1Economics and Culture in the Writing of Financial History*Barry EichengreenDepartments of Economics and Political ScienceUniversity of California, BerkeleyMay 1997Running Head: Economics and CultureWordperfect 7.0Address from July 1, 1997: Research Department International Monetary Fund Washington, DC 20431 USAr:\doc\ai\be\doc\seville.wpd2AbstractRecent research in financial history has taken three approaches to thesubject, focusing on financial institutions, financial systems, and financialnetworks. Functionalist logic from neoclassical economics is both thestrength and the weakness of the first, institution-centered approach. Thislogic provides a coherent framework with which to analyze the development offinancial arrangements but one based narrowly in information economics andagency theory. I argue in this paper that the other two approaches focusingon financial systems and financial networks are richer and more suggestive. They provide microeconomic foundations for arguments emphasizing the path-dependent character of financial development, thereby highlighting thecontribution of financial history to financial economics. By emphasizing theextent to which financial relations are socially embedded and depend oninterpersonal networks, they provide a means of bridging the gap betweeneconomic and cultural history.See Gerschenkron (1962) and Cameron (1967).13Recent research in financial history has benefitted enormously fromconcurrent advances in economics, specifically in information economics andagency theory. Alexander Gerschenkron and Rondo Cameron, to name two of ourformidable forbearers, had an intuitive sense of what distinguished financialmarkets from other markets, namely imperfect information and imperfectcontract enforcement. Recent work in economics, by more rigorously modeling1these market imperfections and drawing out their implications, has helped toclarify the role of financial institutions in the process of economic growth. It has lent rigor to the intuition of earlier financial historians.The theme of this theoretical work is that financial institutionsprovide monitoring and control services useful for overcoming problems ofincomplete information and incomplete enforcability. If borrowers have moreinformation than lenders about the characteristics of projects in which theyinvest (adverse selection) or are in a position to manipulate the outcome(moral hazard), then the price mechanism may not suffice to transfer creditfrom savers to investors. Akerlof's (1969) seminal work on adverse selectiondemonstrated how the interest rate charged on loans might not equilibrate amarket in which the characteristics of participants (in the present context,of those who demand loans) depends on the price charged and where thosecharacteristics are difficult to observe. Were loss-making lenders to raisethe interest rate they charged, low-risk borrowers would drop out of themarket, leaving only customers seeking to finance high-risk projects andtherefore willing to bear the increased interest-rate premium. The level ofrisk (and associated loan losses) having risen with the interest rate, thelender is no better off than before, threatening the market with breakdown. This problem implies the need for an agent (a bank, as in Diamond, 1984, or arating agency, as in Millon and Thakor, 1985) capable of overcoming theinformation asymmetry, ascertaining the borrower's true characteristics, and4pricing credit accordingly.The question then becomes why banks, rating agencies or other financialinstitutions have a comparative advantage in carrying out these tasks. Diamond's (1984) work on banks as delegated monitors assumes that assemblingand processing the information relevant for loan decisions is costly but thatthis task can be carried out more efficiently through specialization. Ifindividual investors delegate the task of screening borrowers and monitoringtheir actions to specialized entities called banks, the latter can undertakethese functions at relatively low cost. It follows from Diamond's model thatan economy in which banks act as delegated monitors will feature a moreefficient allocation of resources, a higher level of investment, and a fasterrate of growth than one whose the banking system is underdeveloped. Fazzari,Hubbard and Petersen (1988) and Hoshi, Kashyap and Schafstein (1991) show howbanks that establish long-term connections with their customers may have acomparative advantage in using monitoring to solve the adverse selectionproblem, again implying that financially-sophisticated economies should growfaster than their more financially-underdeveloped counterparts.Formal models have also appeared in which financial institutions areassumed or shown to have a special facility to control the actions ofborrowers. In Diamond (1991), banks are assumed to be more efficient atapplying sanctions than bond markets; consequently, borrowers withoutreputational capital at risk will be shunned by the bond market but may stillbe able to borrow from banks, since the latter are better positioned tocontrol their actions. Better control is associated with more lending, longerinvestment horizons, and a more efficient allocation of resources (Mayer,1988; Aghion and Bolton, 1992).This work in economics has lent obvious support to the functionalistinterpretation of financial development. Financial institutions exist, thefunctionalist argument runs, to fill the need created by problems ofasymmetric information, adverse selection, and moral hazard. The precise5nature of the information environment and the scope for opportunistic behaviorwill be different in different historical settings, and so will the efficientinstitutional response. In an agricultural setting, where the problem is thecare and timeliness with which farmers prepare their land, plant their cropand tend their livestock, peer monitoring can be carried out by neighboringfarmers, and rural credit cooperatives will be observed (Guinnane, 1994). Inthe circumstances of early industrialization, where the problem facing theaspiring manufacturer is what to produce, how to organize the factory, and howto market the product, critical assets of the enterprise are intangible,providing scope opportunistic behavior on the part of potential borrowers. Credit is provided by family-based financial institutions with insideinformation and the ability to use