Topic Area: Water Management
Geographic Area: Latin America (Mendoza, Argentina)
Focal Question: How does Mendoza's water management system fare with external market factors?
Sources:
(1) Lee, Terrence Richard. "Managing Water Resources in Latin America." Natural Resources Journal, Vol. 30, (1990): 581-607.
(2) Baars, E., A.P.M. Bastiaansen and M. Menenti. "A User-Oriented and Quantifiable Approach to Irrigation Design." Water Resources Management, Vol. 9 (1995): 95-113.
Reviewer: Tom DeCoff, Colby College '97
Review:
Water resource management in Latin America is currently problematic. The difficulties experienced can be traced back to the initial establishment of water rights. In the fifteenth century the Spanish colonizers replaced the pre-existing social control of water in Latin America with their own European-based system. The system again changed when the countries of Latin America gained their independence. Their newly drafted constitutions declared water to be a public good and a function of local, provincial governments.

Mendoza, Argentina is just one province in Latin America that is experiencing difficulties, to which Mendoza's climate is partially responsible. The area is primarily arid, receiving less than 200 millimeters (around 7.8 inches) of rainfall annually. Four rivers are tributaries to a river that does not even reach the sea. The northern and central parts of Mendoza are not equipped with irrigation regulatory structures, but the southern area has an elaborate series of hydroelectric dams.

Despite the arid climate, Mendoza is economically reliant on agriculture, specifically grape cultivation. It is estimated that grapes, for wine production, constitute 75% of Mendoza's agricultural activity. The grape dependency was powered by two distinct forces: 1) the Mendoza-Buenos Aires railroad constructed at the end of the nineteenth century, bringing in an influx of Spanish and Italian grape cultivation specialists and 2) an increasing demand for wine. By 1930, a restricted water supply was inhibiting agriculture.

The discovery and eventual exploitation of groundwater between 1930 and 1970 eased the water deficit, despite an increasing northern demand for water for waste disposal (deviating from almost exclusive previous use for irrigation). The water "abundance" led to prosperous agriculture which led to sustained economic growth and high regional incomes. This scenario created the illusion that the water management system in Mendoza was both effective and efficient. That illusion soon faded.

One contributory factor was a declining economy. Domestic demand for wine had declined because of economic recession and changing consumer preferences in favor of wine substitutes such as beer and soft drinks. Subsequently, Mendoza's agriculture-intensive economy suffered.

The water administration system in Mendoza worsened the economic suffering. Water management, as a function of provincial government, involved two branches, the Departamento General de Irrigacion (DGI) and the Inspecciones de Cauce. The DGI is the primary branch and is responsible for the complex of works, including the river, the diversion structures (such as dams), the main canals and groundwater basins. Despite being one of the oldest water management institutions in Latin America, the DGI has had tremendous difficulties in adjusting to the economic changes evident in Mendoza.

One such problem is the collection of revenue. The DGI collects revenue by levying taxes on farmers. These lump sum taxes are for water access not for usage and are economically problematic. When farmers suffer depressed production (due to grape demand decreases, for example) the DGI's revenue suffers, reducing its future capacity to improve water diversion and distribution. The inability of the DGI to effectively and efficiently provide to farmers compounds the DGI's inability for effective water management in a positive feedback loop fashion. Less money for the DGI today means less resources for the DGI to provide to farmers in the future, which leads to lower farmer productivity, which leads back to less money for the DGI.

Another problem generated by the DGI lies in groundwater regulation. The use of groundwater is fundamentally uncontrolled and little knowledge about its supply, quality and usage exists. Less than 1% of DGI's staff works in the administration of groundwater.

The rigid water rights structure in Mendoza yields further inefficiency problems. Nontransferable water rights are assigned to a piece of land in a specific amount whether used or not. Efficiency dictates that resources be allocated to the party that places the highest value on the resource; Mendoza's system prevents this. Even if the initial allocation of water were efficient, the current system has little capacity to respond to changing circumstances. Infrastructural problems such as bureaucratic budget controls and primitive means of data collection and storage (registers of water rights are maintained by hand and are separate for surface and groundwater rights) contribute to the DGI's demonstration of inefficient water management. Finally, the aforementioned tax system of farmers is an inefficiency causing agent in itself. Farmers pay for the right to use water (a flat fee) instead of paying for the actual usage (a marginal cost). Hence farmers have no economic rationale to conserve or to use water efficiently.

The diagnosis for the DGI's water management system in Mendoza is unfortunately grim. Ill-defined, rigid property rights, bureaucratic blockades and an uneconomical taxing structure plague the DGI's ability to effectively and efficiently administer water. Although Lee provides a prescription of better management through education and training of staff (especially in groundwater), the recognition of other water uses such as waste removal and a better integration of surface water and groundwater management, these are not likely to solve the problem as long as structural problems persist.

Baars develops a possible solution to the water rights rigidity problem in Mendoza. Interviews with farmers reveal that farmers would favor a water distribution system in which they could have more influence in the way water is delivered (i.e. have influence in the DGI's task). The maximum amount of water to be received annually would still be determined by the DGI. The mean maximum allowance to each farmer would essentially remain the same but would vary over the years only in accordance with total water availability. Total water availability would depend on snowfall in the mountainous regions. The system proposed by Baars would allow farmers to allocate his/her annual share of water over the successive months at their discretion. The sole limitation would be that any one monthly allocation must be limited to a certain volume per farmer, set by the management structure. The farmers would still be able to claim the amount of water to which their water right is entitled to, even if they do not cultivate the total area on which these rights are based. They could also economize on water costs by asking only for a volume that would satisfy their true water needs (based on crop needs) and pay only for what they ask for. Assuming that the costs of water are high enough to accurately reflect scarcity, then this "pay-for-what-you-use" system will provide incentives for water conservation and would make water distribution more flexible as water can be delivered in accordance with farmers' perceptions of water requirements of the cropping patterns. Excess water supply due to better availability can be avoided without under-supplying farms.

Baars' proposal appears to be a feasible solution to an evident problem in Mendoza. However, this solution assumes that some of the other problems are corrected. Knowledge of total water availability, for example, requires the better integration of groundwater and surface water management, a point mentioned by Lee. Additionally, the development of this kind of system would rely heavily on the (lack of) expertise addressed in Lee's request for better training and information. The inefficiencies in Mendoza need to be corrected. Baars' suggestion is a step in the right direction but other facets, such as those mentioned by Lee, must be addressed first.