2015年-世界发展银行全球_Exploring_the_Relationship_between_Broadband_and_Economic_Growth_21页_1mb

Summary of "Exploring the Relationship Between Broadband and Economic Growth"

Core Content

This background paper explores the relationship between broadband and economic growth, analyzing the impact of both fixed and mobile broadband on GDP across different countries and economic development levels. It reviews various econometric models and discusses the challenges in measuring broadband's economic effects.

Main Views

1. Broadband as a General-Purpose Technology

Broadband is characterized as high-speed, always-on Internet connectivity. It has driven significant changes in the IT sector and is influencing innovation in other sectors like health, transport, and government. Its economic impact is a growing area of interest for economists.

2. Broadband Impact on GDP Growth

• Fixed Broadband: Studies show that a 10 percentage point increase in fixed broadband penetration leads to an increase in GDP growth, with varying magnitudes across countries and time periods.
  • In developed economies, the impact is estimated at 1.21%.
  • In developing economies, the impact is estimated at 1.38%.
  • More recent studies using updated data (Scott 2012) found a 1.35% increase in developing countries and 1.19% in developed ones.
  • A non-linear model (Zaballos and López-Rivas 2012) estimated a 3.19% increase in per capita GDP for a 10% rise in fixed broadband penetration in Latin America and the Caribbean.
• Mobile Broadband: The impact is more mixed.
  • In some studies, mobile broadband has a negative or insignificant impact on GDP per household.
  • A study by Thompson and Garbacz (2011) found that a 10 percentage point increase in mobile broadband penetration reduces GDP per household by 0.52%.
  • In contrast, a study by Deloitte (2012) found that a 10% increase in 3G penetration leads to a 0.15 percentage point increase in GDP per capita growth.
  • Mobile data usage was also found to raise GDP by 0.5 percentage points when doubled in 14 countries.

3. Country-Specific Studies

Several country-level studies have been conducted using different models and data sources:

• China: A simultaneous equation model found that a 10% increase in fixed broadband penetration leads to a 2.14% increase in real GDP.
• Germany: A study using administrative data found a 0.255% increase in GDP growth for a 10% increase in broadband penetration growth.
• United States: A study using non-traditional variables like temperature and tax rates found a 0.161% increase in GDP for a 10% rise in broadband penetration, but the result was not statistically significant.
• Ecuador: A micro-level study found that broadband access increases household income by $25.76 per year (3.67%).
• Panama: A 10% increase in fixed broadband penetration was associated with a 0.44% increase in GDP.
• Philippines: Fixed broadband had no significant impact, while mobile broadband contributed 0.32% to GDP growth.
• Senegal: Fixed broadband had a negative impact, but mobile broadband had a positive and significant impact, raising GDP growth by 0.44%.

4. Model Variations

• Cross-sectional models are used to analyze the impact across multiple countries.
• Panel models track changes over time within a group of countries.
• Time series models focus on the impact within a single country.
• Some studies use non-linear models, while others rely on logarithmic transformations or instrumental variable approaches to address endogeneity and model specification issues.

5. Challenges in Analysis

• Endogeneity: Broadband adoption is often influenced by economic conditions, which can distort the causal relationship.
• Model Specification: There is no consensus on the best model to use, with variations in variable selection and econometric techniques.
• Data Limitations: Data availability affects the choice of model, with cross-sectional studies being common for groups of countries and time series for single countries.
• Causation: Most studies do not establish clear causality and may suffer from omitted variable bias or reverse causality.
• Threshold Effects: Broadband's economic impact is generally noticeable only after a certain level of penetration, though the exact threshold is unclear.
• Diminishing Returns: Some evidence suggests that the impact may increase with higher penetration due to network effects, while other studies are inconclusive on this point.

Key Information

Broadband Penetration Levels

• By 2014, there were 748 million fixed broadband subscriptions and 2.7 billion mobile broadband subscriptions globally.
• In developing countries, mobile broadband penetration was over four times higher than fixed broadband penetration.

Economic Impact Magnitudes

Methodological Considerations

• Instrumental variable approaches are used to address endogeneity.
• Logarithmic transformations are common in econometric models.
• Control variables such as capital, labor, education, and government spending are often included.
• Time lags and threshold effects are considered in some models.

Recommendations

The paper concludes with the following recommendations:

• Further research is needed to better understand the causal relationship between broadband and economic growth.
• More robust models should be developed to account for endogeneity and model specification issues.
• The choice of variables and data sources should be standardized to improve the reliability of estimates.
• There is a need for longitudinal data and more detailed indicators to capture the nuanced impact of broadband on different sectors and populations.

Conclusion

Broadband has been shown to have a positive economic impact, particularly in developed economies. However, the magnitude and significance of this impact vary across countries and models. Mobile broadband's effect is more mixed, with some studies suggesting it may have non-productive applications or negative impacts. The lack of a unified framework and data limitations continue to pose challenges in accurately measuring broadband's economic contributions.