摘要：Agriculture conventionally supplies food, fiber and fuel that consumers can purchase through the market. With the right incentives, farmers can also provide ecosystem services such as wildlife habitat, climate regulation, surface water flows and waste absorption and breakdown. Such incentives have so far come almost entirely from government-sponsored programs that rely on financial assistance to farmers to encourage them to alter agricultural practices or input mix to enhance ecosystem services. Programs recently implemented in Costa Rica and Columbia rely on payments by the beneficiaries of the ecosystem services, such as municipal water companies and water users (Pagiola et al. 2002). Few of these programs, however, have attempted to establish a market for ecosystem services in which the beneficiaries of such services pay the suppliers their personal values of ecosystem services in an actual market. Markets for ecosystem services must overcome two major challenges. In order to set prices for ecosystem services at the right level, it is imperative to understand consumers preferences. Farmland, however, has multiple attributes such as wildlife habitat services and landscape view; the marginal rate of substitution among those attributes must be understood to design marketable products for ecosystem services. Moreover, many ecosystem services are public goods for which traditional markets are ill-suited, because many individuals can receive benefits simultaneously regardless of whether they have paid part of the cost of provision. Therefore, consumers have an incentive to free-ride on others. Evidence from previous research on public goods clearly suggests that under-contribution is typical (e.g., Ledyard 1995). The overall goal of this study is to explore the potential to establish an actual market in which the public can purchase ecosystem services generated by agricultural land. More specifically, this paper evaluates the performance of alternative elicitation methods that are intended to reduce individuals incentives to free-ride on others payments for ecosystem services. Using a choice experiment involving a large-scale mail survey, we first estimate the marginal rate of substitution consumers place on various attributes of farmland including the ecosystem services such land can provide. We further compare the results across different payment mechanisms and examine which ones are capable of revealing demand that is closer to consumers true value. Second, we attempt to establish an actual market in which individuals are asked to purchase a share of a farm contract to provide ecosystem service with real money under different payment mechanisms. We compare the market outcomes with the choice experiments. To the best of our knowledge, this is the first study to examine the performance of different payment mechanisms for provision of ecosystem services using field experiments both within a hypothetical setting and by developing an actual market. The ecosystem service in question in this study is habitat for a grassland-nesting bird called the Bobolink (Dolichonyx oryivorus). Yellow and black Bobolinks establish ground nests in hay fields from mid-May into early June. Their visibility and entertaining character, combined with evidence that many birds, including bobolinks, are experiencing population declines (Sauer et al. 2004), make the bird a leading candidate to attract public interest in efforts to manage farmland for vulnerable wildlife. Previous studies have established that hay harvesting conducted during the birds five to six week nesting period is devastating to fledgling success (e.g., Mitchell et al. 2000). A fairly moderate shift in the harvest schedule could provide significant refuge for nesting birds while causing some losses of the quantity and quality of the hay harvested. If a market developed that paid farmers acceptable compensation to protect grassland birds, then farmers would have an incentive to add an ecosystem service to their revenue base while enhancing environmental quality for wildlife. In this study, we compare four payment mechanisms applied in field experiments: 1) voluntary contribution mechanism, 2) provision point with a money-back guarantee and proportional rebate of excess contributions (PPMBG-PR), 3) uniform-price, multi-buyer auction and 4) pivotal mechanism. Voluntary contribution mechanism has no provision point but has a money-back guarantee if enough money is not raised. Under PPMBG-PR, the public good is supplied only if a pre-specified amount of money (the provision point) is raised, and contributors receive their money back if the market fails to raise that amount. Under a multi-buyer auction, everyone who is willing to pay above a certain price will pay a price such that the total sum will be enough to cover the cost for a farmer to change harvest practices. Under a pivotal mechanism only those consumers whose payments make a difference in the provision of the good would pay. The pivotal mechanism is incentive compatible and is used as the baseline. We test the following hypotheses on the WTP, market participation rate and total revenue collection: (1) WTP: Voluntary contribution < Multi-buyer Auction < PPMBG-PR < Pivotal (2) Participation rate: Voluntary contribution < PPMBG-PR < Multi-buyer auction < Pivotal (3) Revenue collection: Pivotal < Voluntary < Multi-buyer auction < PPMBG-PR We perform a large-scale cross-mechanism comparison using two types of field experiments, a survey involving hypothetical questions and a functioning market for an ecosystem service. In the survey, the subjects are randomly assigned to one of the elicitation methods. We posit a hypothetical situation involving a market for contracts with farmers, describe the assigned elicitation method and ask respondents to compare several sets of two alternative farm wildlife contracts, each with five attributes presented in a stated-choice format. One of the attributes is to change the timing of hay harvesting to enhance the success of a specified number of bobolink fledglings on a given size of farm. The survey data were collected in fall of 2006. Survey questionnaires were sent to all of the 2,987 households in Jamestown, Rhode Island. The response rate was about 37% after accounting for undelivered surveys. Although still preliminary, a key result from the choice experiment is that consumers are least willing to pay for farm wildlife contracts under the voluntary contribution mechanism. The payment method that generated the greatest total purchases of farm wildlife contracts was the multi-buyer auction, followed by the pivotal mechanism and PPMBG-PR, respectively. This result suggests that some features of the payment mechanisms (such as setting of a provision point, fairness and a money-back guarantee) encourage consumers to reveal demand that is closer to each consumer true value of the good than the demand revealed by a simple voluntary contribution mechanism in a field experiment using a hypothetical situation. We are currently in the process of launching an ecosystem-service market in Jamestown by establishing actual farm wildlife contracts with farmers and selling shares of those contracts to consumers. The market will be open to the public for two months during March and April, 2007. Each farm wildlife contract will be tied to one or more of the payment mechanisms and whether each contract will remain effective during the breeding season will depend on the market outcome under each mechanism. Consumers will be randomly assigned to one of the elicitation methods. Using market data, we will compare the actual consumer behavior across alternative methods and also to their willingness to pay as estimated in the survey choice experiment.
