Methodology
The Government's Role in Encouraging the Depth of Finanical Markets
Governments play an important role in creating, structuring, and regulating the deep financial markets crucial for provision of credit and access to banking services to citizens.
The Financial Depth indicator proxies for government policies encouraging the private provision of credit and widespread public use of the formal banking sector.
The indicator is computed as arithmetic mean of two series:
- Domestic Credit in the Private Sector (WB_CREDIT):
- Financial System Deposits (WB_DPOSIT):
Proxy
Governments with a high levels of Domenstic Credit in the Private Sector and high levels of Financial System Deposit must have policy packages which encourage these outcomes.
Specific Policies to Encourage Deep Finanicial Markets.
...
Potential Counfounding Factors in Policy Proxy: Other Factors
...
Data Coverage and Imputations
Comprehensive data is available for of country-year pairs. When either WB_CREDIT or WB_DPOSIT data is missing ( of pairs), we impute the missing data for the particular intermediate, then compute the overall indicator by averaging the imputed value with the available value for the other series.
When neither WB_CREDIT nor DEPOSIT data is available for a given year ( of country-year pairs), we impute the value for the FDEPTH directly using the most proximate value(s) of FDEPTH.
Imputation Methods
For all countries except United Kingdom (GBR), all SSPI countries have an observation in at least one year for both series, so all imputations are anchored to a valid observation in the dataset. The United Kingdom has no observations for WB_DPOSIT
- Backward Extrapolation: We use backward extrapolation when there exist no observation prior to the year imputed. The imputed value is equal to the first observed value in the dataset (i.e. the most proximate value).
- Forward Extrapolation: We use forward extrapolation when there exist no observation after the year to be imputed. The imputed value is equal to the last observed value in the dataset (i.e. the most proximate value).
- Linear Interpolation: We use linear interpolation when there exist observations in sample on either side of the missing observation (i.e. the two most proximate values).