Abstract
Purpose
Ethical consumption is on the rise amidst concerns about the environmental and health impacts of industrial agriculture. In light of increasingly complex food choices, alternative food networks have emerged. However, their success depends on a deeper understanding of the product attributes that guide (ethical) consumer decisions. This study focuses on the preferences of consumers when choosing and buying fresh vegetables in Romania.
Design/methodology/approach
This study employs a discrete choice experiment to determine how consumers make trade-offs across a set of product attributes, such as local origin and production method.
Findings
The study analysis sheds light on the importance of food attributes relevant to ethical consumers. The main barrier to making an ethically driven choice is convenience. While local production remains of lower importance than the production method, the authors show that the Romanian consumers surveyed strongly prefer non-certified “traditional” vegetables over certified organic products.
Originality/value
This study is pioneering with a state-of-the-art discrete choice setting looking at a set of product attributes that reflect the demand of ethical consumers in an understudied transitional context. The authors go beyond the current debate on the trade-off between organic vs local food labels by introducing traditional small-scale production as a separate attribute level. The food attribute preferences of different consumer segments and a market simulation offer relevant insights how to market fresh vegetables to health- and environmentally-conscious urban people.
Keywords
Citation
Möllers, J., Bäuml, T. and Dufhues, T. (2022), "Understanding the market potential of products from alternative food networks in a transition economy—a discrete choice experiment", British Food Journal, Vol. 124 No. 13, pp. 183-199. https://doi.org/10.1108/BFJ-08-2021-0925
Publisher
:Emerald Publishing Limited
Copyright © 2022, Judith Möllers, Theresa Bäuml and Thomas Dufhues
License
Published by Emerald Publishing Limited. This article is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of this article (for both commercial and non-commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at http://creativecommons.org/licences/by/4.0/legalcode
1. Introduction
Food choices are becoming increasingly complex as consumers grow more aware of the link between food production and the environment, globalisation and food safety, and move away from convenience consumption (Maehle et al., 2015). Alternative food networks (AFN) emerged in the 1990s as a response to perceptions that the industrial food system had become increasingly unethical (Edwards, 2016). AFNs involve both alternative food practices and means of distribution and include, for example, farm sales, farmers' markets, vegetable box schemes and community supported agriculture (Poças Ribeiro et al., 2021). Their shared property is that they appeal to ethical consumerism by offering food with certain attributes such as locally sourced, environmentally friendly and organically produced (Bazzani et al., 2017).
The rising importance of ethical consumerism can be observed, for example, in the growing consumption of organic food and willingness of consumers to pay a premium for organically labelled produce (Connolly and Klaiber, 2014). While organic agriculture was initially associated with local production and direct relationships between farmers and consumers, the global standardisation and industrialisation of organic food have significantly weakened this link. This has led to the rising interest in AFNs and triggered a shift in consumer preferences from organic to locally produced foods (e.g. Adams and Salois, 2010; Darby et al., 2008; Winterstein and Habisch, 2021; Bazzani et al., 2017) Consequently, food from local small-scale family farms may become a substitute for organic produce (Mugera et al., 2017).
Ethical consumerism is most often found in wealthier consumer segments of developed markets (Maye and Kirwan, 2010), but the number of more critical consumers is also rising in less-developed markets, such as in Europe's transition economies. Poças Ribeiro et al. (2021), for instance, found that consumers in Poland held critical views of mass-produced food. The corresponding rise of AFNs, described, for example, by Möllers and Bîrhală (2014) for Romania or by Birtalan et al. (2021) for Hungary, is reinforced by two factors. First, in many transition economies, demand for organic food is only just emerging and immature compared to Western Europe (Willer et al., 2021). Sometimes AFNs are therefore the sole supplier (e.g. Bryła, 2018, for Poland; Melović et al., 2020a, for Montenegro; Möllers and Bîrhală, 2014, for Romania). Second, if alternative suppliers are available, organic food is often still cheaper to obtain via AFNs than, for example, from specialised stores (see, e.g. Slavuj Borčić, 2020, for Croatia; Möllers and Bîrhală, 2014, for Romania).
Our paper is positioned in the strand of research exploring consumers' willingness to pay a premium for organic and locally produced vegetables. While this topic has been covered by others (e.g. Joya et al., 2022; Yeh et al., 2021; Meyerding et al., 2019), we aim to close a gap in these studies by providing a better understanding of consumers' preferences for food attributes relevant to AFNs and ethical consumerism in a transitional context. Taking into account the small-scale agricultural structure of many Eastern and South-Eastern European countries, one central argument of this study is that traditional small-scale production should be considered as a self-standing production mode that competes with certified local and organic production. We thus not only pick up the recent notion that consumers may prefer local production over organic production (Winterstein and Habisch, 2021), but respond to the call of Lang et al. (2014) to go beyond a simple organic vs local food production trade-off by including the question of whether the product comes from a small family farm or not. We hypothesise that traditionally produced products may outcompete certified local or organic production, as they implicitly (on a trust basis) combine their underlying qualities.
In our analysis of urban consumers, we apply a discrete choice experiment (DCE) of preferences for attributes of fresh vegetables. We ask how AFN relevant product attributes influence consumer choice and willingness-to-pay, and how the weight given to these attributes differs between (ethical) consumer segments. Our analysis thus reveals how consumers make trade-offs across a set of product attributes, which in our case are the production mode, local origin, convenience, visual appearance and price. Our empirical case refers to Romania, a country that has been considered a growing market for organic agro-food production and marketing. During the last decade, the cultivated organic area has more than doubled. However, in 2019, certified organic production still covered only 2.9% of agricultural area (Willer et al., 2021). Retail sales for organic products also remain at a low level because consumption is constrained by low incomes (Drăghici et al., 2016). In contrast, local produce from smallholder farms offered at farmers' and town markets for a reasonable price is a common source of fresh food products for Romanians. Often labelled as traditional production, Romanians, like many other consumers in Eastern and South-Eastern Europe, value this source of fresh food for its tasty, healthy and environmentally friendly products (Möllers and Bîrhală, 2014).
The high originality value and contribution of this study is thus threefold: First, the study is pioneering with a state-of-the-art DCE setting looking at a set of fresh vegetable product attributes that reflect the demand of ethical consumers in an understudied transitional context. Second, the study goes beyond the current debate on the trade-off between organic vs local food labels by introducing traditional small-scale production as a separate attribute level. Third, in their complex food choices, consumers vary in their preferences concerning product attributes. Hence, by grouping consumers according to their ethical-value profiles, we identify the food attribute preferences of different consumer segments. This part of the analysis is highly relevant for a better understanding of the market potential of the emerging AFNs in Romania. A market simulation that provides concrete insights into the market potential of higher-priced local products of certified organic quality vs products from traditional production completes the analysis.
2. Brief review of the literature: food product attributes and their relevance for fresh vegetable food choices
We used the 11 food values that influence food choices identified by Lusk and Briggeman (2009) as a starting point for this review and to select attributes for our DCE. We structured them into three categories: visual and sensual appeal (appearance and taste); access-related factors (price and convenience); and health and ethical consumerism-related factors (health, naturalness, safety, nutrition, tradition, origin and fairness).
2.1 Visual and sensual appeal
Visual appearance and taste are important categories that guide “hedonic” food choices (Maehle et al., 2015). Taste is often revealed only after purchase, but consumers tend to believe that organic products have a more natural taste (Hjelmar,2011). Visual appearance, according to Moser et al. (2011), is a relevant attribute that shapes purchasing decisions for vegetables, although less so for consumers with an ethical consumerism profile. In Romania, too, consumers consider visual appearance an important factor in their daily food choices and an unattractive appearance was identified as a possible barrier to the consumption of organic food in the country (Popa and Dabija, 2019).
2.2 Access-related factors
Ethical consumers tend to be less concerned about food prices. Despite this, price usually remains an essential purchasing criterion, especially for food products such as vegetables, where the emotional response involved is small, but the price sensitivity is higher (Maehle et al., 2015). In Romania, most people prefer to buy conventional food products at low prices and price is a crucial barrier to organic food purchases (Bozga, 2015; Popa and Dabija, 2019).
Unlike Lusk and Briggeman (2009), we link convenience to the ease with which the product is accessed instead of how it can be cooked and/or consumed. This is in line with other studies, which found that consumers prioritised fitting their shopping into their daily schedules and preferred nearby, convenient shops (Hjelmar, 2011). For AFNs, the literature stresses that the typical features of these networks cause inconvenience: both the limited choice of products (Cone and Myhre, 2000) and the inconvenience of having to pick them up at certain times and places (Flora and Bregendahl, 2012) have been identified as disadvantages. However, as shown for Romania by Möllers and Bîrhală (2014), there are certain segments of ethical consumers who are willing to sacrifice the convenience of supermarkets to access healthy food through AFNs. This is all the more relevant given that fresh organic produce is still difficult to find on Romanian shelves (Petrescu et al., 2017).
2.3 Health and ethical consumerism-related factors
Health concerns have become one of the most decisive factors in making food choices (Ghvanidze et al., 2017). The key attributes here are food safety and nutrition. For fresh vegetables, it is difficult for the consumer to judge these attributes, but decisions are guided by assumptions or labels that promise, for example, a pesticide-free product. Health concerns also drive organic food purchases in transition economies (Melović et al., 2020b; Birtalan et al., 2021). In Romania, health reasons are an important factor in organic and AFN food purchases (Möllers and Bîrhală, 2014).
Naturalness directly refers to the production method and is described by Lusk and Briggeman (2009) as the extent to which food is produced without modern technologies. Often, this attribute is closely linked to health and the environment; organic production, for example, comes with the promise of fewer toxic substances. The production method is sometimes indicated by labels, but it may also be simply linked to the source of the product. In both cases, consumers usually do not know the details, and conclusions about naturalness are mostly trust-based (Hjelmar, 2011). In Romania, products from smallholder farms are often labelled as traditional or natural, implying that they stem from low-input production with a high naturalness value (Möllers and Bîrhală, 2014). Indeed, much of the production in Romania is close to organic standards, because small farms still use natural fertiliser as their main input (Simon and Borowski, 2007).
The attribute origin indicates where agricultural commodities are grown. A local origin, similar to an organic or low-input production method, is linked to a perception of freshness and better taste compared to imported products. At the same time, consumers appreciate the environmental benefits of less transportation (Hjelmar, 2011). However, how consumers define local varies by their level of use and experience with local foods. When discussing local foods it is therefore important to understand that the concept is evolving and expanding (Lang et al., 2014). Some Romanian consumers perceive local products as superior to certified organic products (Petrescu et al., 2017). One explanation may be that they are often sceptical of certification processes (Bozga, 2015). This is consistent with consumers in other transition countries such as Slovakia (Musova et al., 2018) or Poland (Bryła, 2018). Goszczyński and Wróblewski (2020) explain this as a lack of trust in experts and institutions.
Tradition is linked to the desire to preserve traditional values in the countryside (Hjelmar, 2011). In Romania, such traditional values are supported by the close family links that many urban consumers maintain to the countryside, and through which they are also supplied with home-grown foods (Bozga, 2015). Another aspect that is interlinked with affection for the countryside is fairness (the extent to which all parties involved in the production of the food equally benefit). Fairness in a wider sense reflects the wish to make buying decisions that contribute to leaving a better world for the next generations or supporting animal welfare. In a narrower sense, people may wish to support small farmers as shown by Slavuj Borčić (2020) in Croatia and Möllers and Bîrhală (2014) in Romania.
3. Methodology, study design and data
We apply a DCE to a consumer dataset from Romania. With this, we go beyond most research in the region, which has applied qualitative descriptive or factor or cluster approaches.
3.1 Discrete choice experiment
In his landmark article from 1966 on consumer theory, Lancaster postulated that consumer choices are directed towards a combination of product attributes, rather than the products themselves, to maximise their utility. In line with this, DCEs model consumers' preferences regarding product attributes in a random utility framework, whereby the experiment measures stated preferences and allows for willingness-to-pay to be obtained indirectly (McFadden and Train, 2000).
DCEs are widely used in marketing research to investigate consumer preferences. They are designed to measure the strength and relative importance of attributes in goods and services (Hauber et al., 2016). In the experiment, respondents choose their most preferred option—the one with the highest utility—from a given set of alternatives. In doing so, they assess the total value of a given product by combining the individual values provided by the particular level of each product attribute relevant to consumers. This approach realistically resembles an actual purchasing situation (Rao, 2014). Because it is neither possible nor useful to use all existing attributes of a product in a DCE, attributes need to be reduced to a manageable size. Otherwise, the DCE would become overly complex and easily lead to respondent fatigue, which in turn results in inconsistent and random choices (Bennett and Blamey, 2001).
In our experiment, we distinguished five key attributes that are relevant for the purchase of carrots, potatoes and bell peppers, and that reflect the assumed desires of ethical consumers according to the food values discussed in Section 2. The production method and origin are the most important attributes in this regard. Furthermore, in addition to the price, we considered appearance and convenience. As shown in Section 2, AFN consumers are not always happy with inconvenient shopping arrangements (such as when boxes have to be picked up from faraway locations) or with non-standard product appearances (resulting, for example, from not using pesticides or wasting non-perfect products (Cone and Myhre, 2000; Möllers and Bîrhală, 2014)). For the five attributes, 13 attribute levels were identified (see Table 1).
The method of production has three attribute levels: conventional, traditional and certified organic production. Although products sold through AFNs are mostly linked with organic production, in Romania they are often sold as traditionally produced products, a label that consumers are fully familiar with. Traditional production implies that, although there is no formal certification or strict production rules, the products are produced in a traditional way by small-scale family farms using only natural fertiliser and no chemical pesticides. The inclusion of traditional production into the experiment is an innovative extension that covers important food values related to health and ethical consumerism such as, in particular, naturalness, origin and tradition.
Although there is no clear definition of the term “local food”, it often refers to food produced and sold within a geographical area, regional border or delimited to a radius of up to a few hundred kilometres (Meyerding et al., 2019; Bazzani et al., 2017). As discussed in Section 2, some consumers perceive local food as organic and vice versa, while others relate it to small-scale production (Winterstein and Habisch, 2021; Bimbo et al., 2020). Defining the origin attribute was therefore a challenge. Based on discussions and feedback we received during the pre-test phase, we decided to use a simple indicator of distance in kilometres to define a local origin. We included two attribute levels: (1) Locally produced food produced within a range of 100 km and (2) Non-locally produced food coming from beyond this threshold.
Convenience was measured using a time indicator that reflects the time consumers need to reach their respective points of sale. We distinguished three levels: 45 min, because some products sold through AFNs must be picked up from distant places (Möllers and Bîrhală, 2014); 30 min, which is the typical distance to a hypermarket at the edge of town or a specialised organic shop, which are rare in Romania; and 15 min, which is the typical length of time needed to get to a local farmers' or town market or supermarket, the most convenient choice in close proximity to most urban consumers.
The attribute appearance was represented by two pictures of each vegetable. One picture showed a perfect appearance and one a less perfect appearance (see Table 1). The three chosen price levels reflect the range of prices found in different outlets for vegetables in urban Romania. Prices were collected at farmers markets, supermarkets, hypermarkets, specialised organic shops, and AFN selling points. For the pooled analysis of the three vegetables, the levels were later aggregated into categorical values (low, middle and high).
The relevance of the chosen attributes, the empirical realism of the levels and the visualisation of the choices were crosschecked and validated with a small sample of urban consumers. An example of the presented choice tasks is shown in Table 1.
We used XLstat to generate a d-optimal fractional factorial design with nine profiles for each vegetable studied. Each respondent was randomly assigned to one of the three vegetables: potato, carrot or bell peppers. The nine profiles were divided into 10 choice tasks with three profiles each. Thus, each of the 318 participants responded to 10 choice sets [1] with three product profiles (out of nine product profiles), resulting in 9,540 observations. A DCE simulates the purchasing process and assumes that consumers choose the alternative that provides the highest utility to them. The overall utility can then be decomposed into separate utilities for the attributes of the presented product. Our estimation was performed with effects-coded attributes based on a conditional logit model, also known as random utility model, calculated using the clogit command in STATA 17 [2]. To confirm the robustness of the results, we compared the results with a mixed logit and generalised multinomial approach (available on request). For more information on the theory and technicalities of the choice based random utility modelling we refer to the works of Lancaster (1966), McFadden and Train (2000), Hauber et al. (2016) and Lancsar et al. (2017).
3.2 Data
We collected data from a sample of 318 urban consumers in spring 2016 in two major cities in Romania: the capital Bucharest and Timisoara, the third largest city in the country. Urban adult consumers (at least 18 years of age, who regularly do the grocery shopping for their household) were randomly approached at various public places in and around city centres, such as public green spaces or shopping areas. With this approach we aimed to target the typical urban consumer. When several potential respondents were available at the same time, the interviewer favoured the respondent that differed from the previous respondent with regard to age and gender.
The survey started with the DCE, followed by a basic questionnaire on economic and demographic information and attitudinal statements on health, the environment and convenience. Table A1 in Appendix presents an overview of the structure of the sample along with key socio-economic characteristics as well as the questionnaire itself.
4. Consumer preferences for fresh vegetable products: DCE results
4.1 Estimation results
Table 2 presents the conditional fixed effects logit base model together with the relative importance of each of the five attributes. We also calculated the same models separately for the three vegetables used in the experiment as well as for income groups (Table 2).
In the base model, the convenience attribute is most important in determining consumers' purchasing decisions for vegetables. The relative importance of time convenience was 30%. As shown in Table 2, a short time to the selling point has a high utility, while the inconvenient option of 45 min turned the utility heavily negative. Upper income groups are more sensitive to inconvenient shopping times.
The production method was almost as important as convenience (27%). When looking at the three vegetables separately, the production method was more important for carrots and bell peppers than for potatoes; for carrots, this attribute was the most important (Table 2). Conventional production has a strong negative utility. It is interesting to note, however, that consumers clearly preferred traditional over certified organic production. The preference for traditional production was particularly high for the poorest income group (Table 2). This reflects the fact that traditionally produced products are not only valued for their quality but their affordable price.
The visual appearance of the vegetables was the least relevant attribute, with a relative importance of 8%. Price and origin were of medium importance, with a relative importance of 20 and 15% respectively (Table 2). Interestingly, when it comes to the individual vegetables, the importance of price drops significantly for carrots, to the level of appearance, the least important attribute (Table 2). However, it is not clear why price is of less importance for carrots.
This study is specifically interested in ethical consumerism as a driver of demand for organic, traditional and local produce. Table 3 depicts three (partly overlapping) ethical consumer segments according to their stated interest in health, the environment and fairness towards producers. It can be seen that, across the three models, the production method was the most important attribute for ethical consumers. For the comparison group of less conscious consumers, convenience remained the most important attribute, followed by the production method and price. Organic production was of no interest to the less conscious consumer segment. For the ethical consumer segment, however, the utility of organic production was closer to, but still lower than traditional production. The highest utility for organic products was found for consumers who stated an interest in environmentally friendly production. Conventional production had a heavy negative utility for all consumer groups, especially ethical consumers.
For ethical consumers, convenience was the second most important attribute. Price was the third most important attribute for health and environmentally conscious consumers. Here, however, the relative importance was almost the same as for the origin attribute, which plays a much bigger role in the ethical consumer segment. This is particularly true for the group that holds fairness high in their ethical considerations, with 21% relative importance for the attribute origin, but only 12% for the attribute price. The choices made by this group in the experiment clearly reflect their desire to act in a fair way towards producers. If these consumers can buy local products, the price is less important. Fairness-oriented consumers were also least interested in the appearance of the vegetables. For all consumers, the attribute appearance was ranked lowest, but as expected, it had a very low weight for ethical consumers.
The utility gain derived from obtaining a certified organic vegetable instead of a product from conventional agriculture, calculated from the difference in the coefficients for the two attributes, was 0.754. This utility gain is congruent with the utility loss from a price increase from a low to a high cost (0.753). Hence, according to this simple trade-off calculation, organic carrots or potatoes could cost RON 1/kg (the difference between the low price level and the high price level) more than those from conventional agriculture without causing a utility loss [3]. For bell peppers, the price could increase by RON 5.5/kg from the lowest price level. This price premium is also shown in willingness-to-pay for ethical consumerism attributes, which were calculated for the three vegetable products separately: a premium for organic production was evident mostly for carrots (0.93) and bell peppers (0.81) and much less so for potatoes (0.13) (Table 4). In contrast, the willingness-to-pay a premium for local origin was comparatively high for potatoes (0.52). The willingness-to-pay for local origin was RON 0.31 for carrots and RON 1.11 for bell peppers.
These price premiums cannot be realised if the shopping time becomes highly inconvenient (45 min) due to a significant loss in utility. The possible market shares of hypothetical products with different attributes are listed in Table 5.
While a conventional product from a regular supermarket seems rather unattractive, local and traditional products offered on a nearby market could compete best. As expected, organically certified but high-priced and inconvenient to access food has only a limited market potential, estimated at 12%.
4.2 Discussion
Similar to findings from other countries (e.g. Bimbo et al., 2020), our analysis shows that ethical consumerism food attributes—in our case, local origin and organic or traditional production methods—are highly relevant to the vegetable purchasing decisions of urban Romanians and, in particular, ethical consumer segments. However, contrary to other studies (e.g. Yeh et al. (2021), we found that the main barrier to making an ethically driven choice was not price, but convenience. Indeed, the key constraint in all of our market scenarios was inconvenience of access to the point of sale. The disutility of this could easily neutralise the possible price premium for both organically certified and local origin products. In contrast, the product's appearance played only a minor role in purchasing decisions. This is consistent with results from other transition countries, such as Poland, where the perfect appearance of agricultural products can have a negative connotation due to its association with the use of preservatives (Barska and Wojciechowska-Solis, 2018).
The production mode was found to be a very important purchasing criterion. The heavily negative utility of industrial production, compared to organic or traditional production, corresponds to the strong negative perception of mass-produced foods in other transition economies (Poças Ribeiro et al., 2021). However, a price premium for certified organic products would only be possible in a relatively small, ethical and in particular environmentally conscious consumer segment, and in higher income classes. This is also mirrored in other transition countries. For instance, farmers' markets in Prague usually do not offer organic food, as market organisers and consumers see organic food as unreasonably expensive (Fendrychová and Jehlička, 2018).
A key result of the study is that the surveyed consumers strongly preferred non-certified “traditional” vegetables over certified organic products. This supports our initial hypothesis that traditionally produced products outcompete certified organic products. While it may seem surprising that non-certified traditional production is preferable to certified organic food, Romanian consumers are known to highly appreciate the naturalness and superior taste of products from low-input small-scale farms sold on local markets or provided by relatives from rural areas. Another reason for this was proposed by Moser et al. (2011), who found that the organic attribute may be less preferred due to trust issues and the complexity of organic certification processes. These trust issues are also found in Western Europe but to a lower degree (Yeh et al., 2021). However, Güney and Giraldo (2019) show that consumers often simply do not differentiate, for example, between organic and free-range eggs in their purchasing decisions. Hence, traditional production may also be viewed as “organic enough” for a better price.
While we could not confirm that consumers prefer local production over organic production (Winterstein and Habisch, 2021), the attribute “origin” has significant weight among ethical consumers and, in particular, those that hold fairness towards producers high in their ethical considerations. In Romania, we interpret the preference for traditional production as an indication of the broader meaning of “local”. While the origin is not the most important as a self-standing attribute, it is certainly intrinsically linked to other attributes and, in particular, the traditional production mode. Its relevance must therefore not be underestimated. This was also found in other transition economies, such as Poland (Barska and Wojciechowska-Solis, 2018) and Montenegro (Melović et al., 2020a).
4.3 Limitations and future research
This study, of course, has its limitations. For example, survey respondents were younger, better educated and had higher incomes than the average Romanian consumer. However, while our survey is not representative and generalisations cannot be made beyond the sample, we believe it is typical of urban organic food consumers. Further limitations may lie in biases related to unobserved systematic differences in preferences, self-presentation and social desirability (Auger and Devinney, 2007; Hauber et al., 2016) or a gap between what consumers claim to care about and their actual purchasing behaviour (Auger and Devinney, 2007). Therefore, the DCE results must be interpreted with some caution.
An extended, representative survey could prove the robustness of our results. In particular we recommend conducting a broader investigation into the meaning of traditional agriculture and how it is interwoven with the concept of origin and other important food values. This investigation should go beyond looking at only one country: a cross-cultural study might not only provide a conclusion on the regional transferability of the results, but would gain insights into the impact of cultural, geographical and demographic factors on consumer segmentation (e.g. Ferraris et al., 2019).
5. Conclusions
Our paper offers three important contributions to the study of food purchases and ethical consumerism with implications for stakeholders in the AFN sector. First, this research widens the view on AFN relevant attributes by introducing the traditional production mode into a choice experiment. We demonstrate that food flagged as traditionally produced is highly preferred. The preference and willingness to pay a premium for organic and locally produced foods vary by consumer segments. Second, the study contributes to closing the gap on an under-researched region by looking at a transition economy where interest in ethical consumerism is only just emerging. The high valuation of traditional agriculture among consumers has received little discussion in the literature, but its relevance certainly extends to other (but not only) Eastern and South-Eastern European countries.
Third, the study provides relevant insights into the market potential of emerging AFNs. The market simulations show that higher-priced local products of certified organic quality have a market potential only among a small group of ethical consumers, but products from small-scale farms marketed as traditionally produced are highly competitive. Convenience may be more important than organic certification for gaining market share. This should provide important guidance for producers and retailers of sustainably produced local food on how to market and promote these products via AFNs and other marketing channels.
For Romania's small-scale producers and AFNs, bringing traditionally produced vegetables closer to urban consumers could open an excellent market opportunity with a substantial price premium. It appears that, in Romania, the “new organic” is not simply local but it is found in the wholesome nature of traditional production.
Conjoint study design
 |
Pooled base model (conditional fixed effects logit) and split models along products and income groups
| Attribute/levels | Pooled base model | Subsample split on products | Subsample split on income | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Coeff | Std.Err | P>|z| | 95% Conf interval | Potatoes | Carrots | Bell peppers | Above avg | Average | Below avg | ||||||||
| Coeff | Sig | Coeff | Sig | Coeff | Sig | Coeff | Sig | Coeff | Sig | Coeff | Sig | ||||||
| Production | Relative attribute importance: 27% | 20% | 29% | 29% | 23% | 27% | 41% | ||||||||||
| Organica | 0.158 | 0.047 | 0.001 | 0.065 | 0.251 | 0.080 | 0.209 | *** | 0.161 | * | 0.215 | *** | 0.163 | *** | 0.316 | * | |
| Traditional | 0.438 | 0.047 | 0.000 | 0.359 | 0.517 | 0.395 | *** | 0.426 | *** | 0.508 | *** | 0.320 | *** | 0.467 | *** | 0.595 | *** |
| Conventional | −0.596 | 0.040 | 0.000 | −0.688 | −0.504 | −0.475 | *** | −0.635 | *** | −0.668 | *** | −0.535 | ** | −0.630 | *** | −0.911 | *** |
| Origin | Relative attribute importance: 15% | 17% | 16% | 11% | 17% | 14% | 18% | ||||||||||
| Local | 0.279 | 0.030 | 0.000 | 0.219 | 0.339 | 0.365 | *** | 0.299 | *** | 0.221 | *** | 0.310 | *** | 0.272 | *** | 0.323 | *** |
| Non-locala | −0.279 | 0.030 | 0.000 | −0.339 | −0.219 | −0.365 | *** | −0.299 | *** | −0.221 | *** | −0.310 | *** | −0.272 | *** | −0.323 | *** |
| Convenience | Relative attribute importance: 30% | 32% | 27% | 29% | 40% | 27% | 19% | ||||||||||
| 15 min | 0.562 | 0.041 | 0.000 | 0.481 | 0.643 | 0.709 | *** | 0.509 | *** | 0.514 | *** | 0.703 | *** | 0.551 | ** | 0.251 | ** |
| 30 min | 0.025 | 0.030 | 0.404 | −0.034 | 0.084 | −0.040 | −0.014 | 0.138 | *** | 0.050 | −0.007 | 0.182 | ** | ||||
| 45 mina | −0.587 | 0.045 | 0.000 | −0.674 | −0.499 | −0.669 | *** | −0.496 | *** | −0.652 | *** | −0.753 | *** | −0.544 | *** | −0.433 | *** |
| Appearance | Relative attribute importance: 8% | 13% | 14% | 4% | 4% | 12% | 3% | ||||||||||
| Perfect | 0.150 | 0.037 | 0.000 | 0.078 | 0.223 | 0.282 | *** | 0.260 | *** | −0.088 | 0.069 | 0.238 | *** | 0.052 | |||
| Imperfecta | −0.150 | 0.037 | 0.000 | −0.223 | −0.078 | −0.282 | *** | −0.260 | *** | 0.088 | −0.069 | −0.238 | *** | −0.052 | |||
| Price | Relative attribute importance: 20% | 18% | 14% | 26% | 16% | 20% | 20% | ||||||||||
| Low | 0.338 | 0.038 | 0.000 | 0.263 | 0.413 | 0.305 | *** | 0.227 | *** | 0.517 | *** | 0.242 | *** | 0.362 | *** | 0.426 | *** |
| Mediuma | 0.077 | 0.026 | 0.003 | 0.027 | 0.128 | 0.149 | *** | 0.060 | 0.032 | 0.110 | ** | 0.077 | ** | −0.102 | |||
| High | −0.415 | 0.043 | 0.000 | −0.499 | −0.331 | −0.454 | *** | −0.286 | *** | −0.549 | *** | −0.352 | *** | −0.438 | *** | −0.324 | ** |
| Observations = | 9,540 | 2,550 | 3,630 | 3,360 | 2,580 | 5,460 | 810 | ||||||||||
| Log pseudo-likelihood = | −3066.03 | −806.73 | −1173.03 | −1028.72 | −815.55 | −1748.08 | −250.75 | ||||||||||
| Wald χ2 (8) = | 500.19 | 164.26 | 191.11 | 185.19 | 137.02 | 380.00 | 93.05 | ||||||||||
| Prob > χ2 = | 0.0000 | 0.000 | 0.0000 | 0.000 | 0.000 | 0.000 | 0.000 | ||||||||||
| Pseudo R2 = | 0.1224 | 0.1361 | 0.1176 | 0.1639 | 0.137 | 0.114 | 0.155 | ||||||||||
Note(s): aIndicates the omitted category; it should be noted that an effect-coded model compares differences to the mean attribute effect instead of the omitted category
Source(s): Own data
Results for base conditional fixed effects logit model split along ethical consumer segments
| Health-oriented consumersa | Environment-oriented consumersb | Fairness-oriented consumersc | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Attribute/levels | 0 | 1 | 0 | 1 | 0 | 1 | ||||||
| Production (attribute importance) | 23% | 32% | 23% | 37% | 24% | 35% | ||||||
| Organic | 0.023 | 0.292 | *** | 0.057 | 0.450 | *** | 0.046 | 0.384 | *** | |||
| Traditional | 0.447 | *** | 0.436 | *** | 0.406 | *** | 0.533 | *** | 0.449 | *** | 0.419 | *** |
| Conventional | -0.470 | *** | -0.728 | *** | -0.463 | *** | -0.983 | *** | -0.495 | *** | -0.804 | *** |
| Origin (attribute importance) | 13% | 17% | 13% | 17% | 12% | 21% | ||||||
| Local | 0.262 | *** | 0.303 | *** | 0.252 | *** | 0.348 | *** | 0.234 | *** | 0.367 | *** |
| Non-local | -0.262 | *** | -0.303 | *** | -0.252 | *** | -0.348 | *** | -0.234 | *** | -0.367 | *** |
| Convenience (attribute importance) | 31% | 29% | 31% | 27% | 30% | 31% | ||||||
| 15 min. | 0.615 | *** | 0.510 | *** | 0.560 | *** | 0.543 | *** | 0.578 | *** | 0.545 | *** |
| 30 min. | 0.020 | 0.040 | 0.053 | 0.013 | 0.046 | -0.002 | ||||||
| 45 min. | -0.635 | *** | -0.551 | *** | -0.613 | *** | -0.556 | *** | -0.624 | *** | -0.543 | *** |
| Appearance (attribute importance) | 11% | 4% | 12% | 3% | 11% | 2% | ||||||
| Perfect | 0.219 | *** | 0.080 | ** | 0.227 | *** | -0.053 | 0.209 | *** | 0.033 | ||
| Imperfect | -0.219 | *** | -0.080 | ** | -0.227 | *** | 0.053 | -0.209 | *** | -0.033 | ||
| Price (attribute importance) | 21% | 18% | 20% | 17% | 23% | 12% | ||||||
| Low | 0.358 | *** | 0.325 | *** | 0.341 | *** | 0.326 | *** | 0.425 | *** | 0.176 | *** |
| Medium | 0.125 | *** | 0.022 | 0.081 | ** | 0.026 | 0.074 | * | 0.068 | |||
| High | -0.483 | *** | -0.347 | *** | -0.421 | *** | -0.352 | *** | -0.499 | *** | -0.244 | *** |
| Number of respondents = | 153 | 165 | 223 | 95 | 206 | 112 | ||||||
| Observations = | 4,590 | 4,950 | 6,690 | 2,850 | 6,180 | 3,360 | ||||||
| Log pseudolikelihood | -1479.39 | -1568.13 | -2176.81 | -840.26 | -1987.77 | -1044.03 | ||||||
| Wald chi2 (8) = | 237.58 | 273.49 | 363.79 | 186.99 | 338.08 | 194.41 | ||||||
| Prob > chi2 = | 0.000 | 0.000 | 0.000 | 0.000 | 0.000 | 0.000 | ||||||
| Pseudo R2 = | 0.120 | 0.135 | 0.112 | 0.195 | 0.122 | 0.152 | ||||||
Note(s): The groups refer to answers from the statements listed below, which were measured with a 5 point-Likert-like scale (from strongly agree to strongly disagree). Because answers were skewed towards agreement, the table refers to only two groups: consumers who strongly agree with the respective statement (1) vs all others (0)
a “A healthy diet is important to me.”; health-oriented consumers answered with “strongly agree”
b “To me it is important that the vegetables I buy are produced in an environmentally friendly manner.”; environment-oriented consumers answered with “strongly agree”
c “I would be willing to pay a higher price for vegetables, if the farmers profited directly and got a fair pay.” fairness-oriented consumers answered with “strongly agree”
Source(s): Own data
Willingness-to-pay for ethical consumerism attributes
| Potatoes | Carrots | Bell peppers | |||||||
|---|---|---|---|---|---|---|---|---|---|
| wtp | Conf. interval | wtp | Conf. interval | wtp | Conf. interval | ||||
| Traditional production | 0.482 | 0.241 | 0.723 | 0.698 | 0.225 | 1.171 | 2.582 | 1.601 | 3.563 |
| Organic production | 0.134 | −0.131 | 0.400 | 0.928 | 0.249 | 1.607 | 0.810 | −0.154 | 1.775 |
| Local origin | 0.520 | 0.306 | 0.733 | 0.313 | 0.020 | 0.606 | 1.113 | 0.653 | 1.572 |
Source(s): Own data
Estimated market shares for ethical consumerism products
| Hypothetical fresh vegetable products | ||||||||
|---|---|---|---|---|---|---|---|---|
| “AFN organic” | “Conventional supermarket” | Farmers' market lower quality | Farmers' market higher quality | |||||
| Origin | Local | 0.279 | Non-local | −0.279 | Local | 0.279 | Local | 0.279 |
| Production | Organic | 0.158 | Conventional | −0.596 | Traditional | 0.438 | Traditional | 0.438 |
| Convenience | 45 min | −0.587 | 15 min | 0.562 | 15 min | 0.025 | 15 min | 0.025 |
| Appearance | Good | 0.150 | Good | 0.150 | Bad | −0.150 | Good | 0.150 |
| Price | High | −0.415 | Medium | 0.077 | Low | 0.338 | High | −0.415 |
| Utility | −0.415 | −0.085 | 0.929 | 0.477 | ||||
| exp (Utility) | 0.661 | 0.918 | 2.533 | 1.612 | ||||
| Sum exp (U) | 5.723 | |||||||
| Estimated market share | ||||||||
| Probability | 0.12 | 0.16 | 0.44 | 0.28 | ||||
Note(s): This calculation is based on the pooled model
Source(s): Own data
Sample characteristics
| Total sample (n = 318) | Bucharest (n = 160) | Timisoara (n = 158) | ||||
|---|---|---|---|---|---|---|
| Mean | n | Mean | n | Mean | n | |
| Age of respondent in years | 35.27 | 316 | 35.71 | 158 | 34.82 | 158 |
| Number of household members | 2.75 | 315 | 2.96 | 157 | 2.54 | 157 |
| 2.31 | 315 | 2.41 | 157 | 2.20 | 157 |
| 0.44 | 315 | 0.54 | 157 | 0.33 | 157 |
| Female respondents dummy | 0.59 | 318 | 0.58 | 158 | 0.61 | 158 |
| Education dummies | ||||||
| High school | 0.32 | 318 | 0.36 | 160 | 0.29 | 158 |
| University | 0.65 | 318 | 0.63 | 160 | 0.68 | 158 |
| Other | 0.02 | 318 | 0.02 | 160 | 0.03 | 158 |
| Employment status dummies | ||||||
| Employed | 0.50 | 317 | 0.46 | 160 | 0.53 | 157 |
| Self-employed | 0.16 | 317 | 0.16 | 160 | 0.17 | 157 |
| Unemployed | 0.01 | 317 | 0.01 | 160 | 0.01 | 157 |
| Student | 0.18 | 317 | 0.19 | 160 | 0.18 | 157 |
| Pensioner | 0.10 | 317 | 0.12 | 160 | 0.08 | 157 |
| House wife/husband | 0.05 | 317 | 0.06 | 160 | 0.04 | 157 |
| Income status dummies | ||||||
| Below average | 0.07 | 318 | 0.07 | 160 | 0.08 | 158 |
| Above average | 0.08 | 318 | 0.06 | 160 | 0.11 | 158 |
| Average | 0.57 | 318 | 0.59 | 160 | 0.55 | 158 |
| Not specified | 0.27 | 318 | 0.28 | 160 | 0.27 | 158 |
| Ethical consumer statements* | ||||||
| “A healthy diet is important to me” | ||||||
| 1 | 0.52 | 318 | 0.46 | 160 | 0.58 | 158 |
| 2 | 0.40 | 318 | 0.44 | 160 | 0.37 | 158 |
| 3 | 0.07 | 318 | 0.10 | 160 | 0.04 | 158 |
| 4 | 0.00 | 318 | 0.00 | 160 | 0.01 | 158 |
| 5 | 0.00 | 318 | 0.01 | 160 | 0.00 | 158 |
| “To me it is important that the vegetables I buy are produced in an environmentally friendly manner” | ||||||
| 1 | 0.30 | 318 | 0.28 | 160 | 0.32 | 158 |
| 2 | 0.41 | 318 | 0.40 | 160 | 0.42 | 158 |
| 3 | 0.23 | 318 | 0.25 | 160 | 0.22 | 158 |
| 4 | 0.03 | 318 | 0.03 | 160 | 0.03 | 158 |
| 5 | 0.03 | 318 | 0.04 | 160 | 0.01 | 158 |
| “I would be willing to pay a higher price for vegetables, if the farmers profited directly and got a fair pay” | ||||||
| 1 | 0.35 | 318 | 0.34 | 160 | 0.36 | 158 |
| 2 | 0.45 | 318 | 0.46 | 160 | 0.44 | 158 |
| 3 | 0.12 | 318 | 0.12 | 160 | 0.11 | 158 |
| 4 | 0.06 | 318 | 0.06 | 160 | 0.07 | 158 |
| 5 | 0.02 | 318 | 0.03 | 160 | 0.02 | 158 |
Note(s): *Ethical consumer statements were measured on a 5 point-Likert-like scale (from strongly agree to strongly disagree)
Source(s): Own data
|
Notes
We follow here the rule of thumb of using 8 to 15 choice tasks to avoid survey fatigue but allow for the estimation of part-worths.
The conditional logit relies on the same statistical assumptions as a multinomial logit approach but uses a binary dependent variable. McFadden (1974) introduced the term when he first showed that conditional logit was consistent with random utility theory.
At the time of the survey RON 1 equalled EUR 0.22 or USD 0.25.
References
Adams, D.C. and Salois, M.J. (2010), “Local versus organic: a turn in consumer preferences and willingness-to-pay”, Renewable Agriculture and Food Systems, Vol. 25 No. 4, pp. 331-341.
Auger, P. and Devinney, T.M. (2007), “Do what consumers say matter? The misalignment of preferences with unconstrained ethical intentions”, Journal of Business Ethics, Vol. 76 No. 4, pp. 361-383.
Barska, A. and Wojciechowska-Solis, J. (2018), “Traditional and regional food as seen by consumers–research results: the case of Poland”, British Food Journal, Vol. 120 No. 9, pp. 1994-2004.
Bazzani, C., Caputo, V., Nayga, J.R.M. and Canavari, M. (2017), “Revisiting consumers' valuation for local versus organic food using a non-hypothetical choice experiment: does personality matter?”, Food Quality and Preference, Vol. 62, December, pp. 144-154.
Bennett, J. and Blamey, R. (Eds) (2001), The Choice Modelling Approach to Environmental Valuation, Edward Elgar Publishing, Cheltenham.
Bimbo, F., Russo, C., Di Fonzo, A. and Nardone, G. (2020), “Consumers' environmental responsibility and their purchase of local food: evidence from a large-scale survey”, British Food Journal, Vol. 123 No. 5, pp. 1853-1874.
Birtalan, I.L., Neulinger, Á., Bárdos, G., Rigó, A., Rácz, J. and Boros, S. (2021), “Local food communities: exploring health-related adaptivity and self-management practices”, British Food Journal, Vol. 123 No. 8, pp. 2728-2742.
Bozga, N.-A. (2015), “The perception of Romanian consumer upon organic products' prices”, Procedia Economics and Finance, Vol. 27 No. 1, pp. 323-333.
Bryła, P. (2018), “Organic food online shopping in Poland”, British Food Journal, Vol. 120 No. 5, pp. 1015-1027.
Cone, C.A. and Myhre, A. (2000), “Community-supported agriculture: a sustainable alternative to industrial agriculture?”, Human Organization, Vol. 59 No. 2, pp. 187-197.
Connolly, C. and Klaiber, H.A. (2014), “Does organic command a premium when the food is already local?”, American Journal of Agricultural Economics, Vol. 96 No. 4, pp. 1102-1116.
Darby, K., Batte, M.T., Ernst, S. and Roe, B. (2008), “Decomposing local: a conjoint analysis of locally produced foods”, American Journal of Agricultural Economics, Vol. 90 No. 2, pp. 476-486.
Drăghici, M.C., Popa, E.E., Popa, M.E. and Mitelut, A.C. (2016), “Romanian organic food-study on consumer behaviour”, Journal of EcoAgriTourism, Vol. 12 No. 2, pp. 160-167.
Edwards, F. (2016), “Alternative food networks”, in Thompson, P.B. and Kaplan, D.M. (Eds), Encyclopedia of Food and Agricultural Ethics, Springer, Dordrecht, pp. 1-7.
Fendrychová, L. and Jehlička, P. (2018), “Revealing the hidden geography of alternative food networks: the travelling concept of farmers' markets”, Geoforum, Vol. 95, October, pp. 1-10.
Ferraris, A., Del Giudice, M., Grandhi, B. and Cillo, V. (2019), “Refining the relation between cause-related marketing and consumers purchase intentions: a cross-country analysis”, International Marketing Review, Vol. 37 No. 4, pp. 651-669.
Flora, C.B. and Bregendahl, C. (2012), “Collaborative community-supported agriculture: balancing community capitals for producers and consumers”, The International Journal of Sociology of Agriculture and Food, Vol. 19 No. 3, pp. 329-346.
Ghvanidze, S., Velikova, N., Dodd, T. and Oldewage-Theron, W. (2017), “A discrete choice experiment of the impact of consumers' environmental values, ethical concerns, and health consciousness on food choices: a cross-cultural analysis”, British Food Journal, Vol. 119 No. 4, pp. 863-881.
Goszczyński, W. and Wróblewski, M. (2020), “Beyond rural idyll? Social imaginaries, motivations and relations in Polish alternative food networks”, Journal of Rural Studies, Vol. 76, May, pp. 254-263.
Güney, O.I. and Giraldo, L. (2019), “Consumers' attitudes and willingness to pay for organic eggs: a discrete choice experiment study in Turkey”, British Food Journal, Vol. 122 No. 2, pp. 678-692.
Hauber, A.B., González, J.M., Groothuis-Oudshoorn, C.G.M., Prior, T., Marshall, D.A., Cunningham, C., IJzerman, M.J. and Bridges, J.F.P. (2016), “Statistical methods for the analysis of discrete choice experiments: a report of the ISPOR conjoint analysis good research practices task force”, Value in Health, Vol. 19 No. 4, pp. 300-315.
Hjelmar, U. (2011), “Consumers' purchase of organic food products. A matter of convenience and reflexive practices”, Appetite, Vol. 56 No. 2, pp. 336-344.
Joya, K., Ramli, N.N., Shamsudin, M.N. and Kamarulzaman, N.H. (2022), “Consumers’ willingness to pay for food safety attributes of tomato”, British Food Journal, Vol. 124 No. 3, pp. 707-717, doi: 10.1108/BFJ-02-2021-0164.
Lancaster, K.J. (1966), “A new approach to consumer theory”, Journal of Political Economy, Vol. 74 No. 2, pp. 132-157.
Lancsar, E., Fiebig, D.G. and Hole, A.R. (2017), “Discrete choice experiments: a guide to model specification, estimation and software”, Pharmaco Economics, Vol. 35 No. 7, pp. 697-716.
Lang, M., Stanton, J. and Qu, Y. (2014), “Consumers' evolving definition and expectations for local foods”, British Food Journal, Vol. 116 No. 11, pp. 1808-1820.
Lusk, J.L. and Briggeman, B.C. (2009), “Food values”, American Journal of Agricultural Economics, Vol. 91 No. 1, pp. 184-196.
Maehle, N., Iversen, N., Hem, L. and Otnes, C. (2015), “Exploring consumer preferences for hedonic and utilitarian food attributes”, British Food Journal, Vol. 117 No. 12, pp. 3039-3063.
Maye, D. and Kirwan, J. (2010), “Alternative food networks. Sociology of agriculture and food entry for Sociopedia.isa”, available at: http://www.sagepub.net/isa/resources/pdf/AlternativeFood-Networks.pdf.
McFadden, D. (1974), “Conditional logit analysis of qualitative choice behavior”, in Zarembka, P. (Ed.), Frontiers in Economy, New York.
McFadden, D. and Train, K. (2000), “Mixed MNL models for discrete response”, Journal of Applied Econometrics, Vol. 15 No. 5, pp. 447-470.
Melović, B., Cirović, D., Backovic-Vulić, T., Dudić, B. and Gubiniova, K. (2020a), “Attracting green consumers as a basis for creating sustainable marketing strategy on the organic market—relevance for sustainable agriculture business development”, Foods, Vol. 9 No. 11, p. 1552.
Melović, B., Dabić, M., Rogić, S., Đurišić, V. and Prorok, V. (2020b), “Food for thought: identifying the influential factors that affect consumption of organic produce in today's youth”, British Food Journal, Vol. 122 No. 4, pp. 1130-1155.
Meyerding, S.G.H., Trajer, N. and Lehberger, M. (2019), “What is local food? The case of consumer preferences for local food labeling of tomatoes in Germany”, Journal of Cleaner Production, Vol. 207, January, pp. 30-43.
Möllers, J. and Bîrhală, B. (2014), “Community Supported Agriculture: a promising pathway for small family farms in Eastern Europe? A case study from Romania”, Applied Agricultural and Forestry Research, Vols 3/4 No. 64, pp. 139-150.
Moser, R., Raffaelli, R. and Thilmany, D.D. (2011), “Consumer preferences for fruit and vegetables with credence-based attributes: a review”, International Food and Agribusiness Management Review, Vol. 14 No. 2, pp. 121-142.
Mugera, A., Burton, M. and Downsborough, E. (2017), “Consumer preference and willingness to pay for a local label attribute in Western Australian fresh and processed food products”, Journal of Food Products Marketing, Vol. 23 No. 4, pp. 452-472, doi: 10.1080/10454446.2015.1048019.
Musova, Z., Musa, H. and Ludhova, L. (2018), “Environmentally responsible purchasing in Slovakia”, Economics and Sociology, Vol. 11 No. 4, pp. 289-305.
Petrescu, D.C., Petrescu-Mag, R.M., Burny, P. and Azadi, H. (2017), “A new wave in Romania: organic food - consumers' motivations, perceptions, and habits”, Agroecology and Sustainable Food Systems, Vol. 41 No. 1, pp. 46-75.
Popa, I.D. and Dabija, D.-C. (2019), “Developing the Romanian organic market: a producer's perspective”, Sustainability, Vol. 11 No. 2, p. 467.
Poças Ribeiro, A., Harmsen, R., Feola, G., Rosales Carreon, J. and Worrell, E. (2021), “Organising alternative food networks (AFNs): challenges and facilitating conditions of different AFN types in three EU countries”, Sociologia Ruralis, Vol. 61 No. 2, pp. 491-517.
Rao, V.R. (2014), Applied Conjoint Analysis, Springer, Ithaca.
Simon, S. and Borowski, B. (2007), “Länderbericht Rumänien: Zwischen Tradition und Moderne – chancen Durch Ökolandbau?”, Ökologie und Landbau, Vol. 144 No. 4, pp. 43-45.
Slavuj Borčić, L. (2020), “The production of urban commons through alternative food practices”, Social and Cultural Geography, pp. 1-18, Published online: 2020-07-21, doi: 10.1080/14649365.2020.1795234.
Willer, H., Trávníček, J., Meier, C. and Schlatter, B. (2021), The World of Organic Agriculture. Statistics and Emerging Trends 2021, Research Institute of Organic Agriculture FiBL, Frick, and IFOAM – Organics International, Bonn.
Winterstein, J. and Habisch, A. (2021), “Is local the new organic? Empirical evidence from German regions”, British Food Journal, Vol. 123 No. 11, pp. 3486-3501.
Yeh, C.-H., Hartmann, M., Gorton, M., Tocco, B., Amilien, V. and Steinnes, K.K. (2021), “Looking behind the choice of organic: a cross-country analysis applying integrated choice and latent variable models”, Appetite, Vol. 167 No. 1, pp. 1-13.
Acknowledgements
The authors would like to thank Marcel Kalkkuhl and Brînduşa Bîrhală. This paper and the research behind it would not have been possible without their enthusiastic work in the field in Romania.