A paper published in the Ministry of Finance Policy Research Institute's "Financial Review" Issue 160. Professor Takashi Unayama and his research team comprehensively analyzed the characteristics and potential utilization of tax return data as income data.
Key Points
1. Research Overview and Objectives
- Authors: Takashi Unayama (Kyoto University), Shimpei Sano (Kobe University), Kazufumi Yugami (Kobe University), Kazuhiro Inaba (Tax College/Ministry of Finance Policy Research Institute)
- Research Question: Evaluating the usefulness of tax return data as income statistics
- Problem Awareness: Establishing a new data source to complement limitations of existing income statistics
- Analytical Method: Comparative verification with other income statistics
2. Relationship Between Income Statistics and Tax Returns
- Limitations of Existing Income Statistics:
- Household Survey: Sample size and response bias
- Basic Survey on Wage Structure: Employment income only
- Comprehensive Survey of Living Conditions: Difficulty capturing high-income earners
- Characteristics of Tax Return Data:
- Accuracy based on tax filing
- Comprehensiveness of income sources
- Complete capture of high-income earners
3. Personal Coverage of Tax Return Data
- Scope of Filing Obligations:
- Salary earners: Annual income exceeding 20 million yen
- Business income earners: All in principle
- Multiple income source holders
- Trends in Number of Filers:
- Approximately 22 million people (about 17% of population)
- Increasing trend in number of filers
- Characteristics of Non-Filers:
- Salary earners with annual income below 20 million yen
- Some pension recipients
- Those without income
4. Income Coverage of Tax Return Data
- Captured Income:
- Complete picture of business income
- Real estate income
- Dividend and interest income (comprehensive taxation portion)
- Capital gains
- Partial Capture:
- Salary income (high earners only)
- Pension income (partial)
- Uncaptured Income:
- Interest subject to separate withholding tax
- Small dividends
5. Tax Return Data as High-Income Earner Data
- Completeness:
- Complete capture of annual income exceeding 20 million yen
- Comprehensive grasp of asset-based income
- Detailed analysis of income composition possible
- International Comparability:
- Calculation of top income shares
- Estimation of Pareto coefficients
- Accurate calculation of inequality indicators
- Time Series Analysis: Tracking long-term changes in income distribution
6. Comparative Analysis with Other Statistics
- Consistency with National Tax Agency Statistics:
- Agreement at aggregate level
- Superiority of individual data analysis
- Divergence from Household Survey:
- Large differences in high-income brackets
- Differences in business income capture rates
- Relationship with SNA: Possibility of connection to macro statistics
7. Cautions and Limitations in Data Use
- Representativeness Issues:
- Not a representative sample of entire population
- Bias toward specific income brackets
- Definitional Differences:
- Tax law income concepts
- Divergence from economic income
- Timing Issues: Only annual flow information
8. Research Applications and Future Prospects
- Contribution to Inequality Research:
- More accurate grasp of income distribution
- Deepening of wealth research
- Policy Evaluation:
- Measuring effects of tax reforms
- Verification of income redistribution policies
- Future Extensions:
- Linkage with other data
- Dynamic analysis through panel data
- Development of international collaborative research
This research empirically demonstrates that tax return data is an extremely important data source for income distribution research in Japan, opening new horizons for future income and inequality research.