This focus area has developed survey instruments that enable us to compare residents' actual water consumption through water bills and low-cost, high-resolution water meter loggers with their responses to changes in price, institutions, or education in an experimental setting. Individual research efforts include:

· Experimental analysis of consumer demand
· Survey of disaggregated demand
· Socio-economic factors affecting residential water demand
· Data gaps, hybrid modeling and sustainability

Experimental analysis of consumer demand
D. Brookshire, J. Chermak, K. Krause (UNM)

The overall goal of our demand-side management work is to provide understanding of individual water demand behavior for consumers (agricultural, industrial, institutional, commercial and households). This project specifically focuses on household demand behavior, and is connected to other SAHRA projects through the modeling and the evolving Rio Grand/Rio Bravo Modeling team. The research will add a behavioral component to the integrated modeling for the exploration of scenarios for the Rio Grande/Rio Bravo Scenarios Task Group.

The science question inquires into the behavioral aspects of individual demand behavior. This has four sub-components: 1) Do consumer characteristics underlie and thus drive demand behavior, thus creating a heterogeneous demand for water? 2) Can the historical observed behavior be calibrated with behavior in a context-specific laboratory setting? 3) What price response do we observe for ranges of prices outside current pricing domains? and 4) What will the behavioral response be if long- and short-term considerations are possible? (e.g., convert to xeriscape yard).

We have used historical data and experimental settings to determine the role of characteristics and responses to alternative pricing regimes. We observed that a heterogeneous demand for consumers does indeed exist, and we found an alignment of historical data (household water bills) and behavior in the laboratory in a set of context specific experiments. The project has assembled a unique data set that enables us to evaluate disaggregated consumer demand for the demand side of the demand-side management model. Two critical questions have been addressed. First, what is the nature of heterogeneous preferences? This is an essential element for modeling consumer behavior in a variety of urban settings. Second, the historical data has been successfully but preliminarily calibrated with laboratory behavior, setting the stage for the exploration of alternative behavioral response to differing pricing regimes. The behavioral estimates will eventually be incorporated into the Rio Grande/Rio Bravo Integrated Modeling Team effort.

Activities and Results

Sets of experiments were conducted in the fall of 2001. This data was analyzed in the latter part of our Year 2 efforts and in our Year 3 efforts. The heterogeneous demand preferences efforts have been detailed and will be published this year (Krause, Chermak, and Brookshire, forthcoming). The calibration efforts were documented and presented at the American Economic Association meeting, as well as the 2nd World Congress of Environmental and Resource Economists (see Krause et al, Water Consumption, in Presentations section of Section VIII). Beyond data analysis, a refined set of protocols is being developed for further calibration and the exploration of behavior under differing pricing regimes. The calibration efforts involve linking household use to parameters in the laboratory, thus furthering our exploration of the calibration issue. The pricing protocols are a new design within Year 3 efforts. Specifically, coupling the calibration efforts to a series of price regimes that more likely reflect the scarcity value of the water will be designed. This set of experiments will be conducted in the fall of 2002.

In a link to the TA5 project, " Survey of Disaggregated Demand" (below), census data has been merged with the survey data (see the project report below) with the Albuquerque household data from the historical records and the Albuquerque household participants. This will ultimately provide a data platform for the overall Demand Management Model, as driven by econometric estimates of household disaggregated demand activity.

Several papers are either in press or were published during the past year, including a paper (Krause, et al., in manuscript) in which we suggest a way to augment existing studies with experimental data, a first step in an effort to align experimental responses to real-world data. If experimental results can be generated that are consistent with actual water consumption, we can extend the price, and thus the predictive range of the models, outside the current price ranges.

Some of the research was also presented publicly at the Southern Economic Association Meetings, the American Water Resources Association, and the 2nd World Congress of Environmental and Resource Economists (see Krause, Chermak, and Brookshire, Demand for Water, in Presentations section of Section VIII). In this paper we find differences in demand that are correlated with a variety of social and cultural factors, including age, ethnicity, political affiliation, religious affiliation, and risk preferences. We use these demand estimates to construct an incentive-compatible non-linear pricing schedule for a case in which the regulator is faced with a reduction in available supply.

Plans

The next logical steps for the experimental disaggregated demand effort are as follows. First, further calibrate the laboratory with the experimental data. Specifically, the experiments need a set of parameters whereby the contextual water decisions in the laboratory can be linked to actual household decisions. Essentially, this is a matter of linking decisions - such as yard watering choices - to a choice in the laboratory that is realistic and appropriately scaled. Second, after the calibration is complete, the framework will be used to explore alternative price regimes that are "institutional free" (see project report for "Assessment of Institutional Structures for Demand Side Management Integrated Modeling" for a discussion of this issue as a new initiative). A set of experiments will be designed and implemented in Year 5 that will allow the household to "react" to alternative price regimes in a short- and long-term context. Finally, the preliminary work will be undertaken in Years 4 and 5 for extensions of the household consumer demand effort into Mexico.

At some scale it is anticipated that the SAHRA Rio Grande effort will produce, at a minimum, a dynamic simulation model that can be queried using the scenarios that are to be developed by the "Scenario Group." One of our new proposed efforts is to directly address how to embed the disaggregated behavioral market model into the existing Rio Grande hydrologic model produced by Sandia National Laboratories. The model is currently a lumped model, which will eventually need to be disaggregated into individual river reaches. This effort is proposed as a newly identified research proposal, a new initiative for the integrating within the dynamic simulation model.

The calibration and the exploration of alternative price regimes are already underway. Our goal is to implement the calibration effort in the late fall of 2002 or early spring of 2003 (Year 4). Preliminary testing of the basic price regime protocol will occur in the spring of 2003. By the end of 2003 (Year 4), the overall protocols should be implemented and tested. It is not possible to know what glitches will occur, but our goal is to complete the consumer disaggregated market demand effort by the end of Year 4. Also, integrating the consumer demand estimates into an integrated modeling framework will be part of the efforts in Year 5 and beyond.

The initial efforts with Dr. Robert Varady and the Udall Center will begin in Year 4 and continue into
Year 5. We propose to pursue a study of the disaggregated demand for household water in Mexico. In conversations with Dr. Varady over the last year, he has indicated, based on stakeholder contacts, that the Udall Center has a project that could be undertaken. It is our understanding that Dr. Browning-Aiken will facilitate this effort.

A study of the economics of water consumption patterns in communities on the Mexican side of the San Pedro River basin will be of use in futher understanding dissagregated consumer demand. Any modeling of river systems that cross borders addressing demand management issues will require this information. We know of no such investigations. The existing investigations for consumer demand in other countries are of the same aggregate nature that motivated the current ongoing disaggregated demand study in Albuquerque. Thus, this would be a unique study and be important for integrated modeling.

It will be among the first such analyses conducted in a cross-border setting. It will supplement other ongoing social sciences activities in the basin, thus effectively leveraging other sources of funding currently available to the research team.

Several communities offer wonderful opportunities for the study. The largest and economically most important is the city of Cananea, Sonora, Mexico, near where the river arises. Cananea is the home of North America's largest copper mining operation and a diverse population that relies heavily on the mine. From recent surveys administered by the Udall Center, residents are known to be concerned about both the availability and quality of their drinking water. There is little if any documentation of how households supply their water needs or how much they pay. Few homes can count on permanent, full-time, piped water delivery, and what water is provided is of poor quality, partly because of old infrastructure, partly because of contamination from the mining facility. Naco, right on the border, is the next largest community. Like Cananea, it was the subject of a Udall Center water-use perception survey in 2000-2001. Finally, a number of ejidos are situated along the river and can serve as sample communities for the valuation survey. The survey will take full account of water-quality issues, recognizing that quality and concomitant concern for human health are important factors in family water-purchasing behavior and strategies.

Survey of disaggregated demand
D. Brookshire, J. Chermak, K. Krause (UNM)

The overall goal of demand-side management is to provide understanding of individual water demand behavior for consumers (agricultural, industrial, institutional, commercial and households). This project focuses upon industrial, commercial and institutional demand behavior. A significant amount of work has been done regarding agricultural demand for water, which we ultimately draw upon at a later date (see project report for "Data Gaps, Hybrid Modeling, and Sustainability" for this effort in the context of the Conchos River).

Using econometric modeling at the firm level, we inquire into the behavioral aspects of industrial, commercial, and institutional demand behavior. The questions are quite similar to the household behavior issues discussed in the project report immediately above, which is a companion project. That is, what are the characteristics that drive water demand by industry, institution and commercial user? The disaggregated market demand for water (as distinct from the disaggregated non-market demand for water to be discussed the project report "Economic Valuation of Riparian Flyways") has five aspects - households, industrial, commercial, institutional and agricultural.

Activities and Results

At this time, we intend to rely on the Albuquerque data set for the institutional, commercial and institutional disaggregated demand estimates, and will merge data where appropriate to enhance the data set (e.g., Census of Manufacturing). The data set has been assembled beyond the 'raw' data given to us by the City. Preliminary econometric regressions have been run. The Albuquerque data set provides information on the industrial, commercial and industrial use of water. Manipulation of this data set has been extremely time intensive. A significant effort has been made to identify this information in the overall data set. This effort started in earnest around January 2002. Initial econometric estimates have resulted in the demand for water from these disaggregated sectors.

The econometric results are preliminary, but promising. As expected, quantity demanded is negatively correlated with price. We also find correlations between quantity demanded and the climatic variables of temperature and precipitation. Also as expected, there are statistical differences in usage between different meter sizes. These results indicate a more micro-level study would be potentially beneficial. Thus, we will expand the data set to include more firm-specific factors, such as SIC codes.

Plans

The behavioral estimates will eventually be incorporated into the Rio Grande/Rio Bravo Integrated Modeling Team effort. The next logical steps are to complete the analysis. Significant data identification and merger issues remain. As mentioned previously, this involves the linkage of the industries to specific SIC codes and other information that will enhance the underlying behavioral model. Inquiries will be made to obtain a similar data set from other areas.

We do not propose at this time to undertake the agricultural sector within this project; instead we propose to address the agricultural sector in the development of the dynamic simulation effort in the Conchos River basin as Year 4 and 5 initiatives. A significant amount of work in the extant literature has been done regarding agricultural demand for water, which we will draw upon within that project.

At some scale it is anticipated that the SAHRA Rio Grande effort will produce at a minimum a dynamic simulation model that can be queried using the scenarios that are to be developed by the "Scenario Group." One of our new proposed efforts is to directly address how to embed the disaggregated behavioral market model into the existing Rio Grande hydrologic model produced by Sandia National Laboratories. The model is currently a lumped model, which will need to be disaggregated into individual river reaches at some point. This effort is proposed as a newly identified research proposal for integrating within the dynamic simulation model.

The econometric estimates will be available for integration into the Rio Grande/Rio Bravo modeling efforts by the end of Year 4. They will be written up in a draft manuscript and as a poster. The papers will most likely be finalized in Year 5. The demand estimates will be incorporated into modeling frameworks (Year 5).
(return to top)

Socio-economic factors affecting residential water demand
G. Woodard (UA-HWR)

Econometric and engineering-based studies over the past few decades have failed to produce a clear, detailed understanding of domestic water uses, and how various socio-economic factors, including price and conservation programs, influence domestic demand. For example, we understand in general that higher prices reduce demand, but not how. Do people cut back more in outdoor uses than indoor? Do they irrigate less frequently, or for shorter periods of time? Are clothes washers and dishwashers run less often, with fuller loads? Do they take shorter showers, or fewer showers, or install low-flow showerheads? This project is developing and deploying low-cost high-resolution water meter loggers to identify and quantify specific domestic water demands within individual households over extended periods of time. By measuring domestic water flow at the 0.1-gallon, 0.01-second resolution, water use traces are generated which reveal particular water uses and details such as frequency of toilet flushing, length of showers, gallons of water used per load of laundry, and frequency and duration of outdoor irrigation. When extended for years, impacts of weather, climate fluctuations, and price changes can be directly observed. Also, household reactions to various water conservation programs can be observed. The overall goal of the demand-side management is to provide understanding of individual water demand behavior for consumers, within a private good setting. Field tests to date suggest that the data gathered by this approach will directly address the questions posed above, and will integrate with the survey work and experimental economic laboratory work to enhance this understanding.
Activities and Results

SAHRA resources were used to develop and prove the concept, produce, and field test data logging equipment, develop and test the necessary software, and compare various battery options and data storage and retrieval mechanisms. The US Bureau of Reclamation support allowed for the purchase and assembly of 60 meter/logger/sensor systems, additional field testing, and development of long-term field power options. Audubon Society support is being used to install some 50-60 meters and sensors in middle-and upper-class ranchettes in the Sonoita/Elgin region of Arizona and gather and analyze the data for two years. Support from Cochise County will allow another 20 low-income ranchettes along the Upper San Pedro River to be logged.

Three field testing sites were established with cooperation of Tucson Water, and some 10 meters have been installed in the Sonoita/Elgin area in cooperation with the Audubon Society's Research Ranch. Data analysis supports the belief that this methodology will provide the degree of resolution and insights necessary to make a major advance in modeling and forecasting residential water demand.

Results to date are limited to the data collected through field testing. We clearly have sufficient resolution to not only distinguish between types of water usage (e.g., toilet flushing, showering, laundry), but also have the resolution to discern important details of usage, such as: which toilet in the household is being used and how often is it double-flushed; what is the duration of the shower and is a low-flow showerhead being used; and is the clothes washer being run with a full or partial load, and with cold or warm water. A sample meter trace, with labeled water uses, is available to view.

Plans

The next steps are installation of the remaining meters and loggers in Sonoita and Elgin, selecting participants in the Upper San Pedro, and establishing the data retrieval and analysis process. Beyond that, there are several potential areas in which the work can be expanded, including:

1. logging of newly constructed homes in three different price ranges in the metropolitan Tucson area (preliminary agreement reached with local water providers);
2. modification of software and hardware to allow wireless relay of data (City of Phoenix support offered);
3. modification of hardware and software to allow the PDAs to operate as intelligent, updated irrigation controllers that open and close irrigation valves based on downloaded weather data (proposed project with Phoenix Water and Tucson Water); and
4. modification of software and hardware to allow the PDAs to serve as data loggers for other field instruments, and trigger samplers on command (being pursued with Jon Petti and summer REU student).
   (return to top)

Data gaps, hybrid modeling and sustainability
D. Brookshire, J. Chermak, S. Burness (UNM)

The science question for this project involves issues of modeling, data gaps, and integrated modeling from an economic perspective across disciplines. Modeling efforts include a dynamic analysis of the interacting incentives for private versus habitat water use in the context of a mountain front recharge system such as the Upper San Pedro river basin. A concurrent effort involves hybrid modeling that attempts to bridge the gap between hydrologic models and economic optimization by imposing economic benefit functions on a finite element model of interactive groundwater/surface water use. Since SAHRA is ultimately concerned with sustainable water use in semi-arid environments, we also investigated the economic implications for sustainable use and the meaning of the word "sustainable" in an operational context. Data gaps were explored in the context of benefit transfers. Finally, the nature of urban water pricing and western water markets were preliminarily explored.

This project explored issues of hybrid modeling, data gaps and integrated modeling. Significant data gaps were identified in the effort. For instance, the efforts on the urban water demand revealed the lack of variation of prices for policy purposes across the country; the water market paper identified how little is really known about water markets in the West; the benefit transfer papers reveal that a central issue as yet unaddressed by SAHRA is the uncertainty effects of population growth relative to other data uncertainties. The hybrid modeling papers raise issues that will have to be addressed in policy design regarding the needed anticipation of future demands. Although the UNM group considers this phase of the project finished, new initiatives are proposed below that involve modeling within the dynamic simulation platform.

Activities and Results

During the past year we completed a series of manuscripts, drawing on efforts achieved during Years 1 and 2. Some have been accepted for publication, some will remain as white papers.

Plans

Efforts on the Economic Modeling and Data Gaps project have led us to propose two new initiatives. Both ultimately involve using the dynamic simulation modeling platform.

Sub Area 1
Traditional economic models normally assume that a public decision maker's objective is to maximize social welfare. Given this objective we can provide water policy makers with integrated management tools to assess the effectiveness of policy alternatives. However, the assumption of social welfare maximization is just one of several objectives that may be relevant. For example, equitable allocation may be more appropriate in some cases, or a combination of equitable distribution up to a specific level of consumption, followed by social welfare maximization after that point, or a sustainable water use plan. Different objectives most likely will result in different efficient policy tools.

The research will provide a tool with which to assess the impact of varying public objectives. The unique aspect of the tool is not only the recognition of varying objectives but also the incorporation of these objectives with the integrated behavioral and physical science models that have and are being developed in SAHRA. This will provide the decision maker with a more accurate tool with which to assess decisions as well as provide an educational tool for the public.

The research program is composed of three parts. Part I will compile the objectives of water utility decision makers. Work already completed within SAHRA in both the U.S. and Mexico will provide the basis for this. Where necessary, a survey of additional policy makers will be completed to provide a comprehensive perspective of policy makers' objectives for water usage in semi-arid climates. Part II will integrate the objectives into socio-scientific models of water allocation developed within SAHRA. These results will provide the basis for Part III, which is the development of a computer-based interactive tool that can be used by policymakers and stakeholders alike to assess resource management plans under the appropriate policy objective.

Sub Area 1 will provide a tool with which to assess the impact of varying public objectives. The unique aspect of the tool is not only the recognition of varying objectives but also the incorporation of these objectives with the integrated behavioral and physical science models that have and are being developed in SAHRA. Without the interdisciplinary platform provided by SAHRA, the work proposed in this area could not be accomplished in a form that would be particularly beneficial. This will provide the decision maker with a more accurate tool with which to assess decisions as well as provide an educational tool for stakeholders in general. The anticipated deliverables include a manuscript as well as the computer-based analysis tool.

Sub Area 2
Conversations with Juan Valdés and Javier Aparicio (IMTA) have resulted in a tentative agreement to collaborate on modeling in the Conchos River region. Dr. Aparicio has a Stella-based model for Mexico with which Dr. Valdés is working. The model would be converted to the dynamic simulation platform and the U.S. side of the Rio Bravo River would be added. This sub-project will serve as vehicle for integrating the substantial literature on agricultural demand for water into a dynamic simulation format.

Sub Area 2 will provide a tool for analyzing alternative international water allocation schemes. The interesting aspect of the area is the heavy emphasis on the agricultural area. SAHRA has to date not fully addressed the issues at the border. This would address one of the more general issues of efficient and equitable distribution of shared waters along an international border. For Year 4 the primary effort would be writing a proposal for possible funding from the U.S. and Mexico. Also the modeling of the Conchos using dynamic simulation mentioned in Section 4 will be continued and will be the basis for the integration.
(return to top)