Why is sector reform so unpopular in Latin America?
[FIGURE 1 OMITTED]
In this article, I present evidence that sector reform is unpopular even when it is beneficial to most actors. This counterintuitive conjunction may occur because sector reforms have not been designed to be politically sustainable. Better design may make reforms more politically viable, even if less economically coherent. Dislike of even beneficial sector reforms, however, may have deeper roots. I advance the premise that many Latin Americans view pro-market reform as an ultimatum game: even when they benefit, they view their gains as unfair because the politicians and businesspeople who decide on the distribution of benefits keep a much larger share of the gains for themselves. If this view is correct, then the unpopularity of sector reforms is rooted in deeper institutional failures.
Some observers assume that reforms such as utility privatization are unpopular because most people are made worse off: However, Nellis (2003) reviews a number of empirical studies of the effects of infrastructure privatization in Latin America, all of which conclude that privatization improved financial and operating performance in most firms, relaxed constraints on new investments, extended coverage and access to services, and generally enhanced the quality of services. McKenzie and Mookherjee (2003) find that privatization of utilities in Argentina, Bolivia, Mexico, and Nicaragua had largely positive consequences for consumers (see table 1). Access increased, in some cases dramatically. Prices did go up in half the cases, but they went down in the other half, and because access increases have a much larger impact on consumer welfare than price increases, the reforms had a largely positive effect on consumer welfare. Sector reforms generally reduced inequality or left it almost unchanged. As for jobs, large layoffs occurred in Argentina and Mexico, but layoffs in these capital-intensive sectors had limited impact on economywide unemployment, especially over the medium term. Where layoffs were large, a significant percentage of the unemployed were reemployed in the same sector within five years: 45-50 percent in Argentina and 80-90 percent in Mexico.
Welfare calculations are vulnerable to assumptions about elasticities and counterfactuals, but other evidence points to benefits. In Mexico and Chile, privatization and regulatory reform of state electricity, telecommunications, and airline industries improved total-factor productivity and service quality, as well as consumer welfare, worker welfare, and total welfare (Galal et al. 1994). Similar outcomes are reported for urban water in Buenos Aires and Santiago (Shirley 2002). Studies of telecommunications reforms in large samples of developing countries find that privatization combined with competition increased access to services (Petrazzini and Clarke 1996; Ros 1999; Wallsten 2000). Ramamurti (1996) finds that privatization of the railroads in Argentina improved productivity and service and reduced costs.
Negative perceptions might have arisen because reforms had adverse effects on the most vulnerable groups, but in fact such adverse effects did not occur. Some observers maintain that privatization and competition leads profit-oriented utilities to "cream skim"--to serve the most profitable customers and ignore the less-profitable poor, who require new investment to give them access to the service. The poor are usually the ones who did not have access before reform, and, because of expanded access, their welfare increased in all the cases shown in table 1. Others suggest that even if access is expanded, cost-recovery pricing makes the service unaffordable to the poor. This view is challenged by Clarke and Wallsten (2003), who analyze household-expenditure surveys and find that consumption of services by the poor increased after reforms. For example, even though electricity prices increased after reform in Brazil, Colombia, and Peru, access by all households and especially by poorer households increased, sometimes substantially. This evidence matches their evidence for water supply. Galiani, Gertler, and Schargrodsky (2005) find that child mortality fell by 5-9 percent in the Argentine provinces that privatized water, thanks to a reduction in infectious and parasitic diseases. These effects were largest in the poorest regions, where access and water quality were the lowest before reforms.
Yet even where the poor seem to benefit, surprisingly large numbers do not view privatization as beneficial (figure 3). It is not surprising that most middle-class consumers-who probably already had service and now must pay more for it--have negative opinions of privatization. More surprising is that the lower-class support for privatization is so small--smallest in Argentina and Chile, the two countries that reformed the most.
[FIGURE 3 OMITTED]
The Buenos Aires water concession provides insights into how outcomes that seem almost Pareto optimal can still generate controversy (sec Alcazar, Abdala, and Shirley 2002). The concession contract signed with a French private operator in December 1992 led to a substantial increase in investment: as a result, 1.5 million additional people had access to piped water, and 583,000 to sewerage by 1998. Unlike in many other cases, water prices did not go up; they went down. Tariffs were reduced by almost 27 percent at signing, and, despite subsequent increases, they probably did not regain real 1992 levels for five years. Total domestic welfare in Buenos Aires increased by $1.7 billion compared to a counterfactual with no reforms, and 80 percent of this gain went to consumers (according to estimates in Alcazar, Abdala, and Shirley 2002). Workers also gained because layoffs were done through voluntary early retirement, and the workers who remained received 10 percent of the shares in the privatized water company. (4) Nevertheless, positive opinions of the concession fell from 36 percent in 1993 to 18 percent in 1997, and negative opinions rose from 33 to 52 percent (Centro de Estudios Union para la Nueva Mayoria, cited in Frade and Sohail Khan 2003, 275). (5)
Sector reforms seem to be unpopular despite, rather than because of, the consequences. McKenzie and Mookherjee (2003) give three reasons why sector reforms might be unpopular:
1. Inadequate public information about the benefits of reform, in part because benefits are widely diffused and less visible, whereas losses are highly concentrated and very visible, combined with humans' psychological tendency to dislike losses more than they value gains.
2. Association of privatization with other unpopular and perhaps painful pro-market reforms, such as fiscal contraction and trade liberalization.
3. Beliefs that basic services such as water or electricity should not be determined by the profit calculus of multinational corporations, combined with pessimism about the ability of the market, media, and regulators to compel private enterprises to serve the public interest.
Perhaps all these factors have contributed to the public's negative perception of sector reform to some extent. The first two are difficult to deal with. Inadequate information may seem to be a tractable problem, but government public-relations campaigns may not be trusted, and government may not be able to overcome some kinds of information imperfections. (6) Diffuse benefits and concentrated losses are inherent in sector reform. Further, as indicated earlier, reforms are unpopular even where there are few losses. The association of sector reform with other painful market reforms is often unavoidable: a macroeconomic crisis is usually necessary to motivate the change in the political status quo that is a prerequisite for sector reforms.
The third reason for the unpopularity of reform seems remediable, on its face. More attention to political sustainability and to safeguarding the public interest in the design of sector reforms can reduce public suspicion of privatization and pro-market reform. Once again the Argentine water concession is instructive. Although the concession's unpopularity arose in part because of some of the intractable reasons listed earlier, its design was also at fault. The provisions of the contract and the structure of the regulator fed customers' view that their interests were threatened and that no politically independent body was protecting them.
One design flaw was the tariff system. The concession did not change Buenos Aires's opaque and inefficient tariff regime based on a flat rate adjusted by five different factors (size of property; location of property; and size, type, and age of dwelling and other constructions). This tariff system made it impossible for customers to decipher, contest, or control their bills. Tariff increases were mandated whenever a detailed and complex cost index increased by more than 7 percent, resulting in lumpy, nontransparent, and controversial rate hikes. Because the concession was awarded to the operator that bid the largest tariff reduction, the public tended to view any tariff increase as evidence that the original bargain was a bad deal for the country.
The politicized regulatory agency was a second problem. The agency's board consisted of representatives of the federal, provincial, and local governments. Contrary to the intent of the agency's charter, board seats were treated as political appointments, and directors changed whenever the political leaders changed. Political rivals controlled the different levels of government, and partisan disputes often paralyzed the board. When directors did cooperate, they often did so only to engage in politically motivated horse trading and sometimes to make legally questionable but politically popular decisions that the executive branch subsequently overturned.
A third problem was the lack of regulatory independence. The executive branch did not just reverse decisions, including overturning fines levied on the operator for failure to meet the terms of the contract; it also on several occasions bypassed the regulator and reached agreement directly with the company. (7) Although executive intervention helped to preserve the credibility of the contract, it also signaled consumers that no independent regulator was adequately safeguarding their interests.
A final flaw was the decision to charge new users for the cost of extending the secondary network to serve them. Under state ownership, all users had shared this cost. Although the company was required to help new customers finance the charge, the operator gave new consumers only two years to pay what was a large sum for low-income households. (8) This "infrastructure charge" roused resentment, and most new customers refused to pay it. The contract was eventually renegotiated, and the charge was eliminated and replaced with rate surcharges through a highly controversial negotiation that fed further public suspicion.
Based on the Buenos Aires water concession, it may be possible to improve acceptance of sector reforms by designing them with greater sensitivity to political sustainability. Phasing in price increases gradually; awarding the concession on the basis of bids for new investment instead of price reductions; designing a transparent and efficient tariff regime and installing meters; spreading the cost of the expansion across all users or requiring better financing terms for poor customers; creating a less politicized, more independent regulator and a separate, more neutral appeals court--all these choices might have made the concession more politically sustainable. Of course, some design flaws that seem evident with hindsight may not be obvious at the time of reform. More important, many key actors are not likely to be sensitive to political sustainability. The technicians and advisors who design reforms are focused on and skilled in economic and technical rationality, not in political sustainability. Foreign advisors and donors are unlikely to understand the demands and compromises required by local circumstances. Politicians may have too short a time horizon to care about long-run acceptance or may believe with some justification that a reform with widespread benefits will eventually be popular regardless of its flaws. These considerations suggest that the reform team needs to include both analysts and former political actors with a deep understanding of the local political economy.
Design flaws notwithstanding, it is still a puzzle why consumers so dislike a reform that provides most with benefits and few with losses. Most political economy models assume that reforms with large positive payoffs, with many winners and few losers, will eventually become popular. Fernandez and Rodrik (1991) show how uncertainty over the distribution of gains and losses can prevent reforms with positive payoffs from being implemented, but they assume that such reforms will become popular if they are implemented. The three reasons given by McKenzie and Mookherjee (2003) seem inadequate to explain the large and growing unpopularity of the Buenos Aires water concession.
The dislike of the Buenos Aires concession and of other sector reforms in Latin America more broadly may stem from deeply rooted and widely generalized distrust of market forces and government safeguards. Although ideology stoked by populist politicians and interest groups may lie at the heart of this distrust, the negative view is also in part a response to reality: even when consumers gain, politicians and their cronies make much bigger gains. My hypothesis is that the source of this distrust is not the gain, but the game. There is wide acceptance of differential gains in market economies when people regard the gain as somehow proportionate or "deserved," especially when they believe that they might also have opportunities for profit. My premise is that many Latin Americans see sector reform as an ultimatum game. (9)
In the ultimatum game, two strangers are given a sum, say $100, to split. Player A, the Proposer, can offer any share he wishes, from nothing to $100, to player B, the Decider. If B accepts A's offer, both players keep their shares of the money; if B rejects A's offer, neither player receives anything. According to rationality assumptions, B should accept any positive offer, however small, and A should offer as little as possible. B had nothing before, and it is rational--by this reasoning--to prefer something to nothing. When the game is actually played, however, the Deciders regularly reject offers they consider "unfair." They turn down surprisingly large offers, and the game converges at about $40 for B and $60 for A. Although variants of the game produce different proposals, the Deciders persist in their preference for an "equitable" division of the pie and in their willingness to forgo a share in order to punish Proposers for "unfair" offers.
It is beyond the scope of this article to test this premise. Nonetheless, there is evidence of a broad sense of mistrust in Latin America, greater than in other developing areas, a mistrust that is not a product of sector reform, but that precedes and surrounds it. Data on trust from the World Values surveys (1995-96) suggest that Latin Americans have much higher levels of generalized mistrust than people do in other developing countries (figure 4). Less than one-third of Latin Americans surveyed in 2002 believed that they can trust the people who run the government of their country to do what is right, and on average those respondents believed that 71 percent of civil servants were corrupt (Latinobarometro 2002).
[FIGURE 4 OMITTED]
If sector reform--or market reform more broadly--is indeed perceived as an ultimatum game, then its unpopularity has deep institutional roots in laws and norms of behavior that give politicians and their business cronies access to large windfalls and curtail less-privileged groups' opportunities for advancement or redress. For sector reform to have a future in Latin America, it is therefore not enough for it to benefit consumers and to safeguard their interests through better regulation and contractual design. It must also allay the deeper sense of injustice. Unfortunately, we have limited understanding of how to change the embedded institutions that foster corruption, reduce economic opportunities, and deny legal protection for many citizens (Shirley 2003). The underlying institutions responsible for Latin America's unequal political and economic rights are deeply rooting in the colonial and even precolonial past (Engerman and Sokoloff 2002). Despite the promises made by some in the foreign-aid community, outsiders probably cannot change these embedded institutions short of occupying the country, if then.
Opportunities for change do arise, however, and at those moments the design of more sustainable reforms will depend on an understanding of formal rules and informal norms and of local institutions' history, failings, and incentives. Local scholars have a comparative advantage in understanding the idiosyncrasies of their country's institutions and in devising sustainable reforms, but they are often isolated and underfunded and tend to leave the country or move away from research. This problem can be dealt with, but only slowly: wholesale schemes to build local knowledge fall victim quickly to the same institutional failings that undermine other well-meaning reforms.
Acknowledgment: This article was prepared originally for presentation at the Stanford Center for International Development (SCID) Conference on Sector Reform, November 13-15, 2003.
Alcazar, Lorena, Manuel A. Abdala, and Mary M. Shirley. 2002. The Buenos Aires Water Concession. In Thirsting for Efficiency: The Economics and Politics of Urban Water System Reform, edited by Mary M. Shirley, 64-102. Oxford: Elsevier Science.
Apoyo, Opinion y Mercado. Various dates. Unpublished opinion polls, April 1989 to December 1997.
Camerer, Colin, and Richard Thaler. 1995. Anomalies: Ultimatums, Dictators, and Manners. Journal of Economic Perspectives 9, no. 2: 209-19.
Clarke, George R.G., and Scott Wallsten. 2003. Universal Service: Empirical Evidence on the Provision of Infrastructure to Rural and Poor Urban Consumers. In Infrastructure for Poor People: Public Policy for Private Provision, edited by Penelope J. Brook and Timothy C. Irwin, 21-76. Washington, D.C.: World Bank.
Engerman, Stanley L., and Kenneth L. Sokoloff. 2002. Factor Endowments, Inequality, and Paths of Development among New World Economies. Economia 3 (fall): 41-109.
Estudios Mora y Araujo, Noguera y Asociados. 2002. Unpublished opinion polls cited in Alcazar, Abdala, and Shirley 2002.
Fernandez, Raquel, and Dani Rodrik. 1991. Resistance to Reform: Status Quo Bias in the Presence of Individual-Specific Uncertainty. American Economic Review 81, no. 5: 1146-55.
Frade, Tomas, and M. Sohail Khan. 2003. Making Regulatory Mechanisms Work: Lessons from Cases of Private Sector Participation. Water Policy 5: 269-88.
Galal, Ahmed, Leroy Jones, Pankaj Tandon, and Ingo Vogelsang. 1994. Welfare Consequences of Selling Public Enterprise: An Empirical Analysis. New York: Oxford University Press for the World Bank.
Galiani, Sebastian, Paul Gertler, and Ernesto Schargrodsky. 2005. Water for Life: The Impact of the Privatization of Water Services on Child Mortality. Journal of Political Economy 113 (February): 83-120.
Guth, Werner, Rolf Schmittberger, and Bernd Schwarze. 1982. An Experimental Analysis of Ultimatum Bargaining. Journal of Economic Behavior and Organization 3: 367-88.
Latinobarometro. 2002-03. Available at www.latinobarometro.org.
McKenzie, David, and Dilip Mookherjee. 2003. Distributive Impact of Privatization in Latin America: An Overview of Evidence from Four Countries." Economia 3, no. 2: 161-218.
Nellis, John. 2003. Privatization in Latin America. Center for Global Development Working Paper no. 31. Washington, D.C.: Center for Global Development.
Petrazzini, Ben A., and Theodore H. Clarke. 1996. Costs and Benefits of Telecommunications Liberalization in Developing Countries. Hong Kong: Hong Kong University of Science and Technology.
Ramamurti, Ravi. 1996. Testing the Limits of Privatization: Argentine Railroads. World Development 25: 1973-93.
Ros, Agustin J. 1999. Does Ownership or Competition Matter? The Effects of Telecommunications Reform on Network Expansion and Efficiency. Journal of Regulatory Economics 15: 65-92.
Shirley, Mary M., ed. 2002. Thirsting for Efficiency: The Economics and Politics of Urban Water System Reform. Oxford: Elsevier Science.
--. 2003. What Does New Institutional Economics Tell Us about Development? Paper Presented at the Annual Meeting of International Society for New Institutional Economics, Budapest, Hungary, September 2003.
Spiller, Pablo T., and Mariano Tommasi. 2000. The Institutional Foundations of Public Policy: A Transactions Approach with Application to Argentina. Working Paper. Presented to the annual meeting of the International Society for New Institutional Economics, Tubingen, Germany, September 22-24, 2000.
Thaler, Richard. 1988. Anomalies: The Ultimatum Game. Journal of Economic Perspectives 2, no. 4: 195-206.
Wallsten, Scott. 2000. Telecommunications Privatization in Developing Countries: The Real Effects of Exclusivity Periods. Stanford Institute for Economic Policy Research (SIEPR) Policy Paper. Stanford, Calif.: SIEPR.
(1.) Data are taken from Latinobarometro, which polls a sample of 18,526 respondents in seventeen Latin American countries annually.
(2.) The opinion polls for Peru are Apoyo, Opinion y Mercado (April 1989-December 1997), and those for Argentina are from Estudios Moray Araujo, Noguera y Asociados (2002); the information on opposition to privation projects is cited in Nellis 2003.
(3.) For example, respondents who hold a negative opinion of privatization in Lima cite increases in utility prices and layoffs as its main negative effects (see, for example, Apoyo opinion surveys, various years).
(4.) There is also evidence that service improved. For example, response times to complaints were much quicker, less than one-third of the time required before privatization, and three years after the concession more than half of all consumers had adequate water pressure, compared to 17 percent before the privatization (Alcazar, Abdala, and Shirley 2002).
(5.) Approximately 27 percent were neutral, and 3 percent had no opinion.
(6.) For example, poor people in Buenos Aires whose homes were not connected to the water and sewer systems often relied on shallow wells and septic tanks or privies and suffered higher disease rates from bacterial contamination. These users may not have been aware that their water was contaminated.
(7.) Decisions could be appealed to the Supreme Court, but in Argentina the Supreme Court is not seen as neutral; since Peron, it has been famously vulnerable to politically motivated interference by the executive (see Spiller and Tommasi 2000). Justices are regularly impeached, and one of newly elected President Menem's first acts was to increase the number of Supreme Court judges from five to nine (Alcazar, Abdala, and Shirley 2002, 83).
(8.) The infrastructure charge was between $615 and $912 on top of the usual connection charge ($492-$616; all numbers are 1995 U.S. dollars). The average monthly income in the poorer areas was $200-45. See Alcazar, Abdala, and Shirley 2002, 85.
(9.) See Guth, Schmittberger, and Schwarze 1982; Thaler 1988; and Camerer and Thaler 1995.
Mary M. Shirley is president of the Ronald Coase Institute.
Table 1 Effects of Privatization of Utilities on Consumers Change Change Effects on in Access in Prices Consumer % % Welfare Argentina Electricity 3.8 -16.1 positive Telephone 33.3 -32.5 positive Water -16 Bolivia Electricity 2.9 -8.3 lgly positive Telephone 21.6 26.2 lgly positive Water 14.3 La Paz, El Alto -10.5 positive Cochabamba 43 negative Mexico Telephone 41.4 47.9 n.a. Water 6.8 9.2 n.a. Nicaragua Electricity 11.7 24.2 mixed Effects on Welfare of Poor Effects on Consumers Inequality Argentina Electricity positive reduced Telephone positive reduced Water Bolivia Electricity positive reduced Telephone positive increased Water La Paz, El Alto positive reduced Cochabamba positive none Mexico Telephone n.a. n.a. Water n.a. n.a.. Nicaragua Electricity lgly positive sm. increase Source: Adapted from McKenzie and Mookherjee, 2003.