Beruflich Dokumente
Kultur Dokumente
3, 2012, 604–626
doi: 10.1111/j.1477-9552.2012.00358.x
Abstract
Understanding farmers’ willingness to participate in agricultural payment-for-envi-
ronmental-services (PES) programmes is an essential precondition for designing
effective and efficient programmes. Willingness to participate is typically examined
via stated preference surveys using the standard hurdle model for whether and how
much to participate. Among respondents who decline to participate, such analyses
cannot distinguish between respondents who declined due to the payment level and
those who were not interested at all. This paper applies a double hurdle model to
incorporate a prior condition for whether a respondent is even willing to consider
participating in the PES market. The model uses a unique stated preference survey
permitting separation of the consideration and enrolment decisions of 1,700 farmers
in Michigan, USA. The first hurdle probit model suggests that farmers’ willingness
to consider PES chiefly depends on farm and farmer characteristics, while the second
hurdle tobit model shows that decisions on whether and how much to enrol depend
more on the payment offer and marginal benefit–cost criteria. This study provides
fresh insights on facilitating farmer participation in PES programmes using tiered
strategies that differ in costs of programme payment and administration.
1
Shan Ma is a post-doctoral scholar, Natural Capital Project, Stanford University. E-mail:
mashan@stanford.edu for correspondence. Scott M. Swinton is a Professor and Frank Lupi
is a Professor at the Department of Agricultural, Food and Resource Economics, Michigan
State University, East Lansing, Michigan. Christina Jolejole-Foreman is a PhD candidate at
the Department of Agricultural and Consumer Economics, University of Illinois at Urbana-
Champaign, Urbana, Illinois. The authors gratefully acknowledge support from the National
Science Foundation under the Human and Social Dynamics and Long-term Ecological
Research programmes. Thanks are also due to Michigan AgBioResearch and the Michigan
State University Environmental Science and Policy Program. For valuable discussions, they
thank Phil Robertson, Natalie Rector, Rob Shupp, Soren Anderson, Sara Syswerda, Lenisa
Vangjel and Daniel Mooney. For data, they thank the 1,700 survey respondents as well as
participants in focus groups and pre-tests. This research was conducted while the lead author
was a PhD student at the Department of Agricultural, Food and Resource Economics,
Michigan State University, Michigan, USA.
1. Introduction
Agriculture is an ecosystem transformed by humans to generate agricultural produc-
tion. It supplies market goods, such as food, fibre and fuel. In addition, agriculture
also provides non-market environmental services (ES) that depend on farmers’
choices of production inputs and management practices (Wossink and Swinton,
2007). However, because only a small portion of the benefits from non-market ES
accrue to farmers, they have little incentive to produce these services.
Various agri-environmental policies have been implemented to motivate the supply
of environmental services. One prominent example is payment-for-environmental-
services (PES)2, which is attracting increasing attention globally as a policy innova-
tion that translates external ecosystem values into real financial incentives for local
providers (Engel et al., 2008). In the United States, the policy focus has recently
shifted from land retirement programmes, such as the Conservation Reserve Pro-
gramme (CRP), to conservation on working land – land used primarily for crop
production and grazing (Cattaneo et al., 2005). Two prominent examples are the
Environmental Quality Incentives Programme (EQIP) initiated in 1996 and the
Conservation Security Programme (CSP) initiated in 2002.3 The focus of these pro-
grammes has evolved from restricting local negative externalities, such as soil ero-
sion and nitrate run-off, to providing public goods, such as greenhouse gas
mitigation and biodiversity. Similar PES programmes that pay land owners for
effective agricultural land management have also been launched in other developed
countries. The agri-environmental schemes were introduced in 1992 in the European
Union’s Common Agricultural Policy (CAP) to compensate farmers financially for
voluntarily providing environmental services beyond good agricultural practices that
meet minimum legal standards (European Commission Directorate-General for
Agriculture and Rural Development (EC, 2004). There are various agri-environmen-
tal schemes in 26 of 44 European countries (Kleijn and Sutherland, 2003). Examples
include: the Environmental Stewardship Scheme (ESS) and Countryside Steward-
ship Schemes (CSS, replaced by an Environmental Stewardship scheme in 2005) in
the United Kingdom (Dobbs and Pretty, 2008); Rural Environment Protection
Scheme (REPS) in Ireland (Hynes and Garvey, 2009); and the Nitrate Reduction
Programme (NRP) in Greece (Giovanopoulou et al., 2011). There are also user-
financed PES programmes, such as the Vittel (Nestlé Waters) watershed protection
programme in Eastern France (Perrot-Maitre, 2006).
An essential precondition for the success of an agricultural PES programme is
that farmers are willing to participate. If they are, then the next question is how
much. Both decisions involve weighing the potential benefits and costs in PES
2
PES is formally defined as a voluntary transaction where a well-defined environmental
service or a land use likely to secure that service is being ‘bought’ by a service buyer from a
service provider if and only if the service provider secures service provision (Wunder, 2005).
3
The EQIP and CSP programmes are classified as PES programmes by Wunder et al. (2008).
Only a proportion of farmers who consider participating will elect to enrol, as the
rest are unsatisfied with the programme payment offered. Explicitly modelling this
additional choice on whether to consider a PES programme is likely to advance
our understanding of farmers’ participation decisions and provide insights on poten-
tial improvements in current PES programmes. Using the same stated preference
survey data as Jolejole (2009), but with additional information permitting separa-
tion of the consideration and participation decisions, we analyse Michigan corn and
soybean farmers’ decisions in four hypothetical PES programmes, in order to exam-
ine the differences in decisions on whether to consider the programme and the
degree of actual enrolment.
2. Conceptual Model
2.1 Utility function
Following Dupraz et al. (2003), farmers are assumed to maximise utility that is
based upon consumption of market goods (Z) and non-market environmental ser-
vices (E), which are co-produced by farming activities. They face a budget con-
straint that the cost of consumption cannot exceed the sum of profit from farm
production and non-farm income (NFI). Farm profit (p) is earned from selling
agricultural products (Y) at price ry minus variable cost (rxX) and fixed cost (FC).
Output Y is a function of inputs X and FC. The variable cost refers to material
and hired labour associated with the level of production, while fixed cost in this
study refers to predetermined resources, including family labour (L), capital (K),
land area (A), biophysical conditions (B) and information (I) available to farmers.
Environmental services (E), which are produced jointly with market goods (Y)
using variable and fixed inputs, may also affect the magnitude and timing of vari-
able input (X) employment (Zhang et al., 2007). F represents farmer characteristics
that condition the production function and hence condition the effects of PES
offers:
accept (WTA) payment, which is the minimum payment that the farm household
would require to provide specified environmental services under the programme.
Assuming the supply of environmental services increases from E0 to E1, WTA
can be represented as the change in farm household expenditure (e) as a result of
the change in the level of environmental services produced, holding utility constant
at the initial level U0 [Equation (6)]:
Based on Equation (5), the influence of the PES programme payment P on the
change of farmer’s utility can be represented by Equation (7). Under common
assumptions that farmers prefer more payment to less but have a decreasing mar-
ginal rate of substitution between payment and other goods, the change of utility is
an increasing and concave function of payment,
Farmer WTA is a payment level that would leave utility unchanged equal to zero,
as shown in Equation (8):
Table 1
Four cropping systems offered to farmers
Cropping system
Practice Description A B C D
Table 2
Environmental services and outcomes from proposed farming practices in cropping systems
quality due to less nitrogen leaching (Borin et al., 1997; Poudel et al., 2001).
Reduced pesticide rates would also mitigate on-site and off-site air pollution and
possible human health risks (Glotfelty et al., 1987; van den Berg et al., 1999).
A main effect orthogonal design framework was constructed for the 16 question-
naire versions. The versions varied by payment levels offered (4), payment provider
(federal government or non-governmental organisation) and sequence of cropping
practices (increasing or decreasing in complexity and expected environmental bene-
fits). For each cropping system, respondents were first offered a specific payment for
adopting the given system for a period of 5 years, and they were asked how many
acres they would enrol in such a programme. Respondents who chose not to enrol
any land were asked whether they would consider enroling in that system if the pay-
ment were higher. Thus, the ‘consider’ group and ‘not consider’ group are distin-
guished based upon this question and assuming that all farmers who chose to enrol
would also consider the programme with a higher payment. Within the ‘consider’
group, ‘enrol’ sub-group and ‘consider but not enrol’ sub-group are further identified
by the first acreage enrolment question. Note that, unlike the conceptual model, the
willingness-to-enrol question was presented ahead of the willingness-to-consider ques-
tion. This approach mimics existing agri-environmental programmes that offer a real
payment, avoiding an abstract question about hypothetical willingness to consider the
programme without a payment offer. The subsequent question on considering enrol-
ment with a higher payment follows naturally from a response declining to enrol.
Figure 1 shows the number of farms falling into each response group in our dataset.
1,400
200
0
System A System B System C System D
Figure 1. Number of farms that fell into the three participation categories for each of the
four cropping systems (N = 1,688 Michigan corn and soybean farms, year = 2008)
Besides farmer choices associated with each cropping system, other survey informa-
tion includes current crop management practices, farmers’ perception of benefits from
changed practices, attitudes on the importance of enhanced environmental services, past
experience with beneficial farming practices, participation in our four hypothetical crop-
ping systems containing those practices, and demographic background. Detailed infor-
mation about data collection and questionnaire design can be found in Jolejole (2009).
y ¼ c a: ð9Þ
Probit estimation is used for the binary choices of consideration. The latent vari-
able c* indicates the farmer’s utility gain per unit land area by considering enroling
in the programme with a suitable payment:
c ¼ x1 c þ e e Nð0; re Þ ð10Þ
1 c >0
c¼ ð11Þ
0 c 0:
The land area (acreage) enrolment variable a is indicated by a latent variable a*
and cornered at zero [Equation (12)]. The independent variable vectors x1 and x2,
weighted by parameter vectors c and b influence the farmer’s utility [Equation (10)]
and area chosen for enrolment [Equation (13)]. The random terms e and u are
assumed to be independent with zero means and constant variances r2. The positive
enroled acreage can be observed only if the farmer considers the programme and
chooses to participate given a specific payment [Equation (14)]:
a ¼ max½a ; 0 ð12Þ
a ¼ x2 b þ u>0 u N 0; r2u ð13Þ
a if c ¼ 1; a>0
y¼ ð14Þ
0 otherwise:
The associated log-likelihood function used for Maximum Likelihood estimation is:
li ðc; bÞ ¼ 1½yi ¼ 0 log½1 Uðx2i b=ru ÞUðx1i c=re Þ
ð16Þ
þ 1½yi >0flog½Uðx1i c=re Þþ logfu½ððyi x2i bÞ=ru Þ=ru gg:
Estimates of the two models are adjusted for stratum-specific sampling probability.
4.3. Variables
There is one probit regression for consideration and one tobit regression for acreage
enrolment in the double hurdle model. The dependent variable for the consideration
model is farmers’ dichotomous choice of considering enrolment in the programme,
which is contingent on the belief that the payment is politically feasible. The depen-
dent variable for the acreage enrolment model is the number of acres that farmers
would enrol in the programme, including both zero and positive acreages. See
Table 3 for descriptive statistics on the dependent variables.
The same set of independent variables is employed in both probit and tobit regres-
sions for comparison. Six broad categories of explanatory variables linked to the
Table 3
Summary statistics for dependent variables
conceptual model are defined as follows (Table 4): First, the design attributes cate-
gory, corresponding to the attributes of the programmes that were part of our experi-
mental design. These include the per-acre programme payment, P*, for each
cropping system, the sequence in which the four cropping systems were presented to
respondents, and whether the payment was provided by government. With the adop-
tion of different cropping practices, farmers are assumed to incur additional direct
costs (e.g. for labour and ⁄ or material inputs) and opportunity costs (e.g. for growing
a less profitable crop). Following exploratory results from farmer focus group inter-
views in 2007, the payment offer ranges for the four cropping systems were: A: $4 to
$17; B: $10 to $36; C: $15 to $55; and D: $20 to $75. The payment offer is hypothes-
ised to have more effect on acreage enrolment than on the willing-to-consider deci-
sion, for which a larger but politically feasible payment is implied. The descending
sequence dummy variable denotes respondents who received questionnaires with the
sequence of cropping systems decreasing in stringency and associated payment offers.
Second, the perception and attitudes category of variables corresponds to environ-
mental services E. These variables depict farmer perceptions of ES benefits from
certain cropping systems, and their attitudes on whether nature provides services
that could benefit their crop production. These variables are measured on a 5-point
Likert scale (1 for strongly disagree, 2 for disagree, 3 for neutral, 4 for agree, and 5
for strongly agree).
The third category describes the biophysical attributes of farms corresponding to
biophysical conditions B, which includes farm size and soil types. Larger farms are
expected to be more likely to enrol in a PES programme because they have a higher
capacity to invest and to withstand risks from changed practices (Knowler and Brad-
shaw, 2007; Prokopy et al., 2008). Soil type refers to dummy variables for soil texture.
Clay soils may be more fertile but less well-drained than the loam soil baseline,
whereas sandy soils are less fertile but better drained. Soil attributes exhibited mixed
effects in different studies depending on the specific practices. In this study, enrolment
in the reduced chemical system is expected to be positively related to clay soil, which
tends to be more fertile than sandy soil and silty soil. Cropping systems with soil
conservation practices, such as cover crops and corn-soybean-wheat rotation are
expected to be positively related to sandy soil, which is more erodible.
The fourth category measures farm management attributes. Different management
practices would require different combinations of variable inputs X, labour L and
Table 4
Summary statistics of independent variables
Questionnaire version
Government Dummy 1,796 0.497 0.500 0 1
Descending sequence Dummy 1,796 0.503 0.500 0 1
Payment offer (System A) Dollars 1,796 10.2 4.78 4 17
Payment offer (System B) Dollars 1,796 23.2 9.05 10 36
Payment offer (System C) Dollars 1,796 36.7 12.5 15 55
Payment offer (System D) Dollars 1,796 51.0 16.7 20 75
Perception and attitudes
Perceived env. performance (A) Likert 1–5 1,245 2.97 0.808 1 5
Perceived env. performance (B) Likert 1–5 1,189 3.24 0.776 1 5
Perceived env. performance (C) Likert 1–5 1,200 3.37 0.784 1 5
Perceived env. performance (D) Likert 1–5 1,245 3.46 0.794 1 5
General attitudes of ES Likert 1–5 1,475 3.12 1.11 1 5
Farm biophysical attributes
Total land Acres 1,521 1,151 1,408 2 21,500
Sandy soil Dummy 1,796 0.274 0.446 0 1
Silty soil Dummy 1,796 0.028 0.165 0 1
Clay soil Dummy 1,796 0.434 0.496 0 1
Farm management attributes
Moldboard tillage land percent Ratio 1,486 0.067 0.178 0 1
No till tillage land percent Ratio 1,486 0.185 0.244 0 1
Conservation land percent Ratio 1,486 0.342 0.286 0 1
Wheat land percent Ratio 1,486 0.083 0.103 0 0.714
Cover crops land percent Ratio 1,489 0.047 0.149 0 1
PSNT land percent Ratio 1,489 0.050 0.170 0 1
Organic land percent Ratio 1,477 0.002 0.025 0 0.63
Irrigation land percent Ratio 1,796 0.048 0.165 0 1
Reduced fertiliser use Dummy 1,444 0.218 0.413 0 1
Reduced pesticide use Dummy 1,442 0.209 0.407 0 1
MAEAP Dummy 1,371 0.142 0.349 0 1
EQIP Dummy 1,379 0.298 0.458 0 1
CRP Dummy 1,421 0.349 0.477 0 1
CSP Dummy 1,324 0.120 0.325 0 1
Operator attributes
Age Years 1,501 54.8 11.6 21 94
Education Years 1,488 13.4 2.56 6 20
Main income source from farm Dummy 1,488 0.718 0.450 0 1
Market prices
Wheat ⁄ corn price Ratio 1,059 1.682 0.323 0.15 3.33
Wheat ⁄ soybean price Ratio 1,049 0.732 0.358 0.2 7
capital K. Farmer’s current practices of tillage, wheat acreage, cover crops, organic
crops, fertiliser and pesticides4 are expected to affect enrolment positively if they are
similar to the new cropping system. The influence of irrigation on the adoption of
4
Reduced fertiliser and pesticides are dummy variables with one indicating currently band
apply fertiliser ⁄ pesticide to 2 ⁄ 3 of full field rate.
5. Results
Although the willingness-to-consider and acreage enrolment decisions are modelled
using the same set of variables, results of the double hurdle model suggest that the
first-round willingness-to-consider decision depends chiefly on non-price farm and
farmer characteristics. By contrast, the second-round enrolment decision depends
more on payment-driven benefit–cost criteria.
Both consideration and enrolment decisions are influenced by two common fac-
tors. First, the perceived environmental performance of each system significantly
contributes to both enrolment and willingness to consider each of the four cropping
systems. This finding is consistent with previous studies (Gould et al., 1989; Purvis
et al., 1989; Traore et al., 1998; Sidibe, 2005; D’Emden et al., 2008; Wei et al.,
2009), as the perceived ES benefits both individual farmers and the society. A new
finding from this study is that the marginal effects of perceived environmental per-
formance increase from System A to D for the willingness-to-consider decision, but
they decrease for the enrolment decision. This suggests that farmers would be more
willing to consider systems with higher requirements and larger environmental bene-
fits given the same unit increase in environmental performance, but would be less
willing to enrol more land in those systems, perhaps due to higher costs associated
with realising these benefits. The second common factor for two decisions is the
sequence of presenting cropping systems to farmers. Farmers who were presented
with the higher-complexity and higher-payment cropping system (System D) first
were less likely to either consider or enrol in the other three cropping systems. This
unexpected survey design effect may reflect farmers being less willing to make small
improvements in their cropping systems or maintain low environmental perfor-
mance practices if a larger return with substantial improvement is an option.
Table 5
Marginal effects from probit estimation of farmers’ consideration, weighted by stratum, by
cropping systems, 1,688 Michigan corn or soybean farmers, 2008
Descending sequence )0.18*** 0.00 )0.13* 0.05 )0.042 0.45 0.072 0.19
Payment offer 0.016** 0.02 )0.0023 0.43 0.00060 0.76 )0.0012 0.47
Perceived env. perf. 0.082** 0.04 0.13*** 0.00 0.18*** 0.00 0.19*** 0.00
General ES attitudes 0.069** 0.01 0.046* 0.10 0.051** 0.04 0.038* 0.08
No till tillage 0.37** 0.04 0.38** 0.03 0.41** 0.02 0.25 0.11
Wheat ratio 0.18 0.49 0.25 0.31 0.44* 0.07 0.41* 0.07
Reduced fertiliser 0.10 0.15 0.15** 0.01 0.12** 0.03 0.11** 0.04
Reduced pesticide 0.082 0.24 0.13** 0.04 0.013 0.84 0.075 0.22
MAEAP )0.24*** 0.01 )0.19** 0.03 )0.14 0.10 )0.19** 0.02
EQIP 0.088 0.17 0.067 0.27 0.11* 0.07 0.14** 0.01
CRP 0.0024 0.97 0.11 0.15 0.052 0.53 0.15** 0.02
CSP )0.12 0.24 )0.18* 0.07 )0.088 0.38 )0.16* 0.06
Education 0.024* 0.07 0.011 0.45 0.033*** 0.00 0.027** 0.01
Wheat ⁄ corn price 0.16 0.28 0.25** 0.04 0.047 0.69 0.12 0.27
Intercept )3.0** 0.02 )2.6* 0.06 )4.5*** 0.00 )4.0*** 0.01
Number of obs. 600 594 604 613
Wald v2(26) 78 66 71 92
Prob. > v2 0 0 0 0
Pseudo R2 0.23 0.22 0.24 0.27
Log likelihood )321 )319 )306 )304
by a single hurdle model or a simple tobit model that lacks information on what
farmers would respond to a higher payment offer. Three categories of variables
drive the consideration decision (Table 5).
First, farmers who believe their production can benefit from nature are 5% more
likely to consider enroling in the programme. Previous studies also found that posi-
tive attitudes tended to promote enrolment in conservation programmes (Lynne
et al., 1988; Sheikh et al., 2003), but this study explicitly identifies the significant
impact of production-supporting ecosystem services on farmer decisions. Public
extension and outreach programmes have great potential to improve farmers’ gen-
eral understanding of ecosystem benefits, which would further enhance participation
when combined with perceived positive environmental performance from specific
systems.
Second, the similarity of current farm management practices to the proposed
cropping system also increases willingness to consider the PES programme, reflect-
ing lower perceived risks and less additional cost. Prior practice of conservation till-
age, wheat planting, and reduced fertiliser input are illustrative examples. Farmers
with an additional 10% of land under no-till are 3% more likely to consider adopt-
ing the programme that requires conservation tillage. Farmers with an additional
10% more of their land planted to wheat are 4% more likely to adopt the two
cropping systems that include wheat in the crop rotation. Those who currently own
equipment to band apply fertiliser at a reduced rate are 10% more likely to consider
System D, which requires this.
Third, information variables such as education and past experience with a gov-
ernmental PES programme generally promote willingness-to-consider. One more
year of education increases the probability of considering the PES programme by
about 3%. Prior research has also found positive effects of education on adoption
of farming practices as education largely links to knowledge (Ervin and Ervin,
1982; Rahm and Huffman, 1984; Warriner and Moul, 1992; Bosch et al., 1995; Wu
and Babcock, 1998). It was widely shown that farmers who are currently or were
previously involved in conservation programmes are more likely to participate in a
new programme (Ervin and Ervin, 1982; Bosch et al., 1995; Wu and Babcock, 1998;
Defrancesco et al., 2008; Wei et al., 2009; Bertoni et al., 2011). We find that the
influence of programme experience depends on the objective and restriction of those
programmes. Past programme experience with the Environmental Quality Incentives
Programme (EQIP), which is a flexible governmental PES-type programme, facili-
tates consideration of systems C and D by an additional 10%. However, experience
with the Michigan Agriculture Environmental Assurance Programme (MAEAP)
reduces the probability of considering enrolment by approximately 20%. This may
be due to differences in programme goals between MAEAP and our hypothetical
programme. MAEAP does not involve adoption of changed practices to benefit the
environment; instead, it is designed to reduce farmers’ legal and environmental risks
by education, farm-specific risk assessment and on-site verification of ‘generally
accepted agricultural practices’. A previous study on early-adopting livestock pro-
ducers found the primary motivation to participate in MAEAP was concern about
existing or potential environmental regulations, rather than a predisposition toward
environmental stewardship (Vollmer-Sanders et al., 2011). The Conservation Stew-
ardship Programme (CSP) also has a slightly negative impact on considering two of
our four systems, presumably because lands or conservation practices paid by other
programmes are not eligible for CSP.
Table 6
Marginal effects from tobit estimation of farmers’ acreage enrolment, weighted by stratum,
by cropping systems, 1,688 Michigan corn or soybean farmers, 2008
Descending sequence )360*** 0.00 )57 0.36 )160*** 0.00 )65 0.14
Payment offer 19** 0.02 11*** 0.01 6.6*** 0.00 3.6** 0.01
Perceived env. perf. 190*** 0.00 150*** 0.01 120*** 0.00 140*** 0.00
General ES attitudes )8.2 0.82 )20 0.54 )36** 0.04 )39** 0.04
Total land 0.26*** 0.00 0.28*** 0.00 0.24*** 0.00 0.33*** 0.00
Sandy soil 240* 0.07 24 0.82 17 0.80 75 0.32
Clay soil 390*** 0.00 34 0.73 23 0.71 85 0.20
Moldboard tillage )880*** 0.00 )760** 0.02 )190 0.16 )110 0.43
No till tillage )120 0.62 240 0.14 88 0.45 310*** 0.01
Conservation tillage 120 0.54 120 0.30 14 0.87 150** 0.05
Wheat ratio )460 0.11 310* 0.10 210 0.19 )4.2 0.98
Irrigation ratio 620** 0.01 570** 0.04 )73 0.66 52 0.73
Reduced pesticide )65 0.53 )13 0.19 )120** 0.03 )130** 0.02
EQIP 60 0.54 45 0.58 15 0.80 120** 0.04
CSP 370*** 0.00 170 0.15 100 0.21 )45 0.63
Age 3.6 0.24 4.6 0.22 1.3 0.41 4.5*** 0.01
Wheat ⁄ corn price 440 0.11 )31 0.85 )140 0.18 )360*** 0.00
Wheat ⁄ soybean price )460 0.42 51 0.88 140 0.52 710*** 0.00
Farm income 82 0.38 )55 0.57 95* 0.06 9.4 0.87
Intercept )2,100*** 0.00 )1,300** 0.01 )510* 0.06 )880*** 0.00
Number of obs. 364 372 430 406
Wald v2(26) 2.91 1.56 4.96 5.66
Prob. > v2 0.00 0.04 0.00 0.00
Pseudo R2 0.06 0.05 0.04 0.05
Log likelihood )733 )897 )1,302 )1,318
contrast, Defrancesco et al. (2008) found that high dependence on farm income con-
strained farms from participating in a northern Italian agri-environmental pro-
gramme because of the potential risk of uncompensated yield reductions. Older
farmers seem to be willing to enrol more acreage if willing to consider enrolment,
though the consideration probit model indicates that they are less likely to consider
enrolment in the programme. The age variable has shown both positive (Warriner
and Moul, 1992; Okoye, 1998) and negative effects in previous studies (Gould et al.,
1989; Neill and Lee, 2001; Damianos and Giannakopoulos, 2002; Lambert et al.,
2006; Giovanopoulou et al., 2011).
6. Conclusion
This study deepens our understanding of farmers’ willingness-to-participate in pay-
ment-for-environmental-services programmes by separating the initial decision on
whether even to consider the programme from the final decision on how many acres
to enrol at a given payment level. Comparison of different econometric models for
estimating and predicting PES enrolment leads to selection of a double hurdle
model, comprised of a probit for willingness to consider and a tobit for acreage
enrolment. The first-stage willingness-to-consider decision is motivated by farm and
farmer characteristics that have non-marginal impacts on land enrolment, such as
environmental attitudes, experience in conservation programmes, education, and
ownership of large equipment. The second-stage land enrolment decision depends
more on payment-driven benefit–cost factors that especially affect the net marginal
benefit of enrolment, such as the per-acre payment offer, total cropland area, irri-
gated land proportion, use of moldboard tillage and whether most income comes
from farming. The results also reveal the price elasticity effect of enrolment. As
expected, the least complex system – a conventional conservation-till corn-soybean
rotation – is most payment responsive. This system also has a large potential
for state-wide adoption due to its suitability to small farms, which are numerous.
The two stages are underpinned by two common factors: perceived environmental
performance of the proposed systems and, importantly, the sequence in which the
systems are presented to respondents.
Understanding farmers’ decision processes is an essential precondition for design-
ing effective and efficient agricultural PES programmes. The two-stage model used
here distinguishes the willingness-to-consider decision from the subsequent enrol-
ment decision, providing fresh insights into how tiered policies can facilitate partici-
pation in PES programmes. For farmers willing to consider PES, the second-stage
land enrolment model reveals factors deterring enrolment that can be overcome
with higher payments. These factors include small land area, high proportion of irri-
gation land, low farm income, young decision maker and high proportion of inten-
sive-tilled land. A small boost in compensation to farms with these traits could
increase the total acreage enrolment from participants and some non-participants
who will consider enroling in PES.
On the other hand, the first-stage willingness-to-consider model identifies farms
with substantial obstacles to participation that are not likely to be resolved by a
higher payment. Three strategies could be employed to raise participation rates at
successively higher cost. First, at relatively low screening cost, PES programmes can
enhance participation by targeting more educated, experienced and properly
equipped farmers who are favourably disposed toward environmental stewardship.
Second, research and outreach that build farmer understanding of environmental ser-
vices from agriculture and environmental improvements from specific conservation
practices could also contribute to the appeal of agricultural PES programmes. As a
common determinant for both the consideration and enrolment decision processes,
farmers’ perception of environmental performance from specific practices would also
contribute to acreage enrolment. Third, substantial changes in programme design,
such as subsidising certain farm equipment purchases and relaxing the eligibility of
applicants and lands that are enroled in other programmes, could further engage
farmers who are not likely to consider the programme. This last strategy can be use-
ful to target enrolment by farms with high environmental impact in sensitive areas.
In addition, high-stringency, high-performance programmes will be best received if
offered directly to farmers without alternative low-complexity, low-performance
options. While the farmers surveyed demanded higher payments for more complex
practices, they were more open to complex options when those were not preceded by
simpler ones.
Our analysis of hypothetical PES programmes focuses on total environmental ser-
vices generated, rather than additional ones. This approach has the advantage of
treating equitably both prior providers of environmental services and new ones who
join under the programme. However, from a cost-effectiveness perspective, tax
authorities prefer not to pay for environmental services that would be provided for
free. A reasonable solution would be a split-payment scheme that offers a higher
rate for improved stewardship practice and a lower (but positive) rate for existing
practices. Future research could measure the cost for different payment schemes
between paying for all environmental services and paying only for additional envi-
ronmental services generated by farms enroling in new practices. By combining such
supply-side information with estimates of demand for environmental service
improvements, it should be possible to assess the potential for a PES market in
agriculturally generated environmental services.
As a closing caveat, we note that the supply of environmental services from PES
programmes depends not only on farmer enrolment (the focus of this article), but
also on other factors post-enrolment. Recent articles have highlighted how strategic
behaviour in compliance (Bartolini et al., 2012; Fraser, 2012) and environmental
uncertainties can affect the actual delivery of improved environmental services from
agriculture (Glenk and Colombo, 2011).
References
Bartolini, F., Gallerani, V., Raggi, M. and Viaggi, D. ‘Modelling the linkages between cross-
compliance and agri-environmental schemes under asymmetric information’, Journal of
Agricultural Economics, Vol. 63, (2012) pp. 310–330.
van den Berg, F., Kubiak, R., Benjey, W. G., Majewski, M. S., Yates, S. R., Reeves, G. L.,
Smelt, J. H. and van der Linden, A. M. A. ‘Emission of pesticides into the air’, Water,
Air, & Soil Pollution, Vol. 115, (1999) pp. 195–218.
Bertoni, D., Cavicchioli, D., Pretolani, R. and Olper, A. ‘Agri-environmental measures adop-
tion: New evidence from Lombardy region’, A. Sorrentino, S. Severini and R. Henke (ed.),
The Common Agricultural Policy After the Fischler Reform: National Implementations,
Impact Assessment and the Agenda for Future Reforms (Surrey, UK: Ashgate Publishing,
Ltd, 2011, pp. 275–294).
Blundell, R. and Meghir, C. ‘Bivariate alternatives to the tobit model’, Journal of Economet-
rics, Vol, 34, (1987) pp. 179–200.
Borin, M., Giupponi, C. and Morari, F. ‘Effects of four cultivation systems for maize on
nitrogen leaching 1. Field experiment’, European Journal of Agronomy, Vol. 6, (1997)
pp. 101–112.
Bosch, D. J., Cook, Z. L. and Fuglie, K. O. ‘Voluntary versus mandatory agricultural poli-
cies to protect water quality: Adoption of nitrogen testing in Nebraska’, Review of Agricul-
tural Economics, Vol, 17, (1995) pp. 13–24.
Burton, M., Dorsett, R. and Young, T. ‘Changing preferences for meat: Evidence from UK
household data, 1973–93’, European Review of Agricultural Economics, Vol. 23, (1996)
pp. 357–370.
Cattaneo, A., Claassen, R., Johansson, R. and Weinberg, M. ‘Flexible conservation mea-
sures on working land: What challenges lie ahead?’ (United States Department of Agricul-
ture, Economic Research Service, Washington, DC, 2005). Available at: http://
www.ers.usda.gov/publications/err-economic-research-report/err5.aspx (last accessed July
18 2012).
Correll, D. ‘Role of phosphorus in the eutrophication of receiving waters: a review’, Journal
of Environmental Quality, Vol. 27, (1998) pp. 261–266.
Cragg, J. G. ‘Some statistical models for limited dependent variables with application to the
demand for durable goods’, Econometrica, Vol. 39, (1971) pp. 829–844.
Damianos, D. and Giannakopoulos, N. ‘Farmers’ participation in agri-environmental
schemes in Greece’, British Food Journal, Vol. 104, (2002) pp. 261–273.
Davey, K. A. and Furtan, W. H. ‘Factors that affect the adoption decision of conservation
tillage in the prairie region of Canada’, Canadian Journal of Agricultural Economics,
Vol. 56, (2008) pp. 257–275.
Defrancesco, E., Gatto, P., Runge, F. and Trestini, S. ‘Factors affecting farmers’ participa-
tion in agri-environmental measures: a northern Italian perspective’, Journal of Agricultural
Economics, Vol. 59, (2008) pp. 114–131.
Delgado, J., Sparks, R., Follett, R., Sharkoff, J. and Riggenbach, R. ‘Conserve soil and
water quality in the San Luis Valley of south central Colorado’, R. Lal (ed.), Soil Quality
and Soil Erosion (Boca Raton, FL, CRC Press, 1999, pp. 125).
D’Emden, F. H., Llewellyn, R. S. and Burton, M. P. ‘Factors influencing adoption of conser-
vation tillage in Australian cropping regions’, Australian Journal of Agricultural and
Resource Economics, Vol. 52, (2008) pp. 169–182.
Dillman, D. A. Mail and Internet Surveys: The Tailored Design Method (Hoboken, NJ: John
Wiley & Sons, 2007).
Dobbs, T. L. and Pretty, J. ‘Case study of agri-environmental payments: the United King-
dom’, Ecological Economics, Vol. 65, (2008) pp. 765–775.
Dupraz, P., Vermersch, D., De Frahan, B. H. and Delvaux, L. ‘The environmental supply of
farm households: a flexible willingness to accept model’, Environmental and Resource Eco-
nomics, Vol. 25, (2003) pp. 171–189.
Engel, S., Pagiola, S. and Wunder, S. ‘Designing payments for environmental services in theory
and practice: an overview of the issues’, Ecological Economics, Vol. 65, (2008) pp. 663–674.
Epplin, F. M. and Tice, T. F. ‘Influence of crop and farm size on adoption of conservation
tillage’, Journal of Soil and Water Conservation, Vol. 41, (1986) pp. 424–427.
Ervin, C. A. and Ervin, D. E. ‘Factors affecting the use of soil conservation practices: Hypothe-
ses, evidence, and policy implications’, Land Economics, Vol. 58, (1982) pp. 277–292.
European Commission Directorate-General for Agriculture and Rural Development (EC).
The common agricultural policy explained. (Brussels, Belgium: Eugène Leguen de Lacroix,
2004).
Fraser, R. ‘Moral hazard, targeting and contract duration in agri-environmental policy’,
Journal of Agricultural Economics, Vol. 63, (2012) pp. 56–64.
Giovanopoulou, E., Nastis, S. A. and Papanagiotou, E. ‘Modeling farmer participation in
agri-environmental nitrate pollution reducing schemes’, Ecological Economics, Vol. 70,
(2011) pp. 2175–2180.
Glenk, K. and Colombo, S. ‘How sure can you be? A framework for considering delivery
uncertainty in benefit assessments based on stated preference methods’, Journal of Agricul-
tural Economics, Vol. 62, (2011) pp. 25–46.
Glotfelty, D. E., Seiber, J. N. and Liljedahl, A. ‘Pesticides in fog’, Nature, Vol. 325, (1987)
pp. 602–605.
Goodwin, B., Offenbach, L., Cable, T. and Cook, P. ‘Discrete ⁄ continuous contingent valua-
tion of private hunting access in Kansas’, Journal of Environmental Management, Vol. 39,
(1993) pp. 1–12.
Gould, B. W., Saupe, W. E. and Klemme, R. M. ‘Conservation tillage: The role of farm and
operator characteristics and the perception of soil-erosion’, Land Economics, Vol. 65,
(1989) pp. 167–182.
Hoben, J. P., Gehl, R. J., Millar, N., Grace, P. R. and Robertson, G. P. ‘Nonlinear nitrous
oxide (N2O) response to nitrogen fertilizer in on-farm corn crops of the US Midwest’, Glo-
bal Change Biology, Vol. 17, (2011) pp. 1140–1152.
Hynes, S. and Garvey, E. ‘Modelling farmers’ participation in an agri-environmental scheme
using panel data: an application to the rural environment protection scheme in Iireland’,
Journal of Agricultural Economics, Vol. 60, (2009) pp. 546–562.
Jensen, H. H. and Yen, S. T. ‘Food expenditures away from home by type of meal’, Cana-
dian Journal of Agricultural Economics, Vol. 44, (1996) pp. 67–80.
Jolejole, M. C. B. 2009. ‘Trade-offs, incentives and the supply of ecosystem services from
cropland’, Michigan State University, Agricultural, Food and Resource Economics.
Available at: http://aec.msu.edu/theses/fulltext/jolejole1_ms.pdf (last accessed June 6,
2011).
Joyce, B., Wallender, W., Mitchell, J., Huyck, L., Temple, S., Brostrom, P. and Hsiao, T.
‘Infiltration and soil water storage under winter cover cropping in California’s Sacramento
Valley’, Transactions of the ASAE, Vol. 45, (2002) pp. 315–326.
Kleijn, D. and Sutherland, W. J. ‘How effective are European agri-environment schemes in con-
serving and promoting biodiversity?’Journal of applied ecology, Vol. 40, (2003) pp. 947–969.
Knowler, D. and Bradshaw, B. ‘Farmers’ adoption of conservation agriculture: a review and
synthesis of recent research’, Food Policy, Vol. 32, (2007) pp. 25–48.
Lal, R., Griffin, M., Apt, J., Lave, L. and Morgan, M. ‘Managing soil carbon’, Science,
Vol. 304, (2004) pp. 393.
Lambert, D., Sullivan, P., Claassen, R. and Foreman, L. Conservation-compatible practices
and programmes: Who participates? Economic Research Report No. 14 (Washington, DC:
Economic Research Service, US Department of Agriculture, 2006). Available at: http://
webarchives.cdlib.org/sw15d8pg7m/http:/ers.usda.gov/publications/err14/ (last accessed July
18, 2012).
Larsén, K. 2006. ‘Farmers’ participation in agri-environmental programmes and impact on
farm performance: An empirical analysis applied to Swedish agriculture’ Paper presented
at 8th Annual Bioecon Conference on ‘Economic Analysis of Ecology and Biodiversity.
Cambridge, UK, 29-30 August. Available at: http://bioecon-network.org/pages/8th_2006/
Lars%E9n.pdf (last accessed Feb 19, 2012).
Lasley, P., Duffy, M., Kettner, K. and Chase, C. ‘Factors affecting farmers’ use of practices
to reduce commercial fertilizers and pesticides’, Journal of Soil and Water Conservation,
Vol. 45, (1990) pp. 132–136.
Lee, L. K. and Stewart, W. H. ‘Landownership and the adoption of minimum tillage’, Ameri-
can Journal of Agricultural Economics, Vol. 65, (1983) pp. 256–264.
Lynne, G. D. and Rola, L. R. ‘Improving attitude-behavior prediction models with economic
variables: Farmer actions toward soil conservation’, Journal of Social Psychology, Vol.
128, (1988) pp. 19–28.
Lynne, G. D., Shonkwiler, J. S. and Rola, L. R. ‘Attitudes and farmer conservation behav-
ior’, American Journal of Agricultural Economics, Vol. 70, (1988) pp. 12–19.
Ma, S. ‘Supply and demand for ecosystem services from cropland in Michigan’, Dissertation,
(East Lansing, Michigan, MI: Michigan State University, Department of Agricultural,
Food and Resource Economics, 2011). Available at: http://aec.msu.edu/theses/fulltext/
ma_phd.pdf (last accessed Feb 19, 2012).
Mathijs, E. ‘Social capital and farmers’ willingness to adopt countryside stewardship
schemes’, Outlook on Agriculture, Vol. 32, (2003) pp. 13–16.
Mazvimavi, K. and Twomlow, S. ‘Socioeconomic and institutional factors influencing adop-
tion of conservation farming by vulnerable households in Zimbabwe’, Agricultural Systems,
Vol. 101, (2009) pp. 20–29.
McSwiney, C. P. and Robertson, G. P. ‘Nonlinear response of N2O flux to incremental fertil-
izer addition in a continuous maize (Zea mays L.) cropping system’, Global Change Biol-
ogy, Vol. 11, (2005) pp. 1712–1719.
McSwiney, C. P., Snapp, S. S. and Gentry, L. E. ‘Use of N immobilization to tighten the N
cycle in conventional agroecosystems’, Ecological Applications, Vol. 20, (2010) pp. 648–662.
Musser, W. N., Shortle, J. S., Kreahling, K., Roach, B., Huang, W.-C., Beegle, D. B. and
Fox, R. H. ‘An economic analysis of the pre-sidedress nitrogen test for Pennsylvania corn
production’, Review of Agricultural Economics, 17, (1995) pp. 25–35.
Negatu, W. and Parikh, A. ‘The impact of perception and other factors on the adoption of
agricultural technology in the Moret and Jiru Woreda (district) of Ethiopia’, Agricultural
Economics, Vol. 21, (1999) pp. 205–216.
Neill, S. P. and Lee, D. R. ‘Explaining the adoption and disadoption of sustainable agricul-
ture: The case of cover crops in Northern Honduras’, Economic Development and Cultural
Change, Vol. 49, (2001) pp. 793–820.
Newman, C., Henchion, M. and Matthews, A. ‘A double-hurdle model of Irish household
expenditure on prepared meals’, Applied Economics, Vol. 35, (2003) pp. 1053–1061.
Norris, P. and Batie, S. ‘Virginia farmers’ soil conservation decisions: an application of tobit
analysis’, Southern Journal of Agricultural Economics, Vol. 19, (1987) pp. 79–90.
Nowak, P. ‘Why farmers adopt production technology’, Journal of Soil and Water Conserva-
tion, Vol. 47, (1992) pp. 14–16.
Oades, J. ‘Soil organic matter and structural stability: Mechanisms and implications for man-
agement.’, Plant and Soil, Vol. 76, (1984) pp. 319–337.
Okoye, C. U. ‘Comparative analysis of factors in the adoption of traditional and recommended soil
erosion control practices in Nigeria’, Soil and Tillage Research, Vol. 45, (1998) pp. 251–263.
Pautsch, G., Kurkalova, L., Babcock, B. and Kling, C. ‘The efficiency of sequestering carbon
in agricultural soils’, Contemporary Economic Policy, Vol. 19, (2001) pp. 123–134.
Peel, M. ‘Crop rotations for increased productivity’ (Fargo, ND: Bulletin. North Dakota
State University Extension Service, 1998). Available at: http://www.ag.ndsu.edu/pubs/
plantsci/crops/eb48-1.htm (last accessed June 6, 2011).
Perrot-Maitre, D. The Vittel Payments for Ecosystem Services: A ‘‘perfect’’ PES Case. (Lon-
don, UK: International Institute for Environment and Development, 2006).
Pimentel, D. and Kounang, N. ‘Ecology of soil erosion in ecosystems’, Ecosystems, Vol. 1,
(1998) pp. 416–426.
Poudel, D., Horwath, W., Mitchell, J. and Temple, S. ‘Impacts of cropping systems on soil
nitrogen storage and loss’, Agricultural Systems, Vol. 68, (2001) pp. 253–268.
Prokopy, L. S., Floress, K., Klotthor-Weinkauf, D. and Baumgart-Getz, A. ‘Determinants of
agricultural best management practice adoption: Evidence from the literature’, Journal of
Soil and Water Conservation, Vol. 63, (2008) pp. 300–311.
Purvis, A., Hoehn, J., Sorenson, V. and Pierce, F. ‘Farmers’ response to a filter strip pro-
gramme: Results from a contingent valuation survey’, Journal of Soil and Water conserva-
tion, Vol. 44, (1989) pp. 501.
Rahm, M. R. and Huffman, W. E. ‘The adoption of reduced tillage: the role of human capi-
tal and other variables’, American Journal of Agricultural Economics, Vol. 66, (1984)
pp. 405–413.
Reganold, J. P., Elliott, L. F. and Unger, Y. L. ‘Long-term effects of organic and conven-
tional farming on soil erosion’, Nature, Vol. 330, (1987) pp. 370–372.
Reicosky, D. and Lindstrom, M. ‘Fall tillage methods: Effect on short-term carbon dioxide
flux from soil’, Agronomy Journal, Vol. 85, (1993) pp. 1237–1243.
Roberts, R. K., English, B., Gao, Q. and Larson, J. A. ‘Simultaneous adoption of herbicide-
resistant and conservation-tillage cotton technologies’, Journal of Agricultural and Applied
Economics, Vol. 38, (2006) pp. 629.
del Saz-Salazar, S. and Rausell-Koster, P. ‘A double-hurdle model of urban green areas
valuation: Dealing with zero responses’, Landscape and Urban Planning, Vol. 84, (2008) pp.
241–251.
Sheikh, A. D., Rehman, T. and Yates, C. M. ‘Logit models for identifying the factors that
influence the uptake of new ‘no-tillage’ technologies by farmers in the rice-wheat and the
cotton-wheat farming systems of Pakistan’s Punjab’, Agricultural Systems, Vol. 75, (2003)
pp. 79–95.
Shrestha, R. A. M., Alavalapati, J., Seidl, A., Weber, K. and Suselo, T. R. I. ‘Estimating the
local cost of protecting Koshi Tappu wildlife reserve, Nepal: a contingent valuation
approach’, Environment, Development and Sustainability, Vol. 9, (2007) pp. 413–426.
Sidibe, A. ‘Farm-level adoption of soil and water conservation techniques in northern Burk-
ina Faso’, Agricultural Water Management, Vol. 71, (2005) pp. 211–224.
Soule, M., Tegene, A. and Wiebe, K. ‘Land tenure and the adoption of conservation prac-
tices’, American Journal of Agricultural Economics, Vol. 82, (2000) pp. 993–1005.
Su, S. B. and Yen, S. T. ‘Microeconometric models of infrequently purchased goods: an
application to household pork consumption’, Empirical Economics, Vol. 21, (1996) pp.
513–533.
Tobin, J. ‘Estimation of relationships for limited dependent variables’, Econometrica, Vol. 26,
(1958) pp. 24–36.
Traore, N., Landry, R. and Amara, N. ‘On-farm adoption of conservation practices: the role
of farm and farmer characteristics, perceptions, and health hazards’, Land Economics,
Vol. 74, (1998) pp. 114–127.
Upadhyay, B. M., Young, D. L., Wang, H. H. and Wandschneider, P. ‘How do farmers who
adopt multiple conservation practices differ from their neighbors?’American Journal of
Alternative Agriculture, Vol. 18, (2003) pp. 27–36.
Uri, N. ‘Conservation tillage and input use’, Environmental Geology, Vol. 29, (1997) pp. 188–201.
Vollmer-Sanders, C., Wolf, C. and Batie, S. S. ‘Financial and environmental consequences of
a voluntary farm environmental assurance programme in Michigan’, Journal of Soil and
Water Conservation, Vol. 66, (2011) pp. 122–131.
Warriner, G. K. and Moul, T. M. ‘Kinship and personal communication network influences
on the adoption of agriculture conservation technology’, Journal of Rural Studies, Vol. 8,
(1992) pp. 279–291.
Wauters, E., Bielders, C., Poesen, J., Govers, G. and Mathijs, E. ‘Adoption of soil conserva-
tion practices in Belgium: an examination of the theory of planned behaviour in the agri-
environmental domain’, Land Use Policy, Vol. 27, (2010) pp. 86–94.
Wei, Y. P., Chen, D., White, R. E., Willett, I. R., Edis, R. and Langford, J. ‘Farmers’ per-
ception of environmental degradation and their adoption of improved management prac-
tices in Alxa, China’, Land Degradation & Development, Vol. 20, (2009) pp. 336–346.
Wossink, A. and Swinton, S. ‘Jointness in production and farmers’ willingness to supply
non-marketed ecosystem services’, Ecological Economics, Vol. 64, (2007) pp. 297–304.
Wu, J. and Babcock, B. ‘The choice of tillage, rotation, and soil testing practices: Economic
and environmental implications’, American Journal of Agricultural Economics, Vol. 80,
(1998) pp. 494–511.
Wunder, S. ‘Payments for environmental services: Some nuts and bolts’, Occasional Paper,
No. 42 (Center for International Forestry Research, Bogor, Indonesia, 2005).
Wunder, S., Engel, S. and Pagiola, S. ‘Taking stock: a comparative analysis of payments for
environmental services programs in developed and developing countries’, Ecological Eco-
nomics, Vol. 65, (2008) pp. 834–852.
Yen, S. and Jensen, H. ‘Determinants of household expenditures on alcohol’, Journal of Con-
sumer Affairs, Vol. 30, (1996) pp. 48–67.
Yen, S. T. and Jones, A. M. ‘Household consumption of cheese: an inverse hyperbolic sine
double-hurdle model with dependent errors’, American Journal of Agricultural Economics,
Vol. 79, (1997) pp. 246–251.
Zbinden, S. and Lee, D. R. ‘Paying for environmental services: an analysis of participation
in Costa Rica’s PSA programme’, World Development, Vol. 33, (2005) pp. 255–272.
Zhang, W., Ricketts, T. H., Kremen, C., Carney, K. and Swinton, S. M. ‘Ecosystem services
and dis-services to agriculture’, Ecological Economics, Vol. 64, (2007) pp. 253–260.