for contemporaneous effect (column (1)), and at the 1 percent level for the joint parameters of zero to 24 months in columns (2) and (3)), and is quite persistent, appearing in the data at it’s original level for up to eighteen months following the vigilance activity, finally becoming statistically indistinguishable from zero two years after the CVC decision or judgement.
This economic effect seems to be sizable for plausible values of the elasticity of gross domestic product with respect to money supply elasticity. For example, if the overall coefficient of .03 were accurate for a bank such as the State Bank of India, which provides approximately a quarter of the credit in the economy, decisions on whether to pursue vigilance cases could have measurable macroeconomic effects.
Columns (5) and (6) of Table 9 present the same specification as in equation 7, but this time with dummies indicating whether a given bank-month is exactly n months before CVC vigilance activity.
Table 9 clearly indicates that there was a reduction in lending in banks prior to the announced vigilance action, as well as after it. This is not surprising, as the formal vigilance activity usually follows a lengthy investigation. The CVC vigilance manual, introduced in 1999 to streamline the process of CVC investigations, outlines a procedure that lists no binding time constraints, but suggests the entire process be completed within twenty months time. Reassuringly there is no discernable effect for vigilance activity farther out than one year ahead.
Conclusion: There seems to be some evidence that the fear of being investigated is reducing lending by a significant extent: banks where someone is being investigated slow down lending relative to their own mean level of lending. This leaves open the question of whether this is a desirable reaction, since it is possible that the loans that are cut are the loans that are unlikely to be repaid. But it does raise the possibility that honest lenders are being discouraged by excessively stringent regulations.
Lending to the government and the easy life
Lending to the government is the natural alternative to lending to firms and offers the loan officers a secure vehicle for their money, with none of the legwork and headaches associated with lending to firms. The ideal way to measure how important high interest rates on government bonds might be in explaining under-lending, would be to estimate the elasticity of bank lending