Exploring consumers' perception and willingness to pay for "non-added sulphite" wines through experimental auctions: A case study in Italy and Spain

Amato, Mario; Ballco, Petjon; López-Galán, Belinda; De Magistris, Tiziana; Vernau, Fabio Article Exploring consumers' perception and willingness to pay for "non-added sulphite" wines through experimental auctions: A case study in Italy and Spain Wine Economics and Policy Provided in Cooperation with: UniCeSV - Centro Universitario di Ricerca per lo Sviluppo Competitivo del Settore Vitivinicolo, University of Florence Suggested Citation: Amato, Mario; Ballco, Petjon; López-Galán, Belinda; De Magistris, Tiziana; Vernau, Fabio (2017) : Exploring consumers' perception and willingness to pay for "non-added sulphite" wines through experimental auctions: A case study in Italy and Spain, Wine Economics and Policy, ISSN 2212-9774, Elsevier, Amsterdam, Vol. 6, Iss. 2, pp. 146-154, https://doi.org/10.1016/j.wep.2017.10.002 This Version is available at: https://hdl.handle.net/10419/194537 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ HOSTED BY Available online at www.sciencedirect.com Wine Economics and Policy 6 (2017) 146–154 Exploring consumers’perception and willingness to pay for “Non-Added Sulphite”wines through experimental auctions: A case study in Italy and Spain $ Mario Amato a, n , Petjon Ballco b , Belinda López-Galán b , Tiziana De Magistris b , Fabio Verneau a a University of Naples Federico II, Department of Political Science, via Rodinò 22, 80138 Naples, Italy b Centro de Investigación y Tecnología Agroalimentaria de Aragón (CITA), Av. Montañana, 930, 50059 Zaragoza, Spain Received 28 June 2017; received in revised form 4 October 2017; accepted 11 October 2017 Available online 18 October 2017 Abstract Although scientific literature is still uncertain towards the real causes that can link headache episodes with moderate wine consumption, a growing proportion of consumers seems to address sulphites as possible culprits. Hence, the objective of this study is to assess consumers` willingness to pay for wine bearing a sulphites-free label in two traditional wine producing countries, Italy and Spain. The methodological framework is based on the specification of the Becker-DeGroot-Marschak (BDM) type of auction applied to consumers purchasing wine in the supermarket stores of both countries. A left censored Tobit model is used to analyse the bidding behaviour for conventional and "Non-Added Sulphite" (NAS) wine. Results in both countries show that consumers who link the headaches with the consumption of sulphite are willing to exchange the habitually consumed bottle of wine with a NAS wine and pay extra premium prices. &2017 UniCeSV, University of Florence. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Keywords: Experimental auctions; Non-added sulphite wines; Consumer behavior 1. Introduction Within the wine markets, consumer behaviour and preferences have been affected with deep structural changes towards higher quality wines, healthier production processes and new ways of tourism (Bregoli et al., 2016). Although it is proven that a moderate wine consumption promotes health benefits such as improving glucose tolerance, insulin sensitivity and hepatic steatosis (Rosenzweig et al., 2017), other discussions related to sulphur and its organic compounds seem to abound in both scientific literature and wine speeches (Laganà et al., 2017;Machado et al., 2009;Vecchio et al., 2017). Wine is not treated with sulphur but, at best, with sulphur dioxide (SO 2 ), which originates from sulphur combustion. A reaction with liquids creates sulphurous acid, which is partially present as its salts, called sulphites. The terms sulphur dioxide, SO 2 , sulfuric acid and sulphites are interchangeable in common usage, while the trivial definition "sulphur" does not reflect in a proper way the chemical. The anti-microbial, antioxidant and preservative effects of sulphites were already known and used in the wine making areas of central Europe since at least the middle ages. Sulphites are formed naturally during the fermentation phase (Chengchu et al., 2006), and common quantities around 30–90 ppm (ppm) are also added throughout the production phase (Burgstahler and Robinson, 1997) to prevent spoilage and enhance aging potential (Goode and Harrop, 2011). With the development of technological and oenological knowledge combined with changes in customer preferences, the use of additional sulphites is being subjected to increasingly tight constraints, and significantly being reduced. The maximum permitted values of SO 2 were increasingly reduced especially on the basis of toxicological considerations. In many cases the www.elsevier.com/locate/wep http://dx.doi.org/10.1016/j.wep.2017.10.002 2212-9774/&2017 UniCeSV, University of Florence. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer Review under the responsibility of UniCeSV, University of Florence n Corresponding author. E-mail addresses: [email protected] (M. Amato), [email protected] (P. Ballco), [email protected] (B. López-Galán), [email protected] (T. De Magistris), [email protected] (F. Verneau). effects of sulphite on health are very mild; however, depending on the subjects, sulphites can induce severe reactions, including respiratory and gastrointestinal set of symptoms (Lester, 1995), as well as dermatological signs and headaches (Costanigro et al., 2014). Indeed, a small percentage of the population is sensitive to sulphur dioxide to ingestion but studies revealed that the role of sulphites has been occasionally overestimated, as in the case of wine-sensitive asthmatics (Vally and Thompson, 2001). The practical justification for these reflections, in fact, is poorly founded, because around 2000 mg of SO 2 are created every day in the human body via the decomposition of proteins; this sulphur dioxide is then transformed into sulphate and expelled as such with urine. Daily consumption of a bottle of wine, with an average of 100 mg of SO 2 , increases the natural transformation of SO 2 in the body of 5% more or less, which should not cause toxicological problems. Despite uncertainty, apparently consumers are increasingly convinced that the cause behind their “malaise of the next day”and their headaches could be the chemical in question (Costanigro et al., 2014) and its inorganic compounds. Even though, there is a growth of evidences in newspaper and other related topic articles who announce that consumers keep manifesting episodes of headaches and migraines after drinking moderate amounts of wine, especially red variant (Gaiter and Brecher, 2000;Robin, 2010). Until recently, there is no scientific evidence proving the veracity and it's still unclear what could be the cause. A previous study of Jarisch and Wantke (1996) stated that “histamine as a cause for headache should be considered first”, but since then scholars proved that a large amount of chemicals can be the real cause (see Mauskop and Sun-Edelstein (2009), and Millichap and Yee (2003)). Others believe that the “consumption”of sulphites is the starting point of their malaise (Gaiter and Brecher, 2000) most probably due to the level of knowledge and awareness that seem to be still low. Despite the fact that the difference between non-added sulphite wines and organic wines is still unclear in consumers’eyes (Costanigro et al., 2014), scientific literature on organic wine industry has mainly focused on market trends and on consumer attitudes and purchase intentions (Pomarici and Vecchio, 2014; Pomarici et al., 2015;Remaud et al., 2008;Sogari et al., 2015). In addition, few studies that examine consumer attitudes towards organic wines (e.g., Olsen et al., 2012) provide little information about the consumers’perception of sulphite and willingness-to-pay (WTP), especially since the organic production is linked to higher quality, healthier, tastier and more socially and environmentally responsible attributes (Fotopoulos et al., 2003). In addition, even though numerous studies in the scientific literature have used experimental auctions for sustainability and certification topics (De Magistris et al., 2015;Del Giudice et al., 2016), and taking into account that wines without added sulphites could prove to be a profitable niche market, to the best of our knowledge, scant literature exists on consumers’willingness to pay for sulphite free labelling (Costanigro et al., 2014;D’Amico et al., 2016). A study of Costanigro et al. (2014) in the USA reports that a group of headache-syndrome-suffer (HSS) consumers after being informed on sulphites in wine, are willing to pay a price premium of $1.23 to avoid them, meanwhile only 3.4% of the same group of consumers are more likely to purchase a wine with non-added sulphites (NAS). A more recent study of D’Amico et al. (2016), on the Italian consumer analyzing the interest and the WTP for organic wine with NAS found that the majority of respondents (54,5%) were not willing to pay a premium or only a small premium (10%) for NAS. The study also identified that insufficient information is a barrier to higher WTP for organic NAS wine, and that the organic certification of wines per se is not sufficient to ensure consumers health factors. Hence, there is a need to investigate the role of subjective and objective knowledge in determining consumer willingness to pay related to wine additives and particularly sulphites. Therefore, this study analyses consumer preferences and estimates the willingness to pay for wine bearing a free sulphites additive label in two traditional wine producing countries, Italy and Spain. To achieve the main objective of comparing the marginal preferences and price differentiation between NAS wines and classic wine categories we base the methodological framework on the specification of the Becker-DeGroot-Marschak (BDM) type of auction. This type of mechanism is found to be particularly suitable for estimating the willingness to pay for with a certified absence of sulphites and is applied to 240 purchasing wine consumers in the supermarket stores of both countries. This article is organized as follows: section two reviews the materials and method of the research, including a description of the experimental auctions, the experimental procedure and the econometrical model used; section three illustrates the results obtained and section four presents the discussions and concludes. 2. Materials and methods 2.1. Experimental auction Experimental auctions are useful tools used by economists, psychologists, and marketing experts to determine the possible success or failure of a hypothetical entry of new products in the real market. Experimental auctions are commonly used to elicited consumers’preferences due to their incentive compatibility properties. In other words, they provide an incentive to subjects to state their true preferences in contrast to hypothetical elicitation procedures. In a typical experimental auction, subjects bid to obtain one or more goods and the highest bidder(s) have to buy the auctioned product and pay a price that is determined in the auction. This mechanism is also used to evaluate the effects of: (a) different priming strategies; (b) different information messages and (c) different vehicles of information, upon WTP (De Magistris et al., 2015;Jacquemet et al., 2013;Lusk et al., 2004). With respect to the agri-food marketing sector, this tool is used for the assessment of food safety attributes (Thorne et al., 2017), sensory and organoleptic attributes (Hung and Verbeke, 2017) and sustainability characteristics of food products (Vecchio and Annunziata, 2015). M. Amato et al. / Wine Economics and Policy 6 (2017) 146–154 147 Compared to other methods, the advantage of experimental auctions is that people are placed in an active market environment (although organized); a real market, with real money and real products can involve people revealing their preferences, while keeping the attention of individuals facing the task of evaluation. Experimental auctions help to determine the individual's willingness to pay because the mechanism requires that each participant make an offer that, in theory, reflects the value he attaches to the good auctioned. The incentive-compatible BDM type of auction (Becker et al., 1964) was selected as it is particularly suited to eliciting WTP directly at the point of purchase, thereby enhancing external validity (Wertenbroch and Skiera, 2002;McDaniel and Gates, 2001). In addition, the BDM was selected among other demandrevealing valuation mechanisms (e.g., Vickrey or nth-price auctions) because the experiment setting required one participant at a time, a situation for which only the BDM is appropriate. Therefore, in store, non-hypothetical experimental auctions (Costanigro et al., 2011;Nayga et al., 2006) were conducted in the “wine and liquors”lane of two supermarkets in Zaragoza (Spain) and Naples (Italy) in June and July, 2016. The BDM mechanism equally penalizes overand underbidding (Lusk and Shogren, 2007) and reduces the psychic benefitfrombeing declared the winnerofanauction(Corrigan and Rousu, 2006). Furthermore, the BDM auction tends to provide relatively strong incentives for truthful bidding for all individuals regardless of the magnitude of their true WTP (Lusk et al., 2007). Finally, compared to other auction mechanisms the BDM, in its endowupgrade format, is also very easy to explain (i.e. participants need only to place one bid) and most of the time one example is enough to check for participants’comprehension (Combris et al., 2009). In the endow-upgrade BDM auction format, participants have a dominant strategy in bidding an amount equal to their true valuations for the good. In particular, participants were carefully explained that they had to submit a monetary value that she/he would be willing to pay. In order to win, participants had to submit a bid greater than or equal to a randomly extracted binding value, which was randomly drawn, for every participant, from a distribution of values unknown to the subjects (Bohm et al., 1997) to avoid anchoring and ensure that they stated their real willingness to pay. If the exchange was made, the subject had to pay the binding value to make the exchange. On the contrary, if the subject's bid was lower than the binding value, the transaction did not happen (Lusk and Shogren, 2007). Since previous scholars have noted that for BDM auctions training rounds are not required (Feldkamp et al., 2005), none was used in this research. The distribution of potential “prices”was uniform and based on real market prices, ranging from 0 cents (€)to3,50,in increments of 10 cents. 2.2. Experimental procedure The experiments have been conducted throughout a threeweek period in each country and in three different phases of the day (early morning, afternoon and evening) (East et al., 1994). In each session, an interviewer approached subjects individually. In particular, participants were randomly recruited while they were actually purchasing a bottle of wine from the supermarket shelves. In order to be eligible for the study, each participant had to be aged 18 or above and consume wine at least once a month. Each respondent was rewarded with a bottle of conventional wine, averagely priced 4,5€in both countries, for participating in the experiment. On average, about 70% of the contacted wine consumers agreed to participate. The experimental protocol consisted of a penciland-paper questionnaire and a WTP experiment for those who wanted to upgrade the conventional bottle of wine with a “Non-Added Sulphites”wine. 1 It took approximately 15 min for each participant to complete the experiment, including a careful explanation of BDM auction mechanism. Respondents were randomly assigned in two different treatments. The endowed bottles of wines were red or white variant, depending on customer's habitual wine purchase at that moment, since we also wanted to test whether there is a different perception between these two varieties. Then, following Costanigro et al. (2014), participants were asked if they have ever experienced headache after drinking moderate amounts of wine, without directly referencing sulphites. “Moderate consumption”was emphasized to ensure that participants would not confuse their post-consumption headache with alcohol hangover; in case of affirmative response to the previous question, participants were asked to choose one or more perceived causes of their headache from a randomly ordered list of factors (dehydration, histamines, tannins, organic wines, red wines, tyramine, white wine, sulphites and other). After this task, individuals completed a self-report questionnaire on wine subjective knowledge (following Flynn and Goldsmith (1999)). Subsequently, respondents were handed the NAS wine bottle, prepared in order to avoid any external differences with the conventional one (same format of the bottle, vine variety, vintage, closure and label), and were asked if they were interested in exchanging their endowed bottle of wine (conventional) for a NAS one. In case of affirmative response, the BDM auction mechanism started, otherwise their negative response was recorded as a willingness to upgrade the bottle equals to 0 and proceeded with the second part of the questionnaire. Subsequent to the WTP experiment, participants were asked to complete a second short questionnaire explicitly developed to analyse which are the main personal traits that can explain their willingness to pay a premium price for the “non-added sulphite”wine bottle. Items were shown as it follows: Participants’subjective knowledge on sulphite (e.g., “I’m able to distinguish the presence of sulphite in wine”). Attitude towards naturalness and healthy food (“health” subscale adapted from the Food Related Life Style Scale – Grunert, 1993). Label's use, reading and trust (Krystallis et al., 2012). Objective knowledge about sulphites (i.e. we used questions in order to assess the specific knowledge of participants towards sulphites). 1 The Non-Added Sulphites bottles of wines were presented with a logo impressed on the front label. M. Amato et al. / Wine Economics and Policy 6 (2017) 146–154148 Wine consumption habits. Socio-demographic data. At the very end of the questionnaire participants were asked their perceived value of the endowed bottle. 2.3. Econometric model Since the aim of this research is to investigate factors affecting WTP for NAS wines, the dependent variables in the model are the bids declared by each subject. Considered the nature of the data, a left censored Tobit (Robin, 1958)model was used to analyse the bidding behaviour for Bw i(i¼1and2, conventional and NAS wine). Censoring at zero is a common feature of experimental auctions in some recent studies (Bernard andBernard,2010;Shietal.,2013) in which Tobit is used to address the censoring issue. Since the main interest is focused in the most important needed effects, possible interactions were ignored. In general, the Tobit model can be expressed as: y i¼β ′ xiþui;uieNð0;s2Þ yi¼y iif y i40or0ify i≤0 Therefore, the expected willingness to pay for consumer i can be computed as: Ey  i  ¼Eðyijyi40Þñf yijyi40  þEy ijyi¼0  ñFðyi¼0Þ ¼β′xþsλβ ′xi s  ñΦβ ′xi s  þ0 ¼β ′ xiΦβ ′ xi s  þsφβ ′ xi s  where the inverse Mills ratio λ(β′x/s) is equal to φ(β′x/s)/Φ (β′x/s) and the marginal effect (β) for the continuous variable x j is: ∂EðyiÞ ∂xj ¼Φβ ′ xi s  βj In particular, the independent variables are participants’ socio-demographic and lifestyle characteristics, consumption frequency of wine, subjective wine and sulphite knowledge. 3. Results 3.1. Sample description The socio-demographical results of Italian and Spanish populations are showed in Table 1. With regard to the Italian sample, participants (55,8% of females) were in the range 20– 77 years (32,7 710.8 years), living in a medium-size household (3,8 71,3 members); 10% of the sample had a household income higher than 4000€per month, while for more than 55% it was lower than 2000€per month. About a quarter had a University degree (27,2%), while more than 60% is employed. In the Spanish sample, participants’age (50,8% males) ranged between 18–82 years (45,9 715,2 years) living in medium-size households (2,9 71,3 members), mostly employed (62%) and with an average income ranging between 2000 and 4000€per month. 3.2. Explanatory variables The exploratory variables were captured through the questionnaire who aimed in defining which variables explain consumers’willingness to pay. Scales Cronbach's alpha indicates a consistent internal validity, ranging from 0,63 (subjective sulphite knowledge) to 0,82 (subjective wine knowledge). Observations from Fig. 1 show that subjective knowledge on sulphites in both countries is very low (mean score 2,5 71,3 in Spain, 3,5 71,2 in Italy). This evidence is confirmed in Fig. 2, where the objective knowledge 2 has been analysed. In fact, the mean score resulted is 0,1 70,14 for Spain and 0,15 70,20 for Italy, respectively. Table 1 Descriptive Statistics of the sample. Variables Italy Relative Frequencies (Mean) Official Stats a Spain Official Stats a Age Mean value 32.7 (SD¼10.8/ min-max 20–77) 44,2 45.9 (SD¼15.2/ min-max 18–82) 41,9 18–34 76% 18.1% 25.8% 19.8% 35–54 18% 30.6% 46.7% 32.2% 55–74 5% 23.6% 24.2% 21% More than 74 1% 11.3% 2.5% 9.2% Gender Female 55.8% 51,4% 49.2% 50,9% Male 44.2% 48,6% 50.8% 49,1% Household Members 3.8 (SD¼1.3/ min-max 1–8) 2,35 2.9 (SD¼1.3/ min-max 1–9) 2,52 Education Level At least High school degree 63.3% 59.3% 55.8% 56.6% Bachelor's degree or higher 10.1% / 21.7% / Employment status Employed 65% 56.8% 62% 56.5% Unemployed 35% 38% Monthly Income Mean value (NET) / 2,033 €1,718€ Less than 1,000€10.9% 11.7% Between 1,000 and 2,000€ 45.8% 35.9% Between 2,000 and 4,000€ 33.3% 48.3% More than 4,000€10% 4.2% a OECD and Eurostat databases: data refers to 2015. 2 As a reminder, objective knowledge was measured with 5 questions related to sulphites (e.g., “What is the typical smell that indicates the presence of sulphites in wine?”1 - Correct answer, 0 - wrong or “I don’t know”). M. Amato et al. / Wine Economics and Policy 6 (2017) 146–154 149 3.3. Willingness to pay More than 80% of the total sample showed a positive WTP for exchanging the conventional wine bottle with the NAS one. In particular 91 consumers in Spain and 102 in Italy decided to exchange the endowed bottle. Participants then, were asked their perceived value of the endowed bottle (3,75€for Spanish consumers, 4,90€for Italians), this average price can be considered as the regular price for the conventional bottle which is close to the average real market price. The average WTP for exchanging the bottle resulted between 1,17€7 0,76€for Spain and 0,90€70,50€for Italy. However, since the main interest is the mean WTP for the whole sample, zero WTP values are also taken into account (Reiser and Schechter, 2009). In this later case, the mean WTP for exchanging a conventional wine with the NAS bottle of wine, drops to 0,89€ 70,83€for Spanish consumers and 0,77€70,56€for the Italian ones. According to these results, one NAS bottle of wine (750 ml) leads to a premium price of 25% in Spain and 16% in Italy. The perceived price of the two different varieties of wine does not have a statistically significant difference neither in Italy nor in Spain. In addition, the different variety of wine endowed did not significantly affect WTP in Spain (two sample t-test results, t ¼0,19; prob 4t¼0,84), while in Italy white variant (0,89€70,60€) obtains a significantly different price premium (t ¼2,36; prob 4t¼0,01) compared to the red variant (0,65 70,50€). Fig. 3 indicates that about 15% of the Italian consumers are willing to pay an additional price of 1,5 €while 10% of the sample is willing to pay a price premium of 1 €to purchase a NAS wine. In Spain, according to Fig. 4, about 20% of the sample is willing to pay a premium price of more than 1,5 €, whilst around 20% of the sample pay 1€as an additional premium price. The next step in the analysis was to see whether structural differences existed between headache and non-headache symptoms. Since people report that headaches are triggered from the consumption of sulphites, we anticipated that participants who suffer headache would have statistically different preferences towards the presence of sulphites in wine in comparison to those that do not experience headache issues. Table 2 summarizes the WTP inter and intra country mean values divided in three groups. As expected (Group 1) who do not experience headache after a moderate consumption of wine and (Group 2) who do not associate sulphite as a possible cause of their headache are willing to pay less premium prices in both countries in comparison to (Group 3) that directly link headaches with the presence of sulphites in wine. In particular the difference in terms of WTP almost doubles in (Group 3). Comparing the WTPs of (Group 1) which range between 0,63 70,48€for Italy and those of (Group 3) who link headache with the presence of sulphites 1,19 70,60€we estimate an additional increase of 89% 725% respectively. On the other hand, comparing (Group 1) WTPs with (Group 3) for the Spanish consumer, we estimate an additional increase of 91% 738% respectively. In general terms, the Spanish consumer who associate the headache symptoms with the presence of sulphites in wine is willing to pay 2% more than the same group of the Italian consumer. 3.4. Econometric estimates Before considering which are the main drivers that leads to a premium price, a logit model was performed in order to understand if there are specific variables that could explain consumers’entrance in the market. Table 3 reports the parameter estimates of a logistic regression, aggregated for both countries. A “participation dummy variable”was created taking the value of 1 ¼participating and 0 ¼otherwise. In particular, this choice is being explained by the behavioural and psychographic variables of “naturalness”and “sulphite objective knowledge”, followed by two socio-demographic characteristics (age and gender). Once the variables that determine market entry are discriminated, the next step is to apprehend which variables influence consumers’willingness to pay. For that reason, a Fig. 1. Subjective Knowledge about Sulphites. Fig. 2. Objective Knowledge about Sulphites. M. Amato et al. / Wine Economics and Policy 6 (2017) 146–154150 Tobit model was performed and specified as follows: WTPi¼αþβ1ðHeadache cause :sulphitesÞi þβ2ðNaturalnessÞiþβ3ðSubjective Wine KnowledgeÞi þβ4Subjective Sulphite KnowledgeðÞþβ5ðAgeÞi þβ6ðGenderÞiþβ7ðWine Purchasing FrequencyÞi þβ8ðWine VarietyÞiþεi Where WTP i is the average WTP for the ith consumer. The explanatory variable “Headache cause: sulphites”is introduced as dummy variable taking the value of 1 ¼if sulphites causes headache and 0 ¼otherwise. The variable “Naturalness”enters the model as mean values of 3 items measured in a 7-point Likert scale. The “Subjective Wine Knowledge”and “Subjective Sulphite Knowledge”variables enter the model as mean values of 5 and 6 items measured in a 7-point Likert scale. In addition, the “Gender”socio-demographic variable is introduced as a dummy variable which takes the value of 0 ¼ Female and 1 ¼male. “Age”enters as a continuous variable defined as number of years. The “Wine Purchasing Frequency” variable takes the values of 1 to 4 according to the number of times consumer purchase wine (once per month, 2–3 times per month, 4–5 times per month, 6 or more times per month). Lastly, the “Wine variety”variable that defines the experimental manipulation, takes the value of 0 ¼white and 1 ¼ red. Table 4 present the results of the Tobit estimation parameters for NAS wine. Results show that in Spain five (age, gender, wine purchasing frequency, sulphites as perceived cause of headache and naturalness trait) out of eight parameters are statistically significant. While in Italy, only three parameters are statistically significant (gender, naturalness and sulphites perceived as a cause of headache). Taking into account that the proposed model shows some non-significant variables, Table 5 proposes a more parsimonious model that only uses the previously significant variables of each country. Result from the estimation of the Tobit model (Table 5) indicates that the independent variable “headache cause: sulphites”receives the highest premium price (0,58€in Spain and 0,44€in Italy) followed by the “Gender”dummy variable. In particular, the Spanish consumer pays a premium of 0,38 €. The negative sign means that Spanish women are more willing to pay a premium price in comparison to men. Vice versa, in Italy, this variable assumes the value of 0,25€, meaning that men are more likely to pay a premium price instead of women. “Naturalness”variable was measured on a 7-point Likert scale, we consider estimates related to a onepoint increase in the respondent's stated relevance for the statement examined. In this case, an increase of one rating increases WTP by 0,22€in Spain and 0,11€in Italy, meaning that if a participant's score, for instance, is 5, the increase in premium price is 1,1€in the Spanish case and 0,55€in the Italian one (0,22*5 or 0,11*5). The “Wine Purchasing Frequency”variable was also found to be statistically significant and increases WTP by 0,28€in Spain and 0,09€in Italy. It is worth mentioning that the “Age”variable in the Spanish case shows a decrease of the premium price as age increases by 0,2€for each year increment. This is in contradiction to previous theories who state that older people tend to consume less quantity but higher quality food products in comparison to younger people. Fig. 3. Relation between price premium and willingness to purchase NAS wine among Italian consumers. Fig. 4. Relation between price premium and willingness to purchase NAS wine among Spanish consumers. Table 2 Price premium to purchase NAS wine among Italian and Spanish consumers. Italy Spain Obs. WTP (€) Obs. WTP (€) Group 1 a 76 0.63 70.48 75 0.82 70.77 Group 2 b 21 0.82 70.57 35 0.82 70.84 Group 3 c 23 1.19 70.60 10 1.57 71.06 a Consumers who did not experience headache after moderate wine consumption; b Consumers who experienced headache but did not think sulphite as a possible cause; c Consumers who experienced headache and linked it to the presence of sulphites. M. Amato et al. / Wine Economics and Policy 6 (2017) 146–154 151 4. Discussion and conclusions The main objective of this study was to determine the premium price that Italian and Spanish consumers are willing to pay in order to limit the presence of sulphite in wine, focusing on consumers who experience headache after drinking moderate amounts of wine. The results of this study will be provided to the industry in order to deduce the existence of market niches for wine without sulphites. In our study, experimental BDM auctions were carried out to measure factors affecting WTP for NAS wines in two traditional wine producing countries, Italy and Spain. Results show that consumers seem to appreciate the reduction of sulphite in their wine, especially in the case when they suffer headache, despite the causes of that sore head in the literature are still unknown and regardless of the variety, differently from literature where the red variety has been addressed as main responsible (Gaiter and Brecher, 2000;Robin, 2010). Almost 40% of the total sample (240 respondents) who stated to suffer headache after a moderate wine consumption and who believe that sulphites is the cause of that, are willing to pay an additional price premium of 1,19€in Italy and 1,57€in Spain in order to exchange the bottle endowed with a no-added sulphite wine. Therefore, an important strategy would be to introduce non-added sulphite wines and transmit the right information of a healthier and more natural wine to the groups who associate sulphites as the cause of their malaise. Healthier and natural products are related very closely to the organic production which is a constantly a growing market in the wine industry (Goode and Harrop, 2011). Thus, it can produce significant competitive advantages for companies planning to differentiate their products in this market. Since consumers are still unaware of the difference between non-added sulphite and organic wines (Costanigro et al., 2014) and since the there is no scientific evidence of the difference between low sulphites and organic wines, companies may use the opportunity of the low sulphite wines production process which keep risk and production costs relatively lower than the organic production, taken advantage for differentiation and allow price premiums. However, producers should take into consideration the increased risk of faster oxidation and deterioration in comparison to regular wines. Econometric results also confirm that several socio-demographic factors also exerted a positive and statistically significant effect on WTP for this type of wine. With respect to gender, Spanish women are prone to offer a premium price for a bottle of non-added sulphites wine in comparison to Spanish men. This is an expected result since in terms of alcohol preferences between genders, Spanish women show higher preferences for wine and beer than men (León-Muñoz et al., 2015). It is important to mention that this result is focused on the fact that Spanish women are willing to pay a premium price for higher quality wines with NAS in comparison to conventional wines, but is not a comparison of the wine consumption frequency among men and women. Focusing on higher quality food products this result is in line with Laroche et al. (2001) who found that the female segment of consumer is Table 3 Logit parameter estimates. Variables Coeff. z Headache cause: sulphites 1.47 1.38 Naturalness 0.62 nnn 4.10 Subjective Wine Knowledge 0.09 0.50 Subjective Sulphite Knowledge 0.24 1.31 Age 0.03 nn 2.45 Gender 0.16 nn 0.42 Wine Purchasing Frequency 0.36 1.59 Sulphite Objective Knowledge 2.59 n 1.89 Number of observations 240 Likelihood-ratio chi 2 40.65 nnn po0.01. nn po0.05. n po0.1. Table 4 Tobit parameter estimates on bids for NAS wine. Variables Spain Italy Coeff. t Coeff. t Headache cause: sulphites (DUMMY – 0¼No) 0.58 n 1.84 0.46 nnn 3.46 Naturalness (1–7¼lower to higher) 0.22 nnn 3.00 0.12 nn 2.47 Subjective Wine Knowledge (1–7¼ lower to higher) 0.33 0.38 0.04 0.69 Subjective Sulphite Knowledge (1–7 ¼lower to higher) 0.02 0.26 0.01 0.21 Age (Continuous) 0.02 nnn 3.20 0.005 1.20 Gender (DUMMY - 0¼female) 0.38 nn 2.10 0.28 nnn 2.63 Wine Purchasing Frequency (1–4¼ lower to higher) 0.28 nnn 2.82 0.08 n 1.51 Wine Variety (0 ¼white; 1 ¼red) 0.12 0.68 0.11 1.06 Likelihood-ratio chi 2 31.88 31.20 Number of observations 120 each country nnn po0.01. nn po0.05. n po0.1. Table 5 Tobit parameter estimates on bids for NAS wine. Variables Spain Italy Coeff. t Coeff. t Headache cause: sulphites (DUMMY – 0¼No) 0.58 n 1.83 0.44 nnn 3.26 Naturalness (1–7¼lower to higher) 0.22 nnn 2.98 0.11 nn 2.28 Age (Continuous) 0.02 nnn 3.19 / / Gender (DUMMY –0¼female) 0.38 nn 2.13 0.25 nnn 2.45 Wine Purchasing Frequency (1–4¼ lower to higher) 0.28 nnn 2.88 0.09 n 1.70 Pseudo R 2 0.198 0.226 Likelihood-ratio chi 2 31.73 29.05 Number of observations 120 each country nnn po0.01. nn po0.05. n po0.1. M. Amato et al. / Wine Economics and Policy 6 (2017) 146–154152 more likely to pay premium prices for healthier and environmentally friendly food products in comparison to men. Comparing results with the Italian preferences across gender, we find the opposite effects. In other words, in Italy men are disposed to pay extra premium price for NAS wines in comparison to Italian women, in line with D’Amico et al. (2016). With respect to age, our findings suggest that younger Spanish consumers are willing to pay higher premium prices than the elderly people. This finding can be explained by the frequency of wine consumption at this age. As Garcia et al. (2013) reported, a vast majority of younger Spanish people consume wine only in special occasions and, as a consequence, they prefer to spend more money on high quality wines. In the same line, León-Muñoz et al. (2015) found that alcohol consumption among older adults in Spain is consistent with the traditional Mediterranean drinking pattern (moderate alcohol intake, mainly from wine and during meals), which decreases with age (MSSSI, 2011). Another important factor to mention is that in comparison to the Italian elderly people who are willing to pay premium prices for NAS wine, in Spain one of the most important attribute, beside price, is the origin and more specifically the protected designation of origin quality certification and organic (Brugarolas et al., 2005;De Magistris et al., 2014). Spanish elderly people might value the NAS type of wine as an additional production process as a result they do not give any additional value to it. The “Wine Purchasing Frequency”exerted a positive and significant effect in both countries. Precisely, consumers who purchase more wine are more willing to pay a price premium for wines without sulphites. “Naturalness”variable can be reasonably considered as a proxy for the health attitude of consumers. This parameter was found to be a relevant factor in inducing consumers to pay a premium for NAS wine in both country (an increase of one rating on the scale augment WTP of 0,22€in Spain and 0,11€in Italy), and, moreover, it was one of the variable found significant that explained consumers’participation to the auction. This result is consistent with previous economics studies on organic food, if non-added sulphites wines are perceived so, confirming the positive effect of attitude toward naturalness on consumer choices thus influencing the intention to purchase organic food (Hsu and Chen, 2014; Onyango et al., 2007). Although the presence of consumers who are not willing to pay any premium price in both countries is small, more than 80% of the total sample are disposed to pay average extra premium prices of 1,17€70,76€in Spain and 0,90€70,50€in Italy. This is a promising result demonstrating the existence of a seemingly profitable market for this category of wine in both countries. In addition, results show that about 20% of Italian and Spanish consumers appear to recognize a premium price of at least 2€for this particular category of wines. Even though the size of this group of consumer is small, ad hoc marketing campaigns can be built in order to allow them to enter the market. Therefore, from a practitioner's point of view, this implies the necessity of a market segmentation to identify the characteristics of this group with a high willingness to pay, as it has been highlighted in this study. However, there are several limitations inherent and worth mentioning. First and foremost, there might be social desirability issues as respondents often seek to accomplish researchers rather than reveal their true preferences. The final number of subjects involved in the experimental treatments was quite limited (120 per country), although this is an acceptable sample size in the literature, it would lend more credibility if we had a larger sample. A larger sample would have given a more precise definition of the socio-demographical differences (age and gender) among the two countries. Moreover, the specific characteristics of the auction protocol, the BDM methodology, the absence of monetary endowment, undoubtedly influenced respondents. In addition, this research should foster new studies in other European countries culturally distant from Italy and Spain in order to better categorize and analyse on how the “sulphite-in-wine”issue affects consumers, analyse and define the market share of this niche and most importantly define the psychographic variables that will efficiently segment the market to direct ad hoc promotion campaigns designed to elicit a price premium. Conflict of interest None declared. Acknowledgements This research has been carried out thanks to the financial support of the International Organization of Vine and Wine (OIV). References Becker, G., DeGroot, M., Marschak, J., 1964. Measuring utility by a singleresponse sequential method. Behav. Sci. 9, 226–232. Bernard, J.C., Bernard, D.J., 2010. Comparing parts with the whole: Willingness to pay for pesticide-free, non-GM, and organic potatoes and sweet corn, J. Agr. Resour. Econ. 35 (3), 457–475. Bohm, P., Lindén, J., Sonnegård, J., 1997. Eliciting reservation prices: beckerdegroot-marschak mechanisms vs. markets. Econ. J. 107, 1079–1089. Bregoli, I., Hingley, M., Del Chiappa, G., Sodano, V., 2016. Challenges in Italian wine routes: managing stakeholder networks. Qual. Mark. Res. J. 19 (2), 204–224. Brugarolas, M., Martínez-Carrasco, L., Martínez, A., Rico, M., 2005. Determination of the surplus that consumers are willing to pay for an organic wine. Span. J. Agric. Res. 3 (1), 43–51. Burgstahler, A.W., Robinson, M.A., 1997. Fluoride in California wines and raisins. Fluoride 30 (3), 142–146. Chengchu, L., Ruiying, C., Yi-Cheng, S., 2006. Bactericidal effects of wine on vibrio parahaemolyticus in oysters. J. Food Prot. 69 (8), 1823–1828. Combris, P., Bazoche, P., Giraud-Héraud, E., Issanchou, S., 2009. Food choices: what do we learn from combining sensory and economic experiments? Food Qual. Prefer. 20 (8), 550–557. Corrigan, J.R., Rousu, M.C., 2006. Posted prices and bid affiliation: evidence from experimental auctions. Am. J. Agric. Econ. 88 (4), 1078–1090. Costanigro, M., McFadden, D.T., Kroll, S., Nurse, G., 2011. An in‐store valuation of local and organic apples: the role of social desirability. Agribusiness 27 (4), 465–477. Costanigro, M., Appleby, C., Menke, S.D., 2014. The wine headache: consumer perceptions of sulfites and willingness to pay for non-sulfited wines. Food Qual. Prefer. 31, 81–89. M. Amato et al. / Wine Economics and Policy 6 (2017) 146–154 153