Impact of Food Sustainability Labels on the Perceived Product Value and Price Expectations of Urban Consumers
Round 1
Reviewer 1 Report
Thank you for allowing me to review this paper titled “Impact of food sustainability labels on the perceived product value and price expectations of urban consumers”. The topic is undoubtedly relevant, despite widely covered in the literature, but it is not clear the novelty of this study and which research gaps the authors try to cover to advance the knowledge in this research field.
In particular, the authors should explain clearly the research question(s) that lead this study and the research hypotheses, thus helping readers to understand the study. Authors declared that the study aims to provide insights into the impact of logos also in the consumer’s willingness to pay using a choice experiment, but I have some doubts.
As described in pages 5-6, rows 188-195 (only eight rows?), the experiment seems a sort of contingent evaluation, or a complex logit choice model the have to be well explained in his experimental design, theoretical foundation, model specifications, and data analysis, rather than a Choice Experiment. Moreover, why the respondent has to choose only one product? Thus, this implies that behind there is an experimental design; but what is it? Authors must explain these aspects.
Instead, if the authors had done a choice experiment, the main concerns are about the lack of:
a clear theoretical framework referring to the use of the choice experiment that can offer the readers the possibility to understand the novelty of the study, or how this study can contribute to improving the research knowledge in this field of studies; a clear section describing the whole choice experiment (to now limited to some rows, please see rows 172-196).In particular, referring to this last aspect, I think it should be of relevance to describe clearly:
general issues regarding the CE; attributes and their levels, showing an example of the choice set; the experimental design; the theoretical foundations of the choice experiment; the model to account for consumers’ heterogeneous preferences; eventually, the model with or/and without interactions with the socio-economic aspects that authors considered as results of the PCA.Also, are the sample representativeness?
Moreover, is it not clear which information the authors considered for the cluster analysis; are the scores (or z-score) that result from PCA?
Minor concerns are about results and formal aspect of reporting findings but in the light of what aforementioned are of secondary relevance. For example:
the description of cluster 1 and 2 need a table in which is reported the statistics and the p-values; Table 6 reports some errors (e.g., among others, for EU organic farming (Euro-leaf) – Banana, N=93 instead of 94, Total-10%=65(-30%) instead of 66(-29%) ); Authors must carefully consider the correct use of decimals (e.g., page 6, row 215, 60.8 instead of 60.789; page 6, Table 3 3rd row, “1st (22.6)” instead of “1st (22.604)”; pag 8, row 226, “p-value < 0.000000” instead of “p-value < 0.000”; page 8, row 228, “F value 317.13” instead of “F value 317.1273”, …)I am sorry not to be able to recommend accepting this paper, but the concerns about the experimental design referred to the choice experiment mean that I cannot recommend it.
Author Response
Reply to Reviewer 1
Dear Reviewer,
Thank you for your time in reviewing our manuscript entitled “Impact of food sustainability labels on the perceived product value and price expectations of urban consumers”. We have made every effort to respond to your concerns in satisfactory manner. We have introduced some of your suggested changes in the manuscript.
Below we give a point-by-point reply to comments (in italics) presented in the Reviewer’s report.
Comment: The topic is undoubtedly relevant, despite widely covered in the literature, but it is not clear the novelty of this study and which research gaps the authors try to cover to advance the knowledge in this research field.
Response: We agree that the topic is quite widely covered in the literature however the question „why do consumers pay so little attention to B2C certificates on food products” is still to be answered and also changes in time. This is an important issue in modern food systems and abundant offers, especially in Central and Eastern Europe, where Poland is the biggest market. This is due to several factors:
The level of food expenditure is much higher than in Western countries (amounting to 26% of total households’ expenditures) There is a false opinion/stereotype among consumers that Polish food is “ecological” therefore healthy because farmers use much less chemicals in the production of commodities People are becoming more aware about sustainability issues so the role of such certificates will grow.At this point there are very few WTB and WTP studies in our region so this is novel.
If any comprehensive information activities (campaigns) on sustainability labels would be carried out (what is needed for consumers to understand that they have to pay more for higher quality and its assurance) the willingness to pay for certified products will increase. Then we can use our finding to assess the change.
Comment: In particular, the authors should explain clearly the research question(s) that lead this study and the research hypotheses, thus helping readers to understand the study. Authors declared that the study aims to provide insights into the impact of logos also in the consumer’s willingness to pay using a choice experiment, but I have some doubts.
Response: Thank you for this suggestion. In order to further explain the research questions and help the Reader understand the study, two research hypothesis were included to the manuscript:
1/ Consumers (even city dwellers) are dominated by those who are sceptical about sustainability certificates and this applies especially to men.
2/ Consumers are willing to buy certified food products, but can only pay a little more for them compared to the prices of their conventional counterparts.
Comment: “As described in pages 5-6, rows 188-195 (only eight rows?), the experiment seems a sort of contingent evaluation, or a complex logit choice model that have to be well explained in his experimental design, theoretical foundation, model specifications, and data analysis, rather than a Choice Experiment”
Response: Thank you for this comment. The experimental design to the choice experiment can’t be changed at this point but we will certainly consider your suggestions in our next studies in this area. However, we would like to explain our position. The contingent valuation technique (CVM) (that belonging to stated preference method family) is the primary economic tool for estimating the values of non- material goods, for example estimating model for travel demand, to assess the non-market values of forest. The Behavioural Choice Experiment - CE (that attempts to model the decision process of an individual or segment via revealed preferences or stated preferences) approximates real life situation (in our example online shopping) and asks people to directly report their willingness to pay (WTP) to obtain a specified good (in our study plant food). In our experiment we wanted to observe consumer choice behaviour (revealed preferences). Similar construction of behavioral choice experiment was conduct in European study IPSOS - London Economics EAHC/FWC/2012 86 04 . In our study we use the following attributes and level sets:
Two different product categories (Table 1, page 3): fresh (apples- Polish regional origin, bananas – imported) , non-perishable (beans – Polish regional origin , rice – imported) Three types of food sustainability labels (EU organic farming - Euro-leaf , Protected Geographical Indication –PGI and Fair Trade) which are popular on food products in various grocery shops offer. This is mentioned in lines 113-124. Price combinations – that reflect the real market value and possible price level changes . This is explained in lines 188 -196 : “At the beginning of the experiment, the same base price (adequate to real market values in PLN: apples 2.5 per 1kg, bananas 3.99 per 1kg, beans 4.85 - packaging 0.5 kg, rice 3.99 - packaging 0,5 kg) was shown for each product. Participants were asked to choose the product they would buy while shopping in a store (measurement of WTB). If the respondent chose a labelled product, then in the next choice the price of this item increased by 10%. Then the respondent could maintain the choice of the previously selected product at a higher price or choose another (with or without a different sustainability label), which was kept at lower level (WTP measurement). The price increase was carried out until the respondent changed his or her preferences and chose the product without a logo.”The CE (as well as CVM) share a common theoretical framework - Random Utility Theory postulating that an individual who makes a choice among different alternatives strives to maximize utility [1;2]. Under the random utility model (RUM) framework, the indirect utility function for each respondent can be expressed as:
Uij = Vij + εij, where Uij is person i’s utility of choosing alternative j, Vij is the deterministic component of utility and εij is a stochastic element that represents unobservable influences on individual choice.
The probability that any particular respondent prefers option j in the choice set to any alternative option k, can be expressed as the probability that the utility associated with option j exceeds that associated with all other options. Formally,
Pij= P{Vij+ εij, > Vik+ εik;∀k ∈ C} , where C is the set of all possible alternatives. Assuming a Type I Extreme Value distribution for the error terms, the probability of choosing alternative j is:
Pij = e ωVij / ∑ e ωVik
k∈C
Where ω is a scale parameter, inversely proportional to the standard deviation of the error distribution, and typically assumed to be one.
Bibliography
1.Thurstone, L. A law of comparative judgment. Psychological Review , 1927, 34, 273-286.
McFadden, D.L. Conditional logit analysis of qualitative choice behaviour. (In) P. Zarembka (Eds.). Frontiers in Econometrics, New York Academic Press, 1973, pp. 105-142.
Comment: Moreover, why the respondent has to choose only one product? Thus, this implies that behind there is an experimental design; but what is it? Authors must explain these aspects.
Response: Participants in our research were involved in a virtual online grocery shopping exercise that imitate real life situation. We assumed that consumer in typical shopping situation usually choose between several alternatives (for example bananas with no logo, bananas with Fair Trade Logo, and bananas with Euro-leaf logo) and choose only one product (for example item with Euro-leaf logo) and buy it at a particular price.
Comment: Instead, if the authors had done a choice experiment, the main concerns are about the lack of: a clear theoretical framework referring to the use of the choice experiment that can offer the readers the possibility to understand the novelty of the study, or how this study can contribute to improving the research knowledge in this field of studies; a clear section describing the whole choice experiment (to now limited to some rows, please see rows 172-196).
Response: Thank you for this suggestions. We partially presented the theoretical framework of our research in rows 172-177. To fully clarify the study theoretical foundation and data analysis we have expanded part 2.4. (Measures) by explanation:
“In the data analysis we focus only on the weighted average of willingness to pay for labelled products (as the percent of extra amount consumers would pay for a product carrying a particular label as opposed to a product not carrying label). The weighted average of willingness to pay was calculated and presented as a percentage of the base price.”
Comment: Also, are the sample representativeness?
Response: The sample is not representative for the Polish population and we realize that it may be a limitation of our study. We decided that the area of research should be one region in order to avoid differences in the level of socio-economic development of the cities as well as the standard of living of their residents. Respondents were selected from adults aged 19–65 years living in big cities with more than 50 000 inhabitants, bearing in mind that: (i) the concept of sustainable consumption is known so far to just a small part of society; (ii) the population of city inhabitants usually represents higher level of education, achieves higher income, etc. and therefore is more open to new ideas; and (iii) new consumer trends, also in the field of food and nutrition, spread from the inhabitants of large cities to the rural population, for which they constitute a certain model of future food consumption and behaviour. As it was mentioned in L. 137, the Commercial Research Agency ARC Rynek i Opinia selected the respondents from its ePanel.pl, which is an internet-based consumer panel, the oldest and one of the largest research panels in Poland. It currently gathers more than 70 thousand users from all over Poland.
Comment: Moreover, is it not clear which information the authors considered for the cluster analysis; are the scores (or z-score) that result from PCA?
Response: For cluster analysis calculations we have used non standardized variables (the scores).
Comment: Minor concerns are about results and formal aspect of reporting findings but in the light of what aforementioned are of secondary relevance. For example: the description of cluster 1 and 2 need a table in which is reported the statistics and the p-values;”
Response: Thank you for this suggestion. We decided not to introduce another table to the manuscript because in section 3.1. “Description of consumer clusters” we described very detailed differences between cluster 1 and cluster 2. We would like to avoid a situation where consumer segmentation dominates the main part of our manuscript.
Comment: Table 6 reports some errors (e.g., among others, for EU organic farming (Euro-leaf) – Banana, N=93 instead of 94, Total-10%=65(-30%) instead of 66(-29%).
Response: This comment prompted us to check our data again. This was rounding error. We have introduced the suggested changes in Table 6 of the manuscript.
Comment: Authors must carefully consider the correct use of decimals (e.g., page 6, row 215, 60.8 instead of 60.789; page 6, Table 3 3rd row, “1st (22.6)” instead of “1st (22.604)”; pag 8, row 226, “p-value < 0.000000” instead of “p-value < 0.000”; page 8, row 228, “F value 317.13” instead of “F value 317.1273”, …)
Response: In the section 2.5. “Statistical analyses” we have introduced the suggested rounding of the respective numerical values.
Thank you once more for your time and suggestions and we believe that our manuscript gained after your revision.
Kind regards,
Authors
Author Response File: 
Author Response.docx
Reviewer 2 Report
An interesting and relevant study. The project was well researched. While it is correct that plant foods are generally kinder to the environment, I wonder at the assumption that plant products, including bananas, encourage safer dietary choices since these foods have traditionally been subjected to heavy usage of pesticides. The pre-test questionnaire was a useful unbiased tool to determine participants' buying behavior.
The layout of Table 5 may present some confusion when trying to understand the Total column. I suggest either an asterisk to explain totals or address this in the results.
Author Response
Reply to Reviewer 2
Dear Reviewer,
We kindly thank you for your review of our manuscript entitled “Impact of food sustainability labels on the perceived product value and price expectations of urban consumers” and your valuable comments that helped us improve it. In response, we have made corrections which we hope will meet with your approval.
Below we give a point-by-point reply to comments (in italics) presented in the Reviewer’s report.
Comment: ”While it is correct that plant foods are generally kinder to the environment, I wonder at the assumption that plant products, including bananas, encourage safer dietary choices since these foods have traditionally been subjected to heavy usage of pesticides”.
Response: Bananas are currently the second most popular and consumed fruit in Poland (the first is apples). That is why we decided that bananas will be a good example of imported fruit available in Poland. Bananas are also bought more often because they are perishable food and cannot be bought in stock, as in the case of apples. Stickers with certificate symbols are clearly visible on the yellow fruit skin. We expected that some respondents would be interested in the message of these symbols and would make informed choices. We also expected that consumers concerns about potential heavy usage of pesticides at banana plantations should be reflected in confidence in sustainability labels (EU organic farming and Fair Trade) and in the choice of such labelled products.
Comment: ”The layout of Table 5 may present some confusion when trying to understand the Total column. I suggest either an asterisk to explain totals or address this in the results.”
Response: Thank you for this suggestion. We followed it by adding a brief description of the “Total” column: ** “Percent of all respondents who declared willingness to buy a particular product”
Thank you once more for your time and suggestions and we believe that our manuscript gained after your revision.
Kind regards,
Authors
Reviewer 3 Report
The design assumptions are not clear. Why most of the surveys were conducted in the capital of the region selected for research, instead of evenly dividing the surveys carried out between the 7 selected cities. Did such a choice not affect the results of the research, due to the greater education and awareness of the respondents in a large urban agglomeration. Please better justify this choice.
The designation used in the work for the names 'protected organic farming' in the EU, 'Geographical indication' and the 'fair trade' logo should be placed between quotation marks.
Literature at work should be better used.
The conclusions should be less detailed, and more summative and generalizing.
Author Response
Reply to Reviewer 3
Dear Reviewer,
We sincerely appreciate your review of our manuscript entitled “Impact of food sustainability labels on the perceived product value and price expectations of urban consumers”. We tried our best to improve the manuscript accordingly to your comments and we have hope that after the changes it meet with your approval.
Below, we respond to the specific comments (in italics) that were presented in the Reviewer’s report.
Comment: “The design assumptions are not clear. Why most of the surveys were conducted in the capital of the region selected for research, instead of evenly dividing the surveys carried out between the 7 selected cities (...) Please better justify this choice. ”
Response: That's a good point, thank you very much. In fact, we should write more precisely why there was such a large disproportion in the number of participants in the survey from Warsaw and the other six cities. We have completed the description of the sample selection as follows:
In line 129 page 3 we have added an explanation: “Mazovia is the best developed region in the country in terms of gross domestic product and growth, unemployment rate, personal income and education level. In this region there are four cities with 50 000 to 100 000 people; two cities with 100 000 to 500 000 people; and Warsaw, the capital and the largest city in the country, with 1.7 million inhabitants.”
In line 136 page 4 we have added: “Therefore the sample consisted of 423 people where 250 CAWI were completed by residents of Warsaw, by far the biggest city in the region. The remaining questionnaires were filled out by inhabitants of the other six Mazovian cities with 50 000+ inhabitants - Ostrołęka, Legionowo, Pruszków, Siedlce, Płock and Radom.
Comment: “(…) Did such a choice (i.e. participants number in the groups in quota sampling) not affect the results of the research, due to the greater education and awareness of the respondents in a large urban agglomeration. Please better justify this choice.”
Response: We agree, that “such a choice affects the results of the research”. However, we have intentionally decided to conduct research in large cities of the best developed region of central Poland because of very low awareness of the issue of sustainable development in the Polish population. That's why we wrote in the manuscript in line 133-135 page 4: “The sample choice was based on the assumption that residents of big cities tend to have better education, higher incomes and better access to products’ offer and therefore may be more aware of sustainability issues and more familiar with sustainable food labelling”.
Comment: ”The designation used in the work for the names 'protected organic farming' in the EU, 'Geographical indication' and the 'fair trade' logo should be placed between quotation marks.”
Response: The rules of writing indicate the use of the quotation mark for proper names. However, the names of certification/quality assurance schemes are already popular and recognizable. They do not really designate unique certificates but rather categories of certificates, and therefore are not true proper names. In EU studies and articles, some authors use quotation marks, others do not. We decided not to introduce them because there are many quotation marks in the text anyway: for names of clusters, questions from questionnaire, apple and bean varieties.
Comment: “Literature at work should be better used.”
Response: To better use literature at work we added some changes in line 464 -469 page 14:
“Existing literature on willingness to pay for food products affiliated with sustainability labels indicates that there is a highly price inelastic demand for them [10,44]. Large groups of consumers show preferences towards environmentally friendly products (with organic farming, animal welfare, carbon footprint, food miles, local production, geographical indications and traditional specialties, Fair Trade and etc. logos) but only if their prices are low or at least the same as that of conventional items [49,50,51]”.
Comment: ”The conclusions should be less detailed, and more summative and generalizing.”
Response: Thank you for this comment. We have changed the conclusions as follows:
“Our study shows that urban consumers in Poland vary in regard to the value they attach to sustainability labelled food products, however 1/3 of them is more “mindful” and has higher WTB and WTP for such foods.
We conclude that sustainability labels influence consumer buying behaviour in spite of the fact that consumers perceive the benefits of buying them differently. Consumer organisations, government agencies and food industry organisations should take more action to raise awareness of the idea of food certification and the importance of individual certificates to promote the choice of certified food products more intensively.
Finally our findings confirm that Polish urban consumers are very price sensitive in terms of buying products with sustainability logos. Therefore the value added of sustainable products should be communicated very clearly by producers in order to enable consumers to make informed choices in line with sustainable food consumption recommendations. There is a need to tackle the crucial role of price that can inhibit the growth of demand among consumers. Education in this area should draw attention to making consumers aware that prices of certified products must be higher than those of conventional products, and that surplus prices means particular benefits to them. Otherwise a large consumer group is excluded from sustainable consumption. “
Once again, thank you very much for your comments and suggestions. We are very grateful for your effort in improving our manuscript.
Kind regards,
Authors
Round 2
Reviewer 1 Report
Recommendation: reject
Comments:
The authors have made a good start addressing the comments I raised, but more is needed.
Much of the discussion in response to reviewers should be explicitly integrated into the paper, in some cases justifying the choice, for example:
Authors’ response 1 (to better define what is/are the research gap that authors have tried to covered). Authors’ response 6 Authors’ response 7My main concern is about the Choice Experiment. I am aware that the choice experiment cannot be changed at this point, of course. Nevertheless, as already highlighted, to improve the quality of the paper, Authors should better describe the whole choice experiment in a separate section. Here, the Reader should be found the following information in separate subsections that can help him to understand and eventually replicate the study, for example:
The choice of attributes and their levels, and an example of a choice set; The description of the experimental design, where it can be found informations referring to the choosing of the most relevant choice sets used in the experimental design techniques and others related information (e.g. page 6 rows 193-195); The questionnaire, data collection (see page 4, rows 132-153), and descriptive statistics (see page 4, rows 154-171; Theoretical foundations of the choice experiment and the model specification related to the research conducted; The model adopted to analyse the data (e.g. RPL?, … )What is more, why Authors did not consider to evaluate the model with and without considering socioeconomic aspects founded by the PCA?
Below some comments to Authors’ responses:
I would highlight that rows 172-177 refers only to data collection (Authors’ response: “Thank you for this suggestions. We partially presented the theoretical framework of our research in rows 172-177.”). A table could offer more space to describe better the central part of the manuscript avoiding Readers getting bored (Authors’ response: “Thank you for this suggestion. We decided not to introduce another table to the manuscript because in section 3.1. “Description of consumer clusters” we described very detailed differences between cluster 1 and cluster 2. We would like to avoid a situation where consumer segmentation dominates the main part of our manuscript.”).Unfortunately, Authors’ did not pay sufficient attention. Please see: sub-group “BEANS - EU organic farming (Euro-leaf)” x “Base (% of group), the total should be 90 instead of 89; sub-group “RICE - EU organic farming (Euro-leaf)” x “+20%”, the total should be 41 instead of 42; sub-group “BANANAS – Fair Trade” x “+10%), the total should be 64 instead of 65 (Authors’ response: “This comment prompted us to check our data again. This was rounding error. We have introduced the suggested changes in Table 6 of the manuscript.”).
Unfortunately, Authors’ did not pay sufficient attention to the punctuation along with the test (e.g. Pag 6, row 230: “60.798 %” … one decimal is acceptable) and other aspects as the correct use of for comma instead of the dot (e.g., page 3, row 134; “over 50.000” instead of “over 50,000), and reference errors (e.g. page 14, row 490). (Authors’ response: In the section 2.5. “Statistical analyses” we have introduced the suggested rounding of the respective numerical values.)
In light of the criticisms mentioned above, I am not able to recommend this manuscript for publication.
