سرمایه انسانی و رشد اقتصادی در اسپانیا، 1850-2000

We investigate human capital accumulation in Spain using income- and education-based alternative approaches. We, then, assess human capital impact on labor productivity growth and discuss the implications of its alternative measures for TFP growth. Trends in human capital are similar with either measure but the skill-premium approach fits better Spanish historical experience. As education is a high income elastic good, human capital growth computed with the education-based approach seems upward biased for the recent past. Human capital provided a positive albeit small contribution to labor productivity growth facilitating technological innovation.

The role of human capital in the growth process has been extensively analyzed since Adam Smith and Alfred Marshall, and has interested both theoretical economists and economic historians. However, it was not until the mid-twentieth century that Becker (1964), Schultz (1961), and Mincer (1958) developed a complete theory of human capital, according to which the individual level of education and experience determines future labor income. The United Nations (1997) defines human capital as “productive wealth embodied in labor, skills and knowledge” while the OECD describes it as “the knowledge, skills, competencies and attributes embodied in individuals that facilitate the creation of personal, social and economic well-being” (OECD, 2001). This broad definition does not restrict human capital to education but encompasses all investments in humans which are made to improve their skills, including schooling and informal education provided by parents or other family members, on-the-job training and learning-by-doing (i.e., acquiring skills through work experience) or any other activities that improve the productive use of a person's skills. The measurement of human capital is even more elusive than its definition. Many authors have employed formal education measures, such as enrolment rates or the level of educational attainment, while others resorted to indirect proxies including literacy and numeracy as a way to identifying human capital. However, none of these measures adequately defines human capital since they ignore informal education, vocational training, workers' experience, and on-the-job training. Further, these partial measures do not consider the economic value (benefit) of human capital, the potential differences in rates of return between different types of education, and the acquisition of human capital for individual consumption and not for production. The point is that a “sound” measure of human capital should be not only comprehensive but also consistent with theoretical underpinnings. How to measure human capital will be the first question to be addressed in this paper. The second question we will consider, the contribution of human capital to economic growth, has attracted considerable attention from the literature. There is a degree of consensus among economists and economic historians about the important role of human capital in long run growth and its contribution to convergence and catching-up (Abramovitz, 1986). Schultz (1963) argued that a large share of economic growth comes from further additions to the initial stock of human capital and that human capital accumulation was largely responsible for the “residual” in early growth accounting exercises. Denison, 1962 and Griliches and Jorgenson, 1967 examined this hypothesis empirically and concluded that changes in the quality of the workforce did not account for all total factor productivity (TFP) increases. With the emergence of ‘new growth theory’ in the 1980s and 1990s and, in particular, with the important contribution by Lucas (1988), the relationship between human capital and growth became even more central for those interested in the causes of growth. Again, new empirical studies, in this case based on cross-country regressions, have served to qualify initial theoretical arguments. Benhabib and Spiegel, 1994 and Krueger and Lindahl, 2001 pointed out that it is the level of educational attainment rather than its increase that matters for growth. More recently, Cohen and Soto (2007) have argued, instead, that growth in schooling rates has a statistically significant influence (albeit a relatively modest one) on GDP growth rates. Among economic historians, Sandberg (1979) attributed the successful development of the Swedish economy during the nineteenth century to its comparatively higher literacy levels.1 In a similar vein, several studies related successful and unsuccessful development stories during the nineteenth century to the presence or absence of certain education levels.2 The contribution of this article is twofold. On the one hand, we provide two alternative measures of human capital for Spain from 1850 to 2000: the first is based on the concept of education, and the second on Jorgenson's income-based concept of ‘labor quality’. Then, we review the advantages and shortcomings of each measure and discuss whether they are compatible. As a contrast, we present new empirical evidence on human capital accumulation, and we use a growth accounting framework to explore its contribution to economic growth.3 As a country whose people were poor and ignorant during the nineteenth century, but which had joined the club of rich countries by the late twentieth century, Spain's experience is particularly relevant for the debate. Over the past century and a half, Spain has experienced a sustained expansion of GDP per head at an average rate of 1.9% per year, and GDP per hour worked has increased by an annual average of 2.1% (Fig. 1). Although long-run trends in human capital are similar whichever of the two measures is used, we conclude that the direct, skill premium approach favored by Jorgenson fits better the historical experience of Spain as observed over shorter periods. Human capital provided a positive, albeit small, contribution to labor productivity growth and it could be suggested that human capital accumulation probably had a positive effect on GDP level facilitating technological innovation (Table 1).

A final step in our investigation is to assess the impact of human capital on labor productivity growth and in order to do this we use a growth accounting framework.46 This exercise provides another opportunity to judge which of the two approaches to measuring human capital trends offers a more reasonable picture. The sources of labor productivity growth over the long-run are presented in Table 5, with labor quality estimates using the income-based approach and the resulting TFP estimates in Cols. (IV) and (V), and alternative estimates using the education-based approach and the subsequent TFP estimates in Cols. (VI) and (VII). Hence labor productivity trends are determined by human and physical capital/labor ratios and efficiency gains.A main finding is that TFP accounts for half the increase in labor productivity over 150 years considered, with broad capital (mostly physical capital) for the other half. Nonetheless, there is a clear divide between factor input accumulation as the dominant force (contributing two-thirds of labor productivity growth) pre-1950 — with the exception of the 1920s, and TFP then taking over, with a contribution of two-thirds of labor productivity growth in the Golden Age (1951–74) and about one-half during the democratic transition (1975–86), but losing its importance again in 1987–2000. As regards the contribution of human capital to labor productivity growth, we can observe that, with any measure, human capital contributed less than physical capital and TFP. Moreover, there are only slight differences between the contributions of alternative measures of human capital to labor productivity growth over the long-run. Hence, using alternative measures does not change the broad interpretation of the role of capital accumulation in productivity advance. In fact, differences between alternative measures of human capital are concentrated in short-run periods. A closer look shows that the contribution of human capital to labor productivity growth varies according to whether it is measured through the income-based or the education-based approach. For example, it is only in the 1920s and, particularly, in the last period considered l987–2000 that noticeable differentials in labor quality appear and have, therefore, an impact on TFP growth. In the 1920s, human capital would have contributed one-fourth of labor productivity growth according to the income-based approach estimate, but its contribution would have been negligible under the education-based approach. In 1987–2000, the latter's figure contribution to labor productivity growth was 57%, and the former only about 13%. However, over the period of fast productivity advance, 1953–86, the estimated contribution of human capital was twice as much in the income-based approach. As regards the implications of human capital alternative measures for TFP growth, it is worth noting that during the 1920s, a lower improvement in education-based estimates increases TFP growth to 1.4%, compared to the 1.1% obtained under the income-based approach, which would indicate that TFP contributed more than two-thirds and one-half of labor productivity growth, respectively. Conversely, between 1987 and 2000, the more intense labor quality gains in the education-based estimates suppress the TFP contribution to labor productivity growth (−0.3% compared to 0.2 using the income-based approach). The results for 1987–2000 cast doubt on the use of the education-based approach as a measure of human capital. With per capita income growth, the population invested more in education, but not all education was employed in the production side of the economy. Thus, any measure of human capital based on education would tend to overestimate the contribution of human capital to economic growth and, in consequence, to underestimate TFP growth. In Spain, although human capital contributed little to the increase in labor productivity growth, it could be the case that human capital accumulation had a “level” effect of facilitating technological innovation. In particular, the increase in human capital during the Golden Age correlates well with the spectacular rise in TFP growth rates, which was facilitated by the massive adoption of foreign technologies. However, further research will be required to test this proposition.