Development Economics
How Technology, Innovation, and Development Spread Across Economies
Diffusion in development economics refers to the process by which innovations, technologies, ideas, and practices spread across individuals, firms, regions, and nations over time. This phenomenon is central to understanding economic growth, convergence, and the persistence of global inequality.
The study of diffusion bridges multiple disciplines—economics, sociology, geography, and technology studies—to explain why some innovations spread rapidly while others languish, why some countries adopt new technologies quickly while others lag decades behind, and how policy interventions can accelerate beneficial diffusion.
The intellectual history and key concepts underlying diffusion theory
Everett Rogers published his seminal work synthesizing over 500 diffusion studies. He identified five adopter categories—innovators (2.5%), early adopters (13.5%), early majority (34%), late majority (34%), and laggards (16%)—following a normal distribution curve.
Frank Bass formalized the mathematical structure of diffusion, distinguishing between "innovators" (influenced by external communication) and "imitators" (influenced by word-of-mouth). The Bass model became foundational for forecasting technology adoption.
f(t) = (p + q·F(t))(1 - F(t))
where p = innovation coefficient, q = imitation coefficient, F(t) = cumulative adoption
Paul Romer and Robert Lucas formalized how knowledge spillovers and technology diffusion drive long-run economic growth. Their models showed that ideas are non-rival goods—one country's use doesn't diminish another's—making diffusion the key mechanism for global convergence.
Daron Acemoglu, Melissa Dell, and others documented how institutional barriers, skill mismatches, and "appropriate technology" constraints create persistent technology gaps. Technologies designed for high-skill, capital-abundant economies may not be optimal for developing countries.
The maximum level of technology or productivity achievable given current global knowledge. Developed nations typically operate near the frontier, while developing nations face a "distance to frontier" that shapes their growth strategies.
A country's ability to recognize, assimilate, and apply external knowledge. Determined by human capital, institutional quality, R&D investment, and existing technological base. Low absorptive capacity is a primary barrier to technology transfer.
The value of many technologies increases with the number of users. Network effects create positive feedback loops accelerating diffusion past critical thresholds, but can also create lock-in to inferior technologies.
The phenomenon where developing countries skip intermediate technologies, jumping directly to more advanced solutions. Classic examples include mobile phones bypassing landlines and mobile banking bypassing traditional banking infrastructure.
Technology adoption generates knowledge spillovers—learning that benefits non-adopters. Foreign Direct Investment (FDI) and trade serve as key channels for international technology spillovers, though evidence on magnitude is mixed.
The time between technology invention and widespread adoption varies dramatically across countries and technologies. These lags—often 15-50 years—explain much of the cross-country income divergence observed over the past two centuries.
Formal frameworks for analyzing and predicting technology spread
Technology adoption typically follows a sigmoid (S-shaped) curve characterized by three phases:
The steepness and timing of the S-curve vary significantly across technologies and countries, providing key parameters for policy analysis.
Comin and Hobijn's extensive dataset on technology adoption reveals systematic patterns:
dA/dt = β · A(t) · (N - A(t))
Adoption (A) grows proportionally to contact between adopters and non-adopters. Parameter β captures diffusion speed.
P(adopt) = Φ((π - c)/σ)
Adoption probability depends on profitability (π), cost (c), and heterogeneity (σ). Used for analyzing agricultural technology adoption.
Adopt if: E[B|adopt] > c + θ·(1 - n(t))
Agents adopt when expected benefits exceed costs plus a social threshold (θ) that decreases with adoption share n(t).
Ȧ/A = λ·(A* - A)/A + γ·(A*/A)^φ
Domestic technology grows via catch-up to frontier (A*) and autonomous innovation. Distance to frontier affects growth rate.
Data-driven insights on global technology diffusion patterns
Data: World Bank World Development Indicators, ITU. Shows years from technology introduction to 25% adoption.
Mobile subscriptions per 100 people, 1990-2022
Internet users as % of population, 2000-2023
Comin & Hobijn (2010) found that the average adoption lag fell from 46 years for technologies invented before 1900 to 17 years for post-1975 technologies.
Klenow & Rodriguez-Clare (1997) estimate that technology differences account for 25-40% of cross-country income variation, with the remainder attributed to capital and human capital.
Green Revolution technologies doubled or tripled yields in adopting regions. Evenson & Gollin (2003) estimate 21% of total crop production increases from 1960-2000 attributable to international research.
By 2022, mobile money systems processed over $1 trillion annually, with 1.6 billion registered accounts globally, demonstrating rapid financial technology diffusion in developing markets.
| Technology | Invention Year | High Income | Middle Income | Low Income | Gap (Low vs High) |
|---|---|---|---|---|---|
| Steam Power | 1769 | 25 years | 65 years | 120 years | 95 years |
| Railways | 1825 | 15 years | 45 years | 80 years | 65 years |
| Electricity | 1882 | 12 years | 35 years | 60 years | 48 years |
| Telephone | 1876 | 20 years | 50 years | 90 years | 70 years |
| Automobile | 1886 | 18 years | 40 years | 75 years | 57 years |
| Television | 1926 | 20 years | 35 years | 55 years | 35 years |
| Personal Computer | 1975 | 12 years | 22 years | 35 years | 23 years |
| Mobile Phone | 1983 | 10 years | 15 years | 22 years | 12 years |
| Internet | 1991 | 8 years | 15 years | 25 years | 17 years |
Source: Adapted from Comin & Hobijn (2010), "An Exploration of Technology Diffusion," American Economic Review. Years to 25% adoption threshold.
Real-world examples of technology diffusion in development
Launched by Safaricom in Kenya, M-Pesa enabled mobile phone-based money transfers and payments without traditional banking infrastructure. The service spread through existing social networks—users could send money to non-users, who then had incentive to register.
Success factors included: permissive regulation, strong agent network, leveraging existing trust in Safaricom brand, and network effects from compatibility with non-users. Attempts to replicate M-Pesa elsewhere have had mixed results, highlighting context-dependence of technology diffusion.
High-yielding varieties (HYVs) of wheat and rice, developed through international research (CGIAR system), diffused across Asia and Latin America. Norman Borlaug's semi-dwarf wheat varieties transformed agriculture in Mexico, India, and Pakistan.
Successful diffusion required complementary inputs (irrigation, fertilizer, credit) and institutional support (extension services, procurement systems). Africa largely missed the Green Revolution due to inadequate infrastructure and different crop systems, illustrating that technology diffusion depends heavily on context and supporting investments.
India's emergence as a global IT services hub represents successful diffusion of ICT capabilities. Beginning with software maintenance and data entry in the 1980s, the industry progressively moved up the value chain to complex software development and business process management.
Key enabling factors included: English language education, investments in technical institutes (IITs), diaspora connections to Silicon Valley, and time zone advantages. The case illustrates how absorptive capacity, built through education investments decades earlier, enables technology diffusion.
South Korea's transformation from war-devastated agrarian economy to technological leader exemplifies successful technology absorption and diffusion. Strategic industrial policy guided technology acquisition through licensing, reverse engineering, and eventually indigenous innovation.
Korea combined technology importation with massive human capital investments, creating absorptive capacity for progressively complex technologies. Government played active role in coordinating technology acquisition, providing patient capital, and protecting infant industries. Raises questions about replicability given different global trade regimes today.
Evidence-based strategies for accelerating beneficial technology diffusion
Investments in education and human capital are prerequisites for technology absorption. Returns to technology transfer are much higher when recipient capacity exists.
Technologies rarely diffuse in isolation. Successful adoption requires infrastructure, institutions, and supporting services. Bundle interventions addressing multiple constraints.
Pilot programs and early adopter experiences generate information that reduces uncertainty and accelerates subsequent adoption. Support information dissemination.
Frontier technologies may not be optimal for all contexts. Support adaptation and development of context-appropriate solutions that match local factor endowments.
Should governments pick winners and direct technology acquisition (East Asian model), or focus on horizontal policies creating enabling conditions? Evidence suggests context matters—active policies worked in Korea but failed elsewhere. Newer "smart" industrial policy approaches emphasize experimentation and learning.
Strong IP protection may incentivize innovation but can impede technology diffusion to developing countries. Debates over TRIPS flexibilities, compulsory licensing, and technology transfer continue. COVID-19 vaccine distribution renewed focus on this tension.
Digital technologies offer leapfrogging opportunities but also risk creating new divides. Automation may reduce developing countries' labor cost advantages. Policy must navigate harnessing digital opportunities while managing disruption risks.
The intellectual pioneers of diffusion economics
Father of diffusion research. His 1962 book "Diffusion of Innovations" synthesized findings across disciplines, establishing the S-curve model and adopter categories that remain foundational.
Pioneer of econometric analysis of technology diffusion. His 1957 study of hybrid corn adoption demonstrated how economic factors (profitability, farm size) explain diffusion patterns.
Developed endogenous growth theory showing how ideas and knowledge spillovers drive long-run growth. Demonstrated that technology diffusion is the key mechanism for global convergence.
Created comprehensive datasets on technology adoption across countries and time. Research quantifies adoption lags and identifies determinants of cross-country diffusion differences.
Analyzed how institutions shape technology adoption and innovation direction. Work on "appropriate technology" explains why frontier technologies may not diffuse to developing countries.
Pioneered randomized controlled trials in development economics. Work on technology adoption among the poor revealed behavioral and informational barriers beyond standard economic factors.