Why did Colombian private savings decline in the early 1990s?

V/Ps 7S13 POLICY RESEARCH WORKING PAPER 1713 Why Did Colombian The sharp drop in private savings in the I 990s in Private Savings Decline Colombia can be attributed to a decline in private disposable in the Early 1990s? income and, to a lesser extent, to growth in Alejandro L6pez consumption, The World Bank Policy Research Department Macroeconomics and Growth Division January 1997  I POLICY RESEARCH WORKING PAPER 1713 Summary findings The sharp drop in private savings in the 1990s in * Private consumption's recent behavior can hardly be Colombia can be attributed to a decline in private called a boom. It declined throughout the second half of disposable income and, to a lesser extent, to growth in the 1980s before finally showing an upturn in 1992 consumption. equivalent to 2 percent of gross national product. The permanent decline in private disposable income in * Consumption of durable goods after trade reform Colombia between 1950 and 1990 is closely linked to cannot be blamed for the decline in private savings. In tax increases. This trend was accentuated in the early fact, savings began falling in 1988 and, until 1993, trade 1990s by a reduction in corporations' gross operating reform did not cause a stock adjustment of durable surplus. goods. Contrary to the usual hypothesis, L6pez shows that in the 1990s private consumption had a relatively minor effect on national savings. He highlights two findings: This paper - a product of the Macroeconomics and Growth Division, Policy Research Department - is part of a larger effort in the department to assess the determinants of saving. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Emily Khine, room N11-061, telephone 202-473-7471, fax 202-522-3518, Internet address kkhine@worldbank.org. January 1997. (34 pages) The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. Produced by the Policy Research Dissemination Center  Why Did Colombian Private Saving Decline in the Early 1990s?* Alejandro L6pez** The World Bank * For their useful comments on firsts drafts I would like to thank Juan Carlos Jaramillo, Juan Luis Londoino, Ernesto May, Armando Montenegro, Paulo Neuhaus, Luis Serven, Roberto Steiner and Levy Yeyati. This is a revised version of a paper written with Carolina G6mez and Norberto Rodriguez, researchers from El Banco de la Republica (The Central Bank of Colombia). **This paper was written while the author was a research fellow at The World Bank.   1. Introduction Many have pointed out that prior to 1990 the Colombian econorny was characterized by the presence of many microeconomnic distortions and prudent macroeconomic management (see, among others, World Bank (1991), Urrutia (1994)). As a consequence, growth only had a decent performance and inflation and budget deficits were moderate. For most Colombian policymakers, the lack of a major macroeconornic crisis and the failures of different economic experiments throughout Latin America were sufficient evidence that no serious economic reform was needed. However, during the period 1990-1994 several structural reforms were undertaken in areas such as foreign trade, foreign investment, labor markets, financial markets and public utilities. The ultimate objective of the reforms was to foster growth by strengthening the role of the government in the provision of security and justice and by allowing the private sector to compete in former protected markets and expanding its role in the production of "public goods" (Montenegro (1995)). The structural reforms carried out by the government were accompanied by a significant decline in private savings. Until 1992 this behavior was offset by public savings and consequently the national saving rate remained stable (Figure 1). Although the deterioration of private savings started in 1988, its extension in the 1990s was associated with the implementation of the structural reforms'. Until 1989 around 70% of disposable income accrued to liquidity constrained consumers (L6pez (1994)). Given Figure 1: Savings in Colombia 25 10 1951 1958 1965 1972 1979 1986 1993 Private Saving ---- Public Saving - National Saving * Source: National Accounts XAn additional hypothesis not related with the reforms has been a reassessment of permanent income due to the oil bonanza (see, among many others, L6pez (1995)). l  this result, Urrutia and L6pez (1994, 1995) suggested that the acceleration of the financial liberalization process in the 1990s relaxed liquidity constraints causing an increase in consumption. In addition, Echeverry (1996) explained the drastic fall in the Colombian private saving rate taking a similar approach to that developed in the exchange rate based stabilization literature (i.e. Calvo and Vegh (1993, 1994)). He argued that non- liquidity constrained consumers increased their durable consumption due to their expectations regarding the effects of the trade reform on import prices and the temporariness of the high level attained by the real exchange rate. Finally, Steiner (1995) attributed to fiscal and monetary policies the fall in the private saving rate. On the fiscal front, the tax burden was increased to prevent a deterioration of the fiscal stance caused by the trade and political reforms (i.e. fiscal decentralization and the enhanced role of the government in the provision of justice and security)2. The monetary stance was modified in 1992 when the costs and limits of a two-year sterilization policy were evident. As a consequence, interest rates were drastically reduced, international reserves began to be intermediated by the financial system, credit increased, borrowing restrictions were relaxed and a so called consumption boom seemed to have occurred. Despite the variety of explanations given to the recent decline in Colombian private savings, most analysts have not looked closely to the data and have failed to recognize that the apparently low level attained by private savings in the 1990s might reflect just data inadequacies. The paper is organized in the following way. Section 2 discusses some methodological issues and explains two important corrections that can be easily made to the saving measures presented in the Colombian National Accounts. In particular, it adjusts the saving measures by net capital gains from revaluations of assets and liabilities due to inflation and changes in the real exchange rate. As explained by Serven and Schmidt-Hebbel (1996), this exercise is useful because it constructs a saving measure consistent with net wealth changes and, therefore, with consumption theory. Section 3 makes the comparison between the adjusted and traditional measure of private savings. This shows that both of them have declined in recent years. However, whereas in 1992 the 2 The former had a negative impact on revenues due to the reduction of the import surcharge and the latter increased public consumption from 10% of GDP in 1990 to almost 15% of GDP in 1994. 2  traditional measure attained its lowest level since 1950 and was 3.5% of GNP below the average of the period 1951-1993, the adjusted measure was 0.5% of GNP below that average. Section 4 describes the behavior of the main components of Colombian private savings, an exercise that is very useful to understand its recent behavior. It shows whether consumption is more important than disposable income to explain changes in private savings and identifies which components of consumption/disposable income matter most (i.e. durable or non-durable consumption, interest or non-interest income). The main findings and conclusions are presented in Section 5. 2. Methodological Issues The Colombian National Accounts contain valuable information when compared to those of other developing countries. In fact, they have been constructed since 1950 and consumption can be divided between durable and non-durable goods since 1965. In addition, savings can be divided between corporations and households, and corporate savings can be divided between public and private enterprises ensuring that the coverage of the public sector does not distort the measures of public and private saving and disposable income3. Nevertheless, these accounts suffer from many of the problems mentioned by Schmidt-Hebbel and Serven (1996). Among them it is worth mentioning the following: (i) Private consumption is the residual that balances the income-expenditure identity and saving is the ultimate double residual. Therefore, it reflects all the inaccuracies incurred in estimating income and expenditure. In the case of Colombia these problems might be important due to the presence of drug trafficking. On the one hand, income might be underestimated since it does not take into account a large percentage of the value 3 However, between 1950 and 1969 the National Accounts were constructed by the Colombian Central Bank using a different methodology than those constructed by the National Statistical Department (DANE) for the period starting in 1970. The main differences between the two series are: (i) A wider coverage of the public sector in the National Accounts constructed by DANE; (ii) the private saving measure calculated by the Central Bank and DANE are net and gross measures, respectively. However, the National Accounts constructed by the Central Bank present the depreciation of physical assets and, therefore, the gross saving measure can be easily calculated. The Graphs and Tables presented in this paper are consistent with the gross measure of savings. For a detail description of the differences between the National Accounts constructed by the Central Bank and Dane see: Resumen de la Misi6n de Expertos Franceses (1984), Lora (1987), Agudelo (1991). 3  added generated by that sector. However, elements of expenditure, such as investment, include some of the activities undertaken by drug lords. Given that consumption is a residual, it is likely to be underestimated. (ii) Durable consumption is classified as consumption rather than investment, but theory suggests that an adequate consumption measure should be based on flow of services of presently owned consumer durable goods. This problem could be particularly relevant during trade reforms, such as the undertaken by the Colombian government in the early 1990s. In fact, it might be argued that the trade liberalization caused a once and for all increase in the stock of durable goods since they were repressed before the reform. (iii) The balance of payments suffers from significant classification problems, implying measurement inadequacies of external savings and, consequently, of domestic saving. For example, between 1992 and 1993 many Colombian economists argued that during the period 1990-1992 the current account surplus was overestimated because capital inflows were hidden as current account transactions (i.e. through over-reporting net exports or transfers of Colombians abroad). However, according to their estimations this classification problem never surpassed 1% of GDP (Steiner (1995)). Furthermore, even if the problem was larger, it might have been offset by an underestimation of net exports. In fact, if part of the revenue coming from illegal drug exports is held abroad, Colombian savings could be larger than the actual figures since those exports might not be completely compensated by smuggled goods. (iv) Private (public) saving is under (over) estimated because interests paid on total external debt are subtracted from private disposable income. This problem is accentuated because the interests received by the private sector from the public debt it holds are ignored. (v) In the National Accounts saving is a current account measure but consumption theory is based on a capital account measure of saving in which it is computed as the change in net wealth. To ensure a correct diagnosis of the recent behavior of private savings, it is important to correct these inadequacies of the data. Solving the first three problems is beyond the scope of this paper but the last two can be removed easily. In particular, the 4  current account measure of savings can be consistent with that based on net wealth changes if the former is adjusted by net capital gains from revaluations of assets and liabilities due to inflation and changes in the real exchange rate (Arrau and Oks (1992), Schmidt-Hebbel and Serv6n (1996)). If problems (iv) and (v) are corrected, the traditional macroeconomic identity, which suggests that gross investment is financed by national and external savings, takes the following form: I = (Y+ONFS + TR*r - P[r*(l+E) +c]fpr-i + Prb-i -P(7r/1+ 7r)h-i - T - TR- OR+ S- C) + (T + TR +TR*, -P[r*(1 +E) +