Diego Peréz-González
- 2 June 2026
- WORKING PAPER SERIES - No. 3242Details
- Abstract
- This paper examines whether differences in the composition of investment help explain economic growth disparities in the EU and other advanced economies from 1996 to 2021. While overall investment levels in the EU and the US are broadly similar, the EU invests less in intangible and tangible ICT capital. This difference in composition is associated with part of the EU’s productivity gap with the US. Employing panel fixed effects and local projection methods, we find that intangible and tangible ICT investments -particularly in communications equipment, R&D, and other intellectual property products- are associated with higher GDP per capita growth than other forms of investment. To quantify these differences, we construct a novel investment efficiency ratio that relates the estimated economic growth contribution of each asset to its share in total investment. The results are robust across empirical methods, country samples, and time periods, and reveal substantial heterogeneity: the growth association of ICT-related investment is stronger in countries with higher income levels and greater human capital. Overall, the findings suggest that improving the allocation and efficiency of investment, rather than simply increasing its volume, is key to enhancing long-term growth.
- JEL Code
- E22 : Macroeconomics and Monetary Economics→Consumption, Saving, Production, Investment, Labor Markets, and Informal Economy→Capital, Investment, Capacity
O47 : Economic Development, Technological Change, and Growth→Economic Growth and Aggregate Productivity→Measurement of Economic Growth, Aggregate Productivity, Cross-Country Output Convergence
O50 : Economic Development, Technological Change, and Growth→Economywide Country Studies→General
J24 : Labor and Demographic Economics→Demand and Supply of Labor→Human Capital, Skills, Occupational Choice, Labor Productivity
C23 : Mathematical and Quantitative Methods→Single Equation Models, Single Variables→Panel Data Models, Spatio-temporal Models