For this to be accepted, it must be shown that small savings with the government have replaced bank deposits, at least at the margin, in total household savings. But Chart 3, which shows the proportion of bank deposits in the total financial assets of households, shows no significant time trend over the 1990s. In fact, on average the share of bank deposits has been substantially higher in the last three years of the decade (at more than 38 per cent) than in the first three years of the decade (at only 32 per cent).
Chart 3 >>
 
Meanwhile, as Chart 4 indicates, the share of small savings has indeed increased slightly over the 1990s. However, this increase is not marked when compared to the beginning of the decade (or with the last few years of the 1980s) and represents more of a recovery to the earlier shares. Bu what is more to the point is that such changes have definitely not been at the expense of bank credit. Rather, the share of household savings held as shares and debentures, which had reached 10 per cent in 1992-93 following the stock market boom immediately after the initial liberalisation,  has declined sharply (barring the final year of the decade) to only around 2 per cent in 1998-99 (Chart 5).
Chart 4 >> Chart 5 >>
 
So it is mainly the greater uncertainty in the stock market which has driven small investors back to the more secure forms of savings such as Public Provident Fund and so on, and reduced exposure to the riskier forms of saving. It is worth noting that recent fiscal measures, in terms of reduction in interest rates on small savings and tax concessions such as relief in terms of capital gains tax, are designed to cause such investors to turn back to shares and debentures and reduce their holding of public instruments of small savings. The rates of interest on such instruments, some of which are described in Table 1, are now lower than they have ever been in the 1990s.

The evident public disaffection with riskier savings instruments is also evident from Chart 6, which shows that deposits of the public with non-bank financial companies, which grew as a proportion of bank deposits over the middle of the 1990s, have since declined quite substantially in such terms.
Chart 6 >>
 
This means that the argument that small savings have replaced bank deposits at the margin, or among the preferences of households, is not substantiated by the evidence. Bank deposits are as important as they have been from the beginning of the decade; in fact more than they were in the previous decade. Rather, it is the desire to move out of stock market-based instruments and investments in non-bank financial companies that often promise very attractive if risky returns, that has been responsible for the increase in small savings from the middle of the 1990s. This in turn means that high interest rates on such small savings cannot be the culprit in terms of finding an explanation for why bank lending rates have remained high in real terms.
 
The other explanation, which revolved around high (and possibly increasing) spreads on deposit rates because of bank inefficiency or lack of sufficient competition, also does not find validation in the data. As Chart 8 shows, the spread between deposit rates for term deposits of one to three years' duration, and the State Bank of India's Advance Rate, has actually declined quite sharply since the peak reached in 1993-94. This is corroborated by other evidence which suggests that the net interest income (spread ) as a proportion of total assets of banks declined for al major groups of banks over this period. In fact, the decline was most marked for public sector banks, which implies that, if anything, their "efficiency" increased over this period.
Chart 8 >>