Willingness to Pay for Agricultural Soil Quality Protection and Improvement

Round 1

Reviewer 1 Report

This study estimates the general public’s willingness to pay (WTP) for agricultural soil quality protection and improvement in Spain and the UK. Overall, it presents an interesting narrative. However, there are areas that require improvement.

1. The study should clarify the representativeness of the sample and describe quality control measures taken during data collection.

2. To enhance clarity, consider using asterisks (*) to denote statistical significance in Table 7.

3. It is essential to directly describe the distribution of WTP values.

4. The study should provide a detailed explanation for why the control and treatment results in the UK and Spain are opposite in Table 9.

5. Consider merging Tables 7 and 8 for better presentation.

6. Review and ensure accuracy of references 1, 3, 16, and 17.

Author Response

Reviewer Comment: This study estimates the general public’s willingness to pay (WTP) for agricultural soil quality protection and improvement in Spain and the UK. Overall, it presents an interesting narrative. However, there are areas that require improvement.

1. The study should clarify the representativeness of the sample and describe quality control measures taken during data collection.

We thank the reviewer for this comment. We agree that a clarification on the representativeness of the sample and quality control measures taken during data collection.

We have now added the following text to the manuscript to clarify these points under section 3.1:

 “Regarding the representativeness of the sample, the age, gender, and income level statistics align closely with those reported by the UK Office for National Statistics (ONS) and the Spanish Instituto Nacional de Estadística (INS). The educational categories used in this study are not directly comparable to those used by the ONS or the INS. However, some comparisons are still possible”

 The UK government Office for National Statistics (ONS) reports median gross annual earnings for full-time employees was £34,963 in April 2023 which is similar with our sample median of £35,000 [1]. In the year to mid-2022, the median age of the UK population was 40.7 years, which is below to the median sample age of 46.9 and 45.9 years (control and treatment) for the UK [2].

For education, the ONS provides information on the highest level of education for England and Wales. On education level, the ONS indicates that 18.2% of the population in England and Wales in 2021 had no qualifications;  23.0% of have achieved Level 1/Level 2 education [3]. Our sample does not have respondents with no qualifications, 27.8% of respondents have completed primary and secondary education; 23.9% of respondents have a College degree group which can be comparable in some cases to an ONS apprenticeship and level 3 classifications, 5.3% and 16.9%, respectively (22.2%).  Level 4 or above is comparable with our classification of University/Post graduate degree. The ONS statistic of level 4 is 33.8% whereas the sample includes University/Post graduate degree 48.3% indicating that our sample may be overrepresenting higher educated population in the UK.

 The Spanish Instituto Nacional de Estadística, INE, (National Statistical Institute) reports that  the mean gross annual earnings was £26,949 in 2022 which is similar with our sample median of £29,437 and £29,218 for the control and treatment groups, respectively [4]. The level of education of the Spanish population in 2021 was: 36.1% primary education, 23.2% secondary education and post-secondary non superior (e.g. Bachelor), superior education (e.g. university degree, PhD) 40.7% [5]. This compares to our sample as follows: primary and secondary education, and professional of 56.8% and 53.8 (for control and treatment groups, respectively, which is similar to 36.1%+23.2%=59.3%); and 43.3% and 46.2% of respondents achieving superior education. The average age of Spanish population in 2022 was 44.06 years [6], which is comparable to the sample average which is 43.0 and 42.7 years (control and treatment samples).  

Hence, while our sample adequately reflects the demographics of both the UK and Spanish populations in terms of age, gender, and income levels, it underrepresents people with no qualifications and overrepresents people with higher education in the UK. Therefore, while the demographics match closely, some caution should be exercised in generalizing our findings to the entire UK and Spanish populations. However, although it may not be possible to claim that the sample is fully representative of the entire UK and Spanish populations, it still provides valuable insights into the views and attitudes associated with willingness to pay for soil protection and improvement that, while not universally applicable, may still be indicative of broader views and attitudes within the population as well as a reasonable approximation to the average willingness to pay for soil protection and improving, contributing to the body of knowledge in this area and inform future policy.”

In this regard we also added the following text in lines (552-554): “This result suggests that the impact of overrepresentation of higher educated respondents in the sample is likely limited in terms of influencing the specific findings related to willingness to pay.”

Regarding quality control, we have added the following text to the manuscript (lines 158-168): “It is worth noting that we conducted pre-tests and pilot it to refine questions, identify potential issues, and ensure 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 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 to all questions of the questionnaire and took them more than 4 minutes to respond. The median time to respond all questions was 18 minutes.”

 We also added the following text to clarify that we obtained ethical approval to conduct the survey (lines 145-147): “… for which ethical approval according the procedures specified by Newcastle University Research Ethics Committee”

2. To enhance clarity, consider using asterisks (*) to denote statistical significance in Table 7.

Thank you, we have now used asterisks to denote statistical significance in all relevant tables.

3. It is essential to directly describe the distribution of WTP values.

Thank you for raising this point. We agree that this is an important point. We have not presented the individual’s WTP distribution for control and treatment groups for UK and Spanish respondents. We added the following text to introduce the figure (573-580): “Figure 2 shows the willingness to pay distribution for control and treatment groups for  UK and Spanish respondents. Although the control and treatment monthly average willingness to pay results for UK and Spain appear are opposite the willingness to pay distribution indicates that such differences are not significant at the willingness to pay distributional level.”

4. The study should provide a detailed explanation for why the control and treatment results in the UK and Spain are opposite in Table 9.

Thank you for this comment. The reviewer has raised a very important point here. We have edited a relevant paragraph under the discussion section to provide a more thorough explanation of the results (lines 611-637): “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 an 10.5% and 6.3% higher 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 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 higher influence of the level of trust in stewardship what pushes 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 in the model resulting on the willingness to pay estimates being slightly down compared to the control group. For Spanish respondents’ what explains individual’s WTP moves away from aspects associated with individual’s level of concern for soil quality, their level of pro-social behavior and age towards individual’s appreciation of SAP benefits and their level of income.”

 5. Consider merging Tables 7 and 8 for better presentation.

Thank you for your suggestion to merge Tables 7 and 8 for better presentation. We have carefully considered this option. However, we believe that keeping the tables separate allows for a clearer and more focused presentation of the distinct UK and Spain results. We think merging them could potentially complicate the explanation and interpretation of these results, which we preferred to keep them separately.

 6. Review and ensure accuracy of references 1, 3, 16, and 17.

Thank you, we have reviewed this references.

Finally, please note that during the paper revision we realised that the education categories names were incorrectly displayed in the Tables. These have been amended now. In addition, we also realised that a subsection was needed to briefly explain the socio-demographic data collected. A new section has been created for this purpose, section 2.2.11.

1. ONS. Employee earnings in the UK: 2023. Employee earnings in the UK: 2023 2023 [cited 2024 17/07/2024]; Available from: https://www.ons.gov.uk/employmentandlabourmarket/peopleinwork/earningsandworkinghours/bulletins/annualsurveyofhoursandearnings/latest.
2. ONS. Population estimates for the UK, England, Wales, Scotland, and Northern Ireland: mid-2022. 2024 17/07/2024]; Available from: https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationestimates/bulletins/annualmidyearpopulationestimates/mid2022.
3. ONS, Education, England and Wales: Census 2021. 2023, Office for National Statistics.
4. INE, Encuesta de Estructura Salarial (EES), I.N.d. Estadística, Editor. 2022.
5. INE. Nivel de formación alcanzado por la población adulta. 2022 17/07/2024]; Available from: https://www.ine.es/ss/Satellite?L=es_ES&c=INESeccion_C&cid=1259944520019&p=1254735110672&pagename=ProductosYServicios%2FPYSLayout&param1=PYSDetalleFichaIndicador&param3=1259937499084#:~:text=En%20el%20conjunto%20de%20la,%25%20el%20nivel%205%2D8.
6. INE. Indicadores de Estructura de la Población. 2023 17/07/2024]; Available from: https://www.ine.es/jaxiT3/Datos.htm?t=3198.

Reviewer 2 Report

The topic addressed is very original and interesting; also for a greater understanding of the achievable results with respect to environmental policy objectives.  

The proposed analysis comes to an uneasy synthesis and expresses the assessments with appropriate methodology.

In describing some applications conducted in Italy on similar topics, using stated preference methods could also point to other articles (e.g. around line 115, I suggest: Estimating preferences for controlling beach erosion in Sicily, De Salvo, M., Signorello, G., Cucuzza, G., Begalli, D., Agnoli, L., Aestimum, 2018, 72, pp. 27–38 DOI: 10.13128/Aestimum-23739).

The results are useful for their implications for the public decision maker who can act more effectively and consciously on the agricultural policy decisions that need to be made in a context of high uncertainty

From a formal point of view, I would like to point out the following:

Line 151, reduce spacing between words;

lines 286-298 the text is not justified;

lines 307-319 the text is not justified;

line 383 the second ' is missing

line 392, (RUM) is for Random Utility Model not Theory

lines 423-428 the text is not justified

lines 433-464 the text is not justified

line 490 maybe e is a typo

Author Response

Reviewer Comment: The topic addressed is very original and interesting; also for a greater understanding of the achievable results with respect to environmental policy objectives. 

The proposed analysis comes to an uneasy synthesis and expresses the assessments with appropriate methodology.

We thank the reviewer for the positive comment.

In describing some applications conducted in Italy on similar topics, using stated preference methods could also point to other articles (e.g. around line 115, I suggest: Estimating preferences for controlling beach erosion in Sicily, De Salvo, M., Signorello, G., Cucuzza, G., Begalli, D., Agnoli, L., Aestimum, 2018, 72, pp. 27–38 DOI: 10.13128/Aestimum-23739).

We thank the reviewer for pointing out this reference. We have now added a comment on the use of stated preference methods on studying the public support for beach soil erosion: “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”

The results are useful for their implications for the public decision maker who can act more effectively and consciously on the agricultural policy decisions that need to be made in a context of high uncertainty

We thank the reviewer for the positive comment.

From a formal point of view, I would like to point out the following:

Line 151, reduce spacing between words;

Thank you, this has now been amended.

lines 286-298 the text is not justified;

Thank you, this has now been amended.

lines 307-319 the text is not justified;

Thank you, this has now been amended.

line 383 the second ' is missing

Thank you, this has now been amended.

line 392, (RUM) is for Random Utility Model not Theory

Thank you, this has now been amended.

lines 423-428 the text is not justified

Thank you, this has now been amended.

lines 433-464 the text is not justified

Thank you, this has now been amended.

line 490 maybe e is a typo

Thank you, yes, we wanted to remind the reader this is referring to the treatment group. We have now amended this to make it clear. (line 556)

Round 2

Reviewer 1 Report

All comments have been addressed and revised appropriately.