Developing functional food with increased health benefits and acceptable sensory properties has been one of the main objectives of the food industry for the past 20 years. In Europe, consumers' demand for healthier food products is rising continuously with special concern toward nutritional aspects. Increased awareness in health issues has led to an increase consumption of functional dairy products, and more specifically yoghurts enriched with nutrients. Yoghurt is obtained through a fermentation process of milk active bacterial cultures and by-products ( Serafeimidou et al. , 2012 ) that can be used as a vehicle for probiotic cultures ( Lourens-Hattingh & Viljoen, 2001 ), and it is associated with a healthy dietary pattern ( Cormier et al. , 2016 ). The consumption of sufficient amounts of yoghurt live microorganisms ¹ promotes health benefits ( WHO/FAO, 2001 ). Such benefits include a reduction risk of type 2 diabetes ( Diaz-Lopez et al. , 2015 ), reduction in weight gain ( Mozaffarian et al. , 2011 ) and prevention of certain cardiovascular diseases ( Astrup, 2014 ).

Commercial yoghurt has created a widely segmented market offering a variety of functional products whose nutritional and health benefits are reported in nutritional claim (NC) and health claim (HC). One of the hurdles in the success of these products is that a nutritional or/and a health benefit delivered by a functional food is a credence attribute ² which cannot be easily recognized even after repeated consumption. Products that are characterized by credence attributes may result to asymmetric information. In order to fill this gap, the European Union (EU) has introduced regulations ³ with a main objective to reduce the asymmetric information, guarantee truthful and understandable claims by the 'average consumer' and aim in healthier informed food choices.

Within the European context, recent studies on consumer preferences and willingness to pay (WTP) report distinct preferences among consumers from different countries for NCs and HCs. More specifically, a study of Van Wezemael et al. (2014) investigates consumer preferences for nutritional and health claims (NHCs) on lean beef steak in four EU countries (Belgium, France, the Netherlands and United Kingdom) found that in Belgium, the Netherlands and France, NHCs on saturated fat yielded higher utilities than claims on protein and/or iron, while the opposite was found among consumers in the UK. On the other hand, studies conducted in different countries covering a broad range of issues including consumers' attitudes and perceptions ( Urala & Lahteenmaki, 2007 ; Masson et al. , 2016 ), preferences ( Betchtold & Abdulai, 2014 ; Annunziata & Vecchio, 2016 ) and WTP for functional food products with NHCs ( Hellyer et al. , 2012 ; Hirogaki, 2013 ; Cavaliere et al. , 2015 ; de-Magistris & Lopez-Galan, 2016 ; Jurado & Gracia, 2017 ; Lopez-Galan & de-Magistris, 2017 ) found that health-conscious consumers have positive perception and are willing to pay premium prices for food products with NHCs. Others indicate that healthier perceptions and acceptance of functional foods with NHCs depend on the ingredients and their combination within the product. More specifically, Landström et al. (2009) and Cox et al. (2011) found that consumers have healthier perceptions and positive acceptance of food products when the bioactive ingredient is 'naturally added' or inherited. Results are also confirmed by previous studies ( Chase et al. , 2009 ; Krutulyte et al. , 2011 ) who found that consumers have negative perceptions toward yoghurts enriched with omega 3 because the combination of this ingredient is perceived to be artificial. On the other hand, Krutulyte et al. (2011) and Ares & Gámbaro (2007) found positive attitudes on dairy products enriched with calcium rather than antioxidants and iron since the functional component (calcium) is 'naturally' inherited to this product category. Thus, not all type of functional nutrients within the same product category may be perceived positively and generate premiums.

Concerns for healthier food choices have also derived consumers to pay premium prices for functional food with NHCs. In the case of the Italian yoghurts, Carlucci et al. (2013) found premium prices for the added fiber (+32.33%), probiotic (+24.45%) and calcium (+27.18%) attributes. In the case of fruit beverages, Szathvary & Trestini (2014) found positive effects for NCs (5.7%) and for HCs (20.6%), respectively. In the same line, Bimbo et al. (2015) found premium prices for added fiber (0.183€/L) and added vitamins (0.044€/L) to the UHT-treated milk. Barreiro-Hurle et al. (2010) found higher WTP for HCs related to health risks. This result was also confirmed by Annunziata & Vecchio (2013) and Ares et al. (2010) who reported that in the case of dairy products, consumers mostly prefer those HCs that reduce the risk of osteoporosis and cardiovascular diseases.

In spite of the fact that there are many studies focused on consumer acceptance of NHCs, to the best of our knowledge, there is a scant literature assessing (among other attributes) the analysis of specific NHC premium prices on food products in Spain. Hence, the main objective of this work is to fill these gaps and assess the market valuation of (among other attributes) specific NHCs for yoghurts in Spain. This is the first contribution of this paper. To achieve this objective, we used the hedonic price approach as the most appropriate analytical tool due to high range of yoghurts retail prices in the Spanish market. Moreover, hedonic approach has the advantage to explain how yoghurt prices vary depending on NHCs. In Spain, the hedonic price approach has been used by Cabrera et al. (2015) on the extra virgin olive oil, Gracia & Pérez y Pérez (2004) on veal and Sanjuán-Lopez et al. (2009) on saffron. Nevertheless, the present study is the first that analyzes NHCs on yoghurt in Spain using this methodology. Thus, this is the second contribution of our paper. Finally, through an identification of the individual effect of each NHC on the overall price of yoghurt, our results will provide guidance for food manufactures and distribution presenting investment opportunities in the development and marketing of functional food.

The first step includes the estimation of the Box-Cox regression. Table 4 provides the results indicating that two possible functional forms are not rejected. If we consider that that a joint linear transformation is always rejected then we choose the semi-logarithmic log-lin functional form for further analysis.

Since the Box-Cox transformation might lead to un-conclusive results, very common in the hedonic price empirical literature, in addition, Vuong´s test was applied (Table 5). Likewise, the results of the Vuong test indicate that the semi-logarithmic (log-lin) functional form is suitable.

In line with Muñoz et al. (2014) and Cabrera et al. (2015) additional statistical parameters have been performed to verify the functional form that best fits the model. Two likelihood ratio statistics were performed to verify if the semi-logarithmic (log-lin) functional form was significantly preferred to a semi-logarithmic (lin-log) or a double-logarithmic (log-log) specification, respectively. Results clearly indicated that the adopted functional form was superior to the other two alternatives. Goodness-of-fit ( R ² = 0.66) and the adjusted R ² = 0.64 were higher and significant (F-statistic < 0.01) while the Akaike and Schwarz information criterion was lower than those of the log-lin and log-log model, respectively. Moreover, the model showed no problem with the normality of residuals (probability of Jarque-Bera statistic of 0.00). The heteroscedasticity was tested by the Breusch-Pagan-Godfrey and White test statistic and the null hypothesis of the homoscedasticity in the error term was rejected (probability F-statistic 0.00), that indicates homoscedasticity problems. White's robust estimation strategy to obtain the parameter standard errors was used to solve this problem. The estimated hedonic price function parameters are shown in Table 6.

When analyzing, the magnitude of the coefficients must be understood as the percentage change of the price variable in view of the change in a unit of the independent variable. In the case of a continuous variable this percentage change can be determined as:

The main objective of this study was to measure the market value in terms of implicit prices given to yoghurts with nutritional and health claims in the Spanish market. Results show that yoghurt is a highly differentiated food product. The market competition is based in quality attributes related to quantity, brand, type of retailer, the type of yoghurt ( e.g. , natural, with fruits-flavors, bifidus, Greek etc.) and nutritional and health claims. The applied hedonic function provides a measure of the market value of these attributes and investigates some important features of the Spanish yoghurt industry to offer insights on certain competitive strategies.

Results showed that yoghurt prices are positively affected by private brands purchased at neighborhood stores in comparison to leader brands purchased at hypermarkets. This is an expected result because in the Spanish market the neighborhood store ('Mercadona') has the highest market share (23.6%) in comparison to the rest of supermarkets, including also the hypermarket of 'Carrefour' (8.5%) ( Berengueras, 2017 ). In addition, while in the rest of super-hypermarkets the presence of private brand did not exceed 34.1%, the percentage of private brands in the neighborhood store ascended to 56.6% in 2017 ( San Esteban, 2017 ). With respect to the different types of yoghurts present in the Spanish market, the Greek type received the highest premium price followed by drinking yoghurts. Drinking yoghurts have lately become very popular in the local market for being enriched with different types of vitamins ( e.g. , B6, B12 etc.) and are perceived as healthier yoghurts. Consistent with Bonanno´s (2013) findings in the Italian yoghurt market, consumers seem to prefer drinking yoghurts over regular ones, in particular with regard to functional alternatives. Bifidus, natural and yoghurts with fruits and flavors have negligible effects on the Spanish yoghurt prices.

Surprisingly, negligible effects on yoghurt prices are seen for the nutritional claims related to fat-free, no added sugar and fiber contents. These results are in contrary to consumers' preference growth for low-calorie and free-fat food products, and in contradiction to previous studies who state that consumers have stronger preferences for simple ( Bitzios et al. , 2011 ) and more familiar claims ( Lahtenmaki et al. , 2010 ). The neglected valuation of the free-fat nutritional claim is partially in line with Bimbo et al. (2016) who found negative marginal price for zero-fat (-1.9%) yoghurt attributes in Italy and in line with Carlucci et al. (2013) who found negative but not significant relationship between the low-fat attribute and yoghurt's prices in Italy. In general, our results are consistent with Van Wezemael et al. (2014) and Krystallis & Chrysochou (2011) who found that consumers across five different countries have very heterogeneous preferences on nutritional claims. More precisely, consumers from Belgium, the Netherlands, France and Greece give higher value on NCs related to fat content and saturated fat while it is the opposite for consumers in the UK. With respect to the fiber content nutritional claim our result is in line with Ares & Gambaro (2007) who found that fiber added to yoghurt are perceived as interfering with the naturalness and healthiness of the product, and this may reduce consumers' acceptance and price.

Another reason that might influence the negative and not significant impact of the three most pre­sent NCs (' N_FreeFat ', ' N_NoSugar ' and ' N_Fiber ') in the yo­ghurt Spanish market is that since these types of claims have been introduced long time ago they might be in the maturity stage of the product lifecycle. To the contrary, two nutritional claims related vitamin B6 and protein contents that were later introduced in the national yoghurt market have positive influence on price and receive premiums. This outcome seems understandable since both claims are considered to be innovative, are still in the growth stage of the product lifecycle, are perceived as healthy attributes on yoghurts and have a limited competition in the local market (only 'Danone').

On the other hand, findings report that health claims, outperform nutritional claims leading to higher premium prices in the Spanish market. In particular, the highest premium price is received by yoghurts that bare the cholesterol claim. These estimates are consistent with other studies who found that product claiming to prevent cardiovascular diseases by lowering or controlling cholesterol levels are well accepted by dairy product consumers ( Ares & Gámabro, 2007 ; Landström et al. , 2007 ). Moreover, Marette et al. (2010) found positive WTP for cholesterol HCs even for participants without high cholesterol problems. HCs regarding lactose digestion receive positive valuation being the second mostly valued after the cholesterol claim. This is an expected result due to the fact that in 2015 the Spanish Society of Digestive Pathology in collaboration with the Spanish Society of General and Family Physicians found that between 30 and 50% of the Spanish population suffers from lactose intolerance ( Argüelles-Arias et al. , 2015 ).

In contradiction to the calcium content nutritional claim who negatively affects yoghurt prices, the calcium type of HC was found to have a positive impact and received an important premium. This result is in contrast with Szathvary & Trestini (2014) and Barreiro-Hurle et al. (2010) , who found negative interaction effects when nutritional and health claims are labeled together. Premium prices were also received by yoghurts bearing the vitamin B6 joint NHCs. Lastly, the fiber health claim was totally neglected, therefore the attribute did not receive any premium price. This result was similar to Ares & Gámbaro (2007) who found that consumers show positive attitudes on dairy products enriched with calcium rather than fiber since the functional component (fiber) is 'artificially' inherited to this product category.

Our finding imply that NHC matters in determining a yoghurt's premium price due to a differentiation strategy of processors or manufacturers which should take into account the growing consumer concerns on healthier food products and heterogeneous preferences. Especially in the yoghurt market, health enhancing product differentiated by functional food ingredients seems to be the most profitable way of product differentiation. Even though, certain nutritional claims had no effect on yoghurt product prices, a profitable strategy may be to introduce them accompanied by the corresponding health claim that exactly defines the benefits of that nutrient on our health ( e.g. , the case of enriched with calcium nutritional and health claim). Further research is needed in the future to better understand Spanish consumer preferences towards yoghurts with nutritional and health claims. Future research may analyze to what extent consumer preferences and willingness-to-pay for these specific attributes are related to price structures and provide guidance to food manufacturers in deciding whether or not to invest in the development of marketing strategies. This constitutes our future research. Even though this study's interest is more limited to Spanish market, the methodology used can be replicated in other countries.