Water Journal : Water Journal August 2012
refereed paper governance water AUGUST 2012 51 water authorities in five Australian states and two territories were contacted between April and June 2011 to obtain historical rebate uptake data. Data was obtained from programs in South Australia, Victoria, Western Australia and the Australian Capital Territory. Rebate programs in these states have had success in inducing households to install or retrofit water-saving measures. Over 1 million successful rebate applications were paid, representing at least $115 million of government spending. By June 2011, programs in South Australia, Western Australia, Victoria and the Australian Capital Territory had encouraged 46,357 households to replace their showerheads, 40,755 dwellings to install a dual-flush toilet, 86,068 residences to install a rainwater tank and 344,200 households to purchase a water-efficient washing machine. For finer scale analysis, rebate applications were disaggregated by product type and program, then expressed as cumulative uptake normalised by the number of residential water connections in the state. Analysis reveals that the most popular form of rebate claimed by consumers was for new WELS 4-star or higher-rated washing machines. Over a five- and seven-year program, rebates for washing machines achieved uptake of 25.5% and 30.4% in South Australia and Western Australia respectively. This compares to rainwater tanks -- 2.35% and 5.86%; low-flow showerheads -- 3.44% and 2.66% and swimming pool covers -- 0.52% and 4.50%, all for SA and WA respectively. Interstate comparisons reveal variation in uptake rates for the same product are largely independent of rebate level. For instance, a $30 rebate for low-flow showerheads was offered in SA, while a $10 rebate was paid in WA. Despite this, monthly rates of uptake, normalised for the difference in residential households in each state, were largely similar (0.59 v 0.54%/ mth). Analysis of the timing and intensity of advertising of rebate initiatives revealed strong correlation between uptake rates and marketing effort. Rebate programs may affect the adoption of water-saving measures by simply increasing awareness of water efficiency rather than through explicit financial means. Rebate incentives aimed to induce the installation of rainwater tanks achieved greater relative success in South Australia compared to other states. Monthly rates of uptake for standalone rainwater tanks (1.84%/mth) and indoor reuse connected tanks (0.73%/mth) exceeded programs in Victoria (0.02% and 0.01%, per month) and Western Australia (0.29% per month). Apart from the influence of rebate level -- the SA rebate for standalone tanks was $150 greater than that offered in WA and $50 greater than that offered in Victoria -- the pre-rebate penetration rate of rainwater tanks in South Australian homes compared to the national average (45.4% v 19.3%) potentially explains this difference. Greater acceptability of rainwater harvesting and reuse may be inherent among South Australians. Climatic factors may also explain the increased uptake in South Australia relative to Victoria over the rebate period. Adelaide's Mediterranean rainfall pattern -- long, dry summers and wet winters -- provides incentive for households to 'store' rainwater in the winter for use in summer. We later demonstrate through end-use modelling that this perceived benefit of rainwater tanks is flawed -- more effective rainwater tank utilisation is achieved when larger volumes of water are continually used throughout the year. This leads to potential implications for program formulation. Policymakers in areas with consistent rainfall where the public may perceive little benefit in capturing rainwater may need to do more to encourage uptake. Statistical analysis of rebate applications, paired with information about the household, revealed some socio-economic insights into the adoption of water-efficient equipment. Mirroring the findings of Millock and Nauges (2010), who were limited by their use of a stated preference survey dataset, chi square tests performed on South Australian rebate applications show that home-owners are more likely to adopt water-saving measures than tenants. Normalised for the proportions of home- owners and tenants in South Australia, 30 times as many rebates were paid to home-owners compared to renters for dual-flush toilets and rainwater tanks. Even for low-flow showerheads, which have minimal installation difficulty and low capital costs, uptake was 16 times higher (3.93% v 0.24%) in the home-owner group compared to the tenant group. Washing machines were the only measure with higher uptake in the tenant group compared to the home-owner group. Over the five-year H2OME program, approximately one in three tenants claimed a rebate for the purchase of a new washing machine compared to only one in five home- owners in South Australia. Regression analyses of postcode median aggregate income against the postcodes of rebate applicants accept, at the 5% level, a null hypothesis that income is not an influence on rebate program participation. Together, these results have implications for efficient policy formulation. We infer that the adoption of water- saving measures that do not add to consumer utility beyond a water-saving benefit -- low-flow showerheads and dual- flush toilets -- are limited among groups with minimal financial incentive (landlords, rather than tenants in South Australia, generally pay the water supply and first 136kL of volumetric tariff charge) to save water. On the other hand, rebate uptake for measures that add a utility externality beyond water efficiency -- washing machines -- is relatively independent of the financial motivation for reducing domestic water consumption. It may be that subsidies for washing machines are being paid to consumers who have already made a purchasing decision independent of rebate incentive. This leads to Pareto inefficiency caused by the free-rider problem. Subsidy levels for measures such as washing machines need to be set at the typical price differential between WELS 4-star appliances and more inefficient models. S = low-flow showerhead, WM = front loader washing machine, T = dual-flush toilet, PC = pool cover, T(O) = standalone tank, T(O+T) = tank plumbed into outdoor and toilet, T(O+L) = tank plumbed into outdoor and laundry, T(O+L+T) = tank plumbed into outdoor, laundry and toilet. Figure 1. Levelised unit cost of water to the policymaker for various South Australian rebates.
Water Journal September 2012-1
Water Journal July 2012