Urban Agriculture

Relationship between Consumer Behavior and Success of Urban Agriculture

Abstract

Consumers prefer locally grown food products. One source that provides local food is urban agriculture, the farm- ing in and around cities. A number of urban farmers are selling their products directly to consumers. In addition, consumers have the option to grow their own food on certain urban farms. Given this, we investigate how likely consumers are to purchase or grow their own food at urban farms and what determines this likelihood. Given that millennials are a key stakeholder of sustainable consumption and those with the greatest increase in num- bers of food gardeners, we conducted an online survey with over 300 Generation Y respondents. We investigate whether young consumers perceive the health impacts and environmental benefits provided by urban agricul- ture, and what attitudes they hold towards this source of produce. Empirical results show that both psychological and personal factors affect consumer intentions to participate in urban agriculture. Among others, subjective knowledge regarding urban agriculture and a generally favorable attitude towards urban farms increases the likelihood to buy and grow produce at urban farms. Female and older consumers are more likely to grow their own produce. These findings can be used by stakeholders in urban agriculture to design target-oriented marketing activities.

1-Introduction

Consumer demand for local food continues to grow (e.g., Thilmany et al., 2008; Adams and Salois, 2010; Grebitus et al., 2013a; Meas et al., 2015; Feldmann and Hamm, 2015; Pyburn et al., 2016). By 2014 local food sales in the U.S. reached about $12 billion with an expected in- crease to $20 billion by the year 2019 (USDA, 2016). In response to this, housing development builders and cities started to incorporate local food production and points of sale into new and established com- munities (Hughes and Boys, 2015), catering to the approximately 82% of the population in the United States who are living in urban areas (World Bank, 20161).

The  practice  of  “growing,  processing,  and  distributing  of  food and other products through intensive plant cultivation and animal hus- bandry in and around cities” is called urban agriculture2 (Bailkey and Nasr, 1999). Regarding the distribution of food, about 8% or 163,675 of all U.S. farms sell their food using direct-to-consumer marketing (e.g., farmers’ markets, roadside stands, directly to restaurants), with 70% out of the 8% selling their food exclusively using such channels, e.g., Community Supported Agriculture (ERS, 2015). These producers are competing with national grocers such as Whole Foods and Kroger, who are partnering with local farmers to offer products at their stores that are grown within the state lines. For example, Walmart sources 20% of fresh fruits and vegetables in-state and Wegmans sells 30% of lo- cally grown produce (King et al., 2010). It follows that urban farmers, who choose to sell directly, need to understand what drives consumers to purchase produce at the farm in order to effectively and efficiently market their products.

Knowing more about key drivers might also enable stakeholders to increase the share of citizens that take part in urban agriculture. This is valuable because urban agriculture provides a number of additional benefits, such as, contributing to food security, supporting healthy die- tary patterns (e.g., Warren et al., 2015; Zezza and Tasciotti, 2010), and improving local ecology and sustainability (e.g., Wakefield et  al., 2007). Given these benefits, it seems of interest not only from a business standpoint but also from a societal standpoint to further investigate un- derlying reasons that drive individuals to participate in urban agriculture.

Previous literature on participation in urban agriculture has mainly focused on developing countries. For example, Warren et al. (2015) pro- vide a comprehensive review of the literature that investigates the relationship between urban agriculture and nutrition (food security, nutrition status and variety). Studies referring to developed countries investigated the social missions of urban agriculture focusing on the urban farms, i.e., growers (Dimitri et al., 2016), assessed the needs of urban farmers who face limited resources (Surls et al., 2015), and ana- lyzed the impact of community gardens on communities using mainly exploratory methods (e.g., Armstrong, 2000; Firth et al., 2011; Wakefield et al., 2007). Hence, there seems to be a gap in the literature regarding quantitative research in developed countries that focuses on the urban agriculture consumer. We aim to extend the literature by an- alyzing consumer behavior as it relates to urban agriculture as an outlet for food.

The objective of our study is to evaluate success factors of urban ag- riculture participation. In this regard, we focus particularly on the future consumer, the Generation Y. As pointed out by Hume (2010), one key stakeholder group of sustainable behavior is comprised of young con- sumers, the Generation Y or Millennials. This generation is not only ed- ucated, but also environmentally aware without actively behaving in a pro-environmental manner. They seem to be compassionate but not willing to act upon it (McCrindle Research, 2007). It follows, that we need to investigate what affects their behavior in order to develop strat- egies that encourage them to put their beliefs into action.

In order to analyze how urban farms can be successful, it is impor- tant to understand determinants of consumer behavior that are likely to influence shopper habits, when it comes to produce purchase behav- ior. Behavioral concepts, such as, perception, subjective knowledge and attitudes, play an important role for purchase decision making (e.g., Howard and Sheth, 1969). To start with, consumers have to be able to perceive urban agriculture as a viable source of produce. In this regard, making them aware that urban farms are a source for local food can lead to positive perception, given that purchasing foods from urban farms provides firsthand knowledge on the origin of the products and truly is locally produced food. Apart from the perception, consumers need to be knowledgeable about urban farms. Without having any knowl- edge about urban agriculture, as a potential source of local produce and other benefits associated with it, consumers are hypothesized to be less likely to purchase produce from an urban farm. Furthermore, consumers need to hold positive attitudes towards urban agriculture in order to actually make a purchase. This leads to the following re- search questions:

  • How do consumers perceive urban agriculture?
  • Do consumers feel knowledgeable when it comes to urban agriculture?
  • Do consumers hold positive attitudes towards urban agriculture?
  • Are consumers likely to buy produce from an urban farm, and how is this influenced by behavioral aspects?

In addition to urban farms serving as a point of sale for local produce, they also offer citizens the opportunity to grow their own food. When individuals actively participate in growing their own food at the urban farm, they are likely to experience an improvement in health and their communities are more developed (e.g., Wakefield et al., 2007). For ex- ample, community gardens promote physical, mental and social health as they, among others, restore attention, reduce stress, evoke positive emotions and lead to social integration (Abraham et al., 2010). This sug- gests that growing their own food may improve their lifestyle by being more physically active and mentally and emotionally relaxed. Similarly, community gardening promotes social health and community cohesion in form of stable relationships, which also contributes to a healthy life- style (Wakefield et al., 2007; Abraham et al., 2010). This is important, given the increase in obesity and sedentary lifestyles, particularly in children and young adults (e.g., Caballero, 2007; Sahoo et al., 2015). Therefore, gardening, i.e., growing food, can play a part in improving one’s wellbeing. In addition to the physical component of gardening, nutrition education by means of learning about food production and processing comes more naturally when growing food. At the same time, growing produce encourages fruit and vegetable consumption, since it familiarizes the growers with the foods that might be less pres- ent otherwise (Libman, 2007). In general, an improved access to food and nutrition is related to community gardens (Wakefield et  al., 2007). This leads to the final research question:

  • Are consumers likely to grow their own produce, and how is this influenced by behavioral aspects?

As pointed out above, we focus on Generation Y because Millennials are a key stakeholder of sustainable consumption (Hume, 2010). Fur- thermore, they account for the greatest increase in numbers of food gar- deners (NGA, 2014). Hence, understanding the behavior of Millennials towards urban agriculture, in other words investigating the determinants of purchasing food at an urban farm or growing food at  an urban farm, promises insight into long-term success factors of urban agriculture.

Against this background, the aim of this study is to investigate the impact of consumer perception, knowledge and attitudes towards the likelihood to purchase produce from urban farms and the likelihood to grow their own produce at urban farms. Fig. 1 displays this relationship between the questions:

2-Conceptual Framework

In order to answer our research questions we apply Howard and Sheth’s (1969) theory of buyer behavior. The authors state that stimuly, perceptual constructs and learning constructs, lead to the outputs of at- tention, comprehension, attitudes, intention and ultimately purchase behavior. According to Howard and Sheth (1969), attention influences comprehension, which impacts attitudes, which affects intention and then purchase behavior. In this regard, attention is a necessity for per- ception (Orquin and Mueller Loose, 2013). Given that Kroeber-Riel and Weinberg (2003) show that perception influences related product evaluation and purchase decision making, and considering that Shapiro (1970) links perceived quality to purchase likelihood, we apply the concept of perception, rather than attention, in our study, and refer to purchase likelihood as final concept, rather than consider- ing the actual purchase decision. Comprehension can be reflected by its synonym knowledge. In this regard, multiple frameworks (e.g., Ajzen, 1991) have shown that attitudes and knowledge determine in- tention, which can be expressed as likelihood of purchase.

Taking these theories into account, we arrive at a conceptual model that serves as foundation for our research. We address the influence of the psychological constructs perception, knowledge and attitudes on both, the likelihood to purchase produce from and the likelihood to grow produce at urban farms. We also include personal factors as deter- minants of this likelihood based on Steenkamp (1989). Steenkamp (1989) displays in his conceptual model of quality perception the influ- ence of personal factors on perceived quality and it seems reasonable to assume a similar influence of several personal factors on intentions, i.e., purchase and growing likelihood. Hence, as Fig. 2 shows, we distinguish between psychological and personal factors, as determinants of con- sumer behavior related to urban agriculture.

In the following, we will provide the theoretical background on the psychological constructs and will discuss related empirical findings. Af- terwards, we will discuss the methods used to measure the respective constructs and factors. Specifically, we use mixed methods research with the approach of concurrent nested research, conducting one data collection with a predominantly quantitative approach but also includ- ing a qualitatitve data collection (FoodRisC, 2017). While the qualitative data regards perception, the quantitative data regard knowledge, atti- tudes, socio-demographics, produce purchase frequency and intentions. Though we do not mix the data in a traditional way during the analysis (e.g., including the perception data in the bivariate ordered probit model used to determine what affects the intention), the design allows for a deeper insight towards success factors of consumer behavior as it relates to urban agriculture.

3-Theoretical Background and Previous Literature

3.1-Perception

Every product makes an impression on the consumer, which can be understood as perception, where consumers gather, organize and sub- sequently evaluate product information. This impression is then com- bined with prior experiences as well as experiences made during and after the shopping occasion (Gryna, 1998). These experiences include characteristics of the product itself, the shopper and the shopping loca- tion. Depending on the perception process, the importance of certain characteristics can vary in that some consumers prefer convenience and cheaper prices, while others favor specific ambience and assort- ments of products. Moreover, each consumer differs in how they per- sonally perceive products and retail outlets, which depends on their abilities, preferences and experiences (Grebitus, 2008). Based on this background, the shopping outlet might affect perception through the intended use and/or situational factors, such as, convenience (Oude Ophuis and van Trijp, 1995). Hence, in order to evaluate success of urban agriculture, it is important to investigate whether consumers per- ceive urban farms as a possible point of sale and whether they consider food from urban farms to be local. Consumer perception of what food qualifies as local is of particular interest, since that results in higher will- ingness to pay. For example, previous studies found that consumers have a higher willingness to pay for local compared to organic apples (Costanigro et al., 2011), and for local compared to organic and GMO- Free potatoes (Loureiro and Hine, 2002). In addition, consumers are willing to pay more for locally grown Gala apples and red round toma- toes, compared to domestic but not local, and imported apples and to- matoes (Onozaka and Mcfadden, 2011). The preferences for locally grown food even extend to ingredients. This was shown by Meas et al. (2015) for local versus organic blackberry jam, and by James et al. (2009) for local versus USDA organic, low fat, or no sugar added apple- sauce. For a comprehensive review on local food perception and prefer- ences see Feldmann and Hamm (2015). In this study, we analyze consumer perception of urban agriculture to determine whether they would consider it a viable source of produce. Since the perception af- fects the purchase behavior, it will be an important factor in assessing future success of urban agriculture.

3.2-Subjective Knowledge

 There are distinct categories of consumer knowledge that affect consumer behavior: subjective knowledge and objective knowledge as well as prior experience, i.e., usage experience (Brucks, 1985; Raju et al., 1995). Subjective knowledge has been acquired by direct experience and is interpreted by the experiencer. It is also known as perceived knowledge because it reflects what consumers think they know about a subject as opposed to what they actually, objectively know (Carlson et al., 2009; Raju et al., 1995). An example with regards to urban agricul- ture would be that consumers often believe urban farmers to use organ- ic practices and the food to be more natural even though this does not have to be true. This would be a case of subjective rather than objective knowledge. Previous literature suggests that subjective knowledge im- pacts consumer behavior. For example, using latent class analysis for data from a Canadian and a German consumer survey for beef and potatotes, Peschel et al. (2016) demonstrate the relationship between subjective knowledge and consumer choice of food labeled for carbon and water footprints. Conducting focus group interviews on purchasing and disposal habits, Ellen (1994) provides evidence of the impact subjective knowledge has on precycling and recycling choices and behaviors. Applying a probit model and an analysis of variance to a convenience sample, Aertsens et al. (2011) show how subjective knowledge influences consumer attitudes and motivations towards organic food and its consumption. Furthermore, subjective knowledge can deter- mine the quality of consumer decisions (Moorman et al., 2004). There- fore, we study consumers’ subjective knowledge regarding urban agriculture and attempt to determine the relationship between this type of knowledge and the likelihood of urban farms to become a viable source of produce.

3.3-Attitudes

Attitudes are consumer evaluations of a psychological entity. They are formed through beliefs about the probability of consequences de- pending on behavior and assessments of how good or bad those conse- quences would be if they became a reality (Trafimow and Finlay, 2002; Ajzen, 1991). When attitudes are strong enough, they impact purchase behavior (Trommsdorff, 2003). For example, this was documented when consumers started to increasingly purchase local foods instead of organic foods, displaying a shift in preferences from organic to local foods in the late 1990s (Gallons et al., 1997; Food Processing Center, 2001). Research in the 1990s showed that 86% of consumers considered it an advantage and held a positive attitude towards purchasing locally grown food (Bruhn et al., 1992). Also, a survey in the early 2000s, indi- cated that almost all surveyed consumers (99%) had recently purchased locally grown food (Food Processing Center, 2001). With this research, we determine what attitudes consumers are holding towards urban ag- riculture. This is then used to inform on the opportunity for urban agri- culture to successfully serve as an outlet for produce. Since attitudes shape consumer shopping behavior, this needs to be addressed when considering urban agriculture as point of sale for produce.

4-Methodology

4.1-Design of the Study

 We conducted an IRB approved online survey with N = 325 individ- uals in fall 2015. Participants were recruited from a 400-level class at Arizona State University in the U.S. All respondents received five points towards the final exam (equivalent to 1% of the final grade) as credit for their participation. We used a student sample because when it comes to sustainable behavior, young consumers (Generation Y) are a key stake- holder group (Hume, 2010). These young consumers are making up a great share of total consumption expenditure in developed countries (Bentley et al., 2004), such as the U.S. They are the future public (Smola and Sutton, 2002). Therefore, it is of importance to consider their behavior, particularly when it concerns purchase decision making and consumption that can be understood to be more sustainable. Fur- thermore, given that we are also researching the intention to grow food, this age group is of interest because data from the National Gardening Association (NGA) show that millennials have the largest in- crease in the number of gardeners compared to other age categories.

About 13 million millennials were growing food in 2013 spending $1.192 billion (NGA, 2014). Hence, we focus on upper division under- graduate students that fit the profile of young consumers described as informed, knowledgeable (Heaney, 2006) and most educated (McCrindle Research, 2007; Hume, 2010).

Specifically, the sample consisted of 38% female and 62% male re- spondents. This is lower than the share of adult females in Arizona, which accounts for 50% (Census, 2015). The median age of participants is 22 years (M = 23; SD = 4.4) ranging from 20 years to 61 years, which is younger than the median age of 34 years (Census, 2015). The educa- tion level of the sample ranges from high school diploma (20%), to some college experience (62%), to technical school diploma (2%) to Bachelor’s Degree (16%), while 86% of Arizonans are a high school graduate or higher (note this accounts for persons age 25 years+) and 28% of Arizonans hold a Bachelor’s degree or higher (again for persons age 25 years+) (Census, 2015). Given the current statistics a direct compar- ison, however, is not sensible for education. About 55% of the inter- viewees have an income lower than $50,000 annually before taxes, this is comparable to the median household income in Arizona which amounts to $50,255 (Census, 2015). Some 6% of respondents had chil- dren in the household (SD = 0.25), with an average household size of 3.2 (SD = 1.29) persons, which is slightly higher than the household size for Arizona of 2.69 (Census, 2015). Overall, differences between the sample and the general population are to be expected, given that we purposely surveyed millennials. In the following gender, age and ed- ucation are included in the bivariate ordered probit model as indepen- dent variables to test the effect of personal factors on intentions towards urban agriculture.

4.2-Measuring Perception

 In order to measure how consumers perceive urban agriculture, we used free elicitation technique (e.g., Grebitus, 2008). Results from free elicitation technique encompass consumers’ stored (prior) information, i.e., cognitive structures, with regards to a certain knowledge area as well as related perceptions or misperceptions. Free elicitation technique is an appropriate method to survey consumer perception (Van Kleef et al., 2005; Kanwar et al., 1981; Olson and Muderrisoglu, 1979). The tech- nique is understood as a non-structured form of questioning, where participants are asked to write down their first thoughts when being presented with a key stimulus (Parasuraman, 1990). The research sub- jects can express anything they think of when seeing the stimulus.

In applying the method, we follow Olson and Muderrisoglu’s (1979) and Roininen et al.’s (2006) technique. While in previous studies partic- ipants were at times limited to the number of associations to be written down, our respondents were asked to write down everything that comes to mind when thinking of community gardens/urban farming. We used both stimuli to analyze perception of urban agriculture be- cause urban agriculture entails, among others, community gardening and farms located in metropolitan areas (USDA, 20163; U.  of California, 20164).

The data collected was analyzed by means of content analysis (Mayring, 2000). This analysis allows making assumptions, and investi- gates the intent and motivation regarding a certain topic in a formal way using a specific content observed (Stempel, 1981; Hsia, 1988). Content analysis is utilized by applying quantitative measures that investigate the data source in question in an objective and systematic way. When using this analysis, a so-called universe is defined which serves as infor- mation source. Next, a unit is identified for the universe, determining which information to count (Hsia, 1988; Wimmer and  Dominick, 1983; Stempel, 1981). Since we apply content analysis to free elicitation technique data, our units are the single words or sentences written down by participants.

In order to provide meaning to the units, we classified the content provided into categories that are used for investigation. This provides a framework for analyzing within the determined context, here urban agriculture. All associations provided by participants regarding the key stimulus are organized, categorized and then added up into frequencies (Lamnek, 1995; Bonato, 1990). The categories are the key of the actual analysis to be used for further exploration of the topic (Wimmer and Dominick, 1983), and need to be closely related to the study’s research questions in a theoretical way. Furthermore, these categories need to be reliable, practical, comprehensive and mutually exclusive (Stempel, 1981; Wimmer and Dominick, 1983). In this article we use the three pillars of sustainability (e.g., Krajnc and Glavič, 2005): Economy, Society and Environment, as three categories. In addition, we include Food and Food Attributes, Point of Sale and Others as categories based on re- sults of an analysis of food quality by Grebitus (2008). These categories seem appropriate to categorize the data into groups meaningful for urban agriculture.

4.3-Measuring Subjective Knowledge

 In order to measure subjective knowledge, we asked respondents how informed they feel towards the more general aspect of sustainabil- ity given that urban agriculture “can make important contributions to social, economic and ecological objectives of sustainable urban develop- ment” (FAO, 2007, p. V). In addition, we inquired towards subjective knowledge on community gardens, Community Supported Agriculture, urban gardens and urban agriculture. We used these terms as they are among the many examples of urban agriculture (community gardening in vacant lots and parks, community supported agriculture based in urban areas, family farms located in metropolitan greenbelts, etc.) (USDA, 20165; U. of California, 20164). To evaluate the subjective knowledge, we followed Grebitus et al. (2013b) using a scale ranging from 1 = no knowledge to 5 = very knowledgeable. We include the values for sustainability of food supply as independent variable in the following econometric analysis. We use the four specific measures for urban agriculture by creating an index for each respondent, summing up the individual items into one “urban agriculture subjective knowl- edge” measure (e.g., Flynn and Goldsmith, 1999; Peschel et al., 2016). After eliciting subjective knowledge we provided a definition on urban agriculture to assure that all participants have sufficient information on the term for the remainder of the survey.

4.4-Measuring Attitudes

 In order to analyze consumers’ attitudes towards urban agriculture we used two sets of questions. The first set considered general attitudes towards urban farms. The second set measured attitudes towards fac- tors that encourage and prevent purchases at urban farms.

Specifically, in the first set respondents received multiple questions to identify what attitudes drive consumers when it comes to participat- ing in urban agriculture via purchasing and growing produce at an urban farm. The questions were answered individually and took those 5 = completely agree to all questions. We then employed exploratory factor analysis to the data. Specifically, we used principal component analysis with varimax as rotational strategy to determine the number of factors. This allowed us to analyze more or less highly correlated items, and enabled us to reduce the attribute space from a larger num- ber of more or less highly correlated individual variables into a select number of unrelated and independent factors. In doing so, a latent structure of the variables was uncovered. To quantify the reliability of the generated factors Cronbach’s alpha was used for each factor, which should be greater than 0.5 to be able to include the respective fac- tor in the following analysis (Kim and Mueller, 1978; Hair et al., 1998). In the second set of questions respondents evaluated a number of reasons that consumers can consider as encouraging or preventing them from purchasing food at urban farms on a scale from 1 (prevent) to 5 (encourage). We then created two indexes by using the median to determine which items should be combined into the “encourage” index (sum of values above the median divided by the number of values) and which items should be combined into the “prevent” index (sum of values equal to and below the median divided by the number of values). The items also stem from the Community Food Project Evaluation Toolkit (National Research Center, Inc., 2006, pages 106, 138, 143, 284) as well as from Pan and Zinkhan (2006), Wakefield et (2007), Grebitus et al. (2013a), and Nilsson et al. (2015).

4.5-Bivariate Ordered Probit Model

 Consumers’ likelihood to purchase produce from an urban farm and their likelihood to grow their own produce at an urban farm are expressed as different likelihood categories on a seven-point Likert scale, where 1 stands for “very unlikely” and 7 is “very likely.” The like- lihood categories are used to measure the corresponding latent utilities. In the survey, the dependent variables are categorical, specifically very unlikely, unlikely, somewhat unlikely, undecided, somewhat likely, likely, very likely. Hence, we estimate a bivariate ordered probit model to find the determinants of the likelihood to purchase from or grow produce at an urban farm at the same time.

We make use of a bivariate ordered probit model that can be derived from a latent variable model (see Sajaia, n.d.). Technically, the univari- ate ordered probit model can be extended to the bivariate ordered prob- it model. In a univariate ordered probit model, the unobserved likelihood for purchasing, i.e., growing produce at an urban farm, is:
where y is the unobserved latent and continuous likelihood to pur- chase/grow produce; β is a vector of parameters to be estimated; xi is the vector of independent variables; εi is a random error term that follows a standard normal distribution. The likelihood is “latent”, since we observe an indicator of the preferred level of being willing to pur- chase or grow the produce, but not the actual behavior itself. In our case, the likelihood to purchase/grow is the observed discrete catego- ries, which are denoted as yi where ui‘s are unknown cut-off values of the latent willingness to purchase/grow which will be estimated; J is the number of frequency cate- gories. As stated above, we had seven categories (1 = very unlikely, 2 = unlikely, 3 = somewhat unlikely, 4 = undecided, 5 = somewhat likely, 6 = likely, 7 = very likely), therefore, J = 7. We assume the error term has a standard normal distribution. We follow Sajaia’s (n.d.) specification of the simultaneous bivariate ordered probit model. For related STATA code as well as specification of the model the interested reader is referred to Sajaia (n.d.).

5-Empirical Results

5.1-Purchase Frequency of Local Produce

 Considering the importance of local food overall and the fact that urban farms can serve as suppliers for local food, we measured how often participants purchase local produce. This can provide information on whether current local purchase frequency could lead to using urban farms as outlet for the food. We measured this on a seven-point scale from never to daily. Results show that 1.5% purchase it daily, 9% shop for it 2–3 times a week and 16% buy local once a week. A total of 21% in- dicated local purchases 2–3 times a month, 16% buy it once a month and the majority of 27% shops for local food less than once a month. About 10% never buy local produce. Given that the majority of the sample pur- chases local food at least once a month this seems promising for urban farms. The purchase frequency of local food is included in the bivariate ordered probit model as independent variable as per our conceptual framework.

5.2-Perception of Urban Agriculture

 Next, we discuss the descriptive findings for the psychological fac- tors displayed in our conceptual framework. We start with the qualita- tive part of the data collection, where we applied free elicitation technique to measure consumers’ perception regarding urban agricul- ture. Our results show that the participants mentioned a total of 478 dif- ferent concepts, including both single terms (e.g., healthy, fresh) and whole phrases (e.g., “A place where people share something like a plot of land to grow items they are interested in eating.”). The concepts were grouped into the six categories, Economy, Society, Environment, Food & Food Attributes, Point of Sale and Other, shown in Table 1. Re- sults indicate that consumers primarily associate specific “food & food attributes” with urban farms followed by the two categories, “economy” and “society”. The category “environment” ranks fourth.

Table 2a and 2b show the associations organized in categories. In order to reduce the large number of associations, they were merged based on similarity. For example, “healthy”, “health” and “healthier” were combined into “healthy”. Overall, the findings show that con- sumers think of, e.g., fresh, healthy, local and organic. They also feel that “life is still fine” and associate earth-friendly and sustainability with urban agriculture, which can be understood as an environmental component. Furthermore, they think of social issues such as “helping local communities,” and “bringing people together.” In addition, they point out the economic perspective displaying contrary opinions of “better quality but more expensive” versus “high prices for little mar- ginal benefit” indicating that not all consumers may be in favor of urban agriculture, and may require additional education. However, neg- ative aspects such as inconvenience and “a waste of time and energy” are also present in the responses.

Specifically, the category “economy” is dominated by associations with urban agriculture being expensive and of higher cost (30% of cate- gory associations), farms/farming in urban areas (22% of category associations) and the local economy (17% of category associations). The ma- jority thinks urban agriculture is expensive, whereas some believe it to be cheap and cost saving. The category “society” is dominated by the as- sociations with positive community building (21% of category associa- tions), helping and supporting the local economy (13% of category associations) and the idea that the community would have access to growing produce (11% of category associations). While the majority thinks urban agriculture will have a positive effect on the local commu- nity, others critically note that it might be challenging to involve com- munity members. The category “environment” is dominated by the associations earth friendliness and sustainability (38% of category asso- ciations), alternative gardening systems (14% of category associations) and the mention of green initiatives (12% of category associations). While the majority thinks urban agriculture will have a positive effect on the environment, it is mentioned that, even though it is a great idea, it will be difficult to realize.

The category “food and food attributes” is dominated by the associations organic (19% of category associations), healthy (17% of category associations) and the mention of certain produce (13% of category associations).

The category “point of sale” is dominated by the associations of farmers markets (76% of category associations), locally sold produce (7% of category associations) and the fact that the produce will not be located at the grocery store (7% of category associations). The category “other” is dominated by the associations with urban agriculture being a good idea (28% of category associations) and not being used enough (13% of category associations), however, some state that it is important to them (10% of category associations). Over all associations, organic and sustainability are associated most frequently with urban farms.

After discussing the psychological factor perception, we continue by describing the findings for the psychological factors that are included in the quantitative measurement, namely subjective knowledge and attitudes.

5.3-Subjective Knowledge Related to Urban Agriculture

As displayed in our conceptual framework we investigate the role of subjective knowledge regarding sustainability and urban agriculture (see Table 3). Results show that participants feel most informed and knowledgeable regarding the general issue of sustainability of food sup- ply with a mean above the midpoint of three. However, they do not feel informed about specific issues, such as community gardens and urban agriculture, as indicated by all means being below the midpoint of three. This leads to the conclusion that more education towards urban agriculture is needed because only if consumers feel knowledgeable they will be confident in choosing a certain outlet for their produce.

We then sum up the values for community gardens, CSA, urban gar- dens and urban agriculture into the subjective knowledge index to be used in the following econometric analysis. The knowledge index has a mean of M = 9.7 (SD = 3.9, Min = 4, Max = 20). The minimum of 4 and maximum of 20 suggest that some participants did not have any knowledge on the tested concepts, while others were very knowledge- able regarding all of them. Subjective knowledge on sustainability and urban agriculture are hypothesized to affect the intentions towards urban agriculture purchase and growing likelihood and are included as independent variables in the econometric analysis.

5.4-Attitudes towards Urban Agriculture

Next, we identify consumer attitudes regarding urban agriculture. To do so, survey participants received 11 questions on specific and general urban agriculture related attitudes. Each item was evaluated individual- ly on a five-point Likert-scale. Table 4 displays the mean and standard deviation. Results show that consumers mostly agree with the notions that urban farms help them to learn more about gardening, eat more or- ganic food, and care about the environment, and that the food from urban farms is fresher. Overall, the mean for all statements is above the midpoint three indicating general agreement.

Afterwards, the data were analyzed using exploratory factor analysis. Table 5 shows the rotated component matrix. The Kaiser- Meyer-Olkin criterion is 0.92, which is considered to be marvelous. Through the factor analysis, the following factors were found to indicate consumer attitudes:

Factor 1 (F1): Urban agriculture (UA) is better for me

Contains generally positive items which express, for example, that urban agriculture helps to care more about the environment and to learn more about gardening. Furthermore, it allows eating more organic food and food that is fresher. The Cronbach’s alpha measures 0.88, which is considered to be good.

Factor 2 (F2): Urban agriculture (UA): new, fit, frugal

Sums up the statements which express more specific reasons to go to urban farms, such as, spending less money on food, being more phys- ically active, allowing to eat new kinds of food and making new friends. The Cronbach’s alpha measures 0.70, which is considered to be questionable.

5.5-Attitudes towards Reasons that Prevent or Encourage Purchases from Urban Farms

 Then, we investigated attitudes towards reasons that prevent or en- courage consumers to purchase produce from urban farms. Results in Table 6 show that freshness, health, taste and support of the environ- ment and local economy are main reasons that encourage consumers to shop at urban farms. At the same time distance traveled, time com- mitment and too much work would prevent them from buying at urban farms.

Similar to the knowledge index we created two indexes by summing up those reasons with a median of four and higher into the index “encourage” and summing up those reasons with a median smaller than four into the index “prevent.” Given that the number of items differs for the two indexes, the sum for each index is then divided by the respective number of the items.

In the next step of the analysis, the factors and indexes are included in the econometric analysis to determine the impact of attitudes on the likelihood to purchase and grow food at urban farms. This tests the in- fluence of psychological factors on intentions.

5.6-Likelihood to Purchase and Grow Produce from Urban Farms

 In our econometric analysis we use two variables as dependent var- iables: the likelihood to purchase from and grow produce at urban farms. We measure both on a seven-point scale from 1 = very unlikely to 7 = very likely. Please note that this is a hypothetical measure given that we did not account for consumers’ actual behavior. Table 7 shows that 60% are likely, at least to some extent, to buy produce from urban farms, whereas about 18% are undecided and about one-fifth is rather unlikely to shop at urban farms.

Compared to this, 44% are likely, at least to some extent, to grow their own produce at urban farms, while 15% are undecided and nearly 40% do not think that they would consider growing produce themselves at urban farms. Table 8 shows a cross-tabulation of the two variables. Results show that there is a correlation between buying and growing produce at urban farms.

Those who are unlikely to grow their own produce are also unlikely to buy from urban farms, while those who would likely use urban farms as an outlet for fresh produce would also consider growing their own fruits and vegetables. These results lead us to consider a bivariate or- dered probit model rather than two independent ordered probit models.

5.7-Determinants of Produce Buying and Growing Likelihood Related to Urban Farms

 In accordance with our conceptual framework, we finish the analysis by testing how psychological and personal factors influence intentions towards urban agriculture. Table 9 presents results from the bivariate ordered probit model, investigating the influence of subjective knowl- edge, attitudes, purchase frequency and socio-demographics on the likelihood to buy and grow produce at urban farms (two categorical de- pendent variables). First of all, results show that the correlation coeffi- cient Rho measures 0.447, indicating that the likelihood to purchase and to grow food at urban farms are positively correlated. This suggests that the more likely someone is to shop at the urban farm, the more like- ly (s)he would be to grow produce there. This result is also supported by the Wald test of independent equations.

Starting with the psychological factors, the findings show that those consumers who are more knowledgeable regarding sustainability in the food chain are more likely to buy produce at urban farms. This, however, has no impact on growing produce at urban farms. Furthermore, the urban agriculture knowledge index is significant and positive suggest- ing the more knowledgeable consumers feel towards urban agriculture, the more likely they are to buy and grow produce at urban farms.

With regards to consumers’ attitudes results show that a generally favorable attitude towards urban farms (F1, e.g., urban agriculture in- creases produce consumption, helps to care about the environment) in- creases the likelihood of buying and growing produce at urban farms. Those holding more specific attitudes towards urban agriculture (F2, e.g., urban agriculture increases physical activity and saves money) are more likely to participate in growing fruits and vegetables at urban farms. Furthermore, differences among the reasons that encourage or prevent participants from purchasing produce from an urban farm are observed. As one would expect, the reasons that encourage purchase at urban farms indeed increase the likelihood to do so. Nevertheless, those that are believed to prevent purchase from urban farms have a significant and positive effect on growing produce at urban farms.

Looking at the personal factors, the more frequently consumers purchase local produce, the more likely they are to participate in urban farming, whether by shopping or by growing. The results on socio-de- mographics indicate that female and older consumers are more likely to grow their own produce. Also, as the level of education increases, the likelihood of consumers to do both, shop and grow produce at urban farms, increases as well.

6-Discussion and Conclusion

Urban agriculture offers direct access to local food, a promising op- portunity, considering that local food continuous to be very popular. While early studies on local food found that consumers consider tradi- tional food attributes such as price, appearance and quality to be more important than food origin (Kezis et al., 1984; Lockeretz, 1986), studies conducted in the 2000s displayed that consumers shifted their prefer- ences towards local food, to the extent that they are now willing to pay a premium for it (Onozaka and Mcfadden, 2011; Willis et  al., 2016; Boys et al., 2014; Loureiro and Hine, 2002). The reasons for this shift are consumer beliefs that local food is fresher, of higher quality, tastes better, enhances the local economy and benefits the environment (Zepeda and Leviten-Reid, 2004; McGarry-Wolf et al., 2005; Schneider and Francis, 2005). Our findings support this with evidence that a posi- tive perception of urban farms is based on food attributes that are part of product quality, such as, being safe to eat and being healthy. Overall, with regards to perception, we found that the three pillars of sustain- ability – economy, society and environment – all play an important part when it comes to consumer perception. However, we  do  note that the perceptions can be contradictory, for example, some consumers perceive produce from urban farms as more expensive (the majority) and others as cheaper (the minority). Furthermore, while these three categories contain each about 16% of the total associations each, the cat- egory that received the most associations and is the main driver of the perception of urban farms based on our content analysis is “Food & Food Attributes” with 38%. Finally, “Point of Sale” is determining con- sumer perception of urban farms. Looking at points of sale, farmers mar- ket in particular plays an important role. This is promising, given that those urban farms that sell directly to consumers could offer an alterna- tive retail venue. The category “Food and Food Attributes,” on the other hand, is dominated by the association of urban agriculture with organic. This suggests that consumers might assume that food from urban farms is produced organically, which seems to be supported by the fact that consumers purchase more local foods, instead of organic, since  the late 1990s (Gallons et al., 1997; Food Processing Center, 2001). This could be understood as displaying a shift in preferences from organic to local foods, or that consumers assume that local food posseses char- acteristics of organic food (Naspetti and Bodini, 2008; Onozaka et al., 2010), leading them to purchase more local. This phenomenon might warrant more in-depth research in the future.

In addition to consumer perception of urban agriculture, we investi- gated the role of subjective knowledge and attitudes as psychological factors that affect intentions towards urban agriculture. Descriptive findings show that consumers in general do not feel knowledgeable re- garding urban agriculture, which could indicate that consumer ion towards urban agriculture might increase the success of urban farms. This, for example, could be achieved via community cen- ters where classes on nutrition and gardening for different ages could be offered. Results regarding attitudes indicate that consumers general- ly have a positive attitude towards urban agriculture and that the rea- sons that prevent them from using those farms as shopping outlets are mainly associated with cost and inconvenience. Indeed, it has to be acknowledged that cost and inconvenience are considerable obsta- cles for shoppers (e.g., McGarry-Wolf et al., 2005). However,  given that prices at urban farms are not necessarily more expensive and, could rather be lower compared to other retail outlets, it might be that consumers misconceive some of the attributes associated with urban agriculture. This suggests that involved actors (e.g., the farmers) might want to provide information in this regard. Also, if convenience is seen as an obstacle, one could counteract this by highlighting that growing your own produce is associated with being outdoors, being active and becoming more involved and literate with food production. All of which are benefits to the individual given that most lifestyles nowadays are sedentary (Wakefield et al., 2007; Abraham et al., 2010; Caballero, 2007).

Our main econometric analysis investigated these findings further. Results show that several psychological and personal factors that influ- ence the purchase likelihood are different from those that increase the growing likelihood, while others are similar. Our data indicate that con- sumers who feel more knowledgeable would go to urban farms to pur- chase produce and are also more likely to grow their own produce. One can argue that this might be because they are more likely to trust their decision on this produce outlet. In addition, attitudes play an important role in purchasing versus growing, showing that a positive attitude will increase urban farm participation in both directions (buying and grow- ing) (F1), while a specific attitude (F2) will increase the chances of con- sumers deciding to grow produce themselves only. While this result contradicts Bamberg and Möser (2007) who find a weak linkage be- tween attitudes and (pro-environmental) behavior, our results are sim- ilar to Feldmann and Hamm (2015) who conclude from their extensive literature review that local food purchase behavior can be predicted by attitudes. Similar to Feldmann and Hamm (2015) we find a strong rela- tionship between intended behavior – as it relates to urban agriculture – and attitudes. Looking at the socio-demographic characteristics, the main result is that women are more likely to grow their own produce, and the same holds for an increase in age. This indicates that strategies that target male and younger consumers might be beneficial to increase the share of those who would consider participating in urban agriculture.

Nevertheless, this study is not without limitations. For example, the interpretation of the associations with urban agriculture, i.e., content analysis, is subjective and hence, a different set of researchers might cat- egorize some associations differently. Furthermore, direct questions in a survey, such as the ones used in this study, can lead to social desirability bias where participants answer questions in a way they believe will please the interviewer (Norwood and Lusk, 2011). However, past re- search shows that Web surveys lower this bias, particularly when ques- tions ask sensitive information, as compared to computer-assisted telephone interviewing and interactive voice recognition (Kreuter et al., 2008). Thus we are reasonably confident that social desirability bias was minimized, given that we used an online survey. Future re- search could reduce this and obtain results that are potentially less bi- ased using discrete choice experiments (Norwood and Lusk, 2011). Also, it needs to be kept in mind that the results are less generalizable given that the sample cannot be deemed as representative for the gen- eral population. Future studies could use broader, random samples to increase the representativeness of the answers.

Finally, we highlight that this survey is of hypothetical nature and does not test actual shopping behavior. Therefore, future research ave- nues might include estimation of consumer demand and preferences using revealed or (non-hypothetical) stated preference methods, such education towards urban agriculture might increase the success of urban farms. This, for example, could be achieved via community cen- ters where classes on nutrition and gardening for different ages could be offered. Results regarding attitudes indicate that consumers general- ly have a positive attitude towards urban agriculture and that the rea- sons that prevent them from using those farms as shopping outlets are mainly associated with cost and inconvenience. Indeed, it has to be acknowledged that cost and inconvenience are considerable obsta- cles for shoppers (e.g., McGarry-Wolf et al., 2005). However,  given that prices at urban farms are not necessarily more expensive and, could rather be lower compared to other retail outlets, it might be that consumers misconceive some of the attributes associated with urban agriculture. This suggests that involved actors (e.g., the farmers) might want to provide information in this regard. Also, if convenience is seen as an obstacle, one could counteract this by highlighting that growing your own produce is associated with being outdoors, being active and becoming more involved and literate with food production. All of which are benefits to the individual given that most lifestyles nowadays are sedentary (Wakefield et al., 2007; Abraham et al., 2010; Caballero, 2007).

No Comments

Post A Comment