Environmental Economics: Water Demand Research

2011-2012

Research Activities
Economist Kerry Smith continues to lead a dynamic program of research in environmental economics, benefiting from sustained and meaningful interaction with stakeholders from the water management community. This research informs the design and evaluation of alternative price and non-price measures for promoting water conservation and evaluating climate-adaptation policies. In the past year, Smith and his team (including Zhao, Klaiber, Brent) have made progress on several related initiatives focused on economic and non-economic factors that influence residential water demand. They developed a new method for estimating price elasticities that can accommodate the complexities of increasing block pricing structures (Klaiber et al. in review, Land Economics).

Demonstrating his national leadership on this topic, Smith in May 2012 organized a conference held at DCDC on “Understanding Residential Water Use: New Approaches to Analyzing, Projecting, and Managing Demand.” The meeting convened leading economic and social scientists from around the country along with municipal water providers from major western cities including Phoenix, Mesa, Tucson, and Seattle. Presentations focused on cutting-edge research on price and non-price incentives; the economic value of residential water; and understanding behavioral responses to water policy. Scientists and policy makers engaged in interactive dialogue to identify opportunities for innovative collaborative research.

Research Findings
Economic Approaches to Analyzing, Projecting, and Managing Residential Water Demand
Economist Kerry Smith’s recent research has produced new approaches for analyzing, projecting and managing residential water demand. Applying their new model to data obtained through a cooperative agreement from the City of Phoenix, three significant findings emerged (Klaiber et al. in review). First, differences in seasonal weather conditions affect the estimated price elasticities and this effect is separate from the influence of weather in the months directly related to the quantity response. Second, price responses vary with the size of the water customer; large customers are typically less responsive to price changes than small customers. Third, seasonal differences in price responses can be larger than the prior research has implied.

2010-2011

Research Activities
Economist Kerry Smith continues to lead a dynamic research program focused on economic and non-economic factors that influence residential water demand. This research informs the design and evaluation of alternative price and non-price measures for promoting water conservation and evaluating climate-adaptation policies. In the past year, Smith and his team have been developing a structural model of residential land use for the Arizona “Sun Corridor” region, progressing as data have become available. The model will enable an integrated assessment of long-term policies that influence household water behavior. Special attention is being devoted to interactions between public land and water policies.

Two sets of research were completed during the first year of DCDC II. The first study exploits the systematic changes in price schedule with season and the panel structure of water records to measure price responsiveness without requiring detailed consideration of the increasing block rate structure. To our knowledge, this is the first time this strategy has been recognized. A second new contribution involves using the linked housing data and water meter records for Phoenix and Tucson water departments that resulted from cooperative data-sharing agreements with each water department. The research exploits the logic of propensity score matching to control for the structure and landscape influences on water demand and extracts the first “pure” price schedule assessment.

Research in progress has led to several intermediate insights. First, structural-modeling work has demonstrated an important interaction that influences preliminary estimates of households’ tradeoffs for enhancements in amenities. This interaction arises due to linkages between the spatial scale of the definition of a community and the spatial scale of the measure used for landscape-related amenities (including proxies for ecosystem services). A second, intermediate insight concerns the potential for important effects of endogeneity in measures of spatial amenities and un-observable variables describing site attributes due to household residential choices in response to the features of the alternative neighborhoods available. Research to date has adapted the Abbott-Klaiber (2011) framework to test the importance of these effects.

Research Findings
Economist Kerry Smith’s research on structural demand models for residential water demand (Strong and Smith 2010) has been extended in two ways. First, he has developed a quasi-experimental framework for exploiting periodic changes in water rates to estimate price elasticities that vary with size of user and weather conditions. Findings confirm that elasticity varies with scale (amount of water consumption) and weather conditions (both precipitation and temperature) when estimating demand and overall seasonal drought conditions.

Second, postdoc Min Qiang (Kent) Zhao and Smith have developed a propensity score matching estimator to link residential properties based on the structural and land attributes of the homes. This process allows residential water demand to be compared controlling for water-using appliances and proxy measures for outdoor uses in Phoenix and Tucson. This method matches comparable homes, controls for differences in weather conditions, and evaluates the effects of the overall price schedule on water demand. The results indicate the Tucson schedule, after controlling for factors influencing indoor/outdoor uses and differences in weather conditions, leads to a significantly lower level of water consumption.

Research on housing prices in Maricopa County by Klaiber and Smith has established several findings related to water policy. They estimated the effects of mesic (wet) versus xeric (dry) landscapes at the subdivision level on home sales prices from the 1990s through 2006. After controlling for structural attributes, distance from the central business district and other amenities, they find a robust and significant premium for mesic landscapes and discounted prices for xeric.

In a subsequent analysis, Klaiber and Smith used the same data with PRISM temperature data to investigate the heat-island effect. Using an extension of the Abbott-Klaiber (2010) adaptation of random effects modeling structure, Klaiber and Smith found that a significant number of homeowners in Maricopa County are willing to pay ~3% premium for homes with one degree lower nighttime temperatures. This would be about $50 per month. It is larger than estimates developed using stated preference methods with the Phoenix Area Social Survey in 2006.

DCDC Researcher: