The second viewpoint is that the relationship between concentration and deposit rates is not negative.
Jackson (1992) has criticized the Berger and Hannan results, arguing that the relation between
concentration and deposit rates is not linear, but U-shaped. At first, deposit rates decrease as
concentration increases. But at a certain point greater concentration implies higher rates, because of the
intense competition in markets dominated by a few banks.5 However, oligopolistic behavior is constrained
by the possibility of new entrants to the markets, according to the mechanism described by the contestable
market theory.
This approach leads us to consider the effect on the market of potential competition from new
entrants, rather than focusing on the relationship among the incumbents. 6 This position results from
skepticism towards the structure-conduct-performance paradigm, which has been criticized on different
grounds. One criticism is that concentration cannot be considered exogenous because it is influenced by the
prices charged by firms: in the case of deposits, for example, higher interest rates should lead to gains in
market share.
Critics of the relationship between concentration and deposits’ compensation add that banking
deregulation and the wave of mergers have caused the notion of local markets to blur (see Osborne,
1992). The view that geographic markets are local was formed in the 1960’s when unit banking - banks
consisting of a single office - was common and branching was heavily restricted in many US states
(Radecki, 1998). Now it seems that concentration at the local level - usually measured by the metropolitan
statistical areas (MSA) - is no longer relevant to interest rates paid on retail deposits. According to
Radecki, the uniformity of bank deposit and loan rates across an entire state suggests that state boundaries
now approximate the shape and extent of retail markets better than the MSA. In contrast, concentration at
the state level affects deposit rates. The increase in bank size, the development of banking holding
companies and the trend towards interstate banking explain why markets are growing larger in geographical
scope. Moore (1998) also claims that banking market borders are wider today than in the past.
In Italy the debate is similar. Within the earlier framework characterized by entry barriers to local
markets, due to limits in branch openings and mergers, Biscaini, Carosio and Padoa Schioppa (1972) were
the first to find that deposit rates were more upwardly rigid than money market rates. These results were in
5 In response, Berger and Hannan (1992) confirm the robustness of their estimates but recognize that the
relationship between concentration and rates is time-variant.
6 Allen, Saunders and Udell (1991) do not find a statistically significant influence of concentration on banking
fees.