The spatial dimensions of intermediary behavior

How does distance from a bank affect borrowers' ability to obtain credit? Conversely, how does the location of clients affect banks' decisions about where to be based? This dissertation devises new theoretical tools to model effects of distance on bank lending, drawing on concepts from economic geography and recent models of endogenous intermediation. It then tests certain implications using fresh regional financial data from two countries.; Distance between borrowers and lenders creates 'friction' that exacerbates problems of asymmetric information in two ways, set out in Chapter 1. First, communication of any information is costly across distance. Although rapid technological advances reduce these costs substantially, they remain significant in developing economies. Second, and more important, the quality and usefulness of unstandardized information declines as distance between source and users increases.; Chapters 2-4 model endogenous intermediaries under costly state verification, using three concepts of distance. Chapter 2 develops continuous-metric bank location models whose results, such as agglomeration absent loan price competition, parallel old real-sector ('circular city') locational models. Chapter 3 uses a discontinuous concept of distance, with agents located in two regions facing different aggregate shocks. Tradeoffs between diversification through inter-regional lending and the higher monitoring costs of non-local lending vary under different banking systems. Chapter 4 considers how lesser differences in organizational culture--a third type of distance--enable small-sized banks to assist small businesses better than larger banks can.; The last two chapters analyze postwar regional financial data from Australia and Canada. Chapter 5 develops a financial center index ('FINDEX') to compare Melbourne with Sydney, and Montreal with Toronto. The concept of a financial center's 'information hinterland' helps explain the formation, rise and decline of financial agglomerations over time. The two city-pairs illustrate striking reversals in apparently path-dependent locational processes. Chapter 6 uses regional lending data to assess claims that nationwide banks discriminate against peripheral regions in allocating credit. There is no evidence of secular decline in the shares of peripheral regions in bank credit; but national monetary policy does affect dispersion of bank credit across regions.