The study of international technology diffusion is essential for understanding how innovations in one country can drive growth elsewhere. Eaton and Kortum’s model addresses the complexities of this process, examining how new technologies are invented and shared among the world’s leading research economies. By focusing on factors such as productivity dynamics, patenting behaviour, and R&D efforts, they provide insights into how countries like the United States, Japan, Germany, the United Kingdom, and France contribute to global technological progress. This analysis also explores the barriers to technology diffusion and how removing them could enhance productivity worldwide.
What Is the Geographic Scope of Technical Progress?
Innovative countries are typically the most productive, yet their innovations also contribute significantly to growth in other regions. This essay explores a model of global growth driven by innovation, developed, estimated, and simulated by Eaton and Kortum. The microfoundation of their model leads to equations that describe productivity dynamics among countries simultaneously inventing and adopting each other’s innovations. They present two versions of this model: one where world growth is fully endogenous and another where growth is ‘semi-endogenous.’ The use of international patent data in the model provides insights into which countries generate ideas and where these ideas are most likely to be adopted.
Five Leading Research Economies
In the late 1980s, the United States, Japan, Germany, the United Kingdom, and France were considered the leading research economies. These countries employed over 80% of the OECD’s research scientists and engineers and accounted for about 60% of the world’s gross domestic product (GDP). Moreover, 70-80% of each country’s foreign patent applications originated from one of the other four, demonstrating a strong link between the generation and utilisation of innovations. Thus, these five countries represented the bulk of global inventive activity and the primary market for new ideas.
Features of the Data
The focus here is on the generation and dissemination of technology among the technological leaders. Given that these five economies accounted for a substantial share of global research activity and GDP, they provide a useful context for studying the dynamics of international technology diffusion. Each country’s level of productivity reflects not only its innovative efforts but also its ability to adopt foreign innovations. The link between sources and uses of innovations is captured by international patenting data, which is crucial for measuring technological diffusion.
The Model
The model aims to capture key aspects of the data to measure the role of innovation and international technology diffusion in economic growth. By incorporating the relationships between technology, research, patenting, and productivity within and across countries, the model becomes intricate yet insightful.
Production
The model considers a world comprising multiple countries. In each country, output at any given time is produced by combining intermediate inputs, using a Cobb-Douglas production function, which reflects a fixed proportion of inputs and consistent production practices across countries. Productivity growth is driven by research, which results in higher-quality inputs. However, the imperfect diffusion of these ideas contributes to differences in productivity between countries.
The Technological Frontier and Productivity
The model examines how research activities push the technological frontier forward, thereby improving input quality and productivity over time. Productivity growth within each country is influenced by domestic research efforts and the adoption of innovations from abroad. Variations in diffusion rates result in different levels of productivity among countries.
The Value of an Idea and the Decision to Patent
The model evaluates the economic value of new ideas and the factors influencing a firm’s decision to patent. It suggests that patenting serves as an indicator of technological diffusion, providing valuable information about where new ideas are most likely to be adopted and the barriers that may hinder their adoption.
The Return to R&D and Equilibrium R&D
The return to research and development (R&D) is explored by considering both the local and global benefits of innovation. The equilibrium level of R&D in each country is influenced by the potential returns, research costs, and the extent to which ideas diffuse internationally.
Technology, Wages, and Income
The model links technological progress to changes in wages and income distribution across countries. As technology advances, productivity improves, leading to higher wages and increased income levels. However, differences in technology diffusion rates contribute to the disparities in income among countries.
The Steady State
The economy reaches a steady state when key variables grow at a constant rate. This requires specific conditions, including steady workforce growth, proportional productivity levels among researchers, constant patenting costs relative to output, and a consistent interest rate and cost of capital across countries. In this steady state, relative productivity and growth rates are stabilised, reflecting the long-term dynamics of innovation and diffusion.
Steady-State Relative Productivities, Growth, and Patenting
In the steady state, countries reach a balance in productivity levels and growth rates. This balance is shaped by research efforts and the diffusion of technology. The rate of patenting activity indicates equilibrium in the labour market and provides a measure of how incentives for innovation align across countries.
Fitting the Model
The model is structured to incorporate both endogenous and exogenous variables. Endogenous variables include productivity growth, relative productivity, research efforts, and patents, while exogenous variables cover workforce size and patenting costs relative to GDP. By fitting the model to observed data, it becomes possible to simultaneously solve equations that account for productivity and patenting behaviour.
Estimation
In estimating the model, the authors introduce multiplicative measurement errors and convert the equations into logarithmic form. This method allows for the identification of essential parameters and provides a means to quantify productivity dynamics across countries.
Results
The findings include parameter estimates for both fully endogenous and semi-endogenous growth scenarios. The estimates for each scenario show a remarkable level of similarity. The model’s fit, evaluated under the fully endogenous growth scenario, closely matches observed data, suggesting that it accurately represents the dynamics of international technology diffusion.
Diffusion Lags and Adoption
The model interprets diffusion lags in three key ways:
– The average time between the invention of an idea in one country and its arrival in another, regardless of its adoption.
– The average time between invention and arrival, assuming the idea is adopted.
– The proportion of potentially useful ideas that are eventually adopted.
The Sources of Growth
The analysis reveals that the United States, Japan, and Germany are the main contributors to growth in each of the five leading economies. Together, the United States and Japan account for over 65% of growth, underscoring their significant impact on global innovation.
The Rewards to Research
The model suggests that the returns on innovation are more localised than the benefits derived from those innovations. Nonetheless, the United States represents a valuable market for ideas from all countries and provides over 80% of the market for its own innovations.
Counterfactuals
Using the model’s general equilibrium nature, the authors explore hypothetical scenarios with varying diffusion patterns and patent protections. They conduct two types of experiments:
1. In the fully endogenous growth case, the different scenarios lead to changes in steady-state growth rates.
2. In the semi-endogenous growth case, variations in productivity levels emerge, while the long-term growth rate remains unchanged.
Conclusion
The analysis demonstrates that the United States and Japan together drive at least two-thirds of the growth in each of the five leading economies. While these results may suggest that barriers to technology diffusion are relatively minor, they still account for significant productivity differences among countries. Removing international barriers to diffusion could not only bring productivity levels closer together but also raise overall productivity. Governments’ various policies, including tax incentives, research grants, and public research institutions, significantly impact innovation. Eaton and Kortum’s framework provides a valuable tool for evaluating the returns on such policy interventions.
REFERENCES
Eaton, Jonathan, and Samuel Kortum. “International Technology Diffusion: Theory and Measurement.” International Economic Review 40, no. 3 (1999): 537-570.