Consumers’ willingness to pay for ethical consumption initiatives on e-commerce platforms

WANG Er-peng,AN Ning,GENG Xian-hui,Zhifeng GAO,Emmanuel KIPROP

1 School of Economics and Management,Nanjing Tech University,Nanjing 211816,P.R.China 2 College of Economics and Management,Nanjing Agricultural University,Nanjing 210095,P.R.China 3 Food and Resource Economics Department,University of Florida,Gainesville,FL 32611-0240,USA

Abstract Despite China’s fast-growing e-commerce and its great achievement in promoting poverty alleviation through consumption,little is known about Chinese consumers’ online ethical consumption. Using the payment card elicitation method,this paper designs a within-subject survey and a between-subject survey to investigate Chinese consumers’ quality perception and preference for apples from poverty-stricken areas. The results show that before “information shock”,emphasizing that taste and safety attributes of apples from poverty-stricken areas are the same as the conventional ones,Chinese consumers on average are willing to pay a 31% premium for apples from poverty-stricken areas. After “information shock”,both the within-subject and between-subject designs show a minimal drop of the premium,implying that the ethical attribute is the main motivation for buying apples from poverty-stricken areas. The regression results show that quality perception of private attributes has significant effect on consumers’ willingness to pay (WTP) for apples from poverty-stricken areas,and trust in government supervision of e-commerce plays an essential role in motivating online ethical consumption.

Keywords:ethical consumption,apples from poverty-stricken areas,WTP,interval regression

1.Introduction

Ethical consumption is one kind of purchasing behavior for consumers to express their feelings toward society (Pintoet al.2020). Through buying products for their positive qualities (e.g.,green products) or boycotting products for their negative qualities (e.g.,products made by child labor),consumers can translate their consumption behavior to ethical concerns about the society (De Pelsmackeret al.2005). As a typical ethical consumption behavior,fair trade buying is one of the most critical issues of ethical concerns in consumer behavior (Shaw and Clarke 1999). Fair trade means buying products from farmers in developing countries,on terms that are relatively favorable than commercial terms,and marketing them in developed countries at an ethical premium (Bird and Hughes 1997;Yanget al.2014; Zanderet al.2018). As to the most critical issue of ethical concern among consumers (Shaw and Clarke 1999),research has found that consumers in many developed countries are willing to pay a premium for Fair Trade coffee (De Pelsmackeret al.2005; Arnotet al.2006;Basu and Hicks 2008). Other studies have found that ethical consumers tend to have a relatively high level of income,education and social status (Roberts 1996; Balmeret al.2001; Carrigan and Attalla 2001). Researchers have also found that consumers’ perceptions and attitudes influence their ethical consumption behavior (Ferrell and Gresham 1985; Shaw and Clarke 1999; Vitellet al.2001; Hunt and Vitell 2006). While much research on ethical consumption has been conducted in developed countries (Zanderet al.2018; Lappemanet al.2019; Hindsleyet al.2020),little is known about ethical consumption in China.

In fact,China has launched an ethical consumption initiative (poverty alleviation through consumption) since 2019,which has mobilized the public to buy products and services from poverty-stricken areas,in order to help poor people increase their income and fight against poverty.Among all the forces,e-commerce has played a very important role in promoting the development of ethical consumption. For example,29 e-commerce enterprises have set up special columns on their platforms with labels such as “Fupin Guan” and “Xingnong Tuopin”,two poverty alleviation initiatives by Jingdong (JD.com) and Taobao (two e-commerce giants in China),respectively,where agricultural products from poor areas have been sold. The total value of agricultural products from China’s poverty-stricken areas sold on these platforms has reached around 284.5 million USD since their foundation (Xinhua 2019). Considering Chinese online market has expanded considerably to food products in recent years:for instance,agricultural product sales on Alibaba had grown from 3.04 billion USD in 2012 to 14.47 billion USD in 2016 (Jianget al.2019); the average annual growth rate of food sales on JD.com was at 150% between 2013 and 2017 (JD Big Data Research Institute 2018),online ethical consumption of agricultural produce would be growing rapidly in the future.

In order to promote the development of agricultural industry in poverty-stricken area by consumers’ ethical consumption,policymakers and e-commerce firms need to know to what extent consumers pay for private attributes and ethical attributes of agricultural products from povertystricken areas. (i) Private attributes refer to higher perceived intrinsic quality,such as taste or safety (Wanget al.2018). Consumers can derive private benefit from agricultural products that are from poverty-stricken areas.Specifically,some consumers may consider agricultural products from poverty-stricken areas as eco-friendly and safe food products,because they may perceive that a poverty-stricken area usually has lower nitrate residues in food (food safety) or its environment is usually not polluted by industries. Previous studies have shown that consumers are willing to pay a premium for eco-friendly food products because of the perceived private benefits (Thompson and Kidwell 1998; Blend and Van Ravenswaay 1999; Moonet al.2002; Chenet al.2018). (ii) Ethical attributes refer to consumers’ ethical value of helping the poverty-stricken areas. As the ethical attribute is a positive public attribute(Bougherara and Combris 2009),consumers can derive utility from contributing to poverty alleviation. However,despite China’s fast-growing consumption of agricultural products from poverty-stricken areas,the factors motivating such online ethical consumption in China remain unclear.

Take apples from poverty-stricken areas (consumption in “Fupin Guan” or “Xingnong Tuopin”) as example,this paper aims to examine consumers’ willingness to pay for agricultural products from poverty-stricken areas to explore the effectiveness of the ethical consumption initiative. In the survey,we select apples from poverty-stricken areas as our hypothetical product,because many poverty-stricken areas are also prestigious apple producing areas. Among the 122 advantageous apple planting counties in China,33 of them are national poverty-stricken counties.1The data are collated and calculated by authors from the Ministry of Agriculture and Rural Affairs of China and the State Council Leading Group of Poverty Alleviation and Development.Many farmers in poverty-stricken areas,such as Shaanxi (CGTN 2019),Xinjiang (Xinhua 2018) and Yunnan (Xinhua 2020),have shaken off poverty by growing and selling apples.Apples are also typical and major agricultural products in online market,because apples are easy to be packaged and suitable for long-distance transportation. Different ethical consumptions (such as fair trade) in most of the developed countries,fast-growing consumption of agricultural products from poverty-stricken areas provides a unique setting to investigate online ethical consumption in China. This research fills the research gap on consumers’ online ethical consumption in light of China’s fast-growing agricultural e-commerce and its role in poverty alleviation. The findings are also expected to provide valuable insight into promoting the effectiveness of poverty alleviation in China,targeting to completely eradicate extreme poverty by the end of 2020(Luoet al.2020; Huang 2020).

2.Data and methods

2.1.Willingness to pay (WTP) elicitation method

Contingent valuation (CV),a popular stated preference method that is widely used to elicit consumer preferences for non-market goods (Hanemann 1994; Huet al.2011),is used to estimate consumers’ WTP for apples from povertystricken areas. This study uses the payment card approach to elicit consumers WTP,which is prevalent in the current literature (Hackl and Pruckner 1999). WTP values estimated by the payment card approach are more robust than those by the dichotomous choice approach (Readyet al.2001).Respondents are asked to choose one value from a list of values to indicate their maximum WTP (Venkatachalam 2004). The true WTP of the respondent is then assumed to be located between the indicated value and the next one lower than the indicated value if the indicated value is not the lowest on the list (Hu 2006). The interval format is often considered better than the format of a dichotomous choice question with a follow-up question (Alberini 1995).

We designed the survey at the following steps:Firstly,participants were informed that the study was to assess consumers’ perceptions and evaluations of apples from poverty-stricken areas. We informed respondents to assume that they are going to buy apple online and the price for conventional online apples was 6 CNY 500 g–1(the most common price shown in two major e-commerce platforms–Jingdong and Taobao). Next,respondents were requested to choose the maximum price in the price list that they would like to pay for apples from poverty-stricken areas in different scenarios. The price list contains six price intervals:6.0–7.0 CNY,7.0–8.0 CNY,8.0–9.0 CNY,9.0–10.0 CNY,10.0–12.0 CNY,12.0 CNY and above.

We used both within-subject (version A) and betweensubject (version B) designs to develop these two survey versions (Bougherara and Combris 2009) to identify whether there is a purely altruistic reason for consumers’WTP for apples from poverty-stricken areas. In version A,respondents were asked to state two WTP values,one before an “information shock”,which comprises information about apples from poverty-stricken areas,emphasizing that their taste and safety attributes are the same as the conventional ones,and the other after such information(“within”). Participants were requested to choose the maximum price in the price list that they would like to pay for apples from poverty-stricken areas. In version B,participants were fully informed from the outset (“between”)that there is no quality difference between the apples from poverty-stricken areas and the conventional ones. With the between-subject design (before informed in version A and fully informed in version B),the value of private attributes between survey data can also be separated. Table 1 reports the elicitation questions in version A and version B.

2.2.Data collection and description

A professional marketing research company (wjx.cn)collected the data in September 2019. Marta-Pedrosoet al.(2007) and Nielsen (2011) showed that the means of welfare measure estimates from contingent valuation methods are the same for face-to-face interviews and online surveys.Our online surveys are of high quality. Firstly,respondents were randomly selected from 2.6 million panelists. Secondly,wjx.cn could control one IP,one computer,and one account for only one questionnaire. Thirdly,some specified requirements (e.g.,being 18 years and older and online fruit buyers) had to be met. Finally,the “trap question” method(Joneset al.2015; Gaoet al.2016) was used to identify respondents who may not have carefully read the survey questions. The survey stated that “Suppose that the price for conventional apples is 6 CNY 500 g–1. Please choose the price of conventional apples.” Respondents who provided a wrong answer were considered less careful in answering the survey. Removing respondents who failed the “trap questions” and those with any missing responses resulted in 1 238 valid respondents for the final analysis.

The online survey collected data in Beijing,Shanghai,Guangdong,Shandong,Jiangsu and Shaanxi. Beijing is the political and cultural center of China; Shanghai is a global financial hub; Guangdong and Jiangsu represent China’s developed coastal areas; and Shandong and Shaanxi are both famous apple production and consumption areas. The descriptive statistics show that all respondents are apple consumers,with about 54.04% often buying fruits online and 65.27% having bought apples from poverty-stricken areas.

Table 2 reports the sample demographic statistics.The majority of respondents were female (63.57%),aged between 25 and 34 (52.58%),and from a family that has at least one child under the age of 12 (62.12%). The median monthly personal income was between 6 000 and 7 999 CNY.Income varied among study areas,with respondents inBeijing and Shanghai reporting the highest monthly income,while those in Shaanxi reporting the lowest. In terms of online consumption habit,respondents purchasing online once a month,every two weeks,once a week and more than once a week accounted for 6.30,28.03,34.49 and 31.18%,respectively,of the total sample.

Table 1 Willingness to pay (WTP) elicitation method

Table 2 Comparison of demographic and behavioral characteristics for participants from different areas

2.3.Theoretical model

Based on Lancaster’s (1966) consumer demand theory that a consumer maximizes his/her utility by selecting the product attribute combination,we assume there are only two agricultural products on the e-commerce platform:conventional agricultural products and agricultural products from poverty-stricken areas. When purchasing the latter,consumers can derive both utilityu1from private attributes such as higher perceived intrinsic quality (e.g.,taste or safety) and utilityu2from ethical attributes that reflect consumers’ ethical value of helping producers in the povertystricken areas. The total utility of a unit of agricultural products from poverty-stricken areas isu,whereu=u1+u2.

Following the theoretical model of Yuet al.(2014),consumers’ utilities for a unit of conventional agricultural products and agricultural products from poverty-stricken areas areu(p0,0;X) andu(p1,1;X),respectively,where,0 and 1 respectively refer to conventional agricultural products and agricultural products from poverty-stricken areas.Xis a vector of control variables,such as quality perception,trust,demographic variables,and behavioral variables.T=p1–p0=Xβ,whereβis a vector of marginal effects ofXon the WTP,denoted asT.

2.4.Econometric model

The payment card method generated two types of WTP observations:interval and censored. The interval regression,which is a generalized Tobit model (Amemiya 1973) for outcomes with interval censoring,is employed.The differences among censored,truncated,and interval data have been discussed by Cameron and Trivedi (2010).The interval censored regression model is consistent with the continuous Cragg model. The only difference is that,in the second stage of estimation,the interval-censored model specifies the data’s payment card nature (Batteet al.2007).

If the value for thejth individual is somewhere in the interval [y1j,y2j],the known boundaries of WTP,as well as the chosen range indicating the underlying maximum WTP for apples from poverty-stricken areas,are taken into consideration by the interval-censored model. Since the interval data boundaries are already known,the true WTP is assumed to locate in regions of (–∞,y1),(y1,y2),...,(yJ,∞).Assuming a latent variableWTP＊ indicates the true WTP of individuali,the estimation equation is expressed in eq.(1):

3.Results and discussion

3.1.Statistics of WTP

We computed the mean values of WTP by assuming the true value locates at the middle point of the interval (Table 3),according to Yuet al.(2014). In survey version A (“within”),the mean WTP premium for apples from poverty-stricken areas before information given was 1.86 CNY,while it reduced to 1.73 CNY after respondents received more information (the means are statistically different,t-stat=4.24).It implies that there were both selfish and altruistic reasons for consumers’ willingness to pay for apples from povertystricken areas. The results showed that consumers in China,on average,were willing to pay 31% more for apples from poverty-stricken areas than for conventional apples. This result is close to the 40% premium for sustainable milk (Gaoet al.2020) and the 41% premium for rice produced in a rice-fish system,but much lower than the 130% premium for organic food (Yinet al.2010). The value of private attributes was 0.13 CNY,only accounting for 6.99% of the total premium. The premium attributed to altruistic reasons was 1.73 CNY,which accounted for 93% of the total WTP premium.2This was the upper bound of the altruistic attribute,because we cannot tease out all the private benefit.This is consistent with Douadiaet al.(2009) who find that a preference for high quality is not necessarily the main motivation for purchasing eco-labeled products.

In survey version B,the mean WTP premium was 1.80 CNY for fully-informed respondents,which is between the ‘before informed’ and ‘after informed’ values in survey version A. We did not find a statistically significant difference in the mean WTP between ‘full informed’ in version B and‘after informed’ in version A (t-stat=–0.939). The result indicated the robustness of the online survey. When we computed the value of private attributes using data from survey version B and survey version A (before respondents were fully informed),the value of private attributes is still small. The results from both within-subject and betweensubject designs indicated that altruism is the main motivation for buying apples from poverty-stricken areas.

3.2.Factors influencing consumers’ WTP for apples from poverty-stricken areas

Interval regression models were used to estimate the WTP equation in Stata 14.1. The explanatory variables include consumers’ quality perception,ethical value,trust in government supervision of e-commerce,demographic characteristics and behavioral variables. We used the forward,backward,and step wise selection methods to determine the variables that should be kept in the final models,and Akaike’s information criterion (AIC) was used for model comparison and selection. Table 4 reports regression results of the before-informed scenario (model 1)and the after-informed scenario (model 2) using the survey A sample,and the fully-reformed scenario (model 3) using the survey B sample.

Results of model (1) show that the coefficients of quality perception variables (safer,more eco-friendly,and more helpful for poverty alleviation) have significantly positive impacts on consumers’ WTP before the “information shock”.The results indicate that both the perception of private and ethical benefits of apples from poverty-stricken areas would affect consumers’ WTP. However,the coefficient of “safer”is less than those of the ethical attributes. This result is different from Gaoet al.(2020)’s study on sustainable milk,which showed that only the private attribute (better quality)was significantly and positively correlated with WTP. Our result suggests that providing consumers knowledge about both private and ethical attributes of agricultural products from poverty-stricken areas can be effective for market promotion. Promoting the private and ethical attributes ofagricultural products from poverty-stricken areas may be a potentially effective policy on poverty alleviation.

Table 3 Mean of willingness to pay (WTP)

Table 4 Estimation of the willingness to pay (WTP) equation

Results of model (1) also show that consumers’ WTP is positively associated with their trust in government supervision of e-commerce,even though consumers can obtain abundant product information from the videos and/or pictures on the web-page and track the logistic information(Jianget al.2019). This result is consistent with the finding by Liuet al.(2019) that consumers’ trust in government supervision system affects their preferences and WTP.Given a severe trust crisis in the Chinese food market(Wanget al.2019),strengthening government supervision of e-commerce and restoring customers’ trust are critical in promoting agricultural productsviae-commerce that also serve the purpose of poverty-alleviation.

Most demographic and behavioral variables are statistically significant except for education level,which is generally consistent with current literature that shows significant relationships between demographic characteristics and WTP (Yuet al.2014; Wang and Gao 2017; Gaoet al.2020). Female and young consumers are more likely to pay a premium for apples from poverty-stricken areas than their male and elder counterparts,respectively.An explanation might be that young generation develop a stronger sense of social responsibility during their recent school education that advocates social equality. Another explanation might be that the elder are reluctant to change their eating and consumption habits and therefore are less likely to pay for new food attributes (Yuet al.2014). Having children under 12 years old is associated with an increased WTP,suggesting a potential over-generation concern about poverty alleviation (Gaoet al.2020). Regarding behavioral variables,respondents who often shop online are likely to pay more for apples from poverty-stricken areas,while the coefficients of other purchasing experience were insignificant.

Estimations of the within-subject and between-subject design after the “information shock” show similar patterns to those before the “information shock”,mainly due to the evidence in this study that altruism is the main motivation for purchasing apples from poverty-stricken areas. This suggests that Chinese consumers’ ethical value on poverty-alleviation has a significant impact on their WTP for apples from poverty-stricken areas,which is consistent with previous findings on the significant correlation between values and ethical consumption behavior (Dickson 2000).It is also interesting to note that respondents who have had purchased apples from poverty-stricken areas before are more likely to pay a premium after the “information shock”. This may be because only 26.82 and 16.40% of respondents regard agricultural products from povertystricken areas as safer and more tasty,respectively. This suggests that improving quality and safety of agricultural products from poverty-stricken areas would attract more consumers to purchase these products and thus benefit the e-commerce poverty alleviation programs. Model (3) also shows that personal monthly income of the fully-informed respondents significantly increases WTP,which is consistent with the literature that the rich tend to pay more for ethical consumption.

The results provide insights for government and e-commerce firms to increase demand for agricultural products from poverty-stricken areas. Although altruism is the main motivation for buying agricultural products from poverty-stricken areas,quality perception of private attributes also has significant effect on consumers’ WTP for apples from poverty-stricken areas. It implies that the effectiveness of this innovative poverty-alleviation initiative also depends on consumers’ quality perception of agricultural products from poverty-stricken areas. As e-commerce has a great potential advantage to promote quality information by providing more detailed information about agricultural products (e.g.,origin,grower and environment) through videos and/or pictures displayed on the web-page,e-commerce can play an important role in promoting agricultural products from poverty-stricken areas.Hence,China’s fast-growing e-commerce in the agricultural sector offers an enormous opportunity to promote ethical consumption.

4.Conclusion

Despite China’s fast-growing e-commerce and its great achievement in promoting poverty alleviation through consumption,little is known about Chinese consumers’online ethical consumption. This paper studies Chinese consumers’ preference and willingness to pay for agricultural products from the country’s poverty-stricken areas and identifies their determinants. Data from the six representative provinces or municipalities show that 91.44%of the respondents consider purchasing agricultural products from poverty-stricken areas as meaningful and helpful for poverty alleviation,while only 26.82 and 16.40% believe these products are safer and more tasty. The results of statistics of WTP show that,before “information shock”,Chinese consumers on average are willing to pay a 31%premium for apples from poverty-stricken areas. After“information shock”,results of the within-subject design show that the drop of WTP is statistically significant but very small in size. This implies that although consumers take private attributes into consideration,the ethical attribute is the main motivation for apples from poverty-stricken areas.

The regression results further show that perceptions of both private and ethical attributes have statistically significant impacts on WTP. However,the effect of the private attribute perception is smaller than that of the ethical attribute perception. Improving both the private and ethical attributes,especially the latter,of agricultural products from povertystricken areas should be mainstreamed to current policies that promote poverty alleviation through consumption.Interestingly,respondents with experience in purchasing apples from poverty-stricken areas are more likely to pay a premium after “information shock”. This may be due to the low private quality perception of apples from povertystricken areas. This implies that improving the quality and safety of agricultural products from poverty-stricken areas is an important means to attract more consumers.

This study also finds that consumers’ trust in the government supervision system of e-commerce has a significant and positive effect on their WTP. Our results imply that trust in government supervision of e-commerce still plays a vital role in motivating online ethical consumption.Demographic variables and behavioral variables have statistically significant effects on WTP for apples from poverty-stricken areas. Young and female consumers who have children are willing to pay more,suggesting the existence of consumer heterogeneity. The positive coefficient of personal monthly income implies that the policy of promoting poverty alleviation through consumption may be more sustainable by targeting richer consumers.

This paper provides important information for government and e-commerce platforms to promote the demand for agricultural products from poverty-stricken areas. However,several important questions are still unanswered. Although this study shows quality perception matters to ethical consumption,we still do not know how quality perception affects sustainable ethical consumption? As there are not certificated labels for ethical consumption in China,should Chinese government establish labeling system to promote poverty alleviation through consumption on e-commerce platforms? All these questions need to be answered in order to promote ethical consumption in China.

Acknowledgements

We thank the comments from the editor and anonymous reviewers. The authors acknowledged the support from the National Natural Science Foundation of China (71903088)and the Project of Humanities and Social Science of the Ministry of Education of China (19YJC790132).

Declaration of competing interest

The authors declare that they have no conflict of interest.