Abstract The sizable technical potential to sequester atmospheric carbon in soils to mitigate climate change will only be realized where and when there is also economic potential. A choice experiment conducted with a random sample of farmers in the State of Indiana, United States, revealed that farmers who have not previously adopted reduced tillage practices on any of their land require a $40 per acre increase in net revenue to switch from conventional tillage to no-till. We estimate that farmers have a $10.57/acre option value of not signing a multi-year sequestration contract, and find that government payments are preferred to carbon markets. Carbon offsets, tillage, choice experiment, attribute non-attendance, climate change, random parameters logit, option value Wide-reaching international agreements like the Kyoto Accord have thus far failed to achieve global greenhouse gas (GHG) emission targets agreed to by the international community because not all national governments have ratified the treaty. The result has been incomplete international implementation of climate change mitigation strategies. The European Union (EU) and a limited number of smaller countries, many of whom contribute relatively little emissions but are most negatively affected by climate change, more or less adhered to the terms of the Kyoto Accord reached through the United Nations Framework Convention on Climate Change (UNFCCC). Subsequent efforts—in Copenhagen (2009), Cancun (2010) and Durban (2011)—to forge a binding international agreement to extend or replace the Kyoto Accord beyond 2012 have given way to an approach to limiting global GHG emissions that relies on individual countries to select nationally-determined contribution (NDC) amounts. During the UNFCCC meeting in Paris in late 2015, 196 industrialized and developing countries pledged NDCs. The Paris Accord has a more decentralized approach than its predecessor, the Kyoto Accord. This is significantly due to the fact that individualized emissions reduction commitments, rather than uniform reduction levels, follow from the NDCs. This international approach is consistent with the different mechanisms being used in different countries, regions and cities today, with each entity having different emissions limits and permitted means of compliance to incentivize their emissions reductions. The most recent legislative attempt to enact a national limit on GHG emissions in the United States failed in 2009, with the U.S. House of Representatives passing the Waxman-Markey Bill (American Clean Energy and Security Act of 2009) but the U.S. Senate not passing it; the bill would have capped emissions, allowed emissions trading, and allowed for the use of emissions offsets to comply with the declining emissions cap. The Obama Administration then took a more traditional regulatory approach to GHG emissions reduction by mandating an increase in vehicle gas mileage using the Corporate Average Fuel Economy (CAFE) standards for vehicles sold in the United States, and by introducing regulations to cap carbon emissions from electric power plants (Environmental Protection Agency 2014), among other initiatives instituted using Presidential Executive Orders instead of relying on Congress. One policy component common to most carbon pricing schemes enacted to date is the inclusion of certified emissions reductions (CERs under the previous IPCCC protocols), which are commonly referred to as “offsets,” and may be used to comply with these policies. An offset is a particular type of CER that allows firms in those economic sectors covered by emissions limits (or, in theory, with carbon taxes imposed upon them) to purchase emissions credits achieved by reducing emissions that are not limited by an emissions cap (or subject to a carbon tax).1 Though implicitly allowed for under any GHG emissions policy that includes an offset provision, the authors are only aware of specific protocols for agricultural soil carbon sequestration (ASCS) having been implemented by the CCX and Alberta schemes (see footnote 1). Previous economic research has examined different practical and technical aspects of ASCS (Gramig 2012; see citations therein). This research addresses a gap in the literature by investigating farmers’ preferences for different attributes of prospective climate change policies and willingness to accept (WTA) payment to change tillage practices to supply emissions offsets to mitigate climate change. We focus on tillage practices instead of afforestation or other practices in agricultural production because our focus is on working lands in the Corn Belt region, where row crops are the highest valued use of most agricultural land. This study is relevant for policymakers and carbon market participants, and is novel in its focus on measuring preferences using a choice experiment (CE) on the supply side of the carbon offset market. This study estimates producer payment and contract terms to adopt various tillage practices. The main objectives of this analysis are (a) to estimate producer WTA payment to adopt conservation tillage, and (b) assess preferences for attributes of mechanisms to incentivize soil carbon sequestration. These are both important research topics because of the low carbon abatement cost of conservation tillage and no-till, and because neither the UNFCCC nor the EU Emissions Trading Scheme (ETS) have developed soil carbon protocols for certified emissions reductions under the Kyoto Protocol or the EU ETS. Development of policies or more detailed protocols themselves must be scientifically rigorous, but also must consider what contract terms, tillage practices, and payments will be acceptable and/or attractive to farmers. Otherwise, further adoption of no-till to supply certified emissions reductions is unlikely. In order to assess farmers’ willingness to change tillage practices, a CE was designed based on key attributes of different policy or market designs to pay farmers for sequestering atmospheric carbon in agricultural soils by reducing or eliminating tillage. Background The use of choice experiments (CE) to estimate willingness to pay (WTP) for environmental quality, attributes of recreational sites, and reductions in morbidity or mortality risk is well-known in the environmental economics literature. In agricultural economics, CEs have also been widely used to estimate consumer WTP for attributes of food products, and livestock products in particular. Various methods and experimental settings have been utilized, including in-person interviews/auctions (Gracia, Loureiro, and Nayga 2009), some at point of purchase (Lind 2007), mail surveys (Nilsson, Foster, and Lusk 2006; Carlsson, Frykblom, and Lagerkvist 2007; Tonsor, Olynk, and Wolf 2009), phone surveys (Lusk and Norwood 2008), and internet surveys (Gao and Schroeder 2009; Olynk, Tonsor and Wolf 2010; Wolf, Tonsor, and Olynk 2011; Olynk and Ortega 2013). While stated preference methods continue to be used to measure preferences on the demand side of the market in environmental economics, CEs are increasingly being conducted with farmers on the supply side to investigate producers’ willingness to change management practices to supply ecosystem services (ES) from working agricultural land (Christensen etal. 2011; Ma etal. 2012) or satisfy consumer demand for credence attributes derived from the production processes used to grow a crop or raise livestock (Schulz and Tonsor 2010). This research uses a choice experiment to explore contract and payment mechanism attributes to estimate farmers’ WTA payment to change the way they manage farm fields and how they trade net revenue against attributes of ES procurement mechanisms to achieve environmental policy objectives. Closely-related articles have previously used CEs to measure farmer preferences for an agri-environmental policy to incentivize the planting of Nitrogen-fixing crops on marginal dryland in Spain (Espinosa-Goded, Barreiro-Hurlé, and Ruto 2010) and to investigate preferences for afforestation contracts to supply different candidate ES in Denmark (Broch and Vedel 2012). The former work studied a more similar population to the current study, while the CE attributes of the latter study are more similar to this study. Other related research surveyed the general public about their preferences for the structure of agricultural land preservation based on the mechanism used, the level of public access to the preserved land, and other attributes (Johnston and Duke 2007). The present study similarly measures preferences for attributes of the means of achieving an environmental policy, rather than estimating the value that individuals or households place on the policy end, as is normally the objective of non-market valuation of environmental resources or quality. As Espinosa-Goded, Barreiro-Hurlé, and Ruto (2010) point out, agri-environmental policies will not be effective unless there is strong farmer participation. Separating the determinants of farmer willingness to consider participating in payment for environmental services schemes and the level of participation is the subject of another, more recent article (Ma etal. 2012). These issues are especially critical when talking about encouraging farmers to reduce or eliminate tillage in order to sequester atmospheric carbon in agricultural soils because those farmers who have not already adopted conservation tillage techniques are the ones who will have to be convinced to change their practices in order to satisfy what is referred to as additionality—that is, new adoption of practices that farmers did not previously adopt or are not expected to adopt in the future in the absence of an offset market. Related survey-based experimental work has examined whether different ways of framing conservation tillage affects expressed farmer interest in the practice (Andrews, etal. 2013; Andrews, etal. 2017). Conservation tillage is of particular interest because of the large technical potential to expand use of the practice to sequester atmospheric carbon (Smith, etal. 2008) and its cost-effectiveness as a means of carbon abatement (McCarl and Schneider 2001). Sampling Strategy, Survey Instrument, and Experimental Design The target population for our survey is farmers with tillable acres in the state of Indiana. Ideally this would mean drawing a random sample from a list of addresses for all farms with tillable acres in the state, but no such list is readily available. In an effort to obtain the largest list of farmer addresses possible, the researchers filed an electronic Freedom of Information Act (e-FOIA) request with the U.S. Department of Agriculture (USDA) for all federal farm subsidy recipients in the state of Indiana in 2009. Only names and addresses associated with a direct or countercyclical payment for either corn or soybean crops were retained, based on the fact that the vast majority of productive, arable land in the state is used to grow these two crops. Eliminating duplicate addresses for those who received farm payments for both corn and soybeans left 47,107 unique addresses, which compared very favorably with the 46,373 corn and soybean farms reported in the most recent prior agricultural census (USDA, National Agricultural Statistics Service 2007). A simple random sample of 2,000 farmers was drawn from the address list to receive the survey. Detailed respondent demographic information for those who completed the survey (see online supplementary appendix A) reveals a very good match with the target population when comparing our sample to USDA population farm statistics in terms of geographical distribution of farms across crop reporting districts. This approach predictably oversamples larger-sized farms relative to smaller farms because small farms are more likely not to participate in subsidy programs. More research is needed to understand the influence of non-conservation farm programs on the adoption of conservation practices in general, but this is beyond the scope of the current study. A mixed-mode survey with paper or internet response options available was conducted following the methods in Dillman, Smyth, and Christian (2008) in between planting and harvest times from July to September 2010. Internet surveys are becoming more popular due to their low costs and speedy completion times (Louviere etal. 2008). Hudson etal. (2004) found that internet surveys did not exhibit nonresponse bias. Similarly, Fleming and Bowden (2009), as well as Marta-Pedroso, Freitas, and Domingos (2007), found no significant differences when comparing results between web-based surveys, conventional mail, and in-person interview surveys. When looking specifically at choice experiments, Olsen (2009) found no significant differences in mean WTP estimates between internet surveys and mail surveys. There were five waves of contacts beginning with a standard invitation letter with a link to the web address where participants could respond online. The intent of allowing online or paper responses was to get as many early online responses as possible and then be able to afford to send print surveys to additional randomly-drawn addresses from the address list. This did not transpire, however, because of low initial online response rates. After sending three printed surveys by mail, with a reminder postcard in between the first and second printed mailings, as well as eliminating bad addresses, the final response rate was 42%. In addition to demographic questions and the CE, respondents were also asked questions about their beliefs related to climate change and the sources of information they receive about tillage (see Gramig, Barnard, and Prokopy 2013, for additional analysis). The survey instrument was developed after consulting many prior surveys of tillage practice adoption and adoption of other conservation practices. The authors consulted with the Conservation Technology Information Center, certified crop advisors that provide agronomic advice to farmers, tillage researchers, and extension educators in selecting the language used in the survey. The instrument was piloted with 25 farmers and undergraduate and graduate students with farming backgrounds in the College of Agriculture at Purdue University during fall 2009. Choice Experiment A CE asks participants to select an alternative—a product for purchase or a recreational site to visit in the case of consumers, or a management practice in the case of producers—by evaluating the bundle of characteristics or attributes of each alternative. Choice experiments mimic real-world purchasing and production situations by allowing decision makers to trade among the levels of different attributes when selecting from a set of alternative options. A CE was designed to elicit producers’ preferences for different tillage practices and associated attributes of different mechanisms to incentivize broader adoption of reduced tillage systems. The choice sets allowed participants to choose between one of two reduced tillage alternatives or to select conventional tillage (the fixed alternative available in every choice set), based on the tillage practice and three additional attributes listed in table 1: net revenue, the source of a carbon payment, and a multi-year contract requirement (online supplemental appendix B details the attribute definitions/explanations shown to survey respondents in preparation for the CE). Table 1 Attributes and Attribute Levels Evaluated in the Choice Experiment Attribute Attribute Levels Tillage Practice Conservation tillage No tillage (or no-till) Conventional tillage Increase in Net Revenue $10/acre $5/acre $0/acre Source of Carbon Payment Cap-and-trade Market Government Program None Multi-year Contract Requirement Contract Required No Contract Required Attribute Attribute Levels Tillage Practice Conservation tillage No tillage (or no-till) Conventional tillage Increase in Net Revenue $10/acre $5/acre $0/acre Source of Carbon Payment Cap-and-trade Market Government Program None Multi-year Contract Requirement Contract Required No Contract Required Note: Attribute and level definitions provided to respondents reported in online supplemental appendix A. Table 1 Attributes and Attribute Levels Evaluated in the Choice Experiment Attribute Attribute Levels Tillage Practice Conservation tillage No tillage (or no-till) Conventional tillage Increase in Net Revenue $10/acre $5/acre $0/acre Source of Carbon Payment Cap-and-trade Market Government Program None Multi-year Contract Requirement Contract Required No Contract Required Attribute Attribute Levels Tillage Practice Conservation tillage No tillage (or no-till) Conventional tillage Increase in Net Revenue $10/acre $5/acre $0/acre Source of Carbon Payment Cap-and-trade Market Government Program None Multi-year Contract Requirement Contract Required No Contract Required Note: Attribute and level definitions provided to respondents reported in online supplemental appendix A. The three different tillage practices were selected to include two forms of reduced tillage capable of achieving different levels of carbon sequestration (conservation tillage, and no tillage, or no-till), and conventional tillage, which was the most widely-used tillage practice in Indiana at the time the survey was conducted (ISDA 2013). The second attribute was the increase in net revenue resulting from each alternative in a given choice set, and included three levels: $0/acre, $5/acre, or $10/acre. Increases in net revenue were chosen to be consistent with market prices, the cost of tillage operations and crop yields at the time the survey was conducted. Because farmers could potentially sell offsets on a cap-and-trade market or be paid to supply climate regulation ecosystem services through a government program similar to a conservation cost-share program like the Environmental Quality Incentives Program (EQIP) in the United States, a third attribute in the choice experiment is the source of a carbon payment: commodity market, government payment, or no payment is made to the farmer. The final attribute is whether or not a multi-year contract is required. The motivation for including this attribute is based on the few existing soil carbon sequestration protocols that have been developed, which require a minimum number of years with reduced or no tillage being performed and the multi-year contracts required for existing USDA environmental cost-share programs (an example choice set can be found in the online supplemental appendix C). A main effects plus two-way interaction design was used to determine choice scenarios. The SAS OPTEX procedure was utilized to identify an experimental design (zero priors) maximizing D-efficiency (98.9). The final choice design resulted in 36 choice sets that were blocked into four groups of nine choice sets to keep the survey manageable for participants and reduce respondent fatigue. Therefore, each survey respondent was shown nine choice sets, each with two alternatives with varying levels of their attributes and a fixed conventional tillage alternative. The choice set order was randomized for online responses to mitigate any ordering effects (Champ, Boyle and Brown 2003; Loureiro and Umberger 2007), but this was not tractable for paper surveys. Because hypothetical CEs are simulated choices and there is no actual exchange of money in this CE, the following instructions were given to participants: “It is important that you make your selections like you would if you were actually facing these choices in making farm management decisions.” This “cheap-talk” strategy was employed to reduce hypothetical bias by informing participants of the context in which they should make the hypothetical choice before participation in the choice experiment (Cummings and Taylor 1999). Theoretical Framework and Research Methods Random Utility Theory Central to the idea of random utility theory is the assumption that economic agents seek to maximize their expected utility subject to the choice sets they are presented with. Based on Manski (1977), an individual’s utility is a random variable because the researcher has incomplete information. In random utility theory, utility (Uit) is obtained from selecting alternative i from a finite set of alternatives contained in choice set C in situation t. Therefore, utility can be characterized by the sum of a deterministic component dependent upon the attributes of the alternative and a stochastic component that is independently and identically distributed over all alternatives and choice scenarios. An individual n will select alternative i if the utility from selecting i is greater than the utility from alternative j, Unit > Unjt for all i ≠ j. Under the assumption that the deterministic portion, Vit, is linear in parameters, the specification of the general model can be expressed as Vit=βxit, where xit is a vector of attributes found in the ith alternative, and β is a vector of parameters associated with the attributes of the ith alternative. Multinomial logit models assume that individuals have homogeneous preferences for the product attributes; however, this assumption will not hold if individuals possess heterogeneous preferences, as suggested by recent literature that has used CEs to evaluate farmer preferences (Olynk, Tonsor, and Wolf 2010; Schulz and Tonsor 2010). Therefore, employing a model that allows for heterogeneous preferences is appropriate (Lusk, Roosen, and Fox 2003; Alfnes 2004). Random Parameters Logit The random parameters logit (RPL) model allows respondent preferences to be heterogeneous. By using the RPL model, we are able to directly estimate this heterogeneity across the evaluated attributes. In the RPL model, the random utility (Unit) of attribute i to individual n in situation t is Unit= vnit + [uni + ɛnit], where vnit is the deterministic portion of the utility function, uni is an error term that is distributed normally over individuals and attributes (but not choice sets), and ɛnit is the stochastic error that is independently and identically distributed over individuals, attributes, and choice sets. The subsequent model for the systematic portion of utility (for n on choice occasion t) is vi=β1NetRevi+β2Tillagei+β3Convi+β4Contracti+β5Govti+β6CATi. (1) The variable NetRev is the increase in net revenue shown to respondents in the choice set. Tillage is an effects-coded variable that equals one for conservation tillage and negative one for no-till, and Conv is a constant used to describe the utility associated with opting for conventional tillage rather than either of the two reduced tillage options. Contract is an effects-coded variable indicating whether or not a multi-year contract is required by the alternative. The remaining two variables are dummy-coded, with Govt indicating a carbon payment received from the government, and CAT indicating a carbon payment received from a cap-and-trade (CAT) market; the reference case (Govt = CAT = 0) is no carbon payment received from the government or a CAT market. Interaction terms with demographic and behavioral (revealed preference) variables that were included in the experimental design were also considered but not included in the final model because none were found to be statistically significant. It is hypothesized that farmers may have positive or negative preferences towards any of the attributes investigated. In order to allow WTA estimates to be either positive or negative, the random parameters, β, were assumed to be drawn from a normal distribution; the coefficients on all explanatory variables except NetRev were specified to vary normally across farmers. Because RPL does not exhibit the independence from irrelevant alternatives property, general patterns of correlated taste parameters can arise (Train 2003). The data are supportive of dependence in tastes and the RPL model allows for a better understanding of correlations in preferences across attributes if some of the estimates of the Cholesky matrix are statistically significant (Scarpa and Del Giudice 2004). Mean WTA estimates can be calculated as WTAk=(2*βk/β1), where βk is the coefficient on the attribute and β1 is the coefficient on increase in net revenue instead of price in a typical demand model. Instead of the usual binary variable, in effects coding the attributes take on a value of one when applicable, a value of −1 when the base category applies, and zero otherwise. The coefficient on attribute k is multiplied by two to calculate WTP for effects coded variables (Lusk, Roosen, and Fox 2003; Tonsor, Olynk, and Wolf 2009), and the two is excluded when calculating WTA for the binary coded attribute conventional tillage. There are numerous methods available to estimate confidence intervals for WTA estimates including delta, Fieller, Krinsky-Robb, and bootstrap methods. Hole (2007) found all of these methods to be reasonably accurate and yield similar results to one another; thus, the Krinsky-Robb method is used to construct 95% confidence intervals. One thousand observations for each WTA estimate were simulated by drawing from a multivariate normal distribution parameterized with the coefficients estimated using the RPL model and the variance-covariance matrix resulting from the same model (Krinsky and Robb 1986). In addition to mean estimates for the entire sample, the sample was divided into two groups based on revealed preference data indicating which respondents have already adopted either conservation tillage or no-till on their farmed acres, and those farmers who have not adopted any form of reduced tillage. Split sample mean and individual WTA estimates were obtained for each attribute of the choice experiment. The complete combinatorial method proposed by Poe, Giraud, and Loomis (2005) was used to determine if there is a statistically significant difference between the WTA for different contract attributes, given respondents’ revealed preferences for different tillage practices. Recently, concerns in the choice literature have emerged around the potential use of decision heuristics on behalf of respondents in order to simplify choice tasks. Attribute Non-Attendance (ANA) refers to respondents ignoring attributes when choosing between alternatives (Scarpa etal. 2009; Hensher and Greene 2010). Ignoring attributes is especially concerning to researchers, marketers, and industry professionals, as failing to account for ANA may impact the marketing and policy conclusions drawn. Past research has found significant evidence of ANA with meaningful impacts on WTP estimates. Scarpa etal. (2009) identified that 90% of their survey population did not attend to the price variable; this caused unrealistically high WTP estimates for rural landscape valuation. In this study we rely on an inferred method of ANA proposed by Hess and Hensher (2010) that uses the coefficient of variation (the ratio of standard deviation to the mean) on individual specific parameter estimates to measure the degree of noise-to-signal ratio on the variability of preference intensity for a given attribute as exhibited by the individual’s choice behavior. Results The coefficients of the parameters estimated in the RPL model are displayed in column one of table 2. The percentage of respondents exhibiting inferred ANA is displayed in the second column of the table. The levels of ANA found are low relative to the literature on its potential importance (Scarpa etal. 2009), and we estimate the same RPL model after removing the observations reflecting ANA to determine if there is a statistically significant impact on WTA estimates when ANA is ignored. Column three of table 2 presents these estimates, which are quantitatively different with the same signs and high significance as the estimates using all of the data. This demonstrates that correcting for ANA will result in different mean marginal WTA (MWTA) estimates. We explore this further below. Table 2 Estimated Parameters from Random Rarameters Logit Model and Attribute Non-attendance (ANA) Variable Full Dataset Rate of Attribute Non-Attendanceb ANA-Corrected Datac Coefficient Estimates Mean MWTA per acre Coefficient Estimates Mean MWTA per acre (Std. Err.) [95% Conf. interval]a [95% Conf. interval]a NetRev: Net revenue 0.16377*** 10% 0.19277*** (0.01022) (0.01359) Tillage: Conservation Tillaged 0.26093*** $3.21 17% 0.30998*** $3.24 (.05064) [$2.04, $4.63] (.07547) [$1.73, $5.02] Conv: Conventional Tillaged 0.77716*** $4.79 4% 1.61152*** $8.42 (.18272) [$2.70, $6.93] (.20535) [$6.40, $10.62] Contract: Contracte −0.86152*** −$10.57 7% −1.13217*** −$11.80 (.05450) [−$12.31,−$9.11] (.06159) [−$13.92, −$9.94] Govt: Government paymentf −0.20278*** −$2.48 20% −0.40313*** −$4.20 (.04824) [−$3.57, −$1.34] (.07118) [−$5.93, −$2.74] CAT: Cap-and-trade market paymentf −1.04265*** −$12.78 16% −1.36954*** −$14.32 (.06913) [−$15.10, −$10.71] (.09542) [−$17.32, −$11.86] Variable Full Dataset Rate of Attribute Non-Attendanceb ANA-Corrected Datac Coefficient Estimates Mean MWTA per acre Coefficient Estimates Mean MWTA per acre (Std. Err.) [95% Conf. interval]a [95% Conf. interval]a NetRev: Net revenue 0.16377*** 10% 0.19277*** (0.01022) (0.01359) Tillage: Conservation Tillaged 0.26093*** $3.21 17% 0.30998*** $3.24 (.05064) [$2.04, $4.63] (.07547) [$1.73, $5.02] Conv: Conventional Tillaged 0.77716*** $4.79 4% 1.61152*** $8.42 (.18272) [$2.70, $6.93] (.20535) [$6.40, $10.62] Contract: Contracte −0.86152*** −$10.57 7% −1.13217*** −$11.80 (.05450) [−$12.31,−$9.11] (.06159) [−$13.92, −$9.94] Govt: Government paymentf −0.20278*** −$2.48 20% −0.40313*** −$4.20 (.04824) [−$3.57, −$1.34] (.07118) [−$5.93, −$2.74] CAT: Cap-and-trade market paymentf −1.04265*** −$12.78 16% −1.36954*** −$14.32 (.06913) [−$15.10, −$10.71] (.09542) [−$17.32, −$11.86] Note: Sample size n = 648; all standard deviation estimates are significant for both datasets (p < 0.01); ***denotes p < 0.01; aindicates 95% confidence intervals found using the Krinsky-Robb method (Krinsky and Robb 1986); bbased on coefficient of variation metric of Hess and Hensher (2010); cindicates estimated coefficients excluding observations subject to ANA as a subset of the entire sample; reference levels of attributes for interpretation of estimation results are dno-till, eno multi-year contract, fno carbon payment. Table 2 Estimated Parameters from Random Rarameters Logit Model and Attribute Non-attendance (ANA) Variable Full Dataset Rate of Attribute Non-Attendanceb ANA-Corrected Datac Coefficient Estimates Mean MWTA per acre Coefficient Estimates Mean MWTA per acre (Std. Err.) [95% Conf. interval]a [95% Conf. interval]a NetRev: Net revenue 0.16377*** 10% 0.19277*** (0.01022) (0.01359) Tillage: Conservation Tillaged 0.26093*** $3.21 17% 0.30998*** $3.24 (.05064) [$2.04, $4.63] (.07547) [$1.73, $5.02] Conv: Conventional Tillaged 0.77716*** $4.79 4% 1.61152*** $8.42 (.18272) [$2.70, $6.93] (.20535) [$6.40, $10.62] Contract: Contracte −0.86152*** −$10.57 7% −1.13217*** −$11.80 (.05450) [−$12.31,−$9.11] (.06159) [−$13.92, −$9.94] Govt: Government paymentf −0.20278*** −$2.48 20% −0.40313*** −$4.20 (.04824) [−$3.57, −$1.34] (.07118) [−$5.93, −$2.74] CAT: Cap-and-trade market paymentf −1.04265*** −$12.78 16% −1.36954*** −$14.32 (.06913) [−$15.10, −$10.71] (.09542) [−$17.32, −$11.86] Variable Full Dataset Rate of Attribute Non-Attendanceb ANA-Corrected Datac Coefficient Estimates Mean MWTA per acre Coefficient Estimates Mean MWTA per acre (Std. Err.) [95% Conf. interval]a [95% Conf. interval]a NetRev: Net revenue 0.16377*** 10% 0.19277*** (0.01022) (0.01359) Tillage: Conservation Tillaged 0.26093*** $3.21 17% 0.30998*** $3.24 (.05064) [$2.04, $4.63] (.07547) [$1.73, $5.02] Conv: Conventional Tillaged 0.77716*** $4.79 4% 1.61152*** $8.42 (.18272) [$2.70, $6.93] (.20535) [$6.40, $10.62] Contract: Contracte −0.86152*** −$10.57 7% −1.13217*** −$11.80 (.05450) [−$12.31,−$9.11] (.06159) [−$13.92, −$9.94] Govt: Government paymentf −0.20278*** −$2.48 20% −0.40313*** −$4.20 (.04824) [−$3.57, −$1.34] (.07118) [−$5.93, −$2.74] CAT: Cap-and-trade market paymentf −1.04265*** −$12.78 16% −1.36954*** −$14.32 (.06913) [−$15.10, −$10.71] (.09542) [−$17.32, −$11.86] Note: Sample size n = 648; all standard deviation estimates are significant for both datasets (p < 0.01); ***denotes p < 0.01; aindicates 95% confidence intervals found using the Krinsky-Robb method (Krinsky and Robb 1986); bbased on coefficient of variation metric of Hess and Hensher (2010); cindicates estimated coefficients excluding observations subject to ANA as a subset of the entire sample; reference levels of attributes for interpretation of estimation results are dno-till, eno multi-year contract, fno carbon payment. All parameters were found to be statistically significant. Though interpretation of raw individual coefficients in random utility models has its limits and may not be particularly insightful, the coefficients had the expected signs and were used to estimate mean MWTA and confidence intervals. Note that the positive sign on the monetary variable, Increase in Net Revenue, has a MWTA connotation instead of the more familiar MWTP interpretation of a cost or price parameter that is more commonly encountered in the consumer demand and non-market valuation literatures. All of the parameters estimated as being random have statistically significant standard deviations (see table 2). Furthermore, all random parameters had statistically significant diagonal elements in the Cholesky matrix, indicating the presence of preference heterogeneity. Consequently, the reported mean MWTA estimates cannot be interpreted as being representative of the entire sample. The MWTA estimates and the associated 95% confidence intervals from the full dataset and the ANA-corrected data are presented in table 2. At first glance, the MWTA when correcting for ANA are uniformly higher and in some cases seem to be substantially higher as a percentage increase over the MWTA without correcting for ANA. Closer inspection of the 95% confidence intervals reveals these intervals are overlapping for all attributes, indicating that there is not a statistically significant difference in MWTA after correcting for ANA in the choice data. This has been demonstrated to be a more conservative standard than conventional significance testing when the null is true (Schenker and Gentleman 2001). This finding is consistent with Widmar etal. (2014), namely, that after correcting for inferred ANA the MWTA estimates do not exhibit differences that impact the overarching conclusions. The discussion of the MWTA results is based on the values estimated from the full data set (column one of table 2). When interpreting an estimated MWTA value in table 2, this number represents the value that farmers place on this attribute when the effects coded or dummy variable equals one relative tto the reference case (noted in able 1). Farmers' MWTA to implement no-till is $3.21 per acre or $4.79 per acre, relative to conservation tillage or conventional tillage, respectively. These values alone mask the distribution of MWTA among respondents. Histograms of the MWTA payment to adopt no-till relative to both tillage alternatives (see online supplemental appendix D) reveal that those respondents who have not previously adopted any form of reduced tillage have higher mean MWTA to adopt no-till than farmers who have previously adopted reduced tillage. Individual farmers who have not previously adopted conservation tillage or no-till are WTA an average of $14.21/acre to undertake no-till instead of conservation tillage; among prior adopters of reduced tillage, WTA is an order of magnitude lower at $1.40/acre. Individual farmers who have not previously adopted are WTA $39.40/acre to undertake no-till instead of conventional tillage compared to prior adopters not requiring any payment (not statistically different from zero) to switch from conventional tillage to no-till. Intuitively, this makes sense because those farmers who have not already adopted reduced tillage techniques on any of their land can be expected to be most resistant to changing tillage practices; farmers who have already adopted reduced tillage techniques would not require payment to switch to no-till. These mean individual farmer estimates are based on WTA calculated for all respondents given prior adoption of reduced tillage, and thus are not identical to the reported MWTA estimates in the online supplemental appendix E. Not surprisingly, there is a negative MWTA for the contract attribute, indicating that farmers would be WTA $10.57 per acre to enter a multi-year contract that limits their ability to change tillage practices for the duration of the contract term. This estimate can be interpreted as the option value that farmers attach to not being locked into a multi-year contract, and is particularly important because it cannot be estimated using production costs or revealed preferences alone. Note that the magnitude of this MWTA estimate should be interpreted carefully; the significance and sign of the estimate are clear, but a multi-level contract length attribute should be studied in future research to have a more precise estimate of the marginal WTA for an additional year of tillage restriction. This result is distinct from, but nonetheless consistent with, previous research that found Danish farmers value flexibility in contract terms of an agri-environmental policy to promote adoption of pesticide-free buffer zones (Christensen etal. 2011). Krah etal. (2017) found a similar dislike of multi-year contracts in the context of contracting to grow a bioenergy crop. The MWTA carbon payments from either the government or a commodity market for carbon is relative to no explicit carbon payment being made; farmers would rather forego carbon payments of $2.48/acre or $12.78/acre from either respective payment source if an increase in net revenue after adopting reduced tillage practices is possible without any carbon payment. This makes sense in light of the full experimental design (table 1) that includes alternatives with no carbon payment from the government or a cap-and-trade market, and given that mean MWTA is calculated at the mean increase in net revenue from adopting conservation tillage or no-till. An increase in net revenue without receiving a carbon payment is consistent with farmers who have experienced a small or negligible reduction in yield after reducing or eliminating tillage while also saving on machinery, fuel and labor costs, thus increasing net revenue per acre relative to conventional tillage. Indeed, the vast majority of all farmers who have already adopted conservation tillage or no-till (as defined in our survey instrument) likely made the decision to reduce or eliminate tillage in the absence of any government or market payment. Figure 1 is a radar plot of the estimated MWTA payment for different attributes in the CE that compares the mean individual welfare measures based on individuals’ revealed preferences for reduced tillage practices (respondents who have and have not adopted). The plotted WTA values are not exactly equal to those reported in tables 2 and 3, and despite being very similar to WTA estimates for the non-adopters in the online supplemental appendix E, are the means of individual estimates of WTA that result from RPL estimation. For the interested reader, the probability density of the same values in figure 1 are plotted in the online supplemental appendix D. There is a statistical difference between the MWTA of adopters and non-adopters for the tillage practice and multi-year contract variables, but not for the carbon payment source variables. Table 3 Mean Marginal Willingness to Accept (MWTA) Given Revealed Preferences for Reduced Tillage Adoption Choice Experiment Attribute Mean MWTA per acre, given have not previously adopted reduced tillage Equivalent EQIPa Payment Rate Equivalent CCAa Offset Price % Respondents with MWTA ≤ EQIP payment % Respondents with MWTA ≤ CCA price No-till relative to conservation tillage $10.90*** $6.66/Metric ton/acre/yrb Not Applicabled 18.4% ≤ EQIP payment No-till relative to conventional tillage $40.38*** $25.33/Metric ton/acre/yrc $11.95/Metric ton/acre/yr 27.2% ≤ EQIP payment 16.5% ≤ CCA price Multi-year Contract −$7.10*** 2 years Yes Government Payment −$0.15 Yes No Cap-and-trade Market Payment −$15.38 No Yes Choice Experiment Attribute Mean MWTA per acre, given have not previously adopted reduced tillage Equivalent EQIPa Payment Rate Equivalent CCAa Offset Price % Respondents with MWTA ≤ EQIP payment % Respondents with MWTA ≤ CCA price No-till relative to conservation tillage $10.90*** $6.66/Metric ton/acre/yrb Not Applicabled 18.4% ≤ EQIP payment No-till relative to conventional tillage $40.38*** $25.33/Metric ton/acre/yrc $11.95/Metric ton/acre/yr 27.2% ≤ EQIP payment 16.5% ≤ CCA price Multi-year Contract −$7.10*** 2 years Yes Government Payment −$0.15 Yes No Cap-and-trade Market Payment −$15.38 No Yes Note: Asterisks denote the following: ***= p < 0.01; significance of difference in MWTA compared to respondents that have adopted reduced tillage (online supplemental appendix E). Superscript aassumes 0.6 Mg per acre per year carbon sequestration rate based on Chicago Climate Exchange protocols (CCX 2009); EQIP denotes USDA Environmental Quality Incentives Program and CCA denotes California Carbon Allowance; bbased on USDA-NRCS practice code 345 (EQIP 2014) and Indiana 2013 payment rates (USDA-NRCS 2014); cbased on USDA-NRCS practice code 329 (EQIP 2014) and Indiana 2013 payment rates (USDA-NRCS 2014); donly tillage practices under NRCS practice code 329 (no-till and strip till) are eligible under CCX offset protocols (CCX 2009). Table 3 Mean Marginal Willingness to Accept (MWTA) Given Revealed Preferences for Reduced Tillage Adoption Choice Experiment Attribute Mean MWTA per acre, given have not previously adopted reduced tillage Equivalent EQIPa Payment Rate Equivalent CCAa Offset Price % Respondents with MWTA ≤ EQIP payment % Respondents with MWTA ≤ CCA price No-till relative to conservation tillage $10.90*** $6.66/Metric ton/acre/yrb Not Applicabled 18.4% ≤ EQIP payment No-till relative to conventional tillage $40.38*** $25.33/Metric ton/acre/yrc $11.95/Metric ton/acre/yr 27.2% ≤ EQIP payment 16.5% ≤ CCA price Multi-year Contract −$7.10*** 2 years Yes Government Payment −$0.15 Yes No Cap-and-trade Market Payment −$15.38 No Yes Choice Experiment Attribute Mean MWTA per acre, given have not previously adopted reduced tillage Equivalent EQIPa Payment Rate Equivalent CCAa Offset Price % Respondents with MWTA ≤ EQIP payment % Respondents with MWTA ≤ CCA price No-till relative to conservation tillage $10.90*** $6.66/Metric ton/acre/yrb Not Applicabled 18.4% ≤ EQIP payment No-till relative to conventional tillage $40.38*** $25.33/Metric ton/acre/yrc $11.95/Metric ton/acre/yr 27.2% ≤ EQIP payment 16.5% ≤ CCA price Multi-year Contract −$7.10*** 2 years Yes Government Payment −$0.15 Yes No Cap-and-trade Market Payment −$15.38 No Yes Note: Asterisks denote the following: ***= p < 0.01; significance of difference in MWTA compared to respondents that have adopted reduced tillage (online supplemental appendix E). Superscript aassumes 0.6 Mg per acre per year carbon sequestration rate based on Chicago Climate Exchange protocols (CCX 2009); EQIP denotes USDA Environmental Quality Incentives Program and CCA denotes California Carbon Allowance; bbased on USDA-NRCS practice code 345 (EQIP 2014) and Indiana 2013 payment rates (USDA-NRCS 2014); cbased on USDA-NRCS practice code 329 (EQIP 2014) and Indiana 2013 payment rates (USDA-NRCS 2014); donly tillage practices under NRCS practice code 329 (no-till and strip till) are eligible under CCX offset protocols (CCX 2009). Figure 1 View largeDownload slide Marginal willingness to accept (WTA) payment for carbon sequestration scheme attributes Note: Asterisks *** indicate that mean MWTA of adopters (dashed line) and non-adopters (solid line) are significantly different (p < 0.01); values are calculated means of individual-specific random parameter logit coefficient estimates, the distribution of which is plotted in online supplementary appendix D. Values presented on a positive scale to facilitate visual presentation, sometimes reversing interpretation from WTA estimates reported in tables. Figure 1 View largeDownload slide Marginal willingness to accept (WTA) payment for carbon sequestration scheme attributes Note: Asterisks *** indicate that mean MWTA of adopters (dashed line) and non-adopters (solid line) are significantly different (p < 0.01); values are calculated means of individual-specific random parameter logit coefficient estimates, the distribution of which is plotted in online supplementary appendix D. Values presented on a positive scale to facilitate visual presentation, sometimes reversing interpretation from WTA estimates reported in tables. Discussion and Policy Implications The only two active carbon limits in the United States today cover non-agricultural emissions in California and the nine northeastern states that are part of the Regional Greenhouse Gas Initiative (RGGI). Recent secondary market spot prices in these two markets were $11.95 and $4.74 per metric ton, respectively (CMNA 2014). Observed price differences reflect the relative scarcity of allowances and stringency of the carbon caps in each market, with prices in both markets reflecting the underlying abatement costs of firms (or the floor price, whichever is lower). While still in its infancy, the market for California Carbon Allowances (CCA) is the largest carbon market in North America in terms of volume, and carbon emissions regulations in California require more aggressive reductions over time than the RGGI market. Neither market has developed formal protocols for awarding certified emissions reductions (UNFCCC terminology), more commonly referred to as “offsets” specifically for managing agricultural soils. Payments to farmers for offsets sold to participants in these markets are the basis for the cap-and-trade commodity market payment for soil carbon included in the experimental design for this study. In general, it would not be correct to directly compare the MWTA estimates from our survey with carbon prices or conservation payments because the monetary variable in the CE is the increase in net revenue per acre. Equality between the net revenue attribute and the types of payments studied holds if either (a) yield, input quantities, and input and output prices are constant, or (b) any combination of yield, input quantities, and input and output price changes exactly cancel one another out, such that the net revenue change reflected in the CE is due entirely to a change in soil carbon derived income. We assume that at least one of these conditions is satisfied in the following discussion to evaluate farmer WTA payment to change tillage practices under existing payment mechanisms. The existing USDA Natural Resources Conservation Service (NRCS) Environmental Quality Incentives Program (EQIP) includes government conservation payments for implementing reduced tillage (practice code 345) and no-till/strip till (practice code 329) practices. These practices are the motivation for the government payment source included in the CE. The EQIP payment levels in Indiana (for a maximum of two years) for conservation tillage and no-till/strip tillage are $4/acre and $14/acre, respectively (USDA-NRCS 2014). Based on the 0.6 metric ton per acre per year carbon sequestration rate that was the basis for awarding offsets under the Chicago Climate Exchange (CCX 2009), these amounts are equivalent to $6.66 and $25.33 per metric ton sequestered. Because additional or new sequestration is of the greatest interest for climate change mitigation, table 3 takes farmers’ revealed preferences for tillage practices into account to compare individuals’ MWTA for different soil carbon sequestration contract attributes, and to determine the percentage of respondents with MWTA less than or equal to (a) EQIP payments for the same practices and (b) carbon market prices. There is a statistically significant (p < 0.01) difference between the mean MWTA of reduced tillage adopters and non-adopters for the tillage type and multi-year contract attributes (table 3) as determined by the complete combinatorial test (Poe etal. 2005) reported in the online supplemental appendix E. The difference in MWTA from different carbon payment sources, whether from a market or a government conservation program, was not statistically significant between adopters and non-adopters. With an eye toward future policy development and more mature carbon markets that may include soil carbon offset protocols, it is important to examine individual decision makers’ preferences for different attributes of market-based payments or government contracts intended to compensate farmers for environmental services provided to society. The results in table 3 use respondent-specific random parameter estimates to calculate individual MWTA for each contract attribute included in the CE. The two rightmost columns of table 3 report the equivalent EQIP payment rate and CCA carbon price per metric ton per acre of land per year, based on the carbon sequestration rate in the Chicago Climate Exchange no-till offset protocols (CCX 2009). These two payment mechanisms embody different levels of the attributes included in the CE, with EQIP having separate practice standards for conservation tillage and no-till, but the CCX offset protocols only allow contracts for no-till (or strip till) acres. Only 18% of respondents who did not report using any form of reduced tillage had MWTA less than the soil carbon equivalent EQIP payment rates for conservation tillage, and 27% for no-till. Similarly, 16.5% of non-adopters had MWTA low enough to imply carbon offset market participation is possible (before any transaction costs). These percentages are particularly important from a climate change mitigation standpoint because only new or increased adoption of carbon sequestering practices results in additional emissions being offset relative to baseline practices when a new carbon market is created. This is the essence of the “additionality” requirement in the expired Kyoto Protocol. The UNFCCC has yet to approve protocols for soil carbon sequestration, despite a large body of research demonstrating that enhanced management of agricultural soils can achieve significant reductions in atmospheric carbon levels (McCarl and Schneider 2001; Smith etal. 2008) and prior development of protocols in two different carbon markets in North America. This research sheds light on the prices that may be necessary to incentivize farmers—and in particular, those who have not previously adopted conservation tillage—to adopt less intensive tillage practices that sequester carbon in soils. Furthermore, we examine key attributes of policy instruments that can be expected to be more or less attractive to the supply side of the soil carbon offset market, and thus influence farmer participation. Stated preferences reveal that the full cost of switching tillage practices is not captured by the economic opportunity cost alone, and that farmers’ option value associated with not entering a multi-year contract must also be considered. Conclusions This study finds that farmers who have not previously adopted reduced tillage practices would be WTA annual payments of $10/acre or $40/acre (table 3) to switch to no-till from (more intensive) conservation or conventional tillage systems, respectively. Though farmers would prefer to experience an increase in net revenue from adopting conservation tillage without having to receive a government payment or participate in a carbon offset market, they dislike a government carbon payment less (-$2.48/acre) than a carbon market payment (-$12.78/acre). Farmer disutility from being required to enter into a multi-year contract is estimated to be $10.57/acre relative to not being required to sign a contract (table 2), which is assumed to be required if a payment is received for carbon sequestration. This is a new finding not previously reported in the carbon sequestration and no-till literatures that captures an option value component of farmers’ opportunity costs of entering into a carbon sequestration contract. A similar finding was recently reported using similar methods to assess farmer willingness to grow dedicated bioenergy crops (Barham etal. 2016). This has not been accounted for in policy discussions and can be expected to increase the cost of procuring soil carbon sequestration over and above inevitable transaction costs in a carbon market. Attribute non-attendance in our sample of farmers was found at much lower rates (≤20%) than has been found previously in the environmental economics literature (Scarpa etal. 2009), and does not have a statistically significant effect on attribute WTA estimates after correcting for ANA in the choice data (see table 2). One explanation for this finding may be that the choice context is not only familiar but real to the sampled population, thus minimizing hypothetical bias that is more likely in other choice contexts. Considering both the stated preference data from the choice experiment and revealed preference (RP) data on the tillage practices that respondents have actually adopted, we find statistically and practically significant differences between no-till adopters’ and non-adopters’ mean marginal WTA payment to switch to no-till (see table 3). This significant difference is consistent with intuition, and practically serves to provide a quantitative estimate of the payment required to induce adoption of no-till by those farmers who have not previously adopted this practice. Deeper investigation reveals that the share of non-adopters (RP) in the sample with WTA less than or equal to existing government and market payment mechanisms are 27% and 16%, respectively. This indicates that higher prices are necessary in government programs and carbon markets to induce widespread expansion and/or new adoption of soil carbon sequestering practices among row crop farmers (see table 3). Even in a post-Kyoto Protocol world, the practical challenge of operationalizing payments for soil carbon sequestration to offset carbon emissions elsewhere in the economy lies in additional and permanent carbon abatement. This may not be as much of a concern for voluntary offset markets, but is critical for the integrity of offsets sold to entities subject to a regulatory emissions limit or cap. The soil carbon sequestration protocols previously developed for the Chicago Climate Exchange (CCX) utilized a reserve pool of carbon credits that was similar in function to the way that point-nonpoint source trading ratios operate in water quality trading markets. The reserve pool was developed to manage uncertainty in the level of carbon abatement and to protect against offset providers releasing carbon sequestered after signing a contract to undertake continuous no-till. The nascent California carbon market has utilized sequestration benchmarking (tree species, climate, etc.) together with measured higher levels of carbon accumulation in its forestry offset protocols to effectively integrate prior (to market creation) adopters into that market while satisfying additionality tests consistent with previously approved UNFCCC protocols. This structure could technically work for crediting prior adopters of no-till as well, subject to the precision limits of soil organic carbon testing. Permanence remains the single largest challenge associated with using conservation tillage to achieve carbon abatement. If land converted to no-till is, even many years later, disturbed through tillage or other means, then the carbon sequestered to offset emissions elsewhere in the economy is released at a later date. Hence, the GHG emissions previously offset were merely delayed in time. Rapid advances in the quality and availability of remotely-sensed data can potentially reduce the monitoring costs of overseeing many decentralized carbon offset providers, even if carbon credit aggregators operate in between landowners and a carbon market. Penalties for violating permanence requirements in carbon offset contracts could be reliably and cost-effectively enforced using remotely-sensed data going forward, perhaps requiring sequestration reversals to be compensated for by purchasing emissions allowances in the spot market at current rates if there is not a reserve pool to fall back on or its capacity is not sufficient. This would seem to adequately address the environmental integrity of using soil carbon sequestration as certified emissions reductions in carbon markets, but this may come at the cost of low participation on the supply side of the offset market. The prices inferred by the WTA estimates in this study suggest that, even if institutional and contractual terms can satisfy additionality and permanence requirements consistent with the UNFCCC, the carbon prices observed today in the California, RGGI, and EU-ETS carbon markets are not high enough to entice widespread new adoption of continuous no-till. If prices do not increase sufficiently or contract terms (permanence, required abatement contribution to reserve pools, penalties for reversal of sequestration practices) are not acceptable to farmers, then the market for soil carbon offsets can be expected to remain thin or not function at all. This would effectively render this low-cost carbon abatement option infeasible to society when taking actions to meet our commitments under the Paris Accord. Supplementary Material Supplementary material is available online at https://doi.org/10.13012/B2IDB-0410446_V1. Footnotes 1Such credits can be generated many different ways, and protocols for crediting have been developed previously by the now-defunct voluntary Chicago Climate Exchange (CCX), the government of Alberta (https://goo.gl/VGNzy0), the state of California (https://goo.gl/PblCjx), and the UNFCCC (https://goo.gl/UDseqQ). Acknowledgements This research was funded in part by Agricultural Research at Purdue, the Purdue Climate Change Research Center, and the USDA National Institute of Food and Agriculture Multi-State Hatch Project W-3133. Research assistance was provided by Jessa B. Barnard and Seong Do Yun, and survey management was coordinated by Cynthia Salazar in Linda Prokopy’s Natural Resource Social Sciences Lab in the Forestry and Natural Resources Department at Purdue University. Helpful comments were received from participants at the 2014 World Congress of Environmental and Resource Economists, the 2015 Agricultural Economics Society meeting, and seminars at Oregon State University, University of Kentucky, University of Illinois, and Virginia Tech. References Alfnes F. 2004. Stated Preferences for Imported and Hormone-Treated Beef: Application of a Mixed Logit Model. European Review of Agricultural Economics 31( 1): 19– 37. Google Scholar CrossRef Search ADS Andrews A.C., Clawson R.A., Gramig B.M., Raymond L. 2013. Why Do Farmers Adopt Conservation Tillage? An Experimental Investigation of Framing Effects. Journal of Soil and Water Conservation 68 ( 6): 501– 11. Google Scholar CrossRef Search ADS Andrews A.C., Clawson R.A., Gramig B.M., Raymond L. 2017. Finding the Right Value: Framing Effects on Domain Experts. Political Psychology 38 ( 2): 261– 78. Google Scholar CrossRef Search ADS Barham B.L., Mooney D.F., Swinton S.M. 2016. Inconvenient Truths about Landowner (Un) Willingness to Grow Dedicated Bioenergy Crops. Choices 31 ( 4). Broch S., Vedel S. 2012. Using Choice Experiments to Investigate the Policy Relevance of Heterogeneity in Farmer Agri-Environmental Contract Preferences. Environmental & Resource Economics 51 ( 4): 561– 81. Google Scholar CrossRef Search ADS Carbon Market North America. 2014. Vol. 9 (2). March 7, 2014. Reuters-Point Carbon North America. Available at: http://crf.thomsonreuters.com/polopoly_fs/1.4434896!CMNA20140307.pdf. Carlsson F., Frykblom P., Lagerkvist C.J. 2007. Consumer Willingness to Pay for Farm Animal Welfare: Mobile Abattoirs versus Transportation to Slaughter. European Review of Agricultural Economics 34 ( 3): 321– 44. Google Scholar CrossRef Search ADS Champ P.A., Boyle K.J., Brown T.C. 2003. A Primer on Nonmarket Valuation. New York: Springer. Google Scholar CrossRef Search ADS Chicago Climate Exchange. 2009. CCX Offset Project Protocol: Agricultural Best Management Practices – Continuous Conservation Tillage and Conversion to Grassland Soil Carbon Sequestration. Intercontinental Exchange, Inc. Chicago, Illinois. Available at: https://www.theice.com/publicdocs/ccx/protocols/CCX_Protocol_Conservation_Tillage_and_Grassland.pdf. Christensen T., Pedersen A.B., Nielsen H.O., Mørkbak M.R., Hasler B., Denver S. 2011. Determinants of Farmers’ Willingness to Participate in Subsidy Schemes for Pesticide-Free Buffer Zones—A Choice Experiment Study. Ecological Economics 70 ( 8): 1558– 64. Google Scholar CrossRef Search ADS Cummings R.G., Taylor L.O. 1999. Unbiased Value Estimates for Environmental Goods: A Cheap Talk Design for the Contingent Valuation Method. American Economic Review 89 ( 3): 649– 65. Google Scholar CrossRef Search ADS Dillman D.A., Smyth J.D., Christian L.M. 2008. Internet,Phone, Mail, and Mixed-mode Surveys: The Tailored Design Method, Third Edition . Hoboken, NJ: John Wiley & Sons. Environmental Protection Agency. 2014. Carbon Pollution Emission Guidelines for Existing Stationary Sources: Electric Utility Generating Units; Proposed Rule . Washington, DC: Federal Register, pp. 130. Espinosa-Goded M., Barreiro-Hurlé J., Ruto E. 2010. What Do Farmers Want from Agri-Environmental Scheme Design? A Choice Experiment Approach. Journal of Agricultural Economics 61 ( 2): 259– 73. Google Scholar CrossRef Search ADS Fleming C.M., Bowden M. 2009. Web-based Surveys as an Alternative to Traditional Mail Methods. Journal of Environmental Management 90 ( 1): 284– 92. Google Scholar CrossRef Search ADS PubMed Gao Z., Schroeder T.C. 2009. Effects of Label Information on Consumer Willingness-To-Pay for Food Attributes. American Journal of Agricultural Economics 91 ( 3): 795– 809. Google Scholar CrossRef Search ADS Gracia A., Loureiro M.L., Nayga R.MJr. 2009. Valuing Animal Welfare Labels with Experimental Auctions: What Do We Learn from Consumers? Paper presented at the International Association of Agricultural Economists conference, Beijing, China. Gramig B.M. 2012. Some Unaddressed Issues in Proposed Cap-and-Trade Legislation Involving Agricultural Soil Carbon Sequestration. American Journal of Agricultural Economics 94 ( 2): 360– 7. Google Scholar CrossRef Search ADS Gramig B.M., Barnard J.M., Prokopy L.S. 2013. Farmer Beliefs about Climate Change and Carbon Sequestration Incentives. Climate Research 56 ( 2): 157– 67. Google Scholar CrossRef Search ADS Hess S., Hensher D.A. 2010. Using Conditioning on Observed Choices to Retrieve Individual-Specific Attribute Processing Strategies. Transportation Research Part B: Methodological 44 ( 6): 781– 90. Google Scholar CrossRef Search ADS Hensher D., Greene W. 2010. Non-Attendance and Dual Processing of Common-Metric Attributes in Choice Analysis: A Latent Class Specification. Empirical Economics 39 ( 2): 413– 26. Google Scholar CrossRef Search ADS Hole A.R. 2007. A Comparison of Approaches To Estimating Confidence Intervals for Willingness To Pay Measures. Health Economics 16 ( 8): 827– 84. Google Scholar CrossRef Search ADS PubMed Hudson D., Seah L.H., Hite D., Haab T. 2004. Telephone Presurveys, Self-Selection, and Non-Response Bias to Mail and Internet Surveys in Economic Research. Applied Economics Letters 11 ( 4): 237– 40. Google Scholar CrossRef Search ADS Indiana State Department of Agriculture. 2013. Conservation Tillage Data. Available at: http://www.in.gov/isda/2383.htm. Johnston R.J., Duke J.M. 2007. Willingness to Pay for Agricultural Land Preservation and Policy Process Attributes: Does the Method Matter? American Journal of Agricultural Economics 89 ( 4): 1098– 115. Google Scholar CrossRef Search ADS Krah K., Petrolia D.L., Williams A., Coble K.H., Harri A., Rejesus R. 2017. Producer Preferences for Contracts on A Risky Bioenergy Crop. Applied Economic Perspectives and Policy . doi:org/10.1093/aepp/ppx034. Krinsky I., Robb A.L. 1986. On Approximating The Statistical Properties of Elasticities. The Review of Economics and Statistics 68 ( 4): 715– 9. Google Scholar CrossRef Search ADS Lind L.W. 2007. Consumer Involvement and Perceived Differentiation of Different Kinds of Pork–A Means-End Chain Analysis. Food Quality and Preference 18 ( 4): 690– 700. Google Scholar CrossRef Search ADS Loureiro M.L., Umberger W.J. 2007. A Choice Experiment Model for Beef: What U.S. Consumer Responses Tell Us about Relative Preferences for Food Safety, Country-Of-Origin Labeling and Traceability. Food Policy 32 ( 4): 496– 514. Google Scholar CrossRef Search ADS Louviere J.J., Islam T., Wasi N., Street D., Burgess L. 2008. Designing Discrete Choice Experiments: Do Optimal Designs Come at a Price? Journal of Consumer Research 35 ( 2): 360– 75. Google Scholar CrossRef Search ADS Lusk J.L., Norwood F.B. 2008. A Survey to Determine Public Opinion about the Ethics and Governance of Farm Animal Welfare. Journal of the American Veterinary Medical Association 233 ( 7): 1121– 6. Google Scholar CrossRef Search ADS PubMed Lusk J.L., Roosen J., Fox J.A. 2003. Demand for Beef from Cattle Administered Growth Hormones or Fed Genetically Modified Corn: A Comparison of Consumers in France, Germany, the United Kingdom, and the United States. American Journal of Agricultural Economics 85 ( 1): 16– 29. Google Scholar CrossRef Search ADS Ma S., Swinton S., Lupi F., Jolejole-Foreman C. 2012. Farmers' Willingness to Participate in Payment-for-Environmental-Services Programmes. Journal of Agricultural Economics 63 ( 3): 604– 26. Google Scholar CrossRef Search ADS Manski C.F. 1977. The Structure of Random Utility Models. Theory and Decision 8 ( 3): 229– 54. Google Scholar CrossRef Search ADS Marta-Pedroso C., Freitas H., Domingos T. 2007. Testing for the Survey Mode Effect on Contingent Valuation Data Quality: A Case Study of Web-based versus In-Person Interviews. Ecological Economics 62 ( 3): 388– 98. Google Scholar CrossRef Search ADS McCarl B.A., Schneider U.A. 2001. Greenhouse Gas Mitigation in U.S. Agriculture and Forestry. Science 294: 2481– 2. Google Scholar CrossRef Search ADS PubMed Nilsson T., Foster K., Lusk J. 2006. Marketing Opportunities for Certified Pork Chops. Canadian Journal of Agricultural Economics 54 ( 4): 567– 83. Google Scholar CrossRef Search ADS Olsen S.B. 2009. Choosing between Internet and Mail Survey Modes for Choice Experiment Surveys Considering Non-Market Goods. Environmental and Resource Economics 44 ( 4): 591– 610. Google Scholar CrossRef Search ADS Olynk N.J., Ortega D.L. 2013. Consumer Preferences for Verified Dairy Cattle Management Practices in Processed Dairy Products. Food Control 30 ( 1): 298– 305. Google Scholar CrossRef Search ADS Olynk N.J., Tonsor G.T., Wolf C.A. 2010. Consumer Willingness To Pay for Livestock Credence Attribute Claim Verification. Journal of Agricultural and Resource Economics 35: 261– 80. Poe G.L., Giraud K.L., Loomis J.B. 2005. Computational Methods for Measuring the Difference of Empirical Distributions. American Journal of Agricultural Economics 87 ( 2): 353– 65. Google Scholar CrossRef Search ADS Scarpa R., Del Giudice T. 2004. Market Segmentation via Mixed Logit: Extra-Virgin Olive Oil in Urban Italy. Journal of Agricultural and Food Industrial Organization 2 ( 1): 141– 60. Google Scholar CrossRef Search ADS Scarpa R., Gilbride T.J., Campbell D., Hensher D.A. 2009. Modeling Attribute Non-Attendance in Choice Experiments for Rural Landscape Valuation. European Review of Agricultural Economics 36 ( 2): 151– 74. Google Scholar CrossRef Search ADS Schenker N., Gentleman J.F. 2001. On Judging Significance of Difference by Examining the Overlap Between Confidence Intervals. American Statistician 53 ( 3): 182– 6. Google Scholar CrossRef Search ADS Schulz L.L., Tonsor G.T. 2010. Cow‐Calf Producer Preferences for Voluntary Traceability Systems. Journal of Agricultural Economics 61 ( 1): 138– 62. Google Scholar CrossRef Search ADS Smith P., Martino D., Cai Z., Gwary D., Janzen H., Kumar P., McCarl B., et al. 2008. Greenhouse Gas Mitigation in Agriculture. Philosophical Transactions of the Royal Society B-Biological Sciences 363 ( 1492): 789– 813. Google Scholar CrossRef Search ADS Tonsor G.T., Olynk N.J., Wolf C.A. 2009. Consumer Preferences for Animal Welfare Attributes: The Case of Gestation Crates. Journal of Agricultural and Applied Economics 41 ( 3): 713– 30. Google Scholar CrossRef Search ADS Tonsor G.T., Wolf C.A. 2011. On mandatory labeling of animal welfare attributes. Food Policy 36 ( 3): 430– 437 Google Scholar CrossRef Search ADS Train K.E. 2003. Discrete Choice Methods with Simulation . Cambridge, MA: Cambridge University Press. Google Scholar CrossRef Search ADS U.S. Department of Agriculture, National Agricultural Statistics Service. 2007. 2007 Census of Agriculture, State Data. Available at: www.agcensus.usda.gov/Publications/2007/Full_Report/index. U.S. Department of Agriculture, National Agricultural Statistics Service. Indiana EQIP Payment Rates and Eligible Practices. Available at: http://www.nrcs.usda.gov/wps/portal/nrcs/main/in/programs/financial/eqip/. Widmar N.J.O., Ortega D.L., McKenzie A., Paudel K. 2014. Comparing Consumer Preferences for Livestock Production Process Attributes across Products, Species, and Modeling Methods. Journal of Agricultural and Applied Economics 46 ( 3): 375– 91. Wolf C.A., Tonsor G.T., Olynk N.J. 2011. Understanding U.S. Consumer Demand for Milk Production Attributes. Journal of Agricultural and Resource Economics 36 ( 2): 326– 42. © The Author 2017. Published by Oxford University Press on behalf of the Agricultural and Applied Economics Association. All rights reserved. For Permissions, please e-mail: firstname.lastname@example.org
Applied Economic Perspectives and Policy – Oxford University Press
Published: Oct 7, 2017
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera