¹ For an account of the evolution of the Latin American state, from its interventionist past to the underpinnings of the new liberalism, see Fishlow, A., ‘The Latin American State,’ Journal of Economic Perspectives, vol. 4, no. 3 (1990), pp. 61–74.CrossRefGoogle Scholar, This article was written while the author was a visiting assistant professor at the Graduate School of Business at the University of Chicago. It has drawn on some material from T. Glaessner and I. Mas, ‘Incentive Structure and Resolution of Financial Institution Distress: Latin American Experience’, World Bank working paper LACTD no. 12. The views expressed here are the author's and do not represent those of the World Bank. Ignacio Mas is a Financial Economist at the World Bank.

² For a survey of Latin America's banking systems, see Morris, F. et al. , ‘Latin America's Banking System in the 1980s: A Cross-Country Comparison’, World Bank Discussion Paper no. 81 (Washington D.C., 1990)Google Scholar.

³ For a discussion of these issues see Stiglitz, J. E., ‘The Role of the State in Financial Markets’, paper presented at the World Bank's Annual Conference on Development Economics (Washington D.C., 05 1993)Google Scholar.

⁴ Whether a high reserve requirement is directly financing deficits through earmarking or is used to sterilise large central bank credit to the government is a semantic issue, especially if the central bank is not independent of the Treasury as was the case throughout the region prior to 1991.

⁵ The concept of reserve requirement tax is of course intimately linked to – in fact is a part of – the broader concept of inflation tax. This is because poorly remunerated required reserves are not compensated for their loss of purchasing power inflation.

⁶ The term Quasi-fiscal is used throughout the article to refer to public sector expenditures and taxes, whether explicit of implicit, that operate through the central bank and the financial system and do not appear in the national budget.

⁷ For a general discussion of inflationary financing and its macroeconomics impact, see Fry, M., ‘;The Fiscal Abuse of Central Banks’, International Monetary Fund working paper no. 93/58 (Washington, D.C., 1993)Google Scholar. Mas, I., ‘Los Bacons combo Agents Fiscals del Gobierno: el Ca so de Nicaragua, 1986–90,’ Monetarism, vol. 17, no. 3 (1994), pp. 283–314Google Scholar, contains a case study of Nicaragua.

⁸ Of course, such experiences are not restricted to Latin America. Blocking bank accounts as a stabilization measure was used extensively in European countries during and after World War II, e. g., Austria (1945, 1947), Belgium (1944), Bulgaria (1947, 1952), Czechoslovakia (1945, 1953), Denmark (1945), France (1945, 1948), Netherlands (1945), Norway (1945) and Poland (1944, 1950). For a summary of these experiences, see Donuts, R. and Wolf, H., ‘Monetary Overhang and Reforms in the 1940s’, mimeo (09 1990)Google Scholar.

⁹ The freeze affected savings and checking accounts in excess of the equivalent of US? 1,000, and overnight accounts in excess of US$ 500. Frozen assets were to be released to the public in 12 monthly instalments, starting six months after the freeze. For a full description of the plan, and the economic context in which it occurred, see Cards, E. A., ‘Deficit Finance and Monetary Dynamics,’ journal of Development Economics, vol. 37, no. 1–2 (1991), pp. 192–5Google Scholar.

¹⁰ Immediately after the freeze frozen cruzados traded at a value of about 65 % of historic face value, according to Cardoso, 'Deficit Finance and Monetary Dynamics', p. 193. However, the government quickly moved to eliminate this budding market.

¹¹ This was, of course, by design. The scramble for liquidity was meant to impose deflationary pressures which would serve to stabilise the economy.

¹² As of 31 August, the day before the measures were announced; the figure excludes government and inter-bank accounts. Dollar-denominated deposits amounted to 29 % of private sector deposits if converted at the official rate, and 33% if converted at the black-market rate on that date. Of course, the black-market rate immediately shot up with the announcement of these measures. The data are from Instituto Nacional de Estadística, Boletín Mensual de Información Econo'mica, vol. 7, no. 11 (México, 12 1983)Google Scholar.

¹³ SeePérez-López, E., Expropiación Bancaria en México y Desarrollo Desestabiliador (Mexico, 1987)Google Scholar, ch. 3, for exchange rate quotes, and Banco de México, Informe Anual 1983 for the monthly stocks of mexdollar balances and data on balance of payments. For political and economic background on the accompanying nationalisation of banks, see also Tello, C., La Nacionalización de la Banca en Mexico (Mexico, 1984)Google Scholar.

¹⁴ Despite the ruling, no compensation appears to have been paid ex-post,

¹⁵ All figures on exchanges rates, deposits and money supply come from Banco Central de Bolivia, Memoria Anual 1982, which also contains the text of the ‘dedollarisation’ decree of 3 November 1982. For a description of the economic context in which the dedollarisation occurred, see Morales, J. A., Pretios, Salariosy Política Económica Durante la Alta Inflación Boliviano de 1982 a 1985 (La Paz, 1986)Google Scholar, ch. 3.

¹⁶ Foreign currency deposits accounted for 37% of total banking system deposits at the official exchange rate, or 47% at the parallel exchange rate. It is difficult to relate the size of foreign currency deposit confiscation to the national budget. Inflation was running at almost 300% per annum, and hence nominal budget figures are heavily distorted.

¹⁷ This discussion sidesteps the issue of whether directed credit programmes can be justified theoretically on the basis of market failures and information asymmetries; Calomiris, C. W. and Himmelberg, C. P., ‘Directed Credit Programs for Agriculture and Industry: Arguments from Theo'ry and Fact’, paper presented at the World Bank's Annual Conference on Developmental Economics (Washington, 05 1993)Google Scholar provides such a discussion. What is clear in the case of Latin America is that such broadly encompassing, massive and poorly targeted credit subsidies cannot be justified on these grounds.

¹⁸ For a fuller discussion of the incentive problems associated with debt reprogramming, see Glaessner, T. and Mas, I., ‘Incentives and Resolution of Bank Distress,’ The World Bank Research Observer, vol. 10, no. 1 (02. 1995)CrossRefGoogle Scholar.

¹⁹ In actuality, loans in both local currency and dollars were eligible for refinancing, but refinanced credits were denominated in dollars even if the original credit was in local currency. In practice, the maturity of the loan was extended to five years, including one year grace. To be eligible for refinancing under this programme, loans had to be for productive uses, could not have been previously rediscounted at the BCB, and could not be held by bank insiders. The total amount of credit refinanced by a bank to a particular borrower could not exceed 10% of the amount owed by the borrower to that bank as of 30 June 1988. Only privately owned banks that had agreed to a financial strengthening programme with the BCB could refinance loans under this programme.

²⁰ Refinancing non-performing loans is even more attractive financially to the bank as lowering the lending rate may improve the prospects for repayment of the loan.

²¹ The government had adopted a pre-announced schedule of exchange rate devaluations (known as the tablita) with which it sought to contain inflationary expectations. However, prices of non-tradables failed to respond to this environment to the point where over 1979–80 the currency was devalued by only 50% whereas the domestic wholesale price and cost of living indices increased by 3.6 and 4.5 times, respectively. For a description of the macroeconomic backdrop, see The World Bank, ‘Economic Memorandum for Argentina’ (Washington, 05 1983)Google Scholar, ch. 1.

²² Enterprise debts were refinanced for seven years with three years grace, at a rate equal to the reference 30-day deposit rate (índice de ajuste financiero, or IAF) plus 3% per annum. Banks could refinance these debts at the central bank at IAF flat. At the same time, banks had to purchase special equal maturity, non-negotiable government bonds bearing IAF plus 6% per annum. Banks were free to join the scheme, but if they participated they had to meet the refinancing requests of all client firms. See Baliño, T. J. T., ‘The Argentine Banking Crisis of 1980’, IMF working paper no. 87/77 (Washington, D.C., 1987), pp. 48–9Google Scholar.

²³ See Calcagno, E., ‘Los Bancos Transnacionales y el Endeudamiento Externo en la Argentina’, Cuadernos de la Cepal no. 56 (1987), p. 36Google Scholar. At the same time, the government nationalised private external debt through exchange rate guarantees and swaps at favourable rates.

²⁴ Glaessner and Mas (1994) discusses formally why negative net worth needs to be covered with capital (stock) assistance rather than with income enhancement (flow) assistance. See Glaessner, Thomas and Mas, Ignacio, ‘Incentives and the Resolution of Bank Distress,’ The World Bank Research Observer, vol. 10, no. 1, 02 1995CrossRefGoogle Scholar.

²⁵ This assumes that the Treasury could borrow at the same rate as the central bank; however, at that time there were no Treasury bills.

²⁶ FOGADE does receive premiums from banks, but their small size relative to the bank failures it has had to deal with meant that, on the margin, its opportunity cost is given by its borrowing cost.

²⁷ The banks in question were Banco Occidental, Banco Italo-Venezolano and Banco República; they were reprivatised in late 1990, April 1991 and June 1991, respectively.