Jolejole, C.B., S.M. Swinton, and F. Lupi
Presented at the All Scientist and GLBRC Sustainability Meeting (2009-05-05 to 2009-05-07 )
Agriculture is the world’s largest terrestrial ecosystem. Managed to provide food, fuel and fiber to meet humans’ market and subsistence demands, it simultaneously affects and depends on the biophysical and economic settings in which it operates. In so doing, it generates nonmarketed ecosystem services (ES) and disservices, like carbon sequestration into soil or trees, or lake eutrophication from phosphorus runoff. These nonmarketed ES and disservices come as joint products or byproducts w
Unlike natural ecosystems, agriculture is a managed ecosystem. The decisions of its managers drive the mix of ES that it produces. By their choices of production inputs and management practices, farmers shape their impacts on the environment. The supply of marketed farm commodities has been heavily studied. The soil conservation literature shows that private financial motives are the leading driver of adoption of conservation practices. If payments for environmental services were possible, what is the potential supply of ES from agricultural lands? Which farmers are willing to change their practices? How much change is possible and at what cost? Building on these questions, the objectives of this paper are 1) to identify farmer’s willingness to adopt environmental stewardship practices, 2) to investigate the determinants of their willingness to adopt those practices and the relative importance of these factors, and 3) to estimate empirical supply curves for ecosystem services from agriculture.
Conceptual model for provision of ecosystem services from agriculture
We hypothesize that farmer behavior is motivated by utility maximization, where utility is increasing in consumption goods and environmental services. Consumption, in turn, is constrained by net income, which depends on agricultural product revenue minus input costs. Apart from marketed agricultural products, two kinds of ES matter in this model: ES in the utility function and supporting ES that substitute for cash inputs in the agricultural production function (e.g., soil quality, biocontrol of crop pests). Derived input demand – and by extension the demand for changed practices – depends upon prices of products and inputs and perceived ES of both kinds. Note that if society enjoys other ES, it may elect to pay for them, either in the form of an ES output payment or in the “second best” form of an input subsidy expected to generate a certain ES output.
A supply curve for nonmarketed ES can be elicited as the minimum payment that a farmer is willing to accept in order to supply a given amount of ES. If the ES is jointly produced with an agricultural product, then it is possible that a farmer can afford to supply it at lower cost than if he or she were producing that ES alone.
Data and stated preference empirical analysis
We estimate farmers’ potential supply of enhanced ecosystem services by using a mail survey to elicit willingness to adopt various low-input cropping systems under differing payment amounts. Because many nonmarketed ES arise in the production process, and because U.S. farm policy offers precedents for subsidizing production practices, the elicitation of farmer minimum willingness to accept (WTA) payment in order to adopt ES-enhancing practices was presented as a per-acre payment for specified behavior.
In the survey instrument, farmer respondents were faced with a series of four corn-soybean-based cropping systems. The four low-input cropping systems examined have all been the subject of study at the NSF-funded Long-term Ecological Research project in agricultural ecology (KBS-LTER). The four systems differ in the degree of changes to relative to typical systems in use in Michigan, and they offer increasing amounts of ecosystem services compared to typical systems. Scientists at the KBS-LTER have studied each of these systems in long term field trials and have identified their impacts on yields and on ecosystem services. The first, System A, was a corn-soybean crop rotation with chisel plow tillage, pre-sidedress nitrate test in corn, and all agrochemicals broadcast in the field according to University recommendations or label instructions. System B was identical, except that a cover crop was added during winter. System C added wheat to the crop rotation, in addition to the cover crop. Finally, System D was identical to System C, except that fertilizers and pesticides were applied only in bands over the row, resulting in a 1/3 reduction in chemical applications.
The elicitation procedure was conducted via a stated preference mail survey of 3,000 Michigan corn-soybean farmers that received 60% response. The sample for the mail survey was obtained from the agricultural census mailing list of National Agricultural Statistics Service (NASS) office in Michigan. NASS provided the project with a 4-tier, acreage-stratified random sample of 3,000 corn and soybean farmers in Michigan. The sample was selected with 16 replicates within each stratum, with replicates used for the 16 questionnaire versions. The versions followed an experimental design that included four sets of payments (for adoption of the crop systems), two payers (government and non-governmental organization), and two sequences of crop system presentation (increasing vs decreasing effort & degree of ES provision).
Econometric analysis took the form of a hurdle model based on the two stated preference questions that accompanied each cropping system and its payment offer: 1) Would the respondent adopt this system with the assigned payment? And 2) if so, then on how many acres would he or she adopt it? The supply curves being derived from the hurdle model jointly account for the effect that payments have on participation and acreage enrolled conditional on participation.
Results
The empirical findings and predictions of the econometric prohbit model are consistent with the developed conceptual framework that participation decision is greatly affected by the payment offered, farmer’s perception of the new cropping system, and since it is a acreage based enrollment program, total land acreage greatly affects the participation decision as well.
The acreage enrollment decision is estimated separately via truncated regression. The results show that acreage enrollment is more affected by the total land acreage and relative perception. Acreage enrollment is positively affected by subsidy for only two cropping systems.
By doing a separate regression with payment effects by stratum, the study revealed that bigger farmers holding more than 500 acres are more responsive to changes in payment than smaller farms.
Our results suggest that besides technical aspects about environmental benefits and behavioral aspects as proved by earlier studies, the type of cropping system and size of farms have to be taken into account when designing measures. Smaller farms might not be responsive to payment offers but bigger farms proved to be very responsive. It is only by taking into account these elements, can higher acceptance rates and success for agri-environmental policies be achieved.
Back to meeting | Show |