20180213-兴业研究-Assessment_of_the_Credibility_in_Chinese_Local_Governments__Data_14页_710kb

Summary of "Assessment of the Credibility in Chinese Local Governments' Data" (CIB Research, 2018.2.13)

Core Content

This report discusses the credibility of economic data reported by Chinese local governments, focusing on the reasons behind data exaggeration and methods to assess data quality.

Main Reasons for Data Inflation

1. National Bureau of Statistics (NBS) Takeover in 2019
   The NBS will take over data collection from local authorities in 2019, making it easier to detect any data manipulation.

2. Business-Tax-to-VAT Reform (2016)
   The reform reduced the share of VAT revenue that local governments could retain, making it more costly for them to inflate data.

3. End of Local Officials' Tenure
   When local officials complete their term, their successors may revise down inflated data to make future targets more achievable.

4. Transfer Payments from Central Government
   Local governments may revise data to argue for more transfer payments, although past experiences suggest this does not always lead to increased support.

Key Components of Inflated Data

• Fixed Capital Formation and Budgetary Revenue
  These are the most commonly inflated components in local economic data.

Methods to Assess Local Data Quality

The report outlines seven parameters to evaluate the credibility of local economic data:

1. Correlation between GDP growth and budgetary revenue
2. Correlation between GDP growth and night-time light
3. Gap between Li Keqiang index and real GDP growth rate
4. Comparable growth rates of business tax and VAT revenue
5. Share of gross capital formation in GDP
6. Proportion of shared tax revenue in local budgetary revenue
7. Ratio of household disposable income to GDP per capita

Key Findings

• Local data often overstate GDP
  The aggregate GDP of all provinces is typically higher than the national figure, with a gap of 3-4 trillion yuan since 2011.

• Gross fixed capital formation is most vulnerable to manipulation
  This component is frequently inflated due to its ease of measurement and manipulation.

• Weak correlation between budgetary revenue and GDP growth may indicate data issues
  However, it could also reflect structural changes or policy impacts, such as the VAT reform.

• Night-time light data shows a strong correlation with GDP in some provinces
  For example, Zhejiang, but not in Heilongjiang, indicating regional differences in data reliability.

• Li Keqiang index and real GDP growth rate often differ
  The gap suggests potential discrepancies in economic performance reporting.

• Provinces with higher shares of shared tax revenue tend to have more credible data
  This is because they are less incentivized to inflate figures for financial gain.

Composite Score for Local Data Credibility

A composite score is calculated based on the seven parameters. The following provinces have the highest scores:

• Shanghai: 5.50
• Anhui: 5.39
• Zhejiang: 5.37

The lowest scores are:

• Guizhou: 3.64
• Chongqing: 3.53
• Liaoning: 3.53
• Ningxia: 3.39
• Sinkiang: 3.37
• Tibet: 3.31
• Qinghai: 3.19
• Tianjin: 2.95
• Inner Mongolia: 2.75

Key Information

• The report uses data from multiple sources, including Wind and CIB Research.
• Some provinces, like Liaoning and Inner Mongolia, have shown significant declines in budgetary revenue after revising data.
• The report highlights the importance of using alternative indicators, such as night-time light and Li Keqiang index, to cross-validate local economic data.
• It emphasizes the need for investors to independently evaluate data and consult financial advisers before making investment decisions.

Conclusion

The report provides a detailed analysis of the factors influencing local government data inflation and offers a framework for assessing data quality through multiple indicators. It underscores the importance of transparency and the potential risks associated with relying solely on local economic data.