (in the city) increases, the average level of human capital in the city may very well decrease.

Somewhat surprisingly, the received theoretical literature hardly attends to the subject of rural-to-urban migration in developing countries in modern times. As noted in the Introduction, this neglect appears to stem from two interrelated misconceptions: that rural areas produce only agricultural goods, and that the efficient production of manufactured goods can take place only in (large) cities.18 Presumably, it is because of these misperceptions that the role of human capital in the production of the rural sector is largely ignored in the received theoretical literature. We have sought to set up a new framework that bridges the gap. We postulate that a city's productivity is determined by its average level of human capital and by the size of its labor force. The productivity of the rural area is determined by the average level of human capital there. Our analysis yields a rather surprising implication: unrestricted rural-to-urban migration reduces the average income of both rural and urban dwellers in equilibrium. With free inter-area labor mobility, rural-to-urban migration will come to a halt when the urban and rural wages for unskilled workers are equalized: the urban wage will fall continuously with the in-migration of unskilled workers, which reduces the average level of human capital in the city. Furthermore, since the wages of skilled and unskilled workers in the urban area are affected by common productivity factors, the wage of the skilled workers will be driven to a low level by the unrestricted rural-to-urban migration. Thus, our analysis explains the negative consequences of rural-to-urban migration in the developing world. Moreover, our analysis yields several interesting policy insights: the analysis reveals that measures aimed at curtailing rural-to-urban migration by both unskilled and skilled workers can potentially lead to a Pareto improvement for both the urban and rural dwellers, and it shows that without restrictions on rural-to-urban migration, increasing educational expenditures alone may not increase the wage rates in the rural or urban areas. Models of rural-to-urban migration are at the heart of theories of economic development and growth. Ricardo argued that the urban industrial sector can draw away surplus rural labor without causing a rise in wages in either the rural or urban areas. Ricardo's insight was expanded in numerous subsequent writings, including the Nobel-Prize winning treatise of Lewis (1955) and the seminal contribution of Fei and Ranis (1964). This body of work explains nicely the historical demographic transition across regions as well as the economic development of the currently developed countries. Recently, Lucas (2004) has revitalized Ricardo's original insight and developed it further, noting that rural-to-urban migration is essentially a process of the transfer of labor from a traditional, land-intensive technology to a human capital-intensive technology. Despite its prominence in the development economics literature, the current body of research runs into great difficulties in explaining the consequences of the massive rural-to-urban migration flows that have occurred in a great many developing countries in the past few decades. Our theory explains the different impacts of rural-to-urban migration in the past versus nowadays. We posit that the production efficiency of a city depends not only on the size of its labor force (the agglomeration effect), but also on the average level of human capital of its labor force. In a general-equilibrium framework, we demonstrate that unrestricted rural-to-urban migration leads to inefficiency when the average level of human capital plays a significant role in productivity, which might not have been the case in historical times. In the past, production was not knowledge intensive, which suggests that it was the agglomeration effect in and by itself that resulted in high productivity, with urbanization placing the economy on a solid growth path. In the current era of knowledge-intensive production, the average level of human capital is a vital factor in the productivity of both the urban and rural areas. In this setting, unrestricted rural-to-urban migration in developing countries leads to significant negative outcomes for all individuals, in cities and countryside alike.