Targeting Public Utility Services in LAC: Some Lessons from Colombia
Promoting access to public utility services is a common goal of the public sector of most developing countries, but those governments unable to guarantee these services to their population bear very high social costs due to the negative externalities linked to their shortage. In order to promote these policies, most Latin American countries provide supply or demand subsidies under schemes designed to balance their own technical and political economy constraints, usually distorted by inconvenient initial conditions.
Previous work has shown that several countries in Latin America and the Caribbean (LAC) give subsidies to public utility services, but these can take different forms. In some cases, subsidies are similar for the whole population through flat subsidized charges; while there are also cross subsidies conditional on consumption or quality of life levels. Investment subsidies are yet another form. In addition, some countries target their public utility subsidies geographically, as it is the case in Colombia. There are six socioeconomic strata in Colombia. Strata one through three receive subsidized public utility services; the fourth pays the marginal cost of services; and strata five and six, along with the commercial and industrial sectors, pay contributions.
Since the socioeconomic strata were created in Colombia, these have been used to set differential rates for taxes, college and university tuition, social housing, and other social services. The socioeconomic stratification became one of the determining criteria of the poverty and welfare measurement methodologies used to target most public spending in the country.
There are several studies that analyze how well targeted are subsidies to public utility services in LAC countries, including Colombia (See Komives et al. (2006) and the references cited by Medina and Morales (2007), among others.). These studies reveal various limitations of current targeting systems. For instance, targeting systems merely based on consumption levels are among the most deficient, and those that in addition use geographic targeting improve the results to some modest extent. According to these studies, targeting systems that assign subsidies to families based on scores of proxy-means tests perform the best, although even in those cases there are limitations.
Assessing the Actual Incidence of Public Utility Subsidies
Besides the focus placed by these studies in assessing which socioeconomic levels are contributing to fund subsidies to public utility services, and to which are those resources being allocated, there has been a much lower emphasis on how much of those economic resources actually remain in the pockets of households in the form of a subsidy, once distortions on other factors are considered.
That is the question assessed by Medina and Morales (2007), who analyze the Colombian case to quantify the subsidies received by each household net of the distortions such subsidies introduce to relative housing prices. They test the hypothesis that subsidies or contributions to public utility services play a role in determining housing prices in Colombia. To quantify the incidence of these subsidies on housing prices, they estimate hedonic price equations, applying a regression discontinuity approach as their identification strategy. The empirical work uses information from Bogotá, but since the institutional framework that governs the subsidies targeting policy is the same countrywide, they conjecture that their findings apply across Colombia’s main cities, and arguably to other LAC countries also using geographic targeting.
Their approach assumes that the housing market takes into account the subsidies or taxes that residents of certain dwellings will receive or pay. For example, the only difference between two identical houses located in different socioeconomic strata but just across the street from each other would be their stratum, and therefore, the cost to their occupants of residential public utility services. In that case, occupants of the house receiving the highest subsidy would be willing to pay for their house the same amount than their neighbor, plus the net present value of the flow of additional subsidies expected to be received, net of their deadweight loss.
Medina and Morales found that the estimated increment in the house value due to subsidies is similar in magnitude to the present value of the flow of subsidies, discounted at reasonable market rates. They conclude that in Colombia the goal of subsidy financing to the poorest population through government spending on public utility services is not being achieved. The ultimate effect of most of the government spending in this case is the distortion of housing prices.
Even though Colombia’s public sector annually distributes approximately 0.7 percent of its GNP in subsidies for public utility services, they end up being a new feature to houses located in certain areas, which proportionally increases their market price. Households receiving public utility subsidies end up using them to pay a premium for the houses they live in, either in the form of a higher house value, or rent.
The Effect of Eliminating Stratification as a Targeting Mechanism
Who would be the winners and losers if the current system of targeting subsidies for public utility services in urban Colombia was abolished? If a household was a tenant, after the targeting system was abolished it would receive no subsidies, but it would end up paying a lower rent by an amount similar to the subsidy previously received, thus staying relatively unaffected in net terms. If the household was the owner of a rented house, its wealth would decrease (increase) by an amount equivalent to the present value of the subsidies (taxes) on public utility services previously received (payed) through the higher rent paid by its tenant.
The Political Economy of the Reform
Paradoxically, although the current subsidy scheme only distorts housing prices, individuals would not be indifferent to abolishing it given the distribution of owner and tenant households by socioeconomic stratum in Bogotá. The current subsidy scheme has required about half of the households living in strata 1 to 3 who are homeowners, to pay a price for their houses that is higher than the price they would have paid in the absence of the policy. Eliminating the current subsidy scheme would be equivalent to expropriating that part of the value these households paid for their houses under the previous conditions.
Furthermore, since nearly 90 percent of owner households live in strata 1 to 3, the median voter would be a loser if the current subsidy scheme were abolished, creating a political economy constraint to the reform. Put in another way, if the current scheme were abolished, homeowners in the poorest strata would require compensation in an amount equivalent to the distortion that the government caused with the scheme itself, which might be fiscally unaffordable. Clearly, eliminating the current subsidy scheme would be regressive. In addition, the government is not achieving what it sought to achieve through the subsidies: “the subsidies scheme will be provided so that low income people can afford the rates of domiciliary public utility services to cover their basic needs.” Under the current scheme, the goal of generating savings to the poorest via subsidies to public utility services is not being achieved due to the offsetting effect of higher housing prices and rents.
As the current subsidy scheme goes on, the government continues assigning nearly 0.7 percent of GDP in gross subsidies to households in strata 1, 2 and 3--0.3 percent of this number coming from its budget, and the remaining 0.4 percent from households living in strata levels 5 and 6, and the commercial, public, and industrial sectors. The policy ends up doing nothing but distorting relative housing prices and the efficient assignment of factors in the productive sector.
In addition, the current stratification scheme offers individuals a perverse incentive to be targeted by the public authorities in charge of assigning subsidies. According to Colombia’s Quality of Life Surveys of 1997 and 2008, this led, among other things, to an increase between 1997 and 2008 of 200 percent in the number of households living in stratum 1, and of 30 percent in the number of households living in stratum 2; while the number of households living in strata 4 to 6 decreased by 4.4 percent. Furthermore, stratification also leads to segregation by income, which is usually very hard to reverse.
Other schemes, based on instruments like SISBEN could be considered for targeting subsidies for public utility services, as it has been suggested by previous studies. The Colombian government invests about 1.0 percent of GDP a year on the health insurance program for the poor, targeted with SISBEN1, so that resources currently used to subsidize public utility services would be enough to increase that budget by 70 percent.
In addition, according to Medina and Morales (2008), deadweight losses associated with electricity and water subsidies amount to nearly 5 percent and 10 percent, respectively; which is equivalent to US$35 million a year. In a scenario without subsidies, in order to keep households’ welfare unaltered, they will need to receive a transfer of the amount of income necessary to keep their utility as well off as it was in the scenario with subsidies. Even in that case, accounting for the total transfer, Colombia would save up to US$35 million a year previously spent on efficiency losses, much more than what would be required to keep the SISBEN1 mechanism working, estimated at about US$7 million every five years.
On the one hand, if governments of the region wanted to eliminate stratification as its targeting mechanism, an option for dealing with the potential political economy constraints would be to do it over a very long period of time, say twenty years, while simultaneously introducing another mechanism, such as their own proxy-means tests.
Revising targeting schemes of subsidies to public utility services is a task worth pursuing, mostly in LAC countries, many of which have elected leaders that use such schemes to strengthen their popularity and perpetuate their leadership. If they are going to remain in power, let’s try to make sure that it is by employing useful means.
References
Black, Sandra E. (1999) “Do Better Schools Matter? Parental Valuation of Elementary Education” Quarterly Journal of Economics (May) Vol. 114 No. 2, pp. 577-599.
Komives, Kristin; Halpern, Jonathan; Foster, Vivien and Adbullah, Roohi (2006) “Water, Electricity, and the Poor: Who Benefits from Utility Subsidies?” The World Bank, Washington D.C.
Medina, Carlos, and Leonardo Morales (2007) “Stratification and Public Utility Services in Colombia: Subsidies to Households or Distortion of Housing Prices?” Economia, Spring, pp. 41-100.
Medina, Carlos, and Leonardo Morales (2008) “Demanda por Servicios Públicos Domiciliarios en Colombia y Subsidios: Implicaciones sobre el Bienestar” Desarrollo y Sociedad, No. 61, Universidad de los Andes, Bogotá.
1 The System for Selecting Beneficiaries for Social Spending (SISBEN, in Spanish) is a method for targeting social spending based on assessment of living conditions of individual families.
