The scarcity of capital in Russia was such that no banking system could conceivably succeed in attracting funds. . . . Supply of capital for the needs of industrialization required the compulsory machinery of the government.1
Whatever its original objectives, state ownership tends to stunt financial sector development, thereby contributing to slower growth.2
Arthur Lewis, Alexander Gerschenkron, Gunnar Myrdal, and several other prominent development economists writing in the 1950s and 1960s tended to agree that the state should play a key role in the banking sector. Governments appeared to concur: by the 1970s, the state owned 40 percent of the largest banks' assets in industrial countries and 65 percent of the largest banks' assets in developing countries. The 1980s and 1990s witnessed a sea change in the view of the state's role in the economy, and privatization was at the very center of the economic policies codified in the Washington Consensus. Consequently, more than 250 banks were privatized from 1987 to 2003, raising U.S.$143 billion.3 Even after this big privatization wave, however, the presence of the state in the banking sector remained widespread and pervasive. In the mid-1990s, about one-quarter of the largest banks' assets in industrial countries and 50 percent of the largest banks' assets in developing countries were still under state control. [End Page 209]
The key question explored in this paper is whether public presence in the banking sector is justified. Advocates argue that state presence in banking is warranted by market failures and development goals. They point out that financial markets in general, and the banking sector in particular, are different from other markets and that government intervention can improve the working of the financial sector and the overall functioning of the economy. In particular, the social view emphasizes the role of the public sector in making up for market imperfections that leave socially profitable investments underfinanced.4 Also supportive of public participation in the banking sector is the development view, which stresses the need for public intervention in economies where the scarcity of capital, the general distrust of the public, and endemic fraudulent practices among debtors may fail to generate the sizable financial sector required to facilitate economic development.5
Critics argue that banks are not necessarily different from other businesses and that the case for financial market imperfection is often overstated.6 They suggest that market failures can be better addressed through regulation and subsidies than through direct state ownership. This political view contends that politicians create and maintain state-owned (henceforth, public) banks not to channel funds to socially efficient uses, but rather to maximize the politicians' personal objectives.7 Specifically, state ownership of banks is dictated by redistributive politics and the politicians' interest in appropriating the rents that may be derived from the control of bank funds. Somewhere in between the benign assessment of the social and development views and the skepticism of the political view, the agency view highlights the trade-off between allocative efficiency and internal efficiency (namely, the ability of public enterprises to carry out their mandate), asking whether agency costs within government bureaucracies offset the social gains of public participation in the presence of market imperfections.
This paper is divided into three parts. The first part describes the evolution of state ownership of banks in Latin America and the rest of the world. The second part discusses the theoretical justification for the existence of public banks. The third part surveys the existing empirical evidence and presents some new results. [End Page 210]
The Evolution of Public Ownership of Banks
Obtaining consistent time series describing the evolution of public banks around the world is not easy because different authors use different methodologies and sources. Data going back to 1970 are available from La Porta, López-de-Silanes, and Shleifer (who collected data for 1970, 1985 and 1995), while more recent data covering the...