1. Introduction
The soil system provides different types of ecosystem services such as provisioning ecosystem services (e.g., a medium to provide food, natural water reservoir to provide clean water, habitat provision); regulating ecosystem services offering resilience to climate change and extreme weather events, droughts, and floods (e.g., climate regulation through carbon storage, nutrient cycling); and cultural services (e.g., supporting culture, traditions, and practices linked to agriculture and landscapes) [
1]. Consequently, soil management is key to build up resilience to ensure sustainable agricultural systems and the environment through soil quality protection and improvement [
2].
Currently about 60 to 70% of soils in the EU are in an unhealthy state [
3]. The costs of soil degradation in the EU have been estimated in that they may exceed EUR 50 billion annually [
3] based on a previous estimate of EUR 38 billion annually for 25 EU countries that did not include costs from biodiversity decline, sealing, or compaction [
4]. For the UK, soil degradation has been estimated to cost GBP 1.2 billion annually [
5]. The past European Union’s (EU) Common Agricultural Policy (CAP) based on price support in combination with technological change led to agriculture intensification, specialization, and concentration of production with subsequent environmental impacts including habitat loss and decline in wildlife and biodiversity [
6,
7,
8,
9] and possibly detrimental soil quality and yields (e.g., shorter rotations, traffic-induced soil compaction [
10,
11,
12]).
The EU CAP has reacted to these environmental impacts and the constraints to agricultural support set up in the 1994 WTO’s Uruguay Round Agreement in agriculture [
13]. Thus, in 1998, voluntary set-aside was introduced to reduce crop overproduction and deliver environmental benefits, following the 1992 MacSharry reform that made set-aside compulsory and the CAP’s Agenda 2000, which allowed Member States (MSs) to apply cross-compliance. Under Regulation (EEC) 2078/92, MSs were allowed to provide support to farmers for making environmental improvements to their land by changing farming practices [
14]. The Fisher reform in 2003 meant that agricultural policies moved from price support to area-based payments and payments for the supply of environmental goods (e.g., agri-environmental schemes).
More specifically on soil, the EU has developed an EU soil strategy, which has medium- and long-term objectives by 2030 and 2050, respectively. The EU soil strategy’s medium-term objectives include the following: (a) Combat desertification; restore degraded land and soil, including land affected by desertification, drought, and floods; and strive to achieve a land degradation-neutral world (Sustainable Development Goal 15.3). (b) Restore significant areas of degraded and carbon-rich ecosystems, including soils. (c) Achieve an EU net greenhouse gas removal of 310 million tonnes CO
2 equivalent per year for the land use, land use change, and forestry (LULUCF) sector. (d) Reach good ecological and chemical status in surface waters and good chemical and quantitative status in groundwater by 2027. (e) Reduce nutrient losses by at least 50%, the overall use and risk of chemical pesticides by 50%, and the use of more hazardous pesticides by 50% by 2030. (f) Significant progress has been made in the remediation of contaminated sites. As for the long-term objectives, these are (a) reach no net land taking; (b) soil pollution should be reduced to levels no longer considered harmful to human health and natural ecosystems and respect the boundaries our planet can cope with, thus creating a toxic-free environment; (c) achieve a climate-neutral Europe and, as the first step, aim to achieve land-based climate neutrality in the EU by 2035; (d) achieve for EU a climate-resilient society, fully adapted to the unavoidable impacts of climate change by 2050. Also, aiming for that by 2030, at least 75% of soils in each EU Member State are healthy, or show a significant improvement towards meeting accepted thresholds of indicators, to support ecosystem services [
3]. Consequently, the EU has been incorporating specific policy measures to target agricultural soils at regional and national levels [
15]. The CAP has set out soil protection policies via its soil thematic strategy, which includes a proposal for a Directive on soil monitoring and resilience (soil monitoring law) [
16].
Regarding the UK, the 2016 Brexit vote and UK decision to leave the EU meant that the UK could decide on a new agricultural policy, which is currently under development with farm support in the UK changing. Agriculture Act 2020 provided a framework for the UK government to create its own agricultural policy. However, agriculture is a devolved policy area, which means that each administration (England, Scotland, Wales, and Northern Ireland) can shape their own agricultural policies. This offers opportunities to develop policies closer to stakeholders that consider particular environmental, socio-demographic, and geographical characteristics. It also faces challenges such as effective coordination whenever needed. To ensure an effective coordination, the UK government and the devolved administrations agreed to establish the UK agricultural support framework to learn from each other and coordinate policies when needed. Soil management and improved soil quality can be targeted through the environmental land management scheme (ELMS) and its sustainable farming incentive [
17].
These policies in the EU and the UK require investment in soil quality protection and improvement. The funding allocated to the CAP for the 2021–2027 period is EUR 387 billion, which is split into the European agriculture guarantee fund (EAGF), EUR 291.1 billion, and the European agricultural fund for rural development (EAFRD), EUR 95.5 billion. The CAP strategic plans devote EUR 98 billion (EUR 14 billion per year) to deliver specific environmental benefits for climate, water, soil, air, biodiversity, and animal welfare and to encourage practices that go beyond the conditionality [
18]. However, there is no estimate of CAP spending on soil quality protection and improvement apart from an estimate of CAP financing for sustainable soils and manure management to be approximately EUR 85 billion over 2014–2020 (EUR 12 billion per year) [
19], which seems an overestimate considering the CAP strategic plans’ figure.
Regarding the UK budget, agriculture transition plan 2021–2024 states the spending plans across environmental and animal welfare outcomes, improving farm prosperity and direct payments, which amounts to an average of GBP 2.4 billion a year [
17]. However, there are uncertainties on the effectiveness of policy measures and agricultural practices to be applied to protect and improve soil quality (e.g., uncertainties around the effectiveness of measures to improve soil carbon sequestration [
20]). These uncertainties may play a role in the public’s support for public policies that support soil quality.
This study contributes to the scarce literature on the general public’s valuation of agricultural soil quality protection and improvement by (a) evaluating the public support for agriculture soil quality protection and improvement policies, and (b) gaining understanding on what motivates the general public’s support for agricultural soil protection and improvement. In addition, it contributes to the literature that incorporates uncertainty into the economic valuation of agricultural policies by providing a monetary valuation of policies under relatively less and more certain scenarios. To the best of our knowledge, this is the first paper to provide an economic valuation of agricultural soil quality protection and improvement in the UK and Spain and one of the scarce papers estimating the soil’s economic value and how general population values soil functions and their support for public policy action [
21,
22,
23,
24]. The previous literature on willingness to pay (WTP) for soil security has provided WTP estimates for Italian and Australian citizens, indicating that there is public support for these. The average public’s WTP for soil security in the Veneto region in Italy and New South Wales in Australia was estimated to be EUR 244 and AUD 567, respectively, by using a choice experiment approach [
21]. Other stated preference valuation studies of soils [
20,
25,
26,
27,
28] provide useful information for setting specific agri-environmental schemes contributing to manage soil in a sustainable way. However, as pointed out by Bartowski et al. [
29], they are narrow in terms of the soil-based ecosystem services covered (soil erosion, carbon sequestration). Regarding soil erosion, studies using stated preference methods have also addressed other types of erosion control. Specifically, one study examined public support for erosion control programs at a popular beach resort in Sicily, Italy [
30].
The structure of this paper is as follows:
Section 2 covers the material and methods of this study,
Section 3 presents the results of this study,
Section 4 discusses the results, and finally
Section 5 is dedicated to the conclusions of this study.
2. Materials and Methods
2.1. Data Collection: Survey
A cross-sectional survey instrument was designed and administered to a panel using Qualtrics. Qualtrics distributed the survey to a panel, collecting a total of 882 and 910 valid responses from the UK and Spain, respectively, in March–April 2021, for which ethical approval was according to the procedures specified by Newcastle University Research Ethics Committee. Quota restrictions were imposed on age and gender. We divided survey participants into two groups per country: Group 1 UK (n = 449); Group 2 UK (n = 433); Group 1 Spain (n = 462); Group 2 Spain (n = 448).
Both groups in each country were presented with the same background information except that those in group 2 in each country had extra text in their background information, which introduced uncertainty and ambiguity on the environmental scheme presented to improve soil quality. This was based on the fact that it is difficult to see what farmers would decide to carry out (e.g., number and type of soil management practices that will be conducted at each farm; the number, distribution, and location of farms that would join the environmental land management scheme to implement these soil improving measures) [
2].
It is worth noting that we conducted pre-tests and piloted it to refine questions, identify potential issues, and ensure that the questionnaire is clear and understandable. During the data collection process, we initially used a sample of 100 responses to verify that the provided responses were reasonable and coherent. During the main data collection, we implemented real-time supervision and monitoring to oversee the data collected. Time taken to respond to the questionnaire was recorded. Since it was expected that panel data respondents typically respond quicker than non-panel data respondents since they are usually more familiar with the survey process, having participated in previous surveys, we used only responses from respondents who answered all questions of the questionnaire and took more than 4 min to respond. The median time to respond to all questions was 18 min.
2.2. Data Collection: Questionnaire
The survey questionnaire comprised a total of 10 sections including background information, WTP questions, protest views, soil quality attitudes, trust, risk attitudes, time preference, uncertainty, pro-social behavior, and socio-demographics. In addition, the questionnaire included attention check questions to ensure that respondents were focusing on the questions asked.
2.2.1. Background Information
We presented all participants with information about soil quality and its decline in many parts of the world: “Soil quality is in decline in many parts of the world. Agricultural practices and agriculture intensification (e.g., increase in the use of fertilizers per ha) are two of the main drivers of soil quality decline. However, the use of sustainable soil management practices by farmers can help improving soil fertility and soil structure, which can reduce the risk of flooding and erosion. Hence, farmers’ uptake of sustainable soil management practices is seen as key to improve soil quality and the associated environmental benefits”. We also explain to UK and Spanish respondents relevant EU and UK policies. For the UK respondents, we stated that “Under the Agriculture Act 2020 the UK government establishes that it may give financial assistance for environmental protection or improvement. This means that the government may pay farmers in order to achieve these purposes. These payments may be part of the new Environmental Land Management scheme, the cornerstone of the governments’ new agricultural policy. Farmers may be paid for delivering the following public goods: clean air; clean and plentiful water; thriving plants and wildlife; protection from environmental hazards; beauty, heritage and engagement with the environment; reduction of and adaptation to climate change”. To Spanish respondents, we stated that “The soil protection policy of the European Union (EU) is shaped through the EU Soil Thematic Strategy. This soil policy is provided using various instruments such as the common agricultural policy (CAP). The EU can give financial support to farmers for environmental protection or improvement. These payments would be based on farmers achieving environmental benefits: clean and plentiful water; thriving plants and wildlife; protection from environmental hazards; beauty, heritage and engagement with the environment; reduction of and adaptation to climate change”. Then, we asked all respondents to focus specifically on one of the ways in which farmers may contribute to protect and improve the environment, protecting or improving quality of soil, and stated that “Soil is an essential ecosystem that delivers valuable services such as the provision of food, energy and raw materials, carbon sequestration, water purification, nutrient regulation, pest control, and support for biodiversity and recreation”.
Then, we provided all respondents with a table (
Table 1) with information on the ways in which agricultural soil improvement can be achieved by farmers and the related benefits to soil quality and the environment.
Participants in Group 2 were given the following extra information: “However, there is uncertainty on the overall impact/success of the Environmental Land Management scheme (for UK respondents)/EU Soil Thematic Strategy (for Spanish respondents) (i.e., level of provision of public goods) since this depends on: (1) the number, distribution and location of farms which would join the environmental land management scheme to implement these soil improving measures, (2) the number and type of soil management practices that will be conducted in each farm.” This information introduces a lack of certainty about the scheme success/outcome to respondents as well as lack of clarity on how the environmental land management scheme will be implemented (i.e., ambiguity).
2.2.2. Willingness to Pay (WTP) Questions
After the background information was presented, we moved to the WTP questions, providing the following explanation to respondents: “Next you will be asked about your willingness to pay for a policy that aims at improving the soil quality. Under this policy, the government may grant financial assistance to farmers for the protection or improvement of the environment. Payments would be given to farmers who carry out agricultural practices indicated in
Table 1. This information you provide may be useful for government to establish the agri-environmental payments farmers receive in the future associated with soil quality protection and improvement”. We use a double-bounded dichotomous choice contingent valuation approach. Respondents were asked to answer “yes” or “no” to the following question: “Would you be willing to pay an additional £X in your taxes each month for the next 10 years to protect and improve agricultural soil quality?”, which was followed up by another question, which depended on the answer given to the first question (e.g., “You said you would not be willing to pay £5. Would you be willing to pay £1 each month for the next 10 years?”; “You said you would be willing to pay £15. Would you be willing to pay £20 each month for the next 10 years?” (See
Section 2.4 for more detail).
2.2.3. Protest Answers
When respondents answered “no” twice to the WTP question, they were asked about the reasons for their answer.
Table 2 shows a list of possible reasons for respondents answering “no” twice to the WTP question. Respondents were asked to select as many statements as applied as follows: “You answered that you are not prepared to pay for the proposed policy. Could you state the reason for your answer? Click as many as they apply”. Responses driven by the respondent’s economic constraints or failures to derive utility from protecting and improving soil quality are considered as “true zeros” (i.e., statements 1, 3, 7, and 8 are considered “true zeros”). Regarding the “other, please state” option, there were a variety of reasons including “true” and “non-true” zeros. For instance, most respondents expressed dissatisfaction with the scenario presented (e.g., “They all seem to be things the farmers are already doing or should be doing and I fail to see why they need paying to look after the farm properly—also what is it that they feel needs cash another”; “I have never known a poor farmer. Sell one of the Range Rovers and use that money”; ”Farmers and government are well funded already. They don’t need more taxes from already taxed people I think”). These were not considered to be a true zero due to the participant, whereas some respondents expressed economic constraints (e.g., “I am retired and I do not have any spare income”), which were considered to be “true zeros”. For the UK survey, we removed a total of 87 protest responses out of 882 (9.9%) whereas from the Spanish survey, we removed a total of 88 protest responses out of 910 (9.7%), leaving a sample size for the UK of n = 882 and a sample size of n = 910 for Spain.
2.2.4. Views on Sustainable Agricultural Practice Benefits
We collected respondents’ views on how important the benefits related to the implementation of sustainable practices are for them. A set of 10 benefits were stated (
Table 3). To reduce data complexity, we applied a Principal Component Analysis (PCA) to the 10 statements. The PCA produced a single component, named “SAP benefits” (with strong internal consistency—alpha ranging between 0.90 and 0.93; Keiser–Meyer–Olkin (KMO) ranging between 0.92 and 0.94), which was used as an explanatory variable of the individual’s WTP for agricultural soil quality protection and improvement.
2.2.5. Views on How Concerned Participants Are about Soil Quality Decline
We asked participants a question with four possible answers on how serious the issue of declining soil quality was for them. We asked “How serious is the issue of declining soil quality to you?” and the possible answers were “not at all serious”, “not very serious”, “fairly serious”, and “extremely serious”. This allowed us to classify participants into relatively concerned about soil quality decline (those who answered “fairly serious” or “extremely serious”) and relatively unconcerned about soil quality decline (those who answered “not at all serious” or “not very serious”). The variable “Soil quality concern” takes a value of 1 if they are relatively concerned, and 0 otherwise, given the responses given to the question above.
2.2.6. Trust
Regarding trust, UK and Spanish participants were asked to evaluate how much trust they put in the UK/Spanish parliament; UK/Spanish government; UK/Spanish political parties; UK/Spanish politicians; Defra Department for environment and rural affairs (for UK respondents only); Spanish Ministry for Agriculture, Fisheries and Food (for Spanish respondents only); government to monitor farmers’ agricultural practices; farmers to carry out sustainable agricultural practices; and agri-environmental schemes to be successful in improving soil quality. Respondents evaluated their level of trust in these using a 5-point Likert scale (none at all, a little, a moderate amount, a lot, a great deal). We conducted a PCA to identify trust dimensions and how these may affect the public’s WTP. Two components were obtained, “Trust in governance” and “Trust in agriculture stewardship“, which showed strong internal consistency (alpha ranging between 0.87 and 0.92; KMO ranging between 0.87 and 0.91), which were used as explanatory variables in the WTP model. We expect that the more trust respondents have in these, the more likely they are to be willing to pay for the soil quality improvement scheme.
2.2.7. Risk Attitudes
We used an experimental approach to measure risk-taking behavior following Holt and Laury’s ten paired lottery choice decisions [
31] (
Table 4). Each of the lotteries was presented individually to respondents and consecutively (from the top lottery to the bottom lottery shown in
Table 4). Respondents should cross over to option B when the probability of the payoff outcome increases enough as they “move down the table” [
31]. The switching point is used as an estimate for the relative risk aversion (i.e., we use a 10-point risk aversion scale based on the switching point from option A to option B) [
31]. The expected payoff (these were not provided to respondents) indicates that a risk-neutral individual would switch from option A to option B at the fifth question; an extreme-risk-averse individual would switch at the tenth question; and an extreme-risk lover would switch at the first question.
The variable “Risk averse” takes the values at which the respondent has switched. If the respondent switches to B and comes back to A in the next question, the first switch is not considered a definite switch.
2.2.8. Time Preferences
To capture individuals’ time preferences or patience, we use an approach previously used in the literature [
32,
33] where survey participants are given two choices from which they are asked to select one of them: either GBP/EUR 3400 this month or GBP/EUR 3800 next month. Respondents were divided into relatively patient and relatively impatient according to the choice they made. The variable “Time Preference” takes a value of 1 if the respondent is patient and 0 if they are impatient. The relationship between time preference and willingness to pay has been previously analyzed using experiments [
34]. Findings indicated consumers’ WTP for fuel economy improvements to be higher if payments for improvements are made more disperse through time [
34]. The association between risk and time preferences and farmers’ preferences for agroforestry attributes has also been investigated and it has been recommended to be included when studying the adoption of agricultural innovations [
35]. We incorporate individuals’ time preferences to investigate their direct association with individuals’ willingness to pay.
2.2.9. Ambiguity Tolerance
As defined by McLain, ambiguity aversion theory states that decision makers prefer a known risk to an ambiguous risk [
36]. In our study, there is no certainty about the level of success of the policy (i.e., granting farmers financial assistance to carry out agricultural practices in
Table 1) in achieving its objective. To measure individuals’ ambiguity tolerance, we used a 13-item measure of ambiguity tolerance (MSTAT-II) proposed by McLain using a 5-point Likert scale (1 = strongly disagree; 5 = strongly agree) [
37].
Table 5 shows the 13 items used. We do not have any strong prior expectation on the relationship between ambiguity tolerance and individuals’ WTP for granting farmers financial assistance to carry out agricultural practices. Individuals who have a relatively high tolerance to ambiguity are more comfortable with uncertainty and more likely to be willing to pay for soil quality protection and improvement than those with relatively low tolerance to ambiguity. On the other hand, individuals with high ambiguity tolerance may be less willing to pay for soil quality protection and improvement than those with relatively low tolerance to ambiguity since the latter would like to reduce the ambiguity/uncertainty. To reduce data complexity and classify participants by their relative ambiguity tolerance level, we applied a cluster analysis. Two clusters are obtained. The variable “Ambiguity Tolerant” takes a value of 1 if the respondent is ambiguity-tolerant and 0 otherwise. We expect for ambiguity-tolerant respondents to be willing to pay more than ambiguity-intolerant respondents in general, but more in particular under an ambiguous situation such as the one presented to respondents in group 2. Ambiguity tolerance has been found to play a role in explaining organic wine purchase behaviors [
38].
2.2.10. Pro-Social Behavior
We measure pro-social individual behavior by using a set of 27 statements (
Table 6) on social responsibility, empathy, moral reasoning, and self-report altruism (past helpfulness and interpersonal generosity) that participants evaluate using a 7-point Likert scale, following Rapert et al. (2021) [
39]. A cluster analysis is conducted, resulting in 3 groups of respondents. The variable “pro-social behavior” takes value 1 for those with relatively high pro-social behavior; 2 for those with medium pro-social behavior; and 3 for those with relatively low pro-social behavior.
2.2.11. Socio-Demographics
We also collected respondents’ information on their age, gender, highest level of education completed (primary school, secondary school, college qualification (e.g., Diploma), University degree (e.g., BA, BSc, Master’s, PhD, PGCE), and level of income. Regarding the level of income, we asked respondents to indicate their approximate annual household income before taxes by selecting 1 out of 11 income value ranges from less than GBP/EUR 10,000 to GBP/EUR 100,000 or more.
2.3. Conceptualization of the Analysis
Figure 1 illustrates the model variables and how the process was followed to obtain them. We use a set of constructs including views on the benefits associated with sustainable agricultural practices; views on soil quality decline; individuals’ ambiguity tolerance, pro-social behavior, and time and risk preferences; and their trust in governance and agriculture stewardship to investigate the heterogeneity in the WTP for soil quality protection and improvement through payments to farmers to carry out sustainable agricultural practices.
Figure 1 also shows how the constructs have been derived or what information is contained (e.g., socio-demographics). This is applied to both groups of UK and Spanish respondents (those to whom extra background information was provided on uncertainty and ambiguity on the environmental scheme presented to improve soil quality and those to whom this information was not presented).
2.4. Data Analysis: Double-Bounded Dichotomous Choice Contingent Valuation
We use a double-bounded dichotomous choice contingent valuation approach to elicit the general public’s WTP for soil quality improvement through the implementation of policies aiming at farmers to uptake sustainable agricultural practices. The payment vehicle is a monthly tax to be paid in the next 10 years. All participants were first asked whether they would be “willing to pay an additional £/€X in their taxes each month for the next 10 years to protect and improve agricultural soil quality”. The initial bids were GBP/EUR 5, GBP/EUR 15, GBP/EUR 25, GBP/EUR 35, and GBP/EUR 45. These were randomly distributed so that each respondent had the same probability to be shown any of these bids. If the respondent answered “yes” (“no”) to the initial question, then the respondent was asked whether they would be willing to pay a higher (lower) amount. The corresponding higher (and lower) amounts were +GBP/EUR 5 (−GBP/EUR 5) with respect to the original bid. This is GBP/EUR 10, GBP/EUR 20, GBP/EUR 30, GBP/EUR 40, and GBP/EUR 50 for the “yeses” and GBP/EUR 1, GBP/EUR 10, GBP/EUR 20, GBP/EUR 30, and GBP/EUR 40 for the “noes”.
2.5. Data Analysis: Model Estimation
Random Utility Model (RUM) provides the theoretical basis for the double-bound contingent valuation method used. Under RUM, individuals choose alternatives that maximize their utility. We use an interval regression to estimate the WTP model. Our model specification is
where
;
is the WTP (latent variable) of respondent
for a policy program to protect and improve agricultural soil quality;
is a vector of explanatory variables including respondents’ views on the benefits of sustainable agricultural practices (SAP_benefits), soil quality decline views (soil quality concern), respondents’ ambiguity tolerance (ambiguity tolerance), pro-social behavior (pro-social behavior), time preference, risk preference, trust in governance, trust in agriculture stewardship, age, gender, and education level and income;
is a vector of coefficients associated with the explanatory variables to be estimated; and
is the normally distributed error term. The probability of the respondent’s WTP for a policy program to protect and improve agricultural soil quality is
), where
and
are the respondent’s WTP lower and upper bounds;
is the standard normal random variable, and
represents the standard normal cumulative distribution function (cdf) with
and
.
We estimate 4 models: (1) UK respondents—no uncertain and ambiguous background information presented; (2) UK respondents—uncertain and ambiguous background information presented; (3) Spanish respondents—no uncertain and ambiguous background information presented; (4) Spanish respondents—uncertain and ambiguous background information presented.
4. Discussion
We have found that there is significant public support for governments to provide financial assistance to farmers to improve soil quality through the use of sustainable agricultural practices. Our individuals’ average and median WTP estimates are in line with recent research on the WTP estimates for soil security in Italy (EUR 244) and Australia (AUD 567) [
21]. These estimates were obtained using an annual tax to be paid in the next 5 years as vehicle payment. Our vehicle payment was a monthly tax for the next 10 years. We estimated that the median annual WTP for agricultural soil quality protection and improvement is GBP 223 and EUR 293 (GBP 248 at GBP 1 = GBP 0.85 28 June 2024 exchange rate) for the UK and Spain, respectively. The estimates obtained once there was the introduction of uncertainty and ambiguity are GBP 237 and EUR 287 (GBP 243 at GBP 1 = GBP 0.85 28 June 2024 exchange rate) for the UK and Spain, respectively. These estimates allow us to provide an estimate of the capital value of programs supporting farmers financially to conduct sustainable agricultural practices to aim for agricultural soil quality improvement. For the UK, the annual capital value estimate is GBP 15.1–GBP 16.0 billion, whereas the annual capital value of the program for Spain is EUR 13.9–14.2 billion (using median WTP estimates with and without uncertainty and ambiguity and UK population data [
42,
46]). These are significantly higher than the annual GBP 2.4 billion and EUR 14 billion the UK and EU annual budgets allocated to UK agriculture transition plan 2021–2024 and the EU budget to deliver specific environmental benefits for climate, water, soil, air, biodiversity, and animal welfare and to encourage practices that go beyond the conditionality.
We found that providing information about the uncertainty on the overall impact/success of the ELMS scheme/EU Soil Thematic Strategy (i.e., level of provision of public goods) alters the average and median WTP as well as the reasons that individuals have for being (or not) willing to pay for soil quality protection and improvement. We found that introducing information to respondents on the level of uncertainty of the program has an impact on their average WTP, with a 10.5% and 6.3% higher result in the average respondents’ mean and median WTP for UK treatment respondents than UK control respondents and a 2.3% and 2.1% fall in the average respondents’ mean and median WTP for Spanish treatment respondents compared to Spanish control respondents. These opposite results on the effect of introducing uncertainty can be explained by the significant heterogeneity in what motivates respondents’ WTP for soil quality protection and improvement between the two countries. When uncertainty is introduced, the base willingness to pay falls (constant coefficient falls from 12.686 to 6.836), but this is offset by the role of UK respondents’ positive views on sustainable agricultural practices, risk aversion, and time preferences, which become more relevant in their willingness to pay whereas concern for soil quality, ambiguity tolerance, and socio-demographics become less relevant. When estimating the individuals’ willingness to pay, it is UK respondents’ positive views on sustainable agricultural practices, the lesser influence of gender, and the higher influence of the level of trust in stewardship that push the willingness to pay estimate up from the UK control willingness to pay estimate. On the other hand, for Spanish respondents, when uncertainty is introduced, the base willingness to pay falls heavily (constant coefficient falls from 9.887 to −2.034). This is not fully offset by factors accounted for in the model, resulting in the willingness to pay estimates being slightly down compared to the control group. For Spanish respondents, what explains individuals’ WTP moves away from aspects associated with individuals’ level of concern for soil quality, their level of pro-social behavior, and their age towards individuals’ appreciation of SAP benefits and their level of income.
We found that the more tolerant to ambiguity, the less individuals are willing to pay for soil quality protection and improvement. This may indicate that the less ambiguity-tolerant an individual is, the more the individual is keen to remove such ambiguity and therefore would be willing to pay for it.
Under uncertainty, relatively more risk-averse individuals in the UK are less willing to pay than less risk-averse individuals (i.e., relatively more risk-averse individuals value less a program aiming at protecting and improving soil quality when they are informed about the uncertainty of the success of the policy). This effect was not found for Spanish respondents. Results indicating a negative relationship between risk aversion and willingness to pay have been reported in different studies (e.g., for improved agricultural technologies [
47]; energy-efficient insulation and ventilation systems in rental apartments [
48]; insurance contracts [
49]; and functional food [
50]).
Regarding potential limitations of this study, it is worth mentioning that, although widely used, there have been contentious discussions regarding the multiple discrete choice (DCm) approach reliability [
51]; its internal consistency [
52]; and hypothetical bias [
53,
54,
55]. The DCm approach may suffer from hypothetical bias, i.e., a deviation from real market evidence, if the study was conducted in hypothetical situations with no consumption consequences for the participants [
56]. However, on the other hand, Ref. [
57] suggested double dichotomous choice as an efficient approach to obtaining contingent valuation estimates. Carson et al. (2003) [
58] indicated that a careful contingent valuation survey design and development could reduce the concern of biased estimates of DCm, especially when a large-scale survey was employed rather than a small-size experiment. The above literature highlights the still inconclusive justification of using DCm for measuring WTP, and motivates the present study.
This study has developed a conceptual framework that can be altered to answer other types of questions compared to the ones presented here. We recommend that future valuation research incorporates elements of uncertainty and ambiguity into the analysis.
5. Conclusions
There is significant public support for agricultural soil quality protection and improvement in both Spain and the UK. Individuals’ reasons for supporting and being willing to pay for a soil protection program change with increased uncertainty and ambiguity about the program’s impact and results. These reasons may also vary across countries. These findings align with previous literature, which highlights significant spatial variation in responses to policy and biodiversity impacts. [
7].
Results suggest that as the general public, being supportive of public policies, becomes more aware of the level of uncertainty around the success of achieving policy objectives, they can remain supportive of those public policies but their reasons for supporting them may change. Hence, under an increased public awareness on the level of uncertainty about a policy/program outcomes, actions to try to gain/keep public support should differ from actions taken to gain/keep support where public awareness on the level of uncertainty is low. It is also important to acknowledge that since the individual’s reasons to support a public policy may change once the individual becomes more aware of the uncertainty around achieving the policy objectives and that these may vary across geographical locations, any action to gain/keep public support needs to be tailored to the relevant geographical area (e.g., country, region). For instance, for the case of promoting public support for soil protection, illustrating the benefits of SAP is important to gain/keep support in both countries studied, regardless of the level of policy outcome uncertainty. However, using other approaches such as promoting public support for soil protection through messages focusing on soil quality concerns may be less effective, when the public is aware of the uncertainty around the success of soil protection objectives, especially in Spain. Similarly, messages promoting pro-social behavior to gain/keep policy support for agricultural soil quality protection and improvement are less likely to be effective when the public is aware of the uncertainty of the success of these policies.