The link between service privatization and price distribution among consumer types: municipal water services in the Spanish region of Catalonia

In the Spanish region of Catalonia the industrial sector sensibly lobbied for the privatization of municipality-produced water services in the period 1996 ^ 2002. This paper analyzes the extent to which this result is consequent with observed measures of cross-subsidies in favor of residential consumers. I run a treatment-effects regression on a sample of 133 municipalities in 2000 and 2001. Recent privatization has a positive effect on the weight that the local regulator attaches to industrial users' welfare when deciding water tariffs. I conclude that an increase of industrial customers' group size has a twofold effect on the cross-subsidy in favor of residential customers: a direct positive substitution effect and an indirect negative political support effect through lobbying for service privatization. doi:10.1068/c0635 (1)More precisely, my sample covers 133 municipalities in the Spanish region of Catalonia. Results of this paper are then extensive to Catalan municipalities. Other regions in Spain have sensibly different privatization patterns (see Bel, 2006) concerning the local water service. (2) Apart from my own work (Miralles, 2006) the only empirical work I am acquainted with that studies this dynamicity is Hefetz and Warner (2004), for American municipalities. They are particularly interested in the reverse-privatization phenomenon that is growing in the United States. (3) See Buller (1996) for previous European regulation and its effects on water service regulation in several European countries, above all, France and England. It is not clear, though, why an interest group should push for a policy that indirectly affects the tariff structure instead of directly putting pressure on the tariff formation. This motivates the present analysis, which tries to give robustness to the aforementioned results. If previous results are correctly interpreted, the privatization decision of a local water service in, say, the period 1996 ^ 99 must, on average, be linked to an alteration of the tariff policy in favor of industrial customers in further years. This alteration should be more intense than that which is generally observed in the full set of municipalities. As I show in this paper, this appears to be the case. There is a political economy rationale underlying the result. When the local regulator sets water prices, he or she is maximizing a weighted utility function. The lobbies' pressure has an influence on these weights, as suggested by the general theory of regulation of Stigler (1971) and Peltzman (1976). But altering the status quo between industrial consumers and residential consumers has political costs. Households (that is to say, voters) may not understand whey their tariff rise is relatively more than industrial-use tariff rises. Therefore, the local regulator thinks twice before making major changes in the cross-subsidization scheme. In order to minimize the associated political costs, the regulator needs a new environment that could justify major changes.(4) This is due to the temporary preeminence of a discourse in favor of service reforms and to the uncertainty surrounding the process. Local water services that have not recently undertaken major regulatory reforms do not have this environment. And privatization is indeed one of the major reforms in the local water service.(5) The reader might think that the finding of empirical evidence in favor of my hypothesis might be related to other explanations besides industrial lobbying. For instance, the municipality could have devised water-service privatization with the aim of pursuing financial balance of the service. According to the hypothesis that crosssubsidies in favor of residential users existed, the consequent rise in water tariffs would typically be higher for residential consumers than for industrial ones. For the reason described in the previous paragraph, such a change would need a context of major service reforms in order to avoid political suicide. Admittedly, this explanation cannot be empirically disentangled from my first hypothesis with the data at hand. However, it is apparent from Miralles (2006) that the industrial lobbies were influential in the water-service privatization decision during the period under study. This gives reliability to the industrial-lobbying approach. In any case, privatization is used to reduce cross-subsidizations, but the pressure to reduce these cross-subsidies could come from different agents.(6) A second point the reader could be suspicious of is whether there is a reason why privatization should be linked to lower political costs of reducing cross-subsidizations.(7) Indeed, let me be clear, perhaps redundantly, about this. I neither argue for nor contest the fact that cross-subsidies are lower in municipalities where the service is (4) Giuletti and Otero (2002) analyze the timing of major tariff structure changes in regulated industries in England and Wales. Concerning the water industry, it is observed in the majority of regional water authorities that major tariff structure changes started directly after the 1989 privatization. Ordon¬ ez de Haro (2002) considers the case of several local water utilities in the Spanish southern province of Malaga. Concerning privatization and tariffs, the author finds that observed privatization processes are followed by some `̀ unjustified'' price increase. (5) Admittedly, it is difficult to distinguish between these two issues: (1) privatization is a maneuver specifically designed to lead to a costless change in the cross-subsidization scheme, and (2) privatization as a policy that was undertaken for other reasons, but was then taken as a politically costless way to concede to industrial lobby's pressures. (6) I am grateful to a referee for bringing this idea to my attention. (7) Once again, I thank a referee for warning me about this possibility. 160 A Miralles