Skills-biased production makes highly-skilled workers more efficient in wealth

Skills-biased production makes highly-skilled workers more skilled in rich countries

The availability of highly educated labor varies dramatically between rich and poor countries. Consider, for example, India and the United States. In 2010, with some third education, the share of the Indian working population was 9%, compared to 56% in the United States (Barro and Lee 2013). At the same time, the difference in wages premiums for highly educated labor seems small. According to estimates collected by Casselli et al. (2014), an extra year of schooling in both India and the US is associated with a wage increase of about 10%; More generally, returning to school is only poorly related to development.

Combining large differences in the amount of skilled labor and small differences in efficiency premiums, previous works have concluded that the relative skills of highly-skilled workers increase with development (Caselli and Coleman 2006). The reasoning of this argument is simple: under standard estimates, the abundance of skilled workers, everything else being equal, should reduce their relative wages in rich countries, because the services they create are more widely available. The lack of this pattern in the data indicates that the downward pressure on efficiency premiums is offset by the fact that skilled workers in rich countries are compensated by being more productive, which increases their wages.

Literary critics point to two possible reasons for this. On the one hand, productive environments in rich countries may be more ‘efficient-biased’ as firms adopt more suitable technologies for skilled workers, as proposed by Casselli and Coleman (2006), or, more generally, because of institutional arrangements or higher sectoral structures in these countries. -Favorable for skilled labor. On the other hand, as in Jones (2014), highly skilled workers can embody more human capital in rich countries because of differences in educational standards, training or the underlying characteristics of the workers. The difference between these two interpretations is crucial for development accounting: if the ‘human capital approach’ is correct, then the total cross-country diversity in GDP per worker can be calculated by the higher human capital in highly educated labor; Under the alternative approach, the contribution of relative human capital is much lower (Casselli and Secon 2018).

Measuring relative efficiency with micro data

In recent work, I have reconsidered both the measurement and interpretation of the relative efficiency gap across the country (Rossi, 2022). Using micro-level data from 12 countries at different levels of development, I make a comparative arrangement of relative supply of workers and wage premiums with some third education. Unlike previous work, these measures incorporate cross-country variations in employment rates and working hours, and are based on comparative wage data as opposed to the collection of estimated income for schools from different sources. Based on these measures and assuming an ideal production function, I calculate the inherent relative efficiency of high-skilled labor for each country. As shown in Figure 1, there is a strong correlation between GDP per worker. The productivity gap between high- and low-skilled labor is 20 times greater in the United States than in India, with a similar disparity in GDP per worker between the two countries. For the most part, relative efficiency is not driven by differences in cross-country dispersal sectoral structure, self-employment events, or gender and experience returns.

Figure 1

Comments: The plots in the figure log the relative efficiency and GDP per worker for 12 countries with the available micro data. Relative Skills Skills It is normal for the US to take a value of 1 (login 0). The solid line represents the best linear fit.

Explaining Relative Skills Skills: Evidence from International Immigrants

Why are highly-skilled workers more productive in rich countries? I use the premier diversity of skills of international immigrants to shed light on this question. Intuitively, foreign-educated migrants employed in the same labor market are subject to an equally efficient-biased production environment, but have different levels of human capital depending on the quality and characteristics of the educational environment in their country of origin. Comparing the premiers of skills across different nationalities within the same host country allows to distinguish cross-country differences between the relative human capital of high-skilled labor and the bias of production efficiency.

Figure 2 shows the skill premium of U.S. immigrants against the GDP per worker in the country of origin. High-skilled immigrants from wealthier countries receive relatively higher wages in the United States, consistently higher in their relative human capital. However, these cross-national differences are smaller than the cross-country gaps in the relative efficiency skills shown in Figure 1. Indeed, this evidence (combined with similar patterns from other host countries) leads me to conclude that the difference in overall relative efficiency of less than 10% can be explained by the human capital of highly educated labor. For the most part, it is the productive environment that makes high-skilled labor comparatively more efficient in rich countries.

Figure 2

Comments: The figure plots the premium of log skills across the country of origin of U.S. immigrants as opposed to the per capita GDP per worker in the country of origin. The solid line shows the best linear fit.

Implications for cross-country differences in human capital

How can we reconcile these findings with the recent literature that finds a large contribution of human capital in accounting for cross-country income differences (Hendricks and Schweilmann 2018)? My findings in Rossi (2022) suggest that cross-country variation in human capital gap between high- and low-skilled workers is somewhat limited. This means that there is a great need for a large contribution of human capital for income differentiation Uniform Gap across the skill level – that is, all workers, regardless of the skill level, have more human capital in rich countries. This echoes the recent work of Hendricks and Schoellman (2022), which shows that wage gains for high and low-skilled immigrants in the United States are consistent with significant human capital gaps for both groups.

This conclusion places significant limitations on the theory of human capital accumulation and economic development. The large cross-country gaps in human capital for all educational levels urge all workers to pay attention to the factors that affect the structure of human capital, which becomes a natural candidate for explaining the cross-country diversity of human capital. These include the quality of primary education (Fazio et al. 2020), family input (de Phillips and Rossi 2020), cultural characteristics (Hanushek et al. 2020), and opportunities to acquire skills in the life cycle (Lagakos et al. 2017).

What makes the productive environment in rich countries more efficient-biased?

The findings of my research paper open the question of why and through which channels production is more efficient-biased in rich countries. A popular view is that the difference in skills bias is the inherent response to high-skilled labor availability: since highly educated workers abound in rich countries, companies in such countries provide strong incentives for these workers to adopt complementary technologies. The notion that technology adoption factor responds to availability has gained empirical support in a variety of settings (e.g. Beaudry et al. 2010).

Also, cross-country differences in institutional standards, the organization of production, and the proliferation of large and modern corporations may benefit disproportionately high-skilled workers in rich countries, perhaps contributing to explain why more individuals choose to acquire skills. In those countries. Documenting and measuring these channels is important for future work.


Barro, RJ and JW Lee (2013), “A New Data Set of Educational Atainment in World, 1950-2010,” Journal of Development Economics 104 (C): 184-198.

Beaudry, P, M Doms and E Lewis (2010), “Should Personal Computers Be Considered a Technological Revolution? Evidence from the US Metropolitan Area,” Journal of Political Economy 118 (5): 988-1036.

Caselli, F and A Ciccone (2018), “Incomplete Replacement and Development Accounting,”, 9 June.

Caselli, F. and WJ Coleman (2006), “The World Technology Frontier,” American Economic Review 96 (3): 499–522.

Casselli, F., J. Pontiselli and F. Rossi (2016), Minsarian Returns covers a new data set on technological differences across space and timePrinceton University Press.

De Philippis, M and F Rossi (2020), “Parental Influence on Human Capital and the Cross-Country Gap,”, 3 November.

Fazzio, I, A Eble, R Lumsdaine, P Boone, B Bouy, Pt J Hsieh, C Jayanty, S Johnson and F Silva (2020),, 16 December.

Hanushek, EA, L Kinne, P Lergetporer, and L Woessmann (2020), “International Differences in Patience, Risk Taking, and Student Achievement”,, 2 August.

Hendricks, L and T Schoellman (2018), “Human Capital and Development Accounting: New Evidence of Wage Gain in Migration,” Quarterly Journal of Economics 133 (2): 665–700.

Hendricks, L and T Schoellman (2020), “The Difference Between Skilled Labor Productivity and Cross-Country Income,” Working Paper, May.

Jones, BF (2014), “The Human Capital Stock: A Generalized Approach,” American Economic Review 104 (11): 3752–3777.

Lagakos, D, B Moll, T Porzio, N Qian, and T Schoellman (2017), “Life-cycle wage growth across the country,” Journal of Political Economy.

Rossi, F (2022), “The relative efficiency of skilled labor throughout the country: measurement and interpretation,” American Economic Review 112 (1): 235-266.

Leave a Reply

Your email address will not be published.