Project 7: Economic Valuation Willingness to pay for stormwater management: Evidence from a field and conjoint choice experiment Rationale
A large number of benefits from stormwater harvesting and management have already been reported by scientific scholars and realized in some areas in the field. As real choices and behaviour of individuals with respect to stormwater management are difficult to observe, revealed preference methods can be problematic when it comes to understanding how much individuals and society value stormwater. In addition, a number of innovations relating to stormwater harvesting are not implemented in practice yet, making it difficult to infer values from existing markets. In order to elicit individual willingness-to-pay (WTP) for the positive externalities (or the WTP for the prevention of negative externalities) associated with stormwater, a stated preference approach, in particular, a conjoint choice experiment (CE) wil be conducted. The values obtained from stated preference approaches may be subject to a number of biases, including a hypothetical bias that arises from the hypothetical nature of the scenarios presented.
This project applies CE to a new environmental resource, stormwater and its management. The advantage of a CE in this context is three-fold: firstly, it is able to elicit monetary values even for hypothetical projects; secondly, of all stated preference methodologies available, the conjoint choice analysis is best suited to minimise biases, by way of careful survey design, implementation and data analysis; thirdly, by eliciting WTP per attribute one can use CE to evaluate a number of stormwater projects which differ in terms of scale and types of benefits delivered.
To arrive at a near comprehensive value for stormwater harvesting it is useful to distinguish between the various features that make this activity valuable to a community. As customary in conjoint choice experiments these features are referred to as attributes. In the first instance, one can distinguish between use and non-use attributes.
1. Use attributes: All characteristics of stormwater management that are directly or indirectly
beneficial to the community. In the context of stormwater management these attributes can include greater water supply security, recreational benefits, improved physical and mental health, improved landscape, flood protection and liveability.
2. Non-use attributes: All characteristics of stormwater management that relate to the pure
existence of an environmental good or service with no immediate use value to the
1 The revealed preference method relies on inferring values from behaviour or outcomes in related markets. The idea is that individuals reveal their preferences about the environment when making decisions in related markets and that, by examining these related markets, researchers can infer the wil ingness to pay for environmental goods and services.
community. One example of a non-use attribute is the beneficial effect of stormwater harvesting on biodiversity.
The objective of this project is to provide an economic value for the above mentioned attributes of stormwater harvesting. We aim to do this by using an innovative methodology which can take into account risk aversion associated with the different attributes. In the context of stormwater management, most attributes are stochastic in the sense that whether or not a promised outcome is achieved depends not only on the stormwater project itself but also on the weather. Hence risk aversion of individuals can play an important role in determining willingness to pay. For example, while stormwater harvesting may go some way to reducing the need for compulsory water restrictions it may not be sufficient to achieve this outcome during a severe drought. Similarly, investment in stormwater harvesting infrastructure may eliminate the risk of flooding during a one in five year storm event but it may not be sufficient to prevent flooding in a one in one hundred year storm. In contrast, the costs of investing in stormwater infrastructure are more certain.
In addition, we will combine the traditional methods with a state of the art field experiment in order to test for inter-temporal shifts in preferences as a result of natural variation of rainfall, temperatures or flooding.
Key Research Questions We expect to address the fol owing research questions in our project:
• What is the impact on willingness to pay for stormwater harvesting if the respondent has
recently experienced flooding or higher than average rainfall.
• The impact of individuals’ risk aversion and how it influences willingness to pay for different
• Could intrinsic motivation of individuals impact willingness to pay: If intrinsic motivation for the
provision of a public good exists amongst some individuals, then increasing (decreasing) the price of a public good may actually lead these intrinsically motivated individuals to increase (reduce) their provision to the public good.
Proposed Approaches and Methods
This section explains the design of the choice and the field experiments:
Design of the Choice Experiments Stormwater management has a number of benefits that can be traced back to a multitude of attributes. There is a trade-off to be made between valuing all possible attributes and designing
surveys that do not overburden respondents with a questionnaire that is too complex in terms of number of attributes included. Similarly, the number of attributes and attribute levels are directly linked to the sample size required for estimation purposes.
Moreover, one introduces greater sample size requirements for example by allowing for non-linear preference over probabilities. Similarly, labelling options (desalinisation, recycling, stormwater harvesting) or including an attribute (e.g. ‘additional water source’) increases sampling size exponentially. To avoid this, one may use generic choice option design. Hensher et al. (2005, p.113) advocate the use of generic experiments if the focus is on establishing WTP for attributes as opposed to predicting and forecasting product demand. Two additional arguments in favour of generic experimental design relate to the preconceptions of respondents about the water sources used – for example strong dislikes of desalinisation technology or health concerns associated with recycled water. These strong preconceptions may dictate the respondents’ choices. Hence the attributes are no longer independent in the view of respondents, which violates some assumptions important for choice modelling.
To avoid these issues, we plan to design two conjoint choice experiments. Choice experiment 1 uses a generic design to elicit a value for the most important attributes of stormwater management. The objective of choice experiment 2 is to elicit preference for different sources of water supply, which are presented as label ed alternatives to survey respondents.
Design of the Field Experiment Stormwater management is currently proposed or undertaken at a council level. We can expect there to be some variation with respect to heavy rainfall events and subsequent flash flooding.
The research team will conduct 2 survey waves, interviewing an adult representative of owner-occupied households, in a number of councils across metropolitan areas in Australia. To implement the field experiment different councils will be selected, which are all comparable in terms of socio-economic variables. However, the councils will differ in past precipitation patterns.
The experimental setting allows the research team to compare individual WTP along different dimensions: over time and over councils. Combining elicitation methods used in experiments with a choice experiment would provide several methodological and policy innovations. Information col ected from the choice and the behavioural experiments can help us understand the wil ingness to pay for stormwater management and how it is related to risk and time preferences.
Scope of Study
We aim to elicit the value of stormwater management to urban (metropolitan) Australia. In particular, we will run surveys in Victoria and Queensland across two control and two treatment communities. Each state wil have a control group. The treatment groups can either both be in Victoria or in Queensland or one in each state.
This project will provide a monetary evaluation of non-market benefits from stormwater management. Values will be derived by attribute, which will enable the evaluation of diverse stormwater harvesting projects that differ in size and scope and therefore in terms of the benefits they deliver.
The key socio-economic (e.g. income, education, age) and psychological (e.g. risk perception, personal experiences, preconception with respect to water supply source) determinants of the WTP for these benefits will also be identified.
The project will also allow us to explain the variation in WTP between regions and across time.
Integration with other projects
Project 7 aims to evaluate the benefits of storm water management in monetary terms. Of course to do so, we need to identify the different attributes of stormwater. In this context, P7 has benefitted from research conducted within other Projects, including P1, P2, P3 and P4. These comments have been integrated into the design of the choice experiment.
P7 has also some interactions with P6. There would be some overlap between the two projects with respect to the socio-economic determinants of the WTP.
Hensher, D.A., J.M. Rose, and W.H. Greene (2005) Applied choice analysis: a primer, Cambridge University Press.
lagstadgade och frivilliga anmälningar av Avdelningen för epidemiologi Smittskyddsinstitutet (SMI), Solna Versionslogg Version: Logg: Första skarpa versionen. Total omarbetning av specifikationen för exportfilens utformning (XML-Schemat). Utförligare beskrivning av hur exportfilen skall överlämnas. Ny specifikation över kvittensen från webbtjänsten. Omarbetning av dispositi
Special Meeting Minutes May 23, 2011 Page Special Meeting Minutes A Special Meeting of the Board of Education of the Scotia-Glenvil e Central School District was held in the Library Media Center of the Middle School, in said district on May 23, 2011. President Carbone cal ed the meeting to order at 6:15 p.m. Present: Carbone, Conlon, Crapo, Normington, Smith, Yagielski, Superintendent Swartz,