Good deeds revisited: motivation and boundary spanning in formal volunteering

The study investigates the motivation of volunteers to serve at the Federal Association of German Food Banks and volunteers’ interactions with food donors, food pantry managers and users. Social exchange theory is used as a frame to investigate volunteers’ interactions in the context of boundary spanning. Twenty in-depth interviews were recorded, transcribed and analyzed through qualitative content analysis. Volunteers are predominantly socially motivated to work at the German Food Bank, but this is not necessarily reflected all their interactions with food pantry users. Even though the authority in these interactions rests with the volunteers, they still feel uncomfortable in some interactions. Volunteers’ interactions with managers are essential, because managers tell volunteers, which tasks to carry out in which manner. But the volunteers do not necessarily respect the instructions in all cases. The interaction with food donors are negatively affected through a mismatch in the perception of authority within the collaboration. In some interactions, both parties believe they have authority within the interaction, even though they are rather equal partners. The study provides best practice recommendations on how to train volunteers to avoid interaction problems with food pantry users and donors.


Introduction
In Germany, food waste and food insecurity are frequently discussed problems. Each year, approximately 11 million tons of food waste occur (Kranert et al. 2012) and at the same time, 8% of the population suffer from food insecurity (Pfeiffer et al. 2011). The number includes the share of the population living below subsistence level with a food budget of 100-199 Euro per month as well as homeless people. Food insecurity refers to the condition, where people do not have secure access to a sufficient quantity of nutritious food to assure an active and healthy life. Reasons for food insecurity can be the unavailability of food, insufficient financial means, inappropriate distribution or inadequate use of food at the household level (FAO 2015). In developed countries, such as Germany, food insecurity rather occurs due to inequality or poverty than due to food scarcity (Tinnemann et al. 2012;Pfeiffer et al. 2015;Vlaholias et al. 2015;Baglioni et al. 2017).
As both issues are undesired by the German society, government, individuals and organizations work to address them actively. One of the organizations counteracting both food insecurity and food waste is the Federal Association of German Food Banks (abbreviated in the following as German Food Bank). Both issues are featured prominently in the organization's goals (German Food Bank 2017;Lorenz 2012). The German Food Bank is a nonprofit volunteer-based organization collecting and distributing donated food items to people in need (Selke 2011a). Further services are handing out meals in soup kitchens, providing breakfast to schoolchildren, distributing second hand clothing and supporting users' medical, bank and authority visits (Lorenz 2012).
These services are provided by free and associated food pantries listed under the umbrella of the German Food Bank. Free pantries are in contact with the umbrella organization but independent in their operations. Associated pantries are following the organizational standards of the federal association in their operations. According to Von Normann (2011) and Lorenz (2012), the number of food pantries is steadily increasing in Germany since 1993. There are around 940 local food pantries across Germany (German Food Bank 2017) with 50,000 volunteers. Among these volunteers 46,800 are unpaid and approximately 3200 receive a maximum monthly income of 400 Euro for small-scale employment. Volunteers usually work 20 h a month and have been described to be highly committed and competent (Von Normann 2011). In Germany, approximately 1.5 million people rely on food assistance provided by the German Food Bank. Around 30% are children and teenagers, 53% unemployed adults and 17% retirees (Assig 2012).
In free as well as in associated food pantries, food pantry users need to provide documentation of their eligibility to receive food (Lorenz 2012), and volunteers hand out quantities adjusted to the users' poverty level. The food is provided by a wide range of donors, from small bakeries to large food retail chains (Lorenz 2012). Items are usually nonmarketable, such as items close to the best-before-date, excess seasonal produce, items with packaging flaws or incorrectly labeled (Midgley 2014). Since the food reaches the food pantries often in large quantities, volunteers inspect the items and sort them into serving sizes (Von Normann 2011). Each year the German Food Bank collects and redistributes approximately 120,000 tons of food (Lorenz 2012). Food pantry operations are overseen by food pantry managers who typically lead in a cooperative and participative manner. Delegation is supposed to be part of their management approach to motivate volunteers and staff members. However, with respect to food safety, food transportation and food preparation, laws as well as the umbrella organization require managers to provide precise directions and enforce these (Von Normann 2011).
The current body of literature on the German Food Bank as well as on food banks in other European countries and the U.S. is rather diverse. Mostly, the literature discusses food assistance as a societal problem, since food banks are gaining importance due to reductions in social welfare systems (Riches 2002(Riches , 2011Warshawsky 2010;Lutz 2011;Thuns 2011;Dowler 2014;Poppendieck 2014;Silvasti and Riches 2014). In addition, prior studies present political perspectives, for instance, on the establishment of food banks (Koc 2014;Tang et al. 2014). Prior research also has addressed the user perspective (Lorenz 2012;Van der Horst et al. 2014), managerial challenges (González-Torre and Coque 2016) and logistical issues (Baglioni et al. 2017). With respect to the latter two aspects, Poppendieck (1998) and McIntyre et al. (2015) criticized food pantry operations as dysfunctional. The critique addresses barriers to food access, food being unequally distributed, food provided not being nutritious or not meeting cultural needs, using food pantries as a shameful experience, and food pantry services as being less efficient than governmental support such as food stamps. Further studies focused on food insecurity (Davis and Tarasuk 1994;DeLind 1994;Anderson and Cook 1999;Gareau 2004;Vitiello et al. 2015) as well as the perspective of donors and their motivations to cooperate with foodbanks (Lorenz 2012;Vlaholias et al. 2015;Gruber et al. 2016). Research on the volunteer perspective has been limited to demographic profiles and motivations .
Among studies on the German Food Bank, volunteer work has not yet been studied in-depth. German studies were focused on sociological and political aspects of poverty and food insecurity (Selke 2011a, b), public perception of the organization and its service (Witt 2011;Häuser 2011) and the user perspective in the context of dignity (Hoffmann and Hendel-Kramer 2011) and vulnerability (Sedelmeier 2011). The present study explores the perspective of volunteers, including their motivations to serve at the German Food Bank, and the interactions between volunteers, food pantry users, food pantry managers and food donors. In a sense, volunteers take on the role of boundary spanners within the German Food Bank, as their tasks connect the organization with actors in the external environment. The organization relies on volunteers' time, physical and mental work as well as their social, cultural and human capital. Volunteers greatly influence the services provided. Therefore, the present study builds on social exchange theory and the concept of boundary spanning to better understand the nature of the interactions and the effects of volunteering on the German Food Bank as an organization providing services to people in need.

Literature review
Similar to other charitable organizations, the German Food Bank operates mostly with volunteers (Von Normann 2011; Selke 2011c). Most of the responsibilities of the volunteers include connecting the organization with actors in the external environment (Selke 2011c). Therefore, they can be considered as boundary spanners. Boundary spanners are individuals within an organization with the role of linking the organization's internal networks with its external environment (Aldrich and Herker 1977;Williams 2002;Ernst and Yip 2009;Holmes and Smart 2009;Isbell 2012). They have two basic functions, information processing and external representation (Heath and Frey 2004). Information processing relates to evaluating the amount and sources of support for the organization's goals (Holmes and Moir 2007). In addition, boundary spanners have the authority to act and communicate on behalf of the organization, because they act as representatives of the organization in the external environment (Aldrich and Herker 1977). When interacting with the external environment to contribute to the organization's goals, boundary spanners can either show willingness to compromise or choose a manipulative approach to successfully fulfill their duties (Isbell 2012). The latter approach is rather uncommon for boundary spanners when acquiring resources. In resource acquisition, boundary spanners must follow the organizational policies and act in the interest of the organization (Aldrich and Herker 1977;Brown 2005;Isbell 2012). They are required to maintain the organizational image and enhance its social legitimacy when in contact with the organizational environment (Aldrich and Herker 1977). In the case of the German Food Bank, the concept of boundary spanning is of particular relevance for volunteers serving food pantry users as well as picking up food donations from retailers or producers. Up to present, boundary spanning has not been applied and discussed in prior food bank studies.
Independently of volunteers taking on the role of boundary spanners, they provide time and skills to the organization, which results in adding value to its services (Tilly andTilly, 1994 in Wilson andMusick 1997). Since volunteering at the German Food Bank takes place within organizational structures, it is considered formal volunteering (Wilson and Musick 1999;Thoits and Hewitt 2001). Formal volunteering refers to unpaid work without strict obligations. Informal volunteering takes place outside of organizational structures. Wilson and Musick (1997) extended the definition of formal volunteering and characterized it as follows: & Volunteering is a productive activity and must be considered equal to work, because there is a market for volunteers, and similar to any other labor market, qualification and performance matter in the market for volunteering. & Formal volunteering involves collective action. Volunteering within organizational structures is often carried out on behalf of a shared idea or cooperative purpose, where benefits are not limited to those involved, but extend to society as a whole. & The volunteer-recipient relationship is ethical and regulated by incentives. Even though volunteers may have self-centered motives for their involvement, they still freely provide their time and skills. Moreover, volunteering is often framed within social and behavioral norms which have to be followed. & Different types of volunteer work are related to each other. Even though formal volunteering limits volunteers' leisure time, prior studies show that people who are committed to formal volunteering, are likely to also provide informal help (Gallagher 1994).
Since formal volunteering is often unpaid, motivation plays a crucial role (Do Paço and Agostinho 2012). According to Haivas et al. (2012), prior studies on motivation to volunteer found that volunteers are motivated to serve at organizations corresponding to their personal motives. The general motivation to volunteer has been investigated in several studies. These studies showed that people volunteer because of altruistic or egoistic motives, e.g., helping others or achieving goals (Cnaan and Goldberg-Glen 1991;Clary et al. 1996;Thoits and Hewitt 2001;Peterson 2004;Burns et al. 2006;Pajo and Lee 2011;Ferreira et al. 2012;Do Paço et al. 2013;Vázquez et al. 2015). Further motives included personal dispositions, such as empathy or prosocial attitudes (Clary et al. 1998;Davis et al. 1999;Penner 2002;Hustinx and Lammertyn 2003;Peloza and Hassay 2006). One of the most comprehensive approaches to understanding motivation to volunteer was proposed by Clary and Snyder (1999). They identified six motives, namely expressing important values, obtaining a better understanding, enhancing selfesteem, belonging to social groups, developing skills and career opportunities as well as the so-called protective effect, when volunteers strive to take of their mind off personal worries. For example, Agostinho and Paço (2012) found that volunteers in the Portugese Food Bank perceived personal development through volunteering as a motivation. Further, protective motives were found, since some volunteers appreciated working at the Portuguese Food Bank because they could forget their own problems. The main motives stated by volunteers was to help other people and the appreciation of the organization being built on values the volunteers shared.
Motivations influence how people perform tasks and interact with others. The German Food Bank relies on volunteers (Lorenz 2012) and how they carry out their tasks, e.g., collecting food from donors, sorting and repackaging food items and distributing food to the food pantry users (González-Torre and Coque 2016; Baglioni et al. 2017). These tasks involve interaction with donors, managers and users. Due to their role as boundary spanners, the volunteers' motivation and interactions need to be understood to evaluate the quality of service. Lopes et al. (2004) defined interaction as any occasion with at least two people, where both are associated with one another and adjust their behavior in response to each other. Within organizations, such as food banks, numerous interactions take place. Van der Horst et al. (2014) researched interactions in Dutch food banks and found that interactions between food pantry users and volunteers are not always positive. In some cases, interactions led to negative emotional reactions of food pantry users, since they felt volunteers expected them to act gratefully. The expected gratitude resulted in feelings of shame and distress for food bank users. Volunteers also perceived these interactions as negative, since it was a forced interaction with both parties obliged to interact with each other. Prior studies in the U.K. (Lambie-Mumford 2013) and in Germany (Von Normann 2011; Selke 2011a, b, c), also presented examples of negative interactions between food pantry users and volunteers. To use German and British food pantries, users must prove that they are poor to qualify for food assistance. For instance Selke (2011b) and Lambie-Mumford (2013) report cases where volunteers were rather inattentive in the registration situation and acted in a formal manner. This resulted in feelings of humiliation for the users and in users rejecting the services of the food bank, even though they were in need. An earlier study in Canada also showed negative interactions, where dependency and limited choices among food items led to tensions between volunteers and food pantry users (Tarasuk and Eakin 2003).
The interactions between food pantry managers and volunteers have not been researched per se. However, Tarasuk and Eakin (2003) and Lambie-Mumford (2013) shed light on these interactions. Tarasuk and Eakin (2003) showed that volunteers did not necessarily feel responsible if the quantity of food distributed was not sufficient to meet the users' needs. In these cases, volunteers were asked by food pantry managers to provide a small amount to each user as a symbolic gesture. Also, Lambie-Mumford (2013) found that volunteers were required to be attentive to food pantry users' problems, aside from being poor. Volunteers are asked to consider the pantry users' problems and connect them to other institutions that could potentially provide further help to individual problems.
In all types of interactions within the German Food Bank, interpersonal authority is present. As boundary spanners, acting on behalf of the organization, volunteers can be regarded as an authority to food pantry users. Similarly, food pantry managers are an authority to volunteers, as volunteers are hierarchical subordinates to managers. In any task that volunteers fulfill, they must follow the rules and instructions of paid staff members and managers González-Torre and Coque (2016). Accordingly, the volunteermanager interactions are greatly influenced by accountability. The volunteer-donor interactions can be expected to be of a similar nature, given that the volunteer is a representative of the German Food Bank, and the organization depends on donations. Even though the boundary spanner role provides volunteers with authority to represent the organization, food donors are not in a subordinate position to volunteers. Both parties command similar authority because the German Food Bank and the donating party are collaborating with each other.
According to social exchange theory, a theory that serves to understand workplace behavior (Cropanzano and Mitchell 2005); people are motivated by self-interest in their interactions with others. People engage in or end relationships depending on the advantages and disadvantages of being in that relationship compared to alternative options (Blau 1964). Transferring this to organizations, people evaluate and react to authorities and their decisions in terms of what they gain and lose from the authorities' decisions. One important factor regarding the evaluation of authorities is legitimacy (Tyler and Lind 1992). Whether authorities' decisions or procedures are perceived as legitimate is closely connected to whether they are favorable for the individual who will interact accordingly (Cropanzano and Mitchell 2005;Agneessens and Wittek 2012). Social exchange theory and evidence from volunteers' motivations discussed provide the framework for understanding how volunteers' interactions with other actors at food bank effects the services of the German Food Bank.

Methods
As the study seeks to explore motivations and interactions within the German Food Bank and their effects on the organization's service, a perspective which have not been previously studied in detail, a qualitative research approach is employed. According to Bitsch (2005), a qualitative approach is suitable when a theory is developed, an unknown research topic to be explored, or a new perspective added to a wellinvestigated topic. Furthermore, Bitsch (2005) stated that a qualitative research approach is suitable when a study focuses on the perspectives and experiences of actors in their lifeworld. It allows the identification of cultural framings and social realities (Bitsch and Yakura 2007). Since exploring the implications of interactions within the German Food Bank requires an understanding of multiple perspectives (Darbyshire et al. 2005;Perera et al. 2016), and the consideration of a vulnerable population (Hsieh and Shannon 2005), the food pantry users, in depth interviews were used for data collection. A quantitative tool such as a survey questionnaire would have been disadvantageous, since its preset wording would have hindered the interviewees expressing themselves in their own terms (Brand and Slater 2003).
Twenty in-depth interviews were conducted between the fall of 2015 and the spring of 2016. Of the twenty interviewees, five were food pantry managers, four were food pantry users, five were volunteers, one person was a spokesperson of the German Food Bank, and five were retail food donors. All interviewed managers were staff members and receive regular wages. They had been working for the food pantry between three and ten years. In addition, the spokesperson of the German Food Bank is a paid staff member. Volunteers were usually unpaid; only one volunteer interviewed was also a marginally employed staff member, receiving a wage of 400 Euro a month. This interviewee had administrative duties and was involved in food delivery as a truck driver. Two volunteers were short-term volunteers who had served three to twelve months. All other volunteers were long-term volunteers serving for five to ten years. Both groups of volunteers were obliged to volunteer regularly, depending on their agreements with the managers; volunteers came at least twice a week to the pantry. Four food donors were long-term collaboration partners, providing weekly donations, since five to ten years. These donors were large scale food producers or food retail chains. One small scale donor provided food only upon request, but was also a long term collaboration partner. The users interviewed were temporary and permanent users. Among them were migrants, short or long-term unemployed as well as people receiving only small retirement benefits.
Following Suri (2011) and Cleary et al. (2014), a purposeful sampling approach was used for the study. The approach requires researchers to select interviewees with a particular rationale and a specific purpose related to the research question in mind (Collingridge and Gantt 2008). With respect to the German Food Bank this approach was chosen, because demographic profiles of volunteers and other food bank actors that would allow representative sampling are not available. With respect to the selection criteria, interviewees needed to have knowledge on food bank operations, interactions within and related to the organization, and be willing to share their motivation to work at the food bank. Due to interviewees' specific roles and responsibilities there are variations in the information sought from interviewee groups (Table 1).
Interviews lasted 60 to 90 min each and were carried out by the second co-author and a graduate student. All interviewers received interview training, which included active listening, probing and paraphrasing, reflecting on their role and skills as an interviewer as well as being aware of interviewer effects. Interviews were either carried out face-to-face or by phone, depending on the interviewees' preferences. An interview guide outlined the topics of the interview. Topics were addressed through open-ended questions and asked according to the interview flow. The initial interview guide focused on food redistribution and volunteering at the German Food Bank. During the research process, the interview guide was further adjusted based on the input of early interviewees. Further modifications resulted from the interviewees' different backgrounds and activities related to food redistribution by German food retailers or to specific duties within the food pantries. This approach is in line with qualitative research procedures (Corbin and Strauss 2015;Bitsch 2005).
Eighteen of the twenty interviews were audio recorded and transcribed verbatim before data analysis. On two occasions, the interviewees did not agree to audio recording and, therefore, field notes were taken. These field notes as well as all interview transcripts were analyzed through qualitative content analysis. The analysis was carried out in integrative steps, since the process is iterative and recursive. During the analysis process, the interview transcripts were systematically fractured and common themes were extracted through constantly comparing and contrasting the data material. Ultimately motivations, interactions and authority patterns were identified.
The analysis process consisted of two main steps: open coding and the establishment of categories. During open coding, labels were assigned to text fragments. These labels were defined and reflected the key thought behind each text fragment. In the coding process, field notes and transcripts were carefully read several times. Throughout the reading process, codes were reconsidered and relabeled. The coding process linked all relevant interview excerpts with codes and their corresponding definitions (Table 2).
In the next step, categories were established. For this purpose, codes were grouped according to their meaning. According to Elo and Kyngäs (2008), establishing categories cannot be considered as a simple step of bundling codes that are similar or related; it is rather a classification of patterns. Researchers must decide through interpretation, which codes belong into the same category. Accordingly, each category was named by content-characteristic words and then defined. Category definitions do not consist of only one key thought; they are comprised of all related codes and their definitions. These main parts of the qualitative content analysis were carried out using the software package Atlas.ti. The software package allows to systematically analyze text and other documents. It includes tools to locate, code and annotate qualitative data material.

Results and discussion
The first part of this chapter identifies volunteers' motivation to serve at the German Food Bank, as a basis for the analysis of volunteers' interactions with other food bank actors. It is important to understand volunteers' motivation, since motivations influence how volunteers perform their responsibilities and solve problems in interactions. The second part focuses on the interpersonal interactions between volunteers and the respective other groups, food pantry managers, users and food donors, because in their role as boundary spanners they greatly affect the organization's services.

Motivation to volunteer at the German Food Bank
Volunteers highlighted helping people in need, serving the community, being involved in a social network, and continuing to be part of the workforce after retirement as motivation to be involved in the German Food Bank (see Fig. 1). The most dominantly stated motivation by all interviewees was helping people in need.
BMy motivation is not really related to the food, neither to waste, nor the food insecurity issue. I just want to help people who need help^(Food pantry volunteer, male, 50-60 years old, manager of a company).  Authors' elaboration *Denomination of codes follows Clary and Snyder (1999) everything. That is not how it works. It is nice that the state does a lot, but you have to do something as well^(Food pantry volunteer, female, 20-30 years old, student).

BWell
The first statement highlights the explicit desire to help, but rejects interest in food insecurity or waste. The organization's slogan BFood where it belongs^as well as the organizational goals to act against food waste and food insecurity strongly emphasize both issues (German Food Bank 2017). However, the redistribution of unmarketable food that would otherwise become waste is only one of several charitable activities performed by the German Food Bank. According to Lorenz (2012), food pantries also distribute clothing, do counselling and support users with medical, bank and authority visits. Both statements show volunteers have a desire to help, and care about the community where they live. Similar to permanent volunteers at the Portuguese Food Bank , the German volunteers interviewed also had strong social and philanthropic motivations. However, in contrast to the Portuguese volunteers, German Food Bank volunteers expressed no careerrelated motivation. An explanation might be the age of the volunteers interviewed. Many volunteers were older, close to retirement or in retirement, and had already established a career.
BI think it is wonderful. It really sustains me. Working at the German Food Bank really sustains me, because of the structure. It is never boring. I can bring my own ideas. This is very important to me^(Food pantry volunteer, female, 60-70 years old, homemaker). Classification according to Clary and Snyder (1999), White bubble: Own findings BI left my job, and afterwards, I needed some kind of activity. I did not allow myself just to sit around. That is when I started to become interested in the German Food Bank^(Food pantry volunteer, male, 60-70 years old, retiree).
These older volunteers were interested in community involvement and being part of a social network. They wanted to feel that they could still do useful work after retirement. These results confirm Wei et al. (2012) emphasizing subjective wellbeing as an important aspect in providing service among older volunteers. Building on Lorenz (2011), these motivations reflect a social development in the German society. Lorenz (2011) highlighted that parts of the German society are socially not integrated or became redundant as workforce. This development does not only apply to food pantry users as stated by Lorenz (2011), but also to volunteers and others excluded from gainful employment. In contrast, other older volunteers, socially integrated and wealthy, wanted to share part of their wealth with the community.
BI just feel I am really blessed and I would like to share some of my wealth, with other people and just want to do something good and help to make others feel rooted and aware^(Long term food pantry volunteer, male, 50-60 years old, manager of a company).
Besides philanthropic motivation, the statement shows the motive of poverty awareness. The volunteer is aware of food insecurity in Germany while he feels blessed and satisfied with his personal situation. But at the same time he is aware that his situation is not the standard. The statement confirms findings of Vlaholias et al. (2015) that awareness of need is an essential requirement for any type of philanthropic behavior.
Motivations found correspond to the classification proposed by Clary and Snyder (1999). However, volunteers interviewed expressed fewer self-centered motivations than found in prior studies. Their motivations were rather social and community centered. Reasons may be connected to the nature of the organization. As Haivas et al. (2012) stated, the organization where volunteering takes place is of central importance to volunteers, because their personal values and motives are reflected in the organization.

Volunteers' interactions with food pantry managers, food pantry users and food donors
When volunteers carry out their service at the German Food Bank they interact with food pantry users, mangers and food donors. Each type of actor described positive as well as negative interactions affecting the services provided by the German Food Bank.

Volunteers and food pantry users
At the surface, interactions between food pantry users and volunteers interviewed were mostly positive. Both volunteers and food pantry users mentioned emotional interactions with each other. The users interviewed explained that they were happy with the service and products provided and did not explicitly mention feelings of shame or humiliation.
BI do not have any bad experience with the German Food Bank. I am satisfied. I like the people inside. They are very polite. There are very nice people at the German Food Bank^(Food pantry user, female, 30-40 years old, homemaker).
BI was very surprised, everyone is so very nice and friendly. If I were to be treated differently, I would not come to the German Food Bank anymore^(Food pantry user, female, 80-90 years old, retiree).
These statements seemed to contrast with previous studies (Lambie-Mumford 2013; Van der Horst et al. 2014), where interactions between food pantry users and volunteers were described as negative, and dependency and helplessness of food pantry users and unfriendliness or reservedness of volunteers were criticized. The explanation for these differences is the strict code of conduct within the German Food Bank as an organization. Food pantry users must be treated respectfully, similar to guests in a restaurant; also, an open friendly atmosphere is required. However, the second statement implicitly indicates that the code of conduct is not practiced in every food pantry in Germany. The elderly woman says, she was surprised about the friendliness and if she were treated differently she would not continue to use the service. The implicit message is in line with Lambie-Mumford (2013) mentioning that British food pantry users refused to visit food pantries due to improper treatment. Another issue resulting in negative interactions emphasized by both volunteers and users was the understanding of neediness.
BIf they are not coming without any excuses, that they are sick or on holiday, whatever […]. If they are not coming without an excuse, they are dropped out from the list. Then a new person is coming. The waiting list is long. We suppose they are not in need^(Food pantry volunteer, female, 60-70 years old, retiree).
BYou cannot just say 'Yeah or I do not want it.' I do not know […]. I think with a little more charm and being nice they like you better. With a laugh or smile, they are nicer and you can get a potato more for example^(Food pantry user, female, 30-40 years old, homemaker).
The statements emphasize a particular understanding of neediness. Only users that come regularly are considered needy and therefore deserving to receive food assistance. When volunteers refuse to provide food to food pantry users coming irregularly or provide more food to some users, their behavior confirms prior findings of unequal treatment of food pantry users and barriers to food access (Poppendieck 1998;Lambie-Mumford 2013;McIntyre et al. 2015). The particular understanding of neediness and the respective volunteer behavior is also in line with the critique by Poppendieck (1998) and McIntyre et al. (2015) that different degrees of neediness are not considered. The issue of expected gratefulness (Tarasuk and Eakin 2003;Van der Horst et al. 2014) was found, but only implicitly as the user statements emphasized that volunteers appear more responsive to users showing gratitude but expected gratefulness was not addressed by volunteers and users in a direct manner.
In contrast to prior research (Tarasuk and Eakin 2003;Lambie-Mumford 2013;Van der Horst et al. 2014), the present study found examples of volunteers feeling helpless or weak in the interaction with food pantry users. Volunteers stated that they felt helpless, when they were unable to provide requested help or had difficulties to deny food pantry users' requests for extra food, and therefore acted against organizational rules. Volunteers had to negotiate the conflicting interests of the organization and the users, where following the request of one party can lead to dissatisfaction or distrust of the other. Therefore, in situations where volunteers did not follow the rules, they had to cope with the feeling of guilt. According to Van Schie et al. (2014) volunteers' perception of their own role and the perception of their task might have caused the feeling of guilt. Tasks that are subjectively perceived as unnecessary or unreasonable, have a negative emotional effect on volunteers.
Even though Wilson and Musick (1997) claimed relationships and interactions between volunteers and food pantry users as being ethical in every aspect, in practice that is not always the case. Interactions appear to be superficially respectful and polite, and in line with the German Food Bank's code of conduct. However, when explored at a deeper level, interactions were affected by volunteers' decisions and good will. In some cases, volunteers interviewed felt morally guilty and weak during the interactions, especially when acting against the rules, e.g., providing extra food according to user requests. Even though volunteers felt to be the weaker party in the interaction, the authority actually rests with them.
As boundary spanners, volunteers have authority to act on behalf of the organization and are asked to provide information to the German Food Bank concerning the specific needs of food pantry users. In addition, through the lens of the theory of social exchange, the authority rests with the volunteers in all interactions. The social exchange theory can serve as an explanation why interactions were still evaluated positively by food pantry users, even though there were underlying tensions. As the food pantry users (subordinate) benefitted from the volunteers' (authority) decisions, food pantry users considered volunteers as legitimate authorities and were ultimately satisfied.

Volunteers and food pantry managers
Interactions between volunteers and food pantry managers appeared mostly positive. Furthermore, volunteers followed managers' instructions and the German Food Bank's code of conduct in the majority of cases. This included using recommended language and fulfilling the tasks required.
BWe are asked to remind the guests, your card will expire on x of June, so you should bring the documents from the government. So they can prove that they are eligible to get help from our pantry. Then the head of the food pantry will re-issue the card and they can stay on the list, and people can continue to get the food from us^(Food pantry volunteer, male, 20-30 years old, student).
BIf the desk is not clean, there would be a penalty for the station manager. But fortunately I have reliable people [refers to volunteers in charge of an assigned food station, where users get served]^(Food pantry manager, male, 50-60 years old).
Both statements show that volunteers have to follow instructions by food pantry managers, for instance, using the requested terminology for people using the German Food Bank's services. Because food pantry managers wish that food pantry users are treated with respect, volunteers are required to call them customers or guests. Food pantries acting as food distribution centers where the users are provided with groceries call their users Bcustomers.^Food pantries with soup kitchens, where warm meals are handed out, call their users Bguests.^In addition to following such instructions on how to address food pantry users, volunteers need to complete tasks responsibly, this concerns hygiene, food safety and duties where food pantry users depend on them and ignoring instructions potentially has negative consequences for users.
Besides positive interactions, in some case the interactions between food pantry managers and volunteers were affected by inefficient and inappropriate work behavior of volunteers, for instance ignoring managerial instructions, the organizational hierarchy and the chain of command.
BI am kind to our volunteers. They are not paid. But sometimes I need to play hardball with them, if things get out of control^(Food pantry manager, male, 50-60 years old).
BI have a problem right now in [Name of a city]. I was called by a female volunteer who said that there is a customer who receives goods that he is not allowed to have because he isn't eligible. Well, she cannot agree that [Name of a city] carries on distributing that to him. My question to her was, why don't you speak directly with the manager in [Name of a city]? Then she said that she could not, because she has problems with him. Well, she has a problem with her manager; these are the things that I have to deal with as a state representative^(Food pantry manager and also elected state representative, male, 60-70 years old).
Even though food pantry managers command the authority through their hierarchical position in the work relationship, volunteers do not respect that in all situations. This form of defiance led to conflict between both parties. While the first statement addressed conflict between managers and volunteers. The second statement referred to conflict with a third party involved. The statement refers to a situation where a volunteer asked the state representative for help in solving a business problem, because the work relationship with the manager in her pantry was tense. The volunteer required the state representative to make a decision whether to provide food to a user who was not eligible to receive food. Thus, the organizational hierarchy and the chain of command was ignored, since a state representative was asked for a decision and not the food pantry manager in charge. Another aspect of the interviewee's statement is that the he was not only required to help with the business problem, but also with the interpersonal problem between the volunteer and the food pantry manager. Since the state representative held a higher hierarchical position than a local food pantry manager, he was expected to mediate between the two parties.
These findings add to Van der Horst et al. (2014) who presented negative interactions between volunteers and food pantry users. In a similar manner, the interactions between volunteers and food pantry managers were negatively affected, if instructions or the chain of command was not followed. According to social exchange theory, following instructions would be considered beneficial for volunteers, since they can refer back to the authority's word, if they need to perform unpleasant tasks. Furthermore, it would be beneficial for volunteers to be on good terms with the authority, because volunteers on good terms can expect to receive responsible and pleasant tasks. In addition, they do not run the risk of losing their position as a volunteer.

Volunteers and food donors
Interactions between volunteers and food donors showed both positive and negative interactions. Food donors were either satisfied or dissatisfied with the behavior and operational procedures of the volunteers.
BAlso, the problem with the German Food Bank was always that they did not have that many people, and they couldn't collect daily. Because some things must be collected daily. And they were always laying around here for a week. And especially in the summer the goods do not get better then. Because they also cannot always be refrigerated^(Owner of an organic supermarket, male, donates upon request of the German Food Bank).
The statement shows that the collaboration between the German Food Bank and the food donor required an extra effort by the food donor, regarding the selection and storage of food items. The German Food Bank does not accept products containing alcohol or products past the best-before-date. While some donors are dissatisfied with the volunteers' appearance and behavior, other shared positive experiences.
BWe cannot stand them anymore; this is not correctly put. Well, those people came again and again. They always came back and said, 'This is not all right. We only take this with us, but that one we don't.' But it depends on the person [volunteer collecting the food items]. Now we donate it all to the [Name of a city] Youth center. This does not hurt me, because I think, now the food it with someone who is very active in youth work, and the kids are very happy, even though the chocolate is expired^(Manager of a German food retail chain, female, used to donate to the German Food Bank once a week).
BThe collaboration with the German Food Bank is something very positive. All of them are volunteers, very nice people. They usually come on Tuesday mornings, we are used to it. They usually wait until I am available to provide them with the selected food items. Things are great, without any stress^(Manager of a German food retail chain, male, donates to the German Food Bank once a week).
Dissatisfied donors complained about volunteers not maintaining a low profile or that the volunteers collecting the food changed too often and they could not build a trusting relationship. These negative findings contrast with prior studies (e.g., Devin and Richards 2016) presenting mainly positive collaborations between retailers and organizations collecting surplus food. Satisfied donors were happy that volunteers came on the agreed upon day, and adjusted their operations accordingly.
The perceived differences of the interactions can be explained by social exchange theory and both parties' perception of their role and respective authority within the interactions. In the cases of perceived positive interactions between volunteers and food donors, both parties acknowledged each other as equal partners in the collaboration. Volunteers respected the authority of the food donors, because donors provided the resources needed by the German Food Bank. The role as boundary spanners requires volunteers to acquire food items and collect information with respect to the available quantities. Their role provides them the authority to act in the best interest of the organization. Accordingly, they are on time and collect food in a manner that respects donors' needs. Similarly, food donors acknowledged the volunteers as representatives of a collaborative organization. As donors also benefit from the collaboration, they approach volunteers supportively, for instance pre-sorting food items. When the interaction was perceived negative, each party believed that the authority rested only with them and the respective other party was in an inferior position. From the donors' perspective, the authority rested on the donor's side because the donor provides food items that the German Food Bank needs. From the volunteers' perspective, the authority rested with their organization, since the German Food Bank takes items, which would be discarded for costs otherwise. The dissatisfaction occurred due to differing perceptions of authority (Bondy 2008); in terms of authority, both parties are independent. Both parties should be willing to compromise since they depend on each other and their collaboration is beneficial for both.

Conclusions
Results support the notion that volunteers' interactions with food pantry managers, users and donors affect the services provided by the German Food Bank considerably, both in terms of how services are performed and how the organization is perceived by food pantry users and food donors. The theory of social exchange is applicable to all the interactions identified. Volunteers reported having predominantly social motivations, but these motivations are not necessarily reflected in their interactions with food pantry users in all cases; since some interactions seem to result in users' reluctance to continue to take advantage of the Food Bank's services. Further, the understanding of neediness causes tensions in volunteer-user interactions. The mismatch between motivations based on volunteers' good intentions, and interactions that suggest inappropriate use of authority over users, leads to the assumption that volunteers may not be fully aware of the importance of their organizational role as boundary spanners and the ensuing responsibilities.
To improve services provided by the German Food Bank and to remedy the potential interaction problems hinted at in the present study, food pantry managers need to develop more awareness among volunteers on food insecurity and related problems and how to overcome them. Volunteers should receive training on food insecurity and its effects on living conditions, as well as psychological training to better understand users' problems and needs. Such training is necessary since neediness takes different forms. In contrast to the volunteers' perception of need and food bank policies, coming irregularly is rather an indication of being in need, as it shows inability to manage daily routines. Following Kinnane et al. (2011), volunteers should receive formal and practical training. The authors researched training programs for volunteers in the Australian health sector, being exposed to emotionally challenging tasks, and emphasized the value of simulations and role plays in volunteer training. They stated that volunteers value the reality of the training situation as well as the support and safety provided by the training. Similar training practices should be promising for volunteers interacting with food pantry users.
As another way to reinforce the practice of the German Food Bank's code of conduct (respect, dignity and choice), the German Food Bank could follow examples in the U.S., where some food pantries have been transformed into supermarket-like locations where users can choose more freely among available items. Because food pantry users perceive registration and requirements for proof of neediness particularly humiliating, free pantries could consider following the U.S. model where some pantries allow registration on an honorary basis.
In addition to effects from volunteer-user interactions, interactions between food pantry managers and volunteers also affect the services provided. Managers set the rules that govern how volunteers carry out their duties. One point of conflict between the two parties is that volunteers do not necessarily respect the chain of command, and thus may ignore the rules in some cases. Most likely, their role as boundary spanners contributes to these interaction problems, as the role provides extensive authority, which can be misunderstood. Consequently, the role as boundary spanners should be explained to volunteers explicitly when they begin to serve at the Food Bank. The boundary spanner role gives volunteers the authority to act on behalf of the organization, but volunteers are expected to act in the best interest of the organization and follow managerial instructions. A formal memorandum of understanding may be helpful to both parties.
With respect to the interactions between food donors and volunteers, it is important that volunteers perform their boundary-spanning role in a manner that shows willingness to compromise, because tensions can occur when one party imposes on the sensitivities of the other. No party has authority over the other, but the German Food Bank relies on retailers and vice versa, with regard to redistribution. The complexity of this relationship requires sensitivity in interactions. Accordingly, inexperienced volunteers should receive training from more experienced volunteers, and accompany them when collecting food donations. When collecting food, volunteers should make sure that the collection fits the schedules of the retailers involved. Volunteers should avoid collection during busy times as it may be disruptive to retailers' operations. Moreover, having the same volunteers pick up each time at a particular retail outlet could help to develop routines and foster trusting relationships. Volunteers should be trained to be discrete, because not every retail chain or manager wants customers to know that they cooperate with the Food Bank.
Future research should further explore the role of volunteers in food pantries. As boundary spanners, volunteers are granted a certain authority, but in many ways they fulfill a serving role, in which they need to adhere to their authorities' orders. Building on Netting et al.'s (2005) study of how volunteers with multiple role profiles, perceive their duties and roles in faith based organization, further studies should investigate the role perception of volunteers in food pantries, as they are required to fulfill various professional roles. Insights into the volunteer's perception of their own role may help to assign them to tasks in the pantries as well as improve their services. This can help alleviate some of the problems regarding users, managers and donors addressed in this study. Also the focus on volunteer-user interactions can be deepened in future research, following Rombach and Bitsch (2017) a focus on the interaction with refugees in food pantries is suggested, as interactions with this group of users may be particularly difficult for volunteers due to cultural differences and language barriers.
As this work provides insights into interactions of German Food Bank actors and donors, future work could deepen the knowledge on other relevant actors and their interests in the organization through a more complete stakeholder analysis. Stakeholder analysis allows the identification of individuals or groups that affect or are affected by operations of an organization, and classifies them according to their impacts on the organization and the impacts the organization will have on them. In addition to the actors addressed in the present study, further external stakeholders, such as local governments and other societal groups should be investigated. Analysis of interactions with organizations that compete with the German Food Bank for resources, such as the Food Sharing organization, would add to an integrated perspective.