We examine the impact of a comprehensive set of measures of human capital on recently created, direct measures of technology adoption using country-level panel data for the period 1964–2003, covering a wide range of technologies in various sectors of the economy. We consider many dimensions of human capital, using both qualitative and quantitative measures, as well as indirect measures that capture the role of “learning by doing” intrinsic to the process of technological diffusion. Our analysis, which examines the human capital and technological diffusion link more comprehensively relative to previous studies, suggests that the link is a conditional one, resting on various aspects of human capital and the nature of the technology in question. Overall, the results suggest that the type of human capital that is formed via the learning-by-doing mechanism may be the most important determinant of technological diffusion, followed by, to a substantially less degree, qualitative determinants such as cognitive skills (measured using test scores) and quantitative or other measures (such as years of schooling and life expectancy). Our conclusions are robust to the inclusion of institutional variables and other factors that determine technological diffusion.
This is a preview of subscription content, access via your institution.
Buy single article
Instant access to the full article PDF.
Tax calculation will be finalised during checkout.
Subscribe to journal
Immediate online access to all issues from 2019. Subscription will auto renew annually.
Tax calculation will be finalised during checkout.
The Cross Country Historical Adoption of Technology (CHAT) data set captures both the extensive and intensive margins of 104 technologies from 8 sectors for a sample of more than 150 countries, over a period of 1800–2000.
Our paper complements the earlier work by Cinnirella and Streb (2017) who explore the association between different types of quantitative measures of human capital and technological innovation for Prussia in the late nineteenth century, in addition to “training-on-the-job”, an aspect which our results suggest is a very important determinant of technological diffusion. In addition, our paper relates to the literature on growth and innovations which employs measures of human capital such as educational attainment and knowledge as among key factors impacting upon the innovation process and growth of economies (Drivas et al. 2018; Dakhli and De Clercq 2004; Barro 2001). In contrast, however, we employ qualitative constructs of human capital such as cognitive skills. Furthermore, our analysis is more comprehensive in scope, looking at various direct measures of technologies over time and across countries.
This may be justifiable in the sense that the mathematics test consists of basic mathematical knowledge applied to set of analytical problems. The science test, in contrast, is more knowledge specific rather than analytical. Of course, this may be contentious and the reader may not agree with our interpretation. Our choice of the labels “specific” and “generic”, however, proves convenient as well as intuitive in the context of discussing and interpreting the results to follow. Our variable embodying generic skills is a proxy for analytical skills of a very general level.
In addition to our reasoning above, Comin et al (2008) suggest that past level of technology adoption is a strong predictor of current levels; as such a dynamic specification is appropriate. In Comin and Hobijn (2004), which to our knowledge is the only other study analyzing the impact of human capital on technology measures based on the CHAT data set, the lagged variable is not considered and the focus is on quantitative measures of human capital such as secondary school enrollment.
Comin and Hobijn (2009a, b) refer to the definition of “technology” in Merriam-Webster’s Collegiate Dictionary. It defines technology as “a manner of accomplishing a task especially using technical processes, methods, or knowledge”. Given this definition the basic idea behind technological measures in CHAT is to cover these various aspects of technology. For example, it includes the quantity of capital goods required to achieve a specific task (e.g., number of sail ships (measured in tonnage) in use in a country), amounts of times a specific task that have been completed (e.g., metric tonnes of steel produced) and the number of users of a the specific manner in which the task was accomplished (e.g., number of subscribers of cable TV in a household).
Comin and Mestieri (2013) use a different theoretical construct for intensive margin of technology in their theoretical framework.
Comin and Mestieri (2013) suggest use of population or Gross Domestic Product as scaling factors.
As indicated earlier, our sample includes a set of 50 countries and includes both developed and developing countries. Here the choice of countries to illustrate the concept of technological lags is motivated purely by the ease of graphical presentation.
The measure developed in Hanushek and Woessmann (2012) is an extension of Hanushek and Kimko (2000). Details for countries and tests are present in Hanushek and Woessmann (2012). The observations for mathematics and science test scores are sporadically spread across the time period 1963–2003. This is because, firstly, the tests are not conducted every year. Secondly, the countries may not participate in every test that is conducted. Hence, we have missing observations for mathematics and science test scores in our data set. In order to extrapolate observations for these years, we use averages of the available test scores for the countries in our sample.
Details of this qualitative measure of education, from Hanushek and Woessmann (2012), are also presented in part A of the online supplementary appendix.
It is possible that there are some technologies where other countries could be leaders. However, US leads in most cases and is hence chosen for consistency, as well as comparability with the Comin et al (2008) approach to measure usage lags of technologies.
In this estimation procedure, we instrument current variables at time t by their past lags, which eliminates correlation between explanatory variables and the error term. Furthermore, in our GMM estimations the use of these past lags as instruments may (effectively) control for the possible endogeneity of human capital acquisition employed as explanatory variable for adoption and diffusion of technology. However, we do not explicitly control for endogeneity of human capital due to our assumption that it is strictly exogenous, as per the assumptions underlying the difference-GMM approach. Given that the left-hand-side variable is a microeconomic, sectoral variable while our human capital measures are macroeconomic in flavor we believe this is a reasonable assumption. However, we perform robustness checks based on various diagnostic tests and use the system GMM approach of Blundell and Bond (2000) for all cases. The results for these are presented in part B of the online appendix.
We obtain data for life expectancy for the years 1964–2003 from World Development Indicators (WDI) of the World Bank (2015). It is measured as life expectancy at birth in total years. See www.worldbank.org for details.
We thank two anonymous referees for these suggestions.
A complete sector-wise overview of these results including a larger set of technologies is available on request. The more succinct presentation of these results in the form of tables in the paper does not affect the overall findings and interpretation of the analysis.
In the sector-wise analysis, presented in the online appendix, we have 21 technologies each in the mathematics and science panels. The coefficient of mathematics and science skills is significant in 10 and 5 intensity of usage of technologies, respectively. Hence, we could suggest that in a broad sense a labor force embodied with mathematics skills is more suitable for improving the adoption of technology relative to on embodied with science skills. However, there are obvious caveats to such an interpretation as we would expect specialized skills, more directly measured, to impact in sectors where they were relevant. We discuss this caveat in further detail at the end of this section.
While such technologies do not require mathematics skills per se, their prevalence requires human capital in the form of qualified technicians and engineers to provide maintenance and technical support service, who are typically endowed with such skills. It is in this sense that we suggest that the generic nature of mathematics skills is relevant. Following Hanushek and Kimko (2000), we interpret these measures as an indirect proxy of the quality of the labor force of an economy.
Other factors, such as property rights associated with technological transfer, may be of relevance, as suggested by Spielman and Ma (2015).
For example, World Development Indicators (WDI) do not have such indicators for health. However, some country-specific studies do provide this information from their respective national databases (Ceppa et al. 2012).
Supplementary appendix B contains descriptive evidence regarding data used for analysis. A complete sector-wise overview of these results including a larger set of technologies is available on request. The more succinct presentation of these results in the form of tables in the paper does not affect the overall findings and interpretation of the analysis.
We use a system GMM approach given the persistent nature of the lagged variable; the Sargan test in the difference-GMM approach also suggested inappropriate identifying restrictions.
Acemoglu D, Zilibotti F (2001) Productivity differences. Q J Econ 116(2):563–606
Acemoglu D, Johnson S, Robinson JA (2005) Institutions as a fundamental cause of long-run growth. Handbook Econ Growth 1:385–472
Ainsworth M, Over M (1994) AIDS and African development. World Bank Res Obs 9(2):203–240
Aitken BJ, Harrison AE (1999) Do domestic firms benefit from direct foreign investment? Evidence from Venezuela. Am Econ Rev 89(3):605–618
Arellano M, Bond S (1991) Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. Rev Econ Stud 58(2):277–297
Barro RJ (1998) Human capital and growth in cross-country regressions. Harvard University, Cambridge
Barro RJ (2001) Human capital and growth. Am Econ Rev 91(2):12–17
Barro RJ (2013) Health and economic growth. Ann Econ Finance 14(2):329–366
Barro RJ, Lee JW (2010) A new data set of educational attainment in the world, 1950–2010. J Dev Econ 104:184–198
Barro RJ, Lee JW (2013) A new data set of educational attainment in the world, 1950–2010. J Dev Econ 104:184–198
Barro RJ, Sala-I-Martin X (1995) Economic growth theory. Mac Graw-Hill, New York
Basu S, Weil DN (1998) Appropriate technology and growth. Q J Econ 113(4):1025–1054
Benhabib J, Spiegel MM (1994) The role of human capital in economic development evidence from aggregate cross-country data. J Monet Econ 34(2):143–173
Blundell R, Bond S (2000) GMM estimation with persistent panel data: an application to production functions. Econom Rev 19(3):321–340
Branstetter L (2006) Is foreign direct investment a channel of knowledge spillovers? Evidence from Japan’s FDI in the United States. J Int Econ 68(2):325–344
Ceppa DP, Kosinski AS, Berry MF, Tong BC, Harpole DH, Mitchell JD, Onaitis MW (2012) Thoracoscopic lobectomy has increasing benefit in patients with poor pulmonary function: a Society of Thoracic Surgeons Database analysis. Ann Surg 256(3):487
Cinnirella F, Streb J (2017) The role of human capital and innovation in economic development: evidence from post-Malthusian Prussia. J Econ Growth 22(2):193–227
Comin D, Hobijn B (2004) Cross-country technology adoption: making the theories face the facts. J Monet Econ 51(1):39–83
Comin DA, Hobijn B (2009a) The CHAT dataset. National Bureau of Economic Research, Cambridge
Comin D, Hobijn B (2009b) Lobbies and technology diffusion. Rev Econ Stat 91(2):229–244
Comin D, Mestieri M (2013) Technology diffusion: measurement, causes and consequences. NBER Working Paper. 19052
Comin D, Hobijn B, Rovito E (2008) Technology usage lags. J Econ Growth 13(4):237–256
Conley T, Udry C (2001) Social learning through networks: the adoption of new agricultural technologies in Ghana. Am J Agric Econ 83(3):668–673
Conley TG, Udry CR (2010) Learning about a new technology: pineapple in Ghana. Am Econ Rev 100(1):35–69
Dakhli M, De Clercq D (2004) Human capital, social capital, and innovation: a multi-country study. Entrepreneurship Reg Dev 16(2):107–128
Drivas K, Economidou C, Tsionas EG (2018) Production of output and ideas: efficiency and growth patterns in the United States. Reg Stud 52(1):105–118
Foster AD, Rosenzweig MR (1995) Learning by doing and learning from others: human capital and technical change in agriculture. J Polit Econ 103(6):1176–1209
Hanushek EA, Kimko DD (2000) Schooling, labor-force quality, and the growth of nations. Am Econ Rev 90(5):1184–1208
Hanushek EA, Woessmann L (2012) Do better schools lead to more growth? Cognitive skills, economic outcomes, and causation. J Econ Growth 17(4):267–321
Herzer D (2011) The long-run relationship between outward foreign direct investment and total factor productivity: evidence for developing countries. J Dev Stud 47(5):767–785
Hulten CR (2000) Measuring innovation in the New Economy. Unpublished paper, University of Maryland
Jamison DT, Lau LJ, Wang J (1998) Health’s contribution to economic growth, 1965–90. Health, health policy and economic outcomes, 61–80
Jovanovic B, Nyarko Y (1996) Learning by doing and the choice of technology. Econometrica 64(6):1299–1310
Krueger AB, Lindahl M (2001) Education for growth: why and for whom? J Econ Lit 39(4):1101–1136
Lahiri R, Ding J, Chinzara Z (2018) Technology adoption, adaptation and growth. Econ Model 70:469–483
Li X, Liu X, Parker D (2001) Foreign direct investment and productivity spillovers in the Chinese manufacturing sector. Econ Syst 25(4):305–321
Lipsey RG, Carlaw KI (2004) Total factor productivity and the measurement of technological change. Can J Econ/Revue canadienned’ économique 37(4):1118–1150
Madsen JB (2014) Human capital and the world technology frontier. Rev Econ Stat 96(4):676–692
Messinis G, Ahmed AD (2013) Cognitive skills, innovation and technology diffusion. Econ Model 30:565–578
Meyer KE, Sinani E (2009) When and where does foreign direct investment generate positive spillovers? A meta-analysis. J Int Bus Stud 40(7):1075–1094
Nelson RR, Phelps ES (1966) Investment in humans, technological diffusion, and economic growth. Am Econ Rev 56(1/2):69–75
Parente SL, Prescott EC (1994) Barriers to technology adoption and development. J Polit Econ 102(2):298–321
Sinani E, Meyer KE (2004) Spillovers of technology transfer from FDI: the case of Estonia. J Comp Econ 32(3):445–466
Spielman DJ, Ma X (2015) Private sector incentives and the diffusion of agricultural technology: evidence from developing countries. J Dev Stud 52(5):696–717
Sun S (2011) Foreign direct investment and technology spillovers in China’s manufacturing sector. Chin Econ 44(2):25–42
Vandenbussche J, Aghion P, Meghir C (2006) Growth, distance to frontier and composition of human capital. J Econ Growth 11(2):97–127
We would like to thank Vincent Hoang, Janice How, Sandy Suardi, Shrabani Saha, Peter Siminski and participants at various seminars and conferences for thoughtful discussions and comments. We take responsibility for any errors.
This study is not funded by any grant.
Conflict of interest
The authors declare that they have no conflict of interest.
This article does not contain any studies with human participants or animals performed by any of the authors.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
Below is the link to the electronic supplementary material.
About this article
Cite this article
Asif, Z., Lahiri, R. Dimensions of human capital and technological diffusion. Empir Econ 60, 941–967 (2021). https://doi.org/10.1007/s00181-019-01777-3
- Cognitive skills
- Economic growth
- Educational achievements
- Educational attainments
- Human capital