Exploring the determinants for adopting water conservation measures.
What is the tendency of landowners when the resource is already at risk?
Ioanna Grammatikopoulou1, Eija Pouta1, Sami Myyrä1
1
MTT Agrifood Research Finland, Economic Research Unit, Latokartanonkaari 9, 00790
Helsinki, Finland
Paper prepared for presentation at the EAAE 2014 Congress
‘Agri-Food and Rural Innovations for Healthier Societies’
August 26 to 29, 2014
Ljubljana, Slovenia
Copyright 2014 by I.Grammatikopoulou, E.Pouta and S.Myyrä. All rights reserved. Readers
may make verbatim copies of this document for non-commercial purposes by any means,
provided that this copyright notice appears on all such copies.
1
Abstract
In Finland, water conservation policy sets equal incentive regardless of the condition of the
environment. Before any policy reform, it is vital to investigate landowners’ tendency for
adoption. In this study we were particularly interested in examining the tendency for adoption if
the soil quality implies a high leaching risk or water quality is already poor. For this purpose,
survey data were combined with GIS data. Furthermore the aim was to analyse the effect of farm
and farmer characteristics and attitudes on adoption. The findings derived by probit models
indicated that, for active farmers, financial variables were the key determinants. For passive
owners, adoption was also explained by attitudes. Contrary to our expectations, there is no
spontaneous tendency for water conservation in deteriorated areas. Targeted agri-environmental
measures, even though costly, cannot be avoided.
Keywords: water eutrophication, conservation, GIS data, probit model
2
1. Introduction
In Finland current water conservation policy sets equal incentive for water conservation
independently of eutrophication risk caused by the farming activity. This has been criticised for
being an inefficient policy, as greater conservation benefits could be attained from the most risky
areas (Lankoski and Ollikainen, 2003). As a solution for this inefficiency, targeted agricultural
measures have been proposed. However such measures, where different per-hectare payments
could be offered to farmers based on their ability to implement water conservation, (Iho et al.,
2013), have been suggested to cause considerable implementations costs. Before conducting
these expensive policy reforms, evidence is needed that the present water conservation measures
are implemented independently of conservation risks. The benefits of targeted policy would also
be lower if under the current policy there is a spontaneous tendency among land owners to direct
their water conservation actions to those areas where they are most needed. However, there is no
information on the behaviour of the landowners under the current policy. The crucial question is
whether landowners who operate in risky areas are more likely to voluntarily adopt water
conservation measures.
Water conservation in Finland provides mostly off-farm environmental benefits. Although,
soil erosion, nitrogen and phosphorus runoff have an effect on the water bodies next to farms,
they particularly affect water bodies between farms and the Baltic Sea as well as the condition of
the Baltic Sea. Previous studies on soil conservation measures have suggested that voluntary
adoption may be effective in the case of soil erosion. The marginal environmental benefit of
adopting conservation measures on farms with low soil erodibility is likely to be low (Lampert et
al., 2007) and adoption is thus expected to be greater on land prone to erosion (Pautsch et al.,
2001). However, the link between soil characteristics and adoption of conservation in the case
where the benefits of conservation are mostly public is unclear.
In Finland, the share of agricultural land under lease contracts has doubled within the last
15 years. In current form of the leasing practices, the responsibility of the water conservation is
shared between landowner, i.e passive owners who lease their land, and active farmers. With
long term land improvements, such as liming and drainage, or by separating water protection
zones from the field area, landowners can enhance the water conservation. However, this
happens at their own expenses since agri-environmental measures are directed to active farmers.
Past research has shown differences in conservation adoption between landowner types (Soule et
al., 2000). Against this background, it would be interesting to determine how active farmers and
passive owners participate in water conservation.
Awareness of the association between soil characteristics and conservation action is
important for policy design as it enables landowner-focused policies to be better targeted at those
owner groups that are open to policy measures (e.g., Maybery et al. 2005; Ross-Davis and
Broussard, 2007). The identification of those landowners who may be receptive to policy
interventions is a well-justified reason for the study. In addition to the design of policies,
understanding the factors that associate with interest in conservation is a necessity for the viable
implementation of policies and the tailoring of extension services ( Maybery et al. 2005; Kendra
and Hull 2005).
3
In this study our first aim was to analyse whether landowners’ adoption of special
environmental support1 measures is directed to those areas that are most critical in the water
conservation sense. Second, we aimed to define the farm characteristics as well as sociodemographic and attitude factors that affect adoption. Third, we also analysed whether active
farmers and passive landowners have different variables defining their water conservation
tendency. If so, this would imply the need for a different strategy of policy measures for
enhancing the adoption of conservation actions in these two groups.
We analyse the relation between adoption and land characteristics based on a landowner
survey that is complemented with GIS information on land characteristics. Data sets in
landowner studies are typically survey based, containing only landowner assessments of
environmental quality. Lambert (2007) used aggregated measures as a proxy of environmental
sensitivity. Here we test whether national-level GIS databases can offer environmental character
information relevant in landowner studies. Based on probit models we discuss the challenges in
activating the key landowners to engage in water conservation.
2. Theoretical framework
Landowners are traditionally seen as profit maximizing operators but profit-maximization
theory alone doesn’t always provide sufficient evident for correctly predicting behaviour
(Sheeder and Lynne, 2009). Abdulla (2009) has developed a theoretical framework to capture
landowners’ behavior in relation to the adoption of conservation measures. The framework
accounts for both utility and profit maximization merged in the same maximization problem.
Following the model structure proposed by Abdulla (2009) landowners’ utility, considering
the decision to adopt water conservation measures, depends on goods and services consumed, on
the environmental state (e.g. water and soil quality) and on the net profits the landowner acquire
from producing both private and environmental goods. The condition of resource affects the
utility in two ways, i.e. by providing recreation and other opportunities to use good quality
surface water and by affecting the decision to adopt conservation measures within the production
process and ultimately the farm income. Let us consider that a landowner wants to maximize his
or her profit from farm production, and consequently his or her utility given the decision to adopt
water conservation practices. That is,
(1)
(2)
where
,
,
, are the prices of good d, of output m and of input n respectively,
the production units of output m,
is the fixed cost of production,
conservation practice (fixed and variable costs) and
denotes
is the cost of
is the off farm income. Also,
is a
1
Agri-environmental scheme (AES) in Finland is based on a three level system, i.e mandatory basic measures, additional
measures and special measures (Lehtonen et al., 2007). Its main goal is to reduce nutrient runoff and maintain agricultural
landscape as well as biodiversity. Participants have to comply with five mandatory basic measures and at least one additional
measure. Special measures are set on a voluntary basis and are considered environmentally more effective. The present study is
referred to the list of special measures included in the 2000-2006 programming period of Rural Development Policy.
Nevertheless, our findings may be considered relevant also for to the next period 2007-2013, as the structure of Finnish AES has
remained roughly the same (Quizhen et al., 2012).
4
vector of goods and services consumed,
is the condition of resources,
is the
maximum indirect farm profit for using , inputs and , conservation practices at the optimal
level.
Profit maximization leads to the optimal level of farms inputs and conservation practices,
,
. It is assumed that conservation practices affect the production
yield, by either imposing positive synergies between inputs or by causing tradeoffs between
better environmental state and land use. These two optimal levels of ,
can be estimated
either sequentially or simultaneously but in both cases the condition of resource will be a
determinative factor.
The landowner maximizes his/her utility by choosing optimal values of , and
implicitly derived from
and
which are
by profit maximization. The indirect utility function will then
be:
(3), which provides the optimal level of consumption of goods and services
at all levels of total profits. In the case where landowner is willing to adopt conservation
measures the maximum utility obtained is given as
the
choice
of
adoption,
i.e.
landowner
selects
the
, where
adoption
represents
only
if
.
Abdulla (2009) describes the utility maximization problem in case where the landowner
earns his/her farm profit from land rents and the decision process is the same as above.
Nevertheless the case would be different if adopters are entitled to compensation payments, as
landowners who are not engaged in farming are not entitled in these payments, and hence the net
cost of adoption is higher for them.
3. Literature overview of the determinants of adoption
3.1 Lessons learned regarding the profile of adopters
Pro-environmental behaviour has been suggested to be affected by land owner
characteristics such as socio-demographics, financial characteristics, farm characteristics,
external factors, values and beliefs (Ervin and Ervin, 1982; Knowler and Bradshaw, 2007).
Adoption depends on the factors linked to farm characteristics as these influence the indirect
profit function
and on the landowners’ characteristics which are incorporated in the
utility factor
(Vanslembrouck et al., 2002).
Socio-demographic characteristics have been to a certain extent able to predict behaviour.
For example, Langpap (2004), Vanslembrouck et al. (2002), and Wynn et al. (2001) confirmed
the hypothesis of the age affect as young farmers are more likely to adopt conservation measures.
Education may be positively linked to better understanding the value of anticipated measures
(Vanslembrouck et al., 2002). Financial characteristics, namely agricultural and off farm income
affect considerably the decision of landowners (De Francesco et al., 2008; Wossink and van
Wenum 2003; Zbinden and Lee, 2005). The effect of rental income on adoption of
environmental measures have also been investigated by Lynch and Lovell (2003) and it was
anticipated to have a negative sign, though wasn’t found significant. In the list of farm
characteristics, farm size is considered one of the most important determinants of participation
(Lynch and Lovell, 2003; Vanslembrouck et al., 2002), nonetheless the results have been
5
ambiguous (De Francesco et al., 2008). Spatial considerations have in many cases been
incorporated as determinants in behavioural models and it has been shown (Lynch and Lovell,
2003; Wassink and Wenum, 2003; Jongenneel et al., 2008) that participation can be spatially
diversified. Finally, external factors such as access and level of acquired information (Luzar and
Diagne 1999; Bekele and Drake, 2003; Toma and Mathjis, 2006; Zbinden and Lee, 2005)
concerning the anticipated conservation measures found to positively affect farmer decision.
Previous studies have emphasized the strong predictive power of farmer’s values and
attitudes (e.g., Lynne et al., 1988; Bell et al., 1994; Vanslembrouck et al., 2002; van Putten et al.,
2011; Luzar and Diagne, 1999; Defrancesco et al., 2008) and the reason for holding land
(Pannell et al., 2006). Attachment to the land (Ryan et al., 2003) as well as the importance of
owner’s objectives associated with farming or owning land per se (Emtage and Herbohn, 2012;
de Young, 2000). Jongeneel et al., (2008) suggested that the value of land ownership and of
farmership is positively related to the practice of conservation.
3.2. The role of biophysical characteristics
The condition of resource, i.e factor
have been included in participation models, though
not always as a straightforward factor. Farm biophysical characteristics, such as soil
characteristics, have been translated into environmental risk elements (e.g. Bekele and Drake,
2003; Lambert, 2007; Pautch et al., 2001; Toma and Mathjis, 2006). The results of water
conservation risks have been mixed. In some studies sensitivity of soils has not had significant
effect on investing in soil and water conservation (Nyangena, 2008). On the other hand, soil
slope has been found to have an effect on recognition of erosion problems (Pautsch et al, 2001).
The designation of lands as highly sensitive for erosion has had an effect in adopting
conservation practices (Clay et al., 1998), particularly in the case of rented land (Soule et al.,
2000). Farms using land susceptible to water erosion or near water sources were more likely to
install conservation structures (Lampert et al., 2007).
Although a link between risk and conservation participation has been found, the
conservation motive may not be clear. In some studies poor soil type has found to be positively
correlated with entering an environment protection scheme (Hynes and Garvey, 2009), but was
assumed to be related to the lower agricultural productivity and not the environmental risks.
Adopting best management practices has been summarized in a vote count meta-analysis by
Prokopy et al. (2008) from US studies. In the meta-analysis physical characteristics of the land
slope, soil and the proximity of a river have in minority of cases either a positive or negative
effect on adopting best management practices.
Sinden and King (1990) and Bayard and Jolly (2007), comment on how the quality of
resource is integrated in participation via perception and recognition. Perception of
environmental state leads to a level of awareness and attitude towards environmental protection
which in turn results in actual behaviour. The state of biophysical characteristics has, in many
cases, been incorporated as a subjective measure where farmers are asked for their perception of
the environmental problem (Amsallu and Graff, 2007) or about their estimate on certain farm
characteristics related to environmental degradation, e.g. the level of slope (Bekele and Drake
2003; Amsallu and Graff, 2007). However, other studies have attempted to incorporate objective
measurements such as erodibility indeces (Lampert et al., 2007) or to use national resources
inventory data to collect information on resource characteristics such as slope (Pautsch et al.,
2001).
4. Data and methodology
4.1.Mail survey data
6
A mail survey data collected in 2006 was used to provide variables of adoption of agrienvironmental measures as well as owner characteristics (data collection described in Myyrä et
al. 2010). The mail survey data yielded a total of 2,684 observations corresponding to 44% of
the original sample. Of the final sample, 37% represented active farmers while 63% represented
passive landowners. In addition to the mail survey data, information on the respondents was
available from the register of agricultural taxation and income taxation. This made it possible to
compare the respondents to the general farming population. The respondents represented the
population of farmland owners in Finland quite well, as the differences between the data and the
population were all under five percentage units with respect to the demographic profile, farm
size and geographical distribution.
We also collected data on physical measures, i.e water quality and soil characteristics,
acquired from a GIS database at the national level. To obtain GIS variables in our survey data,
each farm was indicated with the geographic middle point with x,y coordinates. The field
characteristics were obtained for the nearest field to this middle point. In this manner the values
of GIS variables represented a sample from the whole spatial distribution of each farm.
The level of water eutrophication depends on the field characteristics. According to
Puustinen et al (2010) clay soil texture was indicated as risk factor as it relates to erosion and
subsequently to nutrient loading, in particular particle phosphorus loading which in turn is linked
to eutrophication. The N leaching can be indicated by peat soil texture (Bergström and Johanson,
1991) as the largest losses of NO3-N ha-1 year-1 occurred on sandy and peat soils. Peat soils in
Finland may also be less liable toto erosion, since the fields where the dominating soil texture is
peat are rather flat (<1%) (Puustinen et al., 2010). The slope itself was excluded from the
analysis considering the large number of missing values and also the fact that landscape is quite
heterogeneous in slope terms. The soil variables for peat and clay soils were obtained from
Surficial deposit database of Geological survey of Finland. We also accounted for the level of
water quality measured for the closest to the farm plot water body. The water quality index
regards a general usability classification of water bodies (lakes, rivers and sea) provided by the
Finnish Environment Institute (SYKE). The index gives an idea about the general suitability of
water bodies for uses such as water supply, fishing and recreation.
4.2.Model framework and variables measurement
We explored the adoption of water conservation measures, as dependent variable. For
adoption elicited by active farmers, a list of measures for which farmers have received
environmental support in 2005 was presented. The measures are related to water eutrophication
and aim to reduce erosion and nutrient load from the fields. Adoption was regarded as positive if
at least one measure was selected from the available list. This simplification was necessary for
the model, even though we acknowledge that farmers adopt measures based on individual
determinants, e.g farm structure, location e.t.c.
Approximately 21% of active farmers reported a positive participation and considering the
specific measures, 33% of them had established buffer water zones, 32% had implemented
manure treatment, 12% leach treatment, 18% practiced organic farming and 6 % implemented
other measures. For passive landowners, adoption was positive if respondents had performed
some voluntary measures to improve water quality. By almost 20% passive landowners
responded positively regarding the application of voluntary measures.
Accounting for landowners’ attitudes we formulated four sets of statements measuring 1)
nature and recreation values derived from farming and land ownership 2) values attached to farm
production 3) social values attached to land ownership and 4) sentimental values attached to the
7
farm district. All statements were measured on a 5-point Likert rating scale. To avoid
multicollinearity problems instead of using summative or mean variables per set of statements
we preferred to include factor loadings. We employed a factor analysis with principal component
analysis and varimax rotation by restricting the number of factors to four. For all sets Cronbach’s
alpha coefficients was found above 0.60 indicating adequate reliability.
Two models were employed; one where region variables are excluded and the other where
water quality and soil characteristics are excluded. This structure was deemed necessary, since
region variables found to be significantly correlated to water quality and soil characteristics.
4.3.Statistical methods
The probit binary response model was employed to examine the probability of the
dependent binary variable, i.e adoption. Binary response models are commonly applied when
respondent faces two alternatives, y=1, 0 where, as applied in our case study, y=1 if respondent
has participated in conservation measures and y=0 otherwise. We assumed that the probability of
participation is given by the relationship: P( y  1 x)  G( 0  x )
where  0 is a constant term and  is a vector with elements corresponding to the
explanatory variables. In the probit binary response model, G is the normal cumulative
distribution function (cdf) which lies between zero and one. Probit parameters are estimated by
Maximum Likelihood Estimation (MLE) method which maximizes the log- likelihood function
and it works by obtaining the estimates of parameters β that maximize the total likelihood of
observing the outcomes as reported.
5. Results and Discussion
The results of the binary probit models addressed to active landowners are presented in
Table 1. Both models (model 1 and 2) performed well in fitting the data, correctly predicting 68
and 69% of cases, respectively. The overall fit measured by the pseudo-R2 was relatively low.
The results indicate that active farmers’ water conservation behavior was not strongly
contributed by the risk factors. The only risk indicator that had positively effect on farmer’s
participation was water quality of the closest water bodies. This is possibly because this indicator
is directly visible by the farmers and clearly associated to water eutrophication. Nevertheless soil
texture indicating eutrophication risk didn’t affect adoption decision as expected. The erosion
and nutrient leaching risk had a negative effect implying that farmers, who operate in areas
where erosion and nitrogen leaching are more likely, were less keen to adopt conservation
measures. The latter may be related to the trend that intensive, expansive farms have recently
cleared peat lands for cultivation.
Model 2 also revealed that the farmers who owned land in the south and east of Finland
were less keen on water protection than those who owned land in the western part. Intensive
farm production is particularly concentrated in the south, which explains the outcome that
farmers there tend to be more reluctant to undertake water conservation measures as they face a
higher trade off between conservation and land use for production. In eastern areas on the other
hand farming activities are restricted and hence environmental problems related to agriculture are
less severe, implying that conservation is less needed there. The west is the region where
husbandry is most developed, and due to manure issues, the adoption of conservation measures
aiming to treat manure seems more probable.
Farmer managerial characteristics, such as agricultural income, off farm income and
productivity level, as well as farm characteristics, namely farm size, were significant in
explaining participation. Higher income classes as well as large scale, farmers were keener in
adoption of measures, reflecting the lower financial risk. Large scale farmers, in particular, tend
8
to be more flexible in participation compared to small scale ones, probably due to the wider
heterogeneity in plot level productivity. For plots were the resource quality and profitability level
are low, investing in conservation measures doesn’t significantly decrease farming income as the
conservation measures are compensated based on average expenses. Conversely, farmers who
earned a high agricultural income per hectare, defined as productivity, were less likely to adopt
any supplement conservation measure, meaning that farmers hesitate to adopt water conservation
measures in cases where land is considerably profitable. The use of contractors positively
affected the probability of adoption implying possibly more versatile decisions in farm
management. If decision making is solely carried out by the farmer then the probability of
adoption is significantly improved. For farmers who operate on farms that are plant production
oriented, adoption was more likely but the parameter was only statistically significant in model
2. Although livestock farms are associated more with nutrient run off, income support through
AES is more vital for plant production farms (Lehtonen et al., 2007) justifing their higher
tendency to participate. The appraisal of information was found to be a significant and positive
determinant of adoption, as expected.
Regarding the socio-demographic profile of farmers, the level of education was positively
related to adoption of measures, while age did not explain dependents’ variable variation.
Farmers who typically live on the farm were less active in adopting any measure. This
contradicts our initial expectations, as one would expect that private benefits associated with
water conservation would be more likely to be taken into account by those who live on or close
to the farm. Among the attitudinal factors only those related to nature and recreation values were
significant and positive.
Marginal effect estimates indicated that actual adoption was more stongly affected by
farmer and financial characteristics than the water quality or soil type. In particular, residency on
the farm and level of productivity showed the highest marginal effect on the probability of
adopting water conservation measures. Moreover the effect of region variable ‘south’ was
significant.
Table 1. Coefficient estimates and marginal effects of the models addressed to active owners (weighted by
‘Region of farm location’)
Model 1
Coef.(S.E)
Model 2
Marginal
Effects
Coef.(S.E)
Marginal
Effects
Constant
-1.811***(0.352)
-
-1.466*** (0.357)
-
Age_Young
Edu_High
Live_farm
Recreational values
Landownership values
Attached to region values
Farm production values
Income_subsidies
Agri_income_high
Off farm_income_high
Productivity_high
Contractors
Decision_owner
Information_importance
Size_farm_large
Plant_production
Water_quality_low
Soil_erosion
-0.143(0.183)
0.217(0.137)
-0.522***(0.176)
0.099*(0.058)
0.031(0.056)
0.009 (0.054)
0.030 (0.072)
0.095 (0.203)
0.359** (0.149)
0.236** (0.117)
-0.544*** (0.153)
0.236*** (0.115)
0.324*** (0.115)
0.214*** (0.074)
0.426*** (0.144)
0.151 (0.127)
0.290*** (0.110)
-0.264* (0.159)
-0.038
0.062
-0.158***
0.027*
0.008
0.002
0.008
0.026
0.098**
0.066**
-0.138***
0.066**
0.089***
0.059***
0.114***
0.041
0.080***
-0.068*
-0.133 (0.183)
0.248* (0.137)
-0.530*** (0.177)
0.122** (0.059)
0.047 (0.056)
-0.003 (0.054)
0.062 (0.072)
0.055 (0.202)
0.396*** (0.151)
0.216* (0.118)
-0.592*** (0.154)
0.285** (0.116)
0.306*** (0.115)
0.199*** (0.074)
0.409*** (0.145)
0.225* (0.132)
/
/
-0.034
0.069*
-0.156***
0.032**
0.012
-0.001
0.016
0.015
0.104***
0.058*
-0.143***
0.078**
0.081***
0.052***
0.105***
0.058*
/
/
9
Soil_nitrogen
South
East
-0.417* (0.224)
/
/
-0.101**
/
/
/
-0.507*** (0.148)
-0.443*** (0.153)
/
-0.130***
-0.107***
North
/
/
-0.147 (0.154)
-0.037
N cases
McFadden Pseudo R-square
Correct predictions
Pred.Pr (=1)
756
0.121
756
0.124
68.25%
0.242
69.05%
0.229
*P<0.10, ** P<0.05, *** P<0.01
Table 2 presents the results of the probit model for passive landowners. The adoption
models (model 1 and 2) performed well in fitting the data, correctly predicting 70.66% and
70.57% of cases, respectively. The overall fit measured by the pseudo R2 is considered relatively
low for both models.
Model 1 revealed a negative and significant relationship between degraded water quality
and adoption. It appears that either passive owners were not well aware of the state of the water
body located closest to their land, or that they did not consider themselves equally responsible
for water quality deterioration. The sign of coefficient may also reflect the private cost of
investing in conservation measures which may be perceived as considerable if the quality of
resource is already degraded. The expectations related to the effectiveness of conservation may
also be at a low level. Moreover, soil type, was found to have a negative but not significant
coefficient. All region variables were found to be positive and significant, in contrast to the
active farmers’ models, indicating that owners of land in the south, east and north of Finland
were more likely to have adopted measures than those owning land in the west of the country.
This may be related to the lower importance of water bodies for recreation in the west.
The model revealed that education and attitudinal factors significantly affected the actions
towards water conservation. Owners with higher level of general education were less likely to
have adopted measures. From the attitude variables, the importance of recreational values and of
farm production values was found to have a positive and significant influence to the actual
adoption. Moreover the level of off-farm income explained much of the variation in adoption and
indicated a positive influence. The size of the farm positively affected the decision to adopt
measures, but the effect was only significant in model 1. Age was only found significant in
model 2, indicating that younger owners were more receptive towards conservation.
The marginal effects of the income level as well as attitudinal factors were shown to
significantly determine the probability of adoption. Interestingly, the marginal effect of resource
quality and in particular that of water was substantial.
Table 2. Coefficient estimates and marginal effects of the models addressed to passive owners (weighted by
‘Region of farm location’)
Model 1
Coef.(S.E)
Constant
Age_Young
Edu_High
Live_farm
Recreational values
Landownership values
Farm production values
Attached to region values
-1.353*** (0.241)
0.178 (0.138)
-0.367** (0.158)
-0.119 (0.126)
0.399*** (0.069)
0.081 (0.063)
0.218*** (0.058)
0.033 (0.060)
Model 2
Marginal
Effects
_
0.047
-0.085**
-0.030
0.101***
0.020
0.055***
0.008
Coef.(S.E)
-1.739*** (0.265)
0.257* (0.138)
-0.403** (0.157)
-0.058 (0.124)
0.422*** (0.068)
0.082 (0.062)
0.223*** (0.057)
0.033 (0.059)
Marginal
Effects
0.069*
-0.094***
-0.015
0.108***
0.021
0.057***
0.009
10
Offarm_Level 2 (9000-20000€)
Offarm_Level 3 (20000-40000€)
Offarm_Level 4 ( more than 40000 €)
Income_leasing
Decision_owner
Information_importance
Size_farm_large
Water_quality_low
Soil_erosion
Soil_nitrogen
South
East
North
N cases
McFadden Pseudo R-square
Correct predictions
Pred.Pr (=1)
0.474** (0.191)
0.130**
0.433** (0.182)
0.119**
0.473*** (0.174)
0.592*** (0.187)
-0.182 (0.117)
0.129 (0.120)
0.083 (0.063)
0.326* (0.188)
-0.521*** (0.128)
-0.037 (0.188)
-0.086 (0.275)
/
/
/
0.124***
0.160***
-0.046
0.032
0.021
0.090
-0.122***
-0.009
-0.021
/
/
/
0.447*** (0.164)
0.521*** (0.177)
-0.177 (0.117)
0.152 (0.117)
0.061 (0.061)
0.283 (0.181)
/
/
/
0.279* (0.157)
0.417*** (0.156)
0.288* (0.168)
0.119***
0.142***
-0.045
0.039
0.016
0.078
/
/
/
0.073*
0.077**
0.112
777
0.139
802
0.132
70.66%
0.214
70.57%
0.216
*P<0.10, ** P<0.05, *** P<0.01
6. Conclusions
We investigated the profile of land owners adopting water conservation measures and the
effect of the condition of resources, namely water and soil, may impose on their decision to
adopt. Active farmers and passive (i.e, who are not engaged in farming) owners, were examined
separately. For active farmers, the model confirmed that income is the key determinant for
adopting water conservation. Furthermore, larger farms have greater possibility to adopt. This
can be explained by the wider spectrum of various plot types; it is easier to find plots on their
estate where productivity is low and that are well suited to water conservation. Conversely,
farmers who operate in more productive areas were less eager to adopt, signaling the importance
of profits versus environmental conservation. For passive landowners, in contrast to active
owners, attitudes were found to be key determinants for voluntary adoption. The attitudes
themselves reveal the reasons why this group of owners still owns land even if they are not
engaged in farming. The higher the off-farm income the more likely it is that the owner would
voluntarily adopt any measure. This is according to our expectation, given that passive owners
receive no compensation for the measures applied.
Both cases showed that the tendency for adoption is spatially diversified. Active farmers
who operate in the south and east are less likely to adopt water conservation measures contrary to
the corresponding findings derived from models for passive owners. This outcome may be of
reliance in guiding policy, as spatially tailored measures may be planned in such a way that
would ultimately increase adoption. Previous studies that have assessed Finnish AES have also
conlcuded that measures and support levels should be adjusted to the need of each region
(Quizhen et al., 2012). We suggest that more incentives should be provided to active farmers
who operate in the south than to those who operate in the west. For passive owners we
recommend information campaigns on opportunities for applying conservation measures,
possibly targeted to western areas.
Regarding the role of resource condition, our analysis provided mixed results. Although we
found some evidence that farmers are more likely to have applied supplementary measures in
areas where water quality is poor, soil texture indicators didn’t affect adoption in the same
manner. The adoption of measures by passive owners was also negatively affected by the fact
11
that water quality and soil type were at risk. All in all, our outcome is, if not ambiguous, weak
supportive of our initial expectation that owners would be naturally motivated to adopt measures
in areas where the environmental state has deteriorated or is at risk of deterioration. Even though
past research has advocated the adoption of conservation measures on lands in a deteriorated
state, e.g with a perceived low soil quality (Zbinden and Lee, 2005) or perceived slope level
(Amsalu and Graff, 2007) or where there is a perception of environmental risk (Tomas and
Mathijs, 2007) our results are not consistent with this. This inconsistency may reflect the type of
data used, given that our analysis was based on objective measures whereas previous researchhas
examined perceived measurements. On the other hand, our findings may simply reflect a lack of
awareness of the condition of environmental resources. Thus, future research needs to be focused
on the formation of environmental perceptions.
Referring back to our initial question, our model indicates that for farmers as well as
landowners, environmentalism and/or stewardship incentives are not strong enough to motivate
adoption for conservation measures. Even if active farmers do care about environmental
protection, they simply need financial incentives to act. Although off-farm benefits may affect
the ‘utility’ part explained in the theoretical framework, the ‘profit maximization’ part prevails
precisely because conservation measures may not be accompanied by direct benefits to the
owners. We conclude that as there is no spontaneous tendency among landowners to adopt water
conservation measures and targeted agri-environmental policy, despite their considerable
implementation costs, cannot be avoided. Spatially tailored measures can attract more adopters
on hot spot areas than homogeneous and/or volunteer ones can. A good example is the green
auctions mechanism where adopters are compensated according to the environmental benefits
they produce, which are greater in areas where environmental condition is poor.
The policy issue for passive landowners is important since they are property owners with a
considerable amount of decision-making power. Lease contracts are short, and landowners have
full responsibility, every sixth year, regarding the next lease holder. Considering that landowners
account for values other than just monetary ones, lease contacts might be enhanced by a
commitment to water conservation. Lease holders, who are pro-water conservation, could later
be compensated. In this way, a group effort towards water conservation, involving both active
and passive landowners, would be attainable.
References
Abdulla, M. (2009). The impact of ownership on Iowa land owners' decisions to adopt
conservation
practices.
Theses
and
Dissertations.
Paper
10918.
http://lib.dr.iastate.edu/etd/10918.
Amsalu, A., and Graaff, J. (2007). Determinants of adoption and continued use of stone terraces
for soil and water conservation in an Ethiopian highland watershed. Ecological
Economics 61: 294-302.
Bayard, B. and Jolly, C. (2007). Environmental behaviour structure and socio-economic
conditions of hillside farmers: A multiple-group structural equation modelling approach.
Ecological Economics 62: 433-440.
Bekele, W., and Drake, L. (2003). Soil and water conservation decision behaviour of subsistence
farmers in the Eastern Highlands of Ethiopia: a case study of the Hunde-Lafto area.
Ecological Economics 46: 437_/451
12
Bell, C.D., Roberts, R.K., English, B.C., Park, W.M. (1994). A logit analysis of participation in
Tennessee's forest stewardship program. Journal of Agricultural and Applied Economics
26 (2): 463–472.
Bergström, L. and Johansson, R. (1991). Leaching of nitrate from monolith lysimeters of
different types of agricultural soils. Journal of Environmental Quality 20: 801–807.
Clay, D., Reardon, T. & Kangasniemi, J. (1998). Sustainable Intensification in the Highland
Tropics: Rwandan Farmers' Investments in Land Conservation and Soil Fertility.
Economic Development and Cultural Change 46: 351-377
De Young, R. (2000). Expanding and Evaluating Motives for Environmentally Responsible
Behavior. Journal of Social Issues 56(3): 509-526.
Defrancesco, E. Gatto, P., Runge, F. & Trestini, S. (2008). Factors Affecting Farmers’
Participation in Agri-environmental Measures: A Northern Italian Perspective. Journal of
Agricultural Economics 59: 114–131.
Emtage, N. and Herbohn, J. (2012). Implications of landowners’ management goals, use of
information and thrust of others for the adoption of recommended practices in the Wet
Tropics region of Australia. Landscape and Urban Planning 107: 351-360.
Ervin, C.A., Ervin, D.E. (1982). Factors affecting the use of soil conservation practices:
hypotheses, evidence and policy implications. Land Economics 58: 277–292.
Jongeneel, A. R.,, Polman, B. P. N., and , Slangen, H. G. L. (2008). Why are Dutch farmers
going multifunctional? Land Use Policy 25, 81–94
Hynes, S. & Garvey, E. (2009). Modelling Farmers’ Participation in an Agri-environmental
Scheme using Panel Data: An Application to the Rural Environment Protection Scheme
in Ireland. Journal of Agricultural Economics 60: 546–562.
Iho, A., Lankoski, J., Ollikainen, M., Puustinen, M. & Lehtimäki, J. 2013. “Agri-environmental
auctions for phosphorus load reduction: Experience from the Finnish Pilot”. Manuscript.
Kendra, A., and R. B. Hull. (2005). Motivations and Behaviors of New Forest Owners in
Virginia. Forest Science 51: 142–154.
Knowler, D., Bradshaw, N. (2007). Farmers’ adoption of conservation agriculture: A review and
synthesis of recent research. Food Policy 32: 25-48.
Lambert M. D., Sullivan P., Claassen R., Foreman L. (2007). Profiles of US farm households
adopting conservation-compatible practices. Land Use Policy 24: 72-88.
Langpap, C. (2004). Conservation incentives programs for endangered species: an analysis of
landowner participation. Land Economics 80 (3): 375–388.
Lankoski, J & Ollikainen, M. (2003). Agri‐environmental externalities: a framework for
designing targeted policies. European Review of Agricultural Economics 30(1): 51-75.
Lehtonen H., Lankoski J., Koikkalainen K., (2007). Economic and environmental performance
of alternative policy measures to reduce nutrient surpluses in Finnish agriculture.
Agricultural and Food Science 16:421-440.
Luzar, E.,J. and Diagne, A. (1999). Participation in the next generation of agriculture
conservation programs: the role of environmental attitudes. The Journal of Socioeconomics, 28: 335–349.
Lynch, L. and Lovell, J., S. (2003). Combining spatial and survey data to explain participation in
agricultural land preservation programs. Land Economics 79 (2): 259–276
Lynne, G.D., Shonkwiler, J.S., Rola, L.R. (1988). Attitudes and farmer conservation behaviour.
American Journal of Agricultural Economics 70: 12–19.
13
Maybery, D., L. Crase, and C. Gullifer. (2005). Categorizing farming values as economic,
conservation and lifestyle. Journal of Economic Psychology 26:59–72.
Myyrä, S. & Pouta, E (2010). Farmland owners’ land sale preferences: can they be affected by
taxation programs? Land Economics 86: 245-262.
Nyangena, W. (2008). Social determinants of soil and water conservation in rural Kenya.
Environment, Development and Sustainability 10: 745-767 .
Pannell D. J., Marshall G.R., Barr N., Curtis A., Vanclay F., Wilkinson R., (2006).
Understanding and promoting adoption of conservation practices by rural
landholders.Australian Journal of Experimental Agriculture 46:1407-1424.
Pautch R. G., Kurkalova A. L., Babcock A. B., Kling K., (2001). The efficiency of sequestering
carbon in agricultural soils. Contemporary Economic Policy 19(2):123-134
Prokopy, L. K. Floress, D. Klotthor-Weinkauf, and A. Baumgart-Getz. (2008). Determinants of
Agricultural BMP Adoption: Evidence from the Literature. Journal of Soil and Water
Conservation 63:300-311.
Puustinen, M., Turtola, E., Kukkonen, M., Koskiaho, J., Linjama, J., Niinioja, R., Tattari, S.
(2010). VIHMA—A tool for allocation of measures to control erosion and nutrient
loading from Finnish agricultural catchments. Agriculture Ecosystems and Environment,
138: 306-317.
Quizhen C., Sipilainen T., Sumelius J., (2012). Assessment of Agri-environmental public goods
provision using fuzzy synthetic evaluation. Discussion papers n:o 6i, Helsinki 2012
Ryan, L.R., Erickson, L.D., de Young, R. (2003). Farmers’ motivations for adopting
conservation practices along riparian zones in Mid-western agricultural watershed.
Journal of Environmental Planning and Management, 26(1): 19-37
Ross-Davis, A. L., and S. R. Broussard (2007). A Typology of Family Forest Owners in NorthCentral Indiana. Northern Journal of Applied Forestry 24:282–289.
Sinden J.A. and King A. D. (1990). Adoption of soil conservation measures in Manilla Shire,
New South Wales. Review of Marketing and Agricultural Economics 58: 179-192.
Sheeder R. and Lynne D. G. (2009). Empathy Conditioned Conservation: "Walking-in-theShoes-of-Others" as a Conservation Farmer. Selected Paper prepared for presentation at
the Agricultural & Applied Economics Association’s 2009 AAEA & ACCI Joint Annual
Meeting, Milwaukee, WI, July 26-28, 2009.
Soule J.M., Tegene A., Wiebe D. K. (2000). Land tenure and the adoption of conservation
practices. American Journal of Agricultural Economics 82(4): 993-1005
Toma, L., and Mathijs, E. (2007). Environmental risk perception, environmental concern and
propensity to participate in organic farming programmes. Journal of Environmental
Management 83: 145–157
Vanslembrouck, I., van Huylenbroeck, G., Verbeke, W. (2002). Determinants of the Willingness
of Belgian Farmers to Participate in Agri-environmental Measures. Journal of
Agricultural Economics 53(3): 489-511
van Putten, E., I., Jennings, M., S., Louviere, J.,J., Burgess, B., L. (2011).Tasmanian landowner
preferences for conservation incentive programs: A latent class approach. Journal of
Environmental Management 92: 2647-2656
Wynn, G., Crabtree, B., Potts, J. (2001). Modelling farmer entry into the environmentally
sensitive area schemes in Scotland. Journal of Agricultural Economics 52 (1): 65-82
14
Wossink, G., A., A. and van Wenum, J., H. (2003). Biodiversity conservation by farmers:
Analysis of actual and contingent participation. European Review of Agricultural
Economics 30: 461–485.
Zbinden, S and Lee, R.D. (2005). Paying for Environmental Services: An Analysis Of
Participation in Costa Rica’s PSA Program. World Development 33(2): 255–272.
15