Residential rebate programs for low-flow water devices have become increasingly popular as a means of reducing urban water demand. Although program specifics vary, low-flow rebates are available in most U.S. metropolitan areas, as well as in many smaller municipalities. Despite their popularity, few statistical analyses have been conducted regarding the effects of low-flow rebates on household water use. In this paper, we consider the effects of rebates from the Albuquerque Bernalillo County Water Utility Authority (ABCWUA). Using panel regression techniques with a database of rebate recipients, we estimate the marginal effects of various low-flow devices on household water demand. Results indicate a negative correlation between household water use and the presence of most low-flow devices, after controlling for water price and weather conditions. Low-flow toilets have the greatest impact on water use, while low-flow washing machines, dishwashers, showerheads, and xeriscape have smaller but significant effects. In contrast, air conditioning systems, hot water recirculators, and rain barrels have no significant impact on water use. We also test for possible rebound effects (i.e. whether low-flow appliances become less-effective over time due to poor rates of retention or behavioral changes) and compare the cost effectiveness of each rebate using levelised-costs. We find no evidence of rebound effects and substantial variation in levelised-costs, with low-flow showerheads being the most cost-effective device under the current ABCWUA rebate program. The latter result suggests that water providers can improve the efficiency of rebate programs by targeting the most cost-effective devices.
Scarce water resources, particularly in semi-arid regions of the United States, are increasingly strained by population growth, economic development, and drought. A major challenge for urban and regional planners is to ensure sufficient resources are available to meet projected demand—and to balance these resources between competing uses (e.g. agriculture, residential, commercial, and industrial). In addressing this challenge, municipal water providers have emphasized the need to reduce residential water use, which accounts for approximately 58% of urban consumption (Barber, 2009).1 Specifically, water providers have turned to demand-side management as a means of reducing per-capita water demand (Arbués and Villanúa, 2006 and Michelsen et al., 1999).
Previous literature distinguishes between two types of demand-side management: price and non-price policies (Kenney et al., 2008 and Krause et al., 2003). Price policies refer to various price structures, most commonly block-rates, used to incentivize water conservation. The effectiveness of these policies largely depends on the price elasticity of water demand. Recent empirical studies indicate that the price elasticity of demand for water is inelastic at current prices, implying that price increases result in only modest declines in the quantity demanded (Arbués et al., 2003 and Dalhuisen et al., 2003). Partially due to this inelasticity, water providers have predominantly utilized non-price policies (Olmstead et al., 2007). Non-price policies refer to a wide range of interventions, including: restrictions on water use, public education campaigns, subsidies for low-flow appliances, and low-flow engineering requirements on new plumbing fixtures. Among the more popular non-price policies are rebate programs for low-flow appliances (e.g. toilets, showerheads, and washing machines). At present, rebate programs can be found in most metropolitan areas, as well as in many smaller municipalities.2 Programs in Austin, Los Angeles, New York, Phoenix, and Tampa distributed nearly 2.3 million low-flow toilets between 1992 and 2000 (U.S. General Accounting Office, 2000). These programs cost $400 million and reduced water consumption by approximately 100 million gallons per day (U.S. General Accounting Office, 2000).
Despite the popularity of rebate programs, few statistical analyses have been conducted regarding their impact on household water use—and questions remain as to their success. Of particular interest is whether actual water-use reductions brought about by low-flow appliances resemble those predicted by engineering estimates. These values may differ if low-flow appliances prompt behavior changes that mitigate or enhance the expected reduction. A related concern is whether the effects of low-flow appliances diminish over time due to poor rates of retention (i.e. the low-flow appliance is replaced) or changing behavior. Finally, given the considerable costs associated with rebate programs, it is of interest to identify the most cost-effective rebates.
In the following analysis, we address these questions using monthly water-use and rebate data from the Albuquerque Bernalillo County Water Utility Authority (ABCWUA). We estimate the marginal impact of various low-flow appliances on household water use, after controlling for the price of water and weather conditions. Results from the analysis contribute to public understanding of conservation programs and offer practical guidance for water providers developing demand-side management policies.
