IMF-加密、腐败和资本控制：跨国关联（英）-2022.3-19页_1mb

Summary: Crypto, Corruption, and Capital Controls: Cross-Country Correlations

Core Content

This IMF Working Paper investigates the relationship between crypto-asset usage and several macroeconomic and institutional indicators across 55 countries. The analysis is based on survey-based data from Statista and explores the correlation of crypto adoption with indicators of corruption, capital controls, inflation, and other factors. The study aims to understand whether crypto-assets are being used to facilitate corruption or circumvent capital controls.

Main Findings

• Crypto-Asset Usage and Corruption: Crypto-asset usage is significantly and positively correlated with the perception of corruption. Countries with higher levels of corruption tend to have greater adoption of crypto-assets.
• Crypto-Asset Usage and Capital Controls: Crypto-asset usage is also significantly and positively associated with more intensive capital controls. This suggests that crypto-assets may be used to bypass these controls.
• Inflation and Economic Development: While there is a correlation between crypto-asset usage and a history of high inflation, it is not statistically significant. Economic development, as measured by real GDP per capita, is negatively correlated with crypto adoption but not statistically significant.
• Financial Development and Internet Infrastructure: The number of commercial bank branches per 100,000 adults is negatively correlated with crypto adoption, but not statistically significant. Secure internet servers are positively correlated with crypto adoption, indicating that digital infrastructure plays a role in enabling crypto usage.

Key Variables and Their Correlations

Methodology and Data

• The primary data on crypto-asset usage comes from Statista, with survey responses from 2,000–12,000 individuals per country.
• The variable used is the logarithm of one plus the crypto adoption rate to reduce the influence of outliers.
• The study also considers alternative data sources, such as the Chainalysis Global Crypto Adoption Index, but notes that these datasets have limitations, including potential misallocation of crypto use due to the use of VPNs and other tools that mask online activity.
• The paper uses a general-to-specific approach for multivariate regression analysis to address multicollinearity, which is a significant issue in the dataset.
• The final regression model includes control of corruption and capital openness as significant explanatory variables, with the other variables being excluded due to their insignificance or redundancy.

Conclusion

The study concludes that crypto-asset usage is empirically linked to higher perceived corruption and more intensive capital controls. Given the rapid macroeconomic relevance of crypto-assets, the authors argue that the evidence supports the case for regulation, including the implementation of know-your-customer (KYC) procedures, rather than a laissez-faire approach. The paper also emphasizes the need for better data to accurately capture the dynamics and key drivers of crypto adoption, while highlighting the potential benefits of the underlying technologies for financial inclusion and government efficiency.

Limitations and Caveats

• The study acknowledges data limitations, including the small sample size and potential measurement errors.
• The use of alternative data sources, such as Chainalysis, leads to inconsistencies in rankings and may not be reliable due to the influence of web traffic and proxy tools.
• The results should be interpreted with caution due to these limitations, but the authors argue that significant findings with low-quality data still warrant attention and regulatory consideration.