In search of information: Investigating source and channel choices in business-to-government service interactions

GOVINF-01139; No. of pages: 13; 4C: Government Information Quarterly xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Government Information Quarterly journal homepage: www.elsevier.com/locate/govinf

In search of information: Investigating source and channel choices in business-to-government service interactions Yvon van den Boer a,⁎, Rex Arendsen b, Willem Pieterson a a b

University of Twente, Center for e-Government Studies, P.O. Box 217, 7500, AE, Enschede, The Netherlands Netherlands Tax and Customs Administration, Centre for Professional Development and Communication, Herman Gorterstraat 75, 3511, EW, Utrecht, The Netherlands

a r t i c l e

i n f o

Article history: Received 30 January 2015 Received in revised form 27 November 2015 Accepted 27 November 2015 Available online xxxx Keywords: Source choice Channel choice Business-government service interactions Information seeking

a b s t r a c t Businesses, far more frequently than citizens, need government information and services to exert their rights and duties. In optimizing their service delivery processes, governments tend to focus on a set of communication channels they offer to businesses to deliver these services. However, in reality, businesses use not only a plethora of different channels but also multiple information sources, extending beyond governments. The focus of this study is to investigate the use of information sources, communication channels and the interaction effects between these sources and channels. A quantitative survey among 1218 representatives of small and mediumsized businesses was conducted. The results show that in most cases, businesses use set combinations of channels and sources; however, driven by factors such as the exact situation, people deviate from these patterns. The results provide novel and extensive insights into how governments can improve service delivery to one of their most important client groups – businesses. © 2015 Published by Elsevier Inc.

1. Introduction For many years, governments have sought suitable service delivery strategies to interact with citizens and businesses concerning various matters (e.g., tax matters). However, due to the rise of new electronic channels, sharing information with others is easier and less expensive than previously (Mulgan, 2004). As a result, people seek a wider variety of appropriate information sources for various situations (Boase, Horrigan, Wellman, & Rainie, 2006). This also seems true for information seeking about, for instance, tax matters; for example, De Vos (2008) found that businesses consult various information sources when searching for governmental information. Furthermore, electronic channels are used not only for socializing but also for seeking information, making decisions, and exchanging experiences about financial matters (Boase et al., 2006). The primary aim of many governments is to develop service delivery strategies that guide information seekers to electronic channels such as websites and portals (OECD, 2012). These channels are assumed to be less expensive than traditional ones such as the telephone (i.e., voice-phone calls) and face-to-face communication, which implies personal help from employees for each individual information seeker. However, the use of these more costly channels remains high (OECD, 2012, 2014). Electronic channels are not fully replacing the traditional channels, as was first expected, but exist in addition to existing channels (e.g., Pieterson & Ebbers, 2008; OECD, 2012, 2014). The availability of electronic channels has increased the ⁎ Corresponding author. E-mail address: [email protected] (Y. van den Boer).

moments of contact between citizens and the government (Pieterson & Ebbers, 2008). Many individuals use sequences of channels rather than making isolated choices (Stephens, Sornes, Rice, Browning & Saertre, 2008). Therefore, whereas businesses and citizens have more potential sources and channels to choose from – which increases the chance that they will choose inappropriate sources and channels to fulfill their tasks – governments must manage more service channels. This limits the efficiency with which governments can provide services to citizens and businesses. Thus, the increasing roles of other information sources and the growing number of available channels that can be used to obtain information from sources have made the information flow between governments and citizens and businesses increasingly complex (Van den Boer, 2014). These notions lead to the question of how governments should anticipate the availability of numerous information sources and channels for more efficient and effective service delivery. Current service delivery strategies of governments are primarily rooted in multichannel management and channel choice theories. The strategies solely focus on organizations' own channels without incorporating the role of the source behind the channels they have at their disposal (Van den Boer, 2014). Furthermore, channel choice theories focus on the moment of choice and its influencing factors (e.g., Media Richness Theory of Daft & Lengel, 1986, 1984, Social Influence Model of Fulk, Schmitz, & Steinfield, 1990). These theories frequently neglect possible influential aspects that occur around the moment of choice (e.g., the use of the channel itself). Although some theories seem implicitly to hint about source choice (e.g., Dual Capacity Model of Sitkin, Sutcliffe, & Barrios-Choplin, 1992), there is no explicit attention given

http://dx.doi.org/10.1016/j.giq.2015.11.010 0740-624X/© 2015 Published by Elsevier Inc.

Please cite this article as: van den Boer, Y., et al., In search of information: Investigating source and channel choices in business-to-government service interactions, Government Information Quarterly (2015), http://dx.doi.org/10.1016/j.giq.2015.11.010

2

Y. van den Boer et al. / Government Information Quarterly xxx (2015) xxx–xxx

to the role of the source, and aspects of sequence are not included. This lack of insight must be addressed. It is necessary to search for theories beyond the channel choice domain that (a) do not use the moment of choice as the point of departure but adopt a broader perspective (e.g., the perspective of information seekers), (b) incorporate both channel and source choice, and (c) provide insight into the flow of information-seeking processes related to source and channel choices. Such a theoretical perspective is presented in the information-seeking behavior literature. As such, insight is provided into the various roles sources and channels play in the networked environment, in which businesses must select relevant sources and suitable channels. This paper addresses the following research question and its sub-questions: Which source and channel choices do businesses make, and in what sequence, during information-seeking processes? RQa: How many sources do they use (on average)? RQb: How many channels do they use (on average) to contact a source? RQc: What reasons exist for selecting more than one source-channel combination? The scope of this paper is limited to the context of business-togovernment (B2G) service interactions. By B2G service interactions, we mean that businesses take the initiative and decide the source from which to obtain information and the channel used to do so. We focus on the channel modes of consultation and conversation (Ebbers, Pieterson, & Noordman, 2008). In consultation, there is single sided interactivity between seeker and source. The information seeker consults a source to obtain the required information. There is two sided interactivity in the conversation mode; the information seeker requests information, and the source provides this information, customized to meet the seeker's needs. These modes are informative, whereas there is also, for instance, a transaction mode in which monetary exchanges occur (Ebbers et al., 2008). The latter mode is beyond the scope of this paper. Although governments deliver services to citizens and businesses, this paper focuses on businesses. The B2G context is far more complex and networked than that faced by citizens, and there is a lack of substantial insights into this group (van den Boer, Pieterson, & Arendsen, 2011; van den Boer, Wijngaert, van de Pieterson, & Arendsen, 2012). Aspects indicating a complex and networked context include the following: 1) businesses have more contact moments with governments than citizens do because the former are subject to a greater number of rules and regulations (Arendsen, van Engers, & Velde, 2006). 2) Some contact moments between governments and businesses are indirect – via intermediaries – whereas others are direct, which facilitates a network character (Jansen, Van de Wijngaert, & Pieterson, 2010). One of the most crucial aspects indicating a complex context is characterized by the notion that businesses can take many forms, from self-employed to businesses with numerous employees (Jansen et al., 2010). A business is a network of individuals, a network that can be partially influenced by and depend on the environment (Jaffee, 2001). Furthermore, Davis (1981) argues that 1) individuals within a business differ in background, perceptions, and experiences, and 2) individuals react differently to situations. This argument suggests that different employees use public service delivery in a variety of ways. This characteristic makes it rather difficult to determine the unique contact point of a business and therefore hinders a government's ability to realize a straightforward service delivery strategy. This lack of clarity hinders the development of valuable insight into businesses' needs, expectations and search behaviors based on easy-to-obtain characteristics (e.g., an organization's size and form). Therefore, additional knowledge about business behavior related to public service delivery is needed. Since we have found in an earlier conducted qualitative study indications for the influence of the position of information seekers on the flow of the information-seeking process (van den Boer, Pieterson, Arendsen, &

van Dijk, 2015), we decided to further study this aspect. So, this paper studies whether the information seeker's position in the organization is a factor allowing governments easily to anticipate the business's search behavior. Therefore, we formulate one additional research question: RQd: Are there differences in source and channel choices related to the information seeker's position in the business? This paper is organized as follows. It starts by providing some theoretical background on source and channel choices in the informationseeking context, followed by an elaboration on the method employed. The fourth section presents the findings of this study, and the paper ends with conclusions, discussion and implications. 2. Theoretical background 2.1. The concepts of source and channel Throughout this paper, source refers to the person or organization storing the information from whom (or which) that information can be obtained by the seeker (adapted from Christensen & Bailey, 1997). Examples of sources are governmental agencies, advisory organizations, industry organizations, friends, family and colleagues. Channel is defined as the means by which information is transferred between the source and the seeker (adopted from Pieterson, 2009). In this paper we consider channels to be equivalent to media; examples include the telephone (i.e., voice-phone calls), e-mail, websites and face-to-face communication. According to Reddick and Anthopoulos (2014) there are three types of channels for government access: 1) traditional channels such as voice-phone calls and face-to-face contact, 2) egovernment channels such as website and e-mail, and 3) new digital channels such as social media and mobile apps. Thus, information can be obtained from information sources by using channels. 2.2. Source and channel choice behavior within businesses The behavior of employees is normally guided by the norms and social context of the organization. This idea of social constructivism (Giddens, 1979) posits that individuals' interactions create these social norms and affect how people behave in social contexts. This also affects how people communicate in organizations. For example, Contractor and Eisenberg (1990) note that both the communication structure and use of media shape one another through interactions. This observation is further expressed in Poole and DeSanctis (1990) adaptive structuration theory, which argues that structure and action in organizations are being produced and reproduced in the process of interacting via communication technology. However, the starting point of this paper is individual choice behavior. One of the main reasons for this approach is our focus on businesses that are relatively small. Addressing obligatory public matters is often a secondary task outside of the core business of small organizations (Bergers, 2003). It is more likely that a small team or a single person is responsible for the majority of contacts with government agencies and the search for public information. Consequently, it is less likely that (large) organizational aspects (e.g., organizational norms) play a major role in an individual's choice behavior. This situation increases the likelihood that choice behavior in the context of this study is predominantly determined by individual, rather than organizational, factors (van den Boer et al., 2015). Therefore, although we argue that seeking tax-related information in B2G-service interactions is primarily an individual activity, we do recognize that some of these organizational factors (e.g., vicarious learning and direct advice of colleagues) exert influence on choices in information-seeking processes. For small organizations, or even individual proprietorships, it is likely that influences come from the direct environment of the organization. For instance, Bergers (2003) argued that

Please cite this article as: van den Boer, Y., et al., In search of information: Investigating source and channel choices in business-to-government service interactions, Government Information Quarterly (2015), http://dx.doi.org/10.1016/j.giq.2015.11.010

Y. van den Boer et al. / Government Information Quarterly xxx (2015) xxx–xxx

friends, family, colleagues, and accountants are the ones who exert the most influence on the behavior of an entrepreneur. The Tax and Customs Administration exerts some influence, together with organizations such as banks, power companies and the Chamber of Commerce. Furthermore, the terms business, SME or company used in this paper all refer to the representatives of the businesses in the context of tax matters. In other words, we focus on the individuals responsible for tax matters in these organizations who act on behalf of the company. 2.3. Source and channel choices in different stages of information-seeking processes Kuhlthau's (1991) information-seeking model describes the information-seeking process in a purely sequential manner. The model adopts the user perspective. It has been mostly applied in education, although it is widely cited and tested (Case, 2012; Ingwersen & Järvelin, 2005) and is universally applicable in any area (Case, 2012). The process flows from initiation to selection, exploration, formulation, collection, and presentation (Kuhlthau, 1991). At each stage, a suitable task is formulated to move into the following stage of the process (e.g., recognizing a need for information is a task for the information seeker in the initiation stage). At every stage of the informationseeking process, source and channel selection seem to occur; however, exactly which sources and channels are selected in what stages remains vague. As the information-seeking process unfolds, actions change from discussing and debating with others in earlier stages to more-specific sources that provide more-focused information later in the seeking process. Although Kuhlthau's model was developed before the rise of electronic channels, it teaches us that information-seeking processes flow from general, undirected searching to more directed seeking activities. A model that employs a slightly different approach is that of Byström and Järvelin (1995), which contains both sequential and causal elements. The information-seeking process starts with the information need and flows to a choice of action, then implementation, and ends with evaluation. Source and channel choices are assumed to occur in the ‘choice of action phase’ after the recognition of an information need. Similar to Kuhlthau (1991), the authors adopt the user's perspective and focus on employees (i.e., employees in a Finnish city secretarial office). Byström and Järvelin (1995) argue that the six stages in Kuhlthau's (1991) model can be divided into a problem formulation and a problem-solving step. The stages initiation, selection, exploration, and formulation are related to the problem-formulation step. The other stages, collection and presentation, are related to problem solving. Task complexity is a crucial factor that affects the flow of information-seeking processes, including what sources and channels are used (Byström & Järvelin, 1995). There is no problem-formulation phase for simple, routine tasks because “the inputs, process and outcomes are a priori known” (p. 984, Byström & Järvelin, 1995). Byström and Järvelin (1995) found that problem formulation and understanding are crucial for solving complex tasks. What follows is the problem-solving step, which is the only step in simple tasks. For complex tasks, different information (in type and complexity) is required. Such information is obtained via different types of channels from different types of sources. Compared with simple tasks, complex tasks involve the selection of more sources and channels. Julien and Michels (2000) studied information seeking in personal and work situations with a focus on preferred (ideal) and actual source choices. They differentiated three types of sources: (a) personal sources, i.e., those contacted via personal channels such as face-to-face and telephone, (b) formal sources that are in printed form, and (c) formal sources that are electronic (e.g., Internet, e-mail, voice-mail, and databases). Their results indicated that each information-seeking process consists of an average of 3.4 source choices. Personal sources are consulted most often as a first step, followed by printed sources and electronic sources. Concerning the latter, Julien and Michels (2000) noted that when started with an electronic source, more sources than

3

average were used, which leads to the notion that these sources were of limited usefulness. Another model, that explains its relevancy for this study in its focus on the business context, is the ‘general model relating information acquisition to the decision-making process’ of Saunders and Jones (1990). They focus on how managers use sources and channels to gain relevant information in decision-making processes. They assume that managers act and make decisions in complex and turbulent environments. Source and channel selection occurs during the informationacquisition component, which is affected by the various phases and routines of the decision-making process and numerous contextual influences (e.g., time pressure and task characteristics). The decisional component flows through three phases, including identification, development, and selection, which are closely related to various routines of recognition, search, and evaluation-choice (Saunders & Jones, 1990). They conclude that different sources and channels are more appropriate at different phases of the process, which is in line with Byström and Järvelin's (1995) understanding that different types of information lead to the selection of different types of sources and channels. Kuhlthau (1991) presented similar findings because they both recognize that in the early stages of the seeking process, channels are selected that allow passive and undirected viewing, which changes as the seeking process unfolds. Although Kuhlthau (1991) and Byström and Järvelin (1995) provide no insight into exactly what sources and channels are selected in each stage, Saunders and Jones (1990) are somewhat more specific. They note that more accessible sources are selected in the identification phase, whereas internal sources (i.e., colleagues within the organization) are more often selected in the later development phase. With respect to the selection of channels, they propose that face-to-face and written channels are used later in the process (i.e., in the selection phase). The next section will further elaborate on accessibility and other factors that affect choices of information seekers in information-seeking processes. 2.4. Factors that shape the flow of information-seeking processes Source-quality and source-accessibility are most likely the moststudied aspects in the field of information behavior that influence source choices (e.g., Agarwal, Xu, & Poo, 2011; Culnan, 1985; Fidel & Green, 2004; Gerstberger & Allen, 1968; O'Reilly, 1982; Lu & Yuan, 2011; van den Woudstra, Hooff, & Schouten, 2012, 2015; Zimmer, Henry, & Butler, 2007). Hertzum (2014) notes, in a literature review, that both quality (i.e., value or reliability of information and its relevance) and accessibility (i.e., time and effort to gain access to the source and access to the information) are important factors in source selection. van den Woudstra et al. (2012) studied the selection of human sources from a social capital perspective and found that source-quality exerts slightly more influence on source selection. However, whether quality or accessibility exerts the most influence on source selection remains a topic of discussion because findings are mixed (Hertzum, 2014). Agarwal et al. (2011) studied source selection of working professionals in Singapore. Compared with face-to-face, the online channel is used more frequently and is found to be more popular in the beginning of the information-seeking process (Agarwal et al., 2011). The email channel is chosen after face-to-face and less frequently. They also found that the telephone (or chat) is used less frequently compared with face-to-face and in a later stage of the process. Overall, their results indicate interaction between task importance and source quality; as the task becomes more important, information seekers place more weight on source quality. In addition, Christensen and Bailey (1997) found that increasing task equivocality interacts with source accessibility and affects media selection. The fact that task characteristics do play a crucial role in source and channel choices is widely recognized in media choice and information-seeking theories (e.g., Byström & Järvelin, 1995; Daft & Lengel, 1986, 1984; Fulk et al., 1990; Leckie, Pettigrew, & Sylvain, 1996; Sitkin et al., 1992). For instance, it is argued that as task complexity increases the need for domain and problem solving information increases,

Please cite this article as: van den Boer, Y., et al., In search of information: Investigating source and channel choices in business-to-government service interactions, Government Information Quarterly (2015), http://dx.doi.org/10.1016/j.giq.2015.11.010

4

Y. van den Boer et al. / Government Information Quarterly xxx (2015) xxx–xxx

and the number of consulted sources increases, while success of information-seeking decreases (Byström & Järvelin, 1995). In addition, Leckie et al. (1996) provided a review of studies related to informationseeking behavior of working professionals. They argue that the work role and the associated tasks of an employee strongly determine and shape the flow of the information-seeking process and the sources and channels selected. This is in line with findings of a qualitative study that we conducted earlier on the exact same topic and in the same context as that of the study presented in this paper (Van den Boer et al., 2015). In that study, we found that employees responsible for coping with financial issues and who occupied various positions in the organization (e.g., managing director or financial specialist) act differently (i.e., select different channels and sources). Thus, position of the information seeker in the organization might be a valuable aspect for governments to consider when designing and tuning their service channels. Furthermore, Hertzum (2014) concludes after an extensive literature review that human sources such as colleagues and other strong ties are consistently among the most frequently consulted sources in work and daily-life contexts. More specifically, Cross and Sproull (2004) argue that patterns of relationships between seeker and source affect source-selection processes. Previously, Cross, Rice, and Parker (2001) found, in a study among information scientists within a global organization, that information seekers receive various benefits from other people (i.e., sources). Some sources provide solutions, whereas others provide meta-knowledge or help with problem reformulation, are consulted for the validation of plans or previously obtained answers, or lead to legitimation from contact with a respected person. Aspects of time and energy play a constraining role in developing and maintaining relationships that provide a variety of information benefits. Cross et al. (2001) argue that people consulted for legitimation are likely highly sought out because they are considered a collective product of expertise or authority. Similarly, trusting relationships, which require great effort to establish and maintain, seem to be a prerequisite for problem reformulation. They further argue that electronic channels such as e-mail are useful for solution and meta-knowledge. However, such electronic channels are not suitable for problem reformulation and validation. Table 1 provides an overview, based on findings in the literature, of what sources and channels are used in the beginning of and later in information-seeking processes. It shows that findings vary across the different studies and contexts. In addition, some provide only general insights into what sources and channels are selected, whereas others are more specific. The work in general is limited in two aspects. The first is that most studies focus on either channels or sources, thereby neglecting potential interaction effects. The second is that if both sources and channels are included, only a very limited set is incorporated in the work. We know of no studies that incorporate a multitude of information sources and communication channels.

In summary, although it is difficult to exactly predict from current insights which sources and channels are selected at what time in the information-seeking process, note that all models (implicitly) support the notion that sources and channels are selected by information seekers according to the various roles they fulfill. Nevertheless, due to differences in (a) definitions and interpretations of the source and channel, (b) context and (c) focus of the studies, consistent and detailed results concerning which sources and channels are selected at what point in information-seeking processes are lacking. Our study tries to fill this gap in the current literature. 3. Method 3.1. Design We designed a questionnaire that is largely based on results of an earlier conducted qualitative study, which was of exploratory nature (Van den Boer et al., 2015). This quantitative study can be considered as the second step in a sequential mixed methods approach (Creswell, 2014). The purpose of the survey was twofold. The first goal was to build a model that provides insight into the factors that influence source and channel choices of businesses during their information-seeking process for information and how source and channel choice are interrelated (Van den Boer, 2014). Second, we aimed to gain a deeper understanding of how the information-seeking process flows (i.e., which sources and channels are selected in what sequence). This paper focuses on the second goal. The questions that contributed to achieve this goal are as follows: • What was your most recent tax-related question? And how complex, specific and important was this question? • Which source(s) did you consult to find an answer to your question? Please indicate in what sequence you consulted the source(s). • Which channel(s) did you choose to obtain information from your first consulted source? (You have indicated source x as your first source.) Please indicate in what sequence you chose the channel(s).

Respondents who indicated that they consulted more than one source or selected more than one channel were confronted with the following questions: • Could you please indicate why you have consulted more than one source to find an answer to your question? (more than one answer allowed) • Could you please indicate why you have obtained information from this source via more than one channel? (more than one answer allowed)

Table 1 Sequences in source and channel choices during information-seeking processes. Authors

Focus/context

(Type of) sources and channels selected in the beginning (Type of) sources and channels selected later in the process

Saunders and Jones (1990)

Organizational decision making (managers)

More accessible sources (easy to consult according to the least effort principle)

Kuhlthau (1991)

Information seeking in an educational context Daily life information seeking, such as work

Julien and Michels (2000) Cross et al. (2001)

Agarwal et al. (2011) Hertzum (2014)

Relationships within an organization with departments across different sites Information seeking of working professionals -

Internal sources (i.e., colleagues within organization) Face-to-face and written Discussing and debating with others Specific sources that provide more focused information Personal sources (i.e., people) contacted via personal Printed and electronic sources (e.g., Internet, channels (e.g., telephone and face-to-face) e-mail, database and voicemail) Electronic channels such as e-mail are suitable for solutions and meta-knowledge. For problem reformulation, validation of previously obtained answers and legitimation sources, with trusting relationships, are consulted Online channel (i.e., seeking for online information, Face-to-face and after that telephone and/or visiting an online forum or chatting) e-mail Human sources such as colleagues and other strong ties are the most frequently selected sources

Note. In this paper we mainly use the definitions for source and channel as presented in Section 2.1 However, when necessary, we make use of the original terms used in the referred studies to clarify specific situations and prevent misinterpretations.

Please cite this article as: van den Boer, Y., et al., In search of information: Investigating source and channel choices in business-to-government service interactions, Government Information Quarterly (2015), http://dx.doi.org/10.1016/j.giq.2015.11.010

Y. van den Boer et al. / Government Information Quarterly xxx (2015) xxx–xxx

Answers to these questions helped to uncover various stages in the information-seeking process and the roles specific sources and channels play in these stages. Furthermore, we were able to study how sourcechannel combinations evolve over time in the process because the respondents indicated which channel(s) they had used to contact their first selected source. 3.2. Sampling procedure and response The study occurred in the context of business-to-government (B2G) service interactions. It was conducted with Dutch Small- and MediumSized Enterprises (individual proprietorships and businesses up to fifty employees) searching for information to answer their tax questions. Due to the skewed distribution of organization size in the population, we choose to use disproportionate stratified random sampling for an efficient and effective sampling process (Foreman, 1991). We invited n = 21,000 businesses via a letter addressed to the management to participate in the survey. We asked the management to pass the invitation on to the person in their business responsible for addressing tax matters. The respondents were randomly selected from the database of The Netherlands Tax and Customs Administration (NTCA), which contains all registered businesses in The Netherlands. It was updated just before we performed our sample survey in January and February 2013. After two weeks, there was a reminder for those businesses that had not yet responded. Ultimately, 6% of all invited businesses completed the survey, which yielded a final number of respondents that completed the survey of n = 1218. A response rate of 6% is quite low. However, the fact that other studies in this context report similar rates makes it less surprising (e.g., Jansen et al., 2010; De Vos, 2008). A first reason that might explain this response rate is that the respondents were not panel members but instead belonged to ‘real’ populations in our target group. It seems plausible that the willingness to participate was less (in percentage terms) compared with panel members who have already shown a commitment to participate (occasionally) in a study. Moreover, the topic we studied is very specific, and only some employees in the invited organizations corresponded to the needed conditions. Businesses in our sample were addressing many different tax issues during the period of data collection and the period before because we asked them to describe their most recently experienced issue. We coded the given answers, which resulted in 49 generic categories. These categories reflect issues varying from specific (obligated) actions such as filing tax returns to (more voluntary or self-initiated) actions related to obtaining information about whether it might be profitable to change the legal form of the business. We observed a variety of all existing taxes in our sample, from taxes that businesses face only once a year in spring (e.g., corporate tax and income tax) to other taxes that must be filed more frequently (e.g., payroll taxes and VAT returns). Most respondents in our sample were men (68.3%) compared with 31.4% women. One-third of the respondents were aged 50–59 (35.1%), one-quarter were 40–49 (27.4%), and 15.5% were younger than 40. One-fifth of the respondents were older than 59 (21.7%). Most respondents were highly educated (54.5%), a third (33.5%) had a medium education, and 12.0% were lower educated. In terms of management, 43.1% of respondents were managing directors of a business without employees, and 40.0% managed a business with employees. The remainder of the respondents (16.9%) were (financial) specialists in the organization. The highly differentiated and specific target group constrains comparison with the entire population in The Netherlands because only more-general numbers are available at Statistics Netherlands (e.g., ratio of women and men among individual proprietorships). Nevertheless, comparison at the business level is possible (e.g., business size and sector). Because we used disproportionate stratified random sampling based on business size, we only compare our sample with the population in the context of sectors (see Table 2).

5

Compared with information from the Dutch Statistical Office (CBS, 2015), there are some slight overrepresentations (e.g., agriculture and manufacturing) and underrepresentations (e.g., building industry and renting, buying and selling real estate). Nevertheless, the largest underrepresentation is observed for financial services. This is not surprising because we did not want to include these types of businesses in our sample. They can act as information sources for other businesses seeking tax-related information and therefore may cause bias in our study. Nevertheless, the findings should be interpreted accordingly. 4. Results This section presents the results. It begins by providing an overview of which sources and channels are generally selected (§4.1), followed by the results concerning the number of selected sources and channels in the single-search process (§4.2). Subsequently, the results concerning the sequences in source and channel choices are presented (§4.3). The section ends by focusing on which source-channel combinations are selected during information-seeking processes (§4.4). 4.1. Chosen sources and channels Respondents were asked to indicate which source(s) they had consulted to find information that answered their latest question. The results are demonstrated in Fig. 1. Formal sources, such as the advisor (67.5%) and The Netherlands Tax and Customs Administration (NTCA) (56.6%), are most often selected. This is not surprising, bearing in mind the formal topic of the studied information-seeking processes. Nevertheless, personal sources such as friends or acquaintances (9.3%) and external contacts (8.0%) also fulfill a role in the informationseeking process. Similarly, respondents were asked to indicate which channel(s) they used to contact their first selected source. Fig. 2 shows that information seekers primarily rely on four channels, of which the traditional telephone (i.e., voice-phone calling) and face-to-face channels seem to be important for obtaining tax-related information. The electronic channels e-mail and website also fulfill a decent role. The written channel is chosen less frequently but remains used by 16.6%. As found in our qualitative study (Van den Boer et al., 2015) taxpayers often choose to confirm answers or solution by letter to get certainty. The next sections will further elaborate on this. We also asked respondents to rate the sources on their expertise, accessibility and level of trust. By doing so, we gained more insight into how the information seekers in our sample perceive the potential sources of issue solutions concerning tax matters (see Fig. 3). We used a 7-point semantic differential scale format. It is obvious that the advisor is perceived as the most trusted and accessible with the highest expertise. Table 2 Representativeness of our sample in the context of sectors (%). Sector

Our sample (2013)

CBS Statistics Netherlands (2015)

Agriculture Mineral extraction Manufacturing Public utility Building industry Trading Transportation and storage Accommodation and food serving Information and communication Financial services Renting, buying, and selling real estate Business services Culture, sports, and recreation Other business services Other Total

8.0 .5 8.2 .2 6.2 16.4 2.5 4.5 4.0 1.8 1.1 26.3 3.1 12.1 5.1 100.0

5.2 .0 4.5 .2 10.4 18.5 2.7 4.0 5.5 8.1 2.6 25.5 6.3 6.6 .0 100.0

Please cite this article as: van den Boer, Y., et al., In search of information: Investigating source and channel choices in business-to-government service interactions, Government Information Quarterly (2015), http://dx.doi.org/10.1016/j.giq.2015.11.010

6

Y. van den Boer et al. / Government Information Quarterly xxx (2015) xxx–xxx

Fig. 3. Perceived trust, accessibility and expertise of government, advisor and personal sources. Fig. 1. Overview of selected sources for answering the most recent question.

4.3. Sequences in source and channel choices The government is much less trusted and accessible. Personal sources (e.g. family/friends, colleagues) have the least expertise. Later in this section, we will see whether these perceptions clarify the observed sequences in source choices.

4.2. Number of chosen sources and channels Fig. 4 shows the number of sources and channels (to contact their first selected source) respondents selected to answer their latest question about tax matters. On average, 1.6 different sources were selected in a single information-seeking process. The average number of channels used to obtain information from the first source is 2.3. More than half (55.7%) of the respondents obtained sufficient information by consulting only one source. Fewer than half of the respondents (41.3%) selected only one channel. From the above information, it can be inferred that seeking tax-related information is a rather ineffective and inefficient process because information seekers use (on average) multiple sources and channels in a single information-seeking process. To gain insight into the impact of task characteristics on the length of the information-seeking process (i.e., number of chosen sources and channels), we asked respondents to rate their tax-question on complexity, specificity and importance. Next to that we asked to assess whether they think the result of their seeking process (i.e., answer or solution) will have small or big consequences for their business. The results of the conducted regression analyses showed that these aspects have no influence on the number of chosen channels (F (4, 1179) = 1.954, p N .05). Further, we found that task characteristics exert some influence on the number of chosen sources (F (4, 1151) = 18.015, p b .001). The results in Appendix A show that an increase in perceived complexity and perceived consequences of the outcome leads to an increase in the number of chosen sources (respectively β = .109, p b .01 and β = .114, p b .01).

Fig. 2. Overview of channels used for answering the most recent question.

Thus far, it is illustrated that the respondents used, on average, multiple sources and channels in their information-seeking process to find answers to their latest question. Fig. 5 represents the sources chosen in the first, second and third instance. Although the exact sequence of source choices cannot be derived from these results, they provide information about the various sources selected in various steps of information-seeking processes (i.e., each selected source represents one step). It is obvious that the NTCA and the advisor fulfill an important role in the beginning of the information-seeking process. Whereas the advisor fulfills a similar role in the second step, the role of the NTCA decreases. Combining this information with the results in Fig. 4 teaches us that accessibility is not of high importance because the NTCA (as a governmental agency) continues to be frequently selected. Personal sources (e.g., friends, family, colleagues, and external contacts) fulfill an increasing role as the information-seeking process progresses. We asked the respondents who indicated that they had consulted more than one source to indicate their reasons for obtaining information from several sources (see Table 3). The main reason for consulting a second source is to gain certainty about the accuracy of the information found (50.8%). Combining this result with the findings presented in Fig. 5 teaches us that the advisor, and to a lesser extent the NTCA, is the main source to gain this certainty because it is often selected as a second source. Furthermore, a quarter of the respondents (25.6%) stated that they consulted a second source because the provided information of the first source was insufficient. This might indicate inefficiency in the information-seeking process and thus also to some extent the design of the service delivery process. Additionally, one-quarter of the respondents (27.1%) continued their information-seeking process

Fig. 4. Number of chosen sources and number of channels (to contact the first selected source) in an information-seeking process.

Please cite this article as: van den Boer, Y., et al., In search of information: Investigating source and channel choices in business-to-government service interactions, Government Information Quarterly (2015), http://dx.doi.org/10.1016/j.giq.2015.11.010

Y. van den Boer et al. / Government Information Quarterly xxx (2015) xxx–xxx

7

Fig. 5. Chosen sources across the seeking process.

because they wanted to compare different answers or views. The desire for different views reflects the complexity of the questions businesses can face when addressing tax matters. There is no single answer or fact available, and careful consideration or interpretation is needed. Furthermore, the respondents were asked to indicate which channel or channels they had used to obtain information from the source they had selected first (depicted in Fig. 6). It obvious that the telephone channel (i.e., voice-phone calling) and to a lesser extent the website channel fulfill an important role in the beginning of the studied informationseeking processes. In the second instance, the role of the website decreases, whereas e-mail and face-to-face fulfill increasing roles, similar to the telephone channel. In the third step, the telephone channel becomes of lesser importance, and e-mail and face-to-face are the main channels selected. The respondents that indicated that they consulted more than one channel were, just as with the source choices, asked why they selected more than one channel (for results see Table 4). Most respondents obtained information from a source via several channels because ‘not all information was available through the first channel’ (56.7%). This indicates that some channels are simply not suitable (or have not been made suitable) to transfer the type of information needed. This is similar to the reason ‘to send extra information’ (24.1%). Furthermore, some channels may be unable to provide certainty, which forces information seekers to use another channel ‘to be sure that the information obtained earlier was interpreted correctly’ (41%) or ‘to confirm exactly what was agreed or discussed via the other channel’ (18.9%). Whereas for source choice, clear tendencies were observed in what role each source fulfills, and it is rather difficult to determine which channel is used for what exact reason because different channels fulfill some role in each step (e.g., e-mail, face-to-face and telephone are all channels selected often in the second instance). Section 4.4 provides more insight into whether and to what extent the role of a channel is related to the selected source.

Table 3 Reasons for consulting more than one source (more than one answer allowed).

To be sure that I have the correct information Sources fulfill various roles; some sources help me with interpretation while I use others for orientation. The desire to compare different answers or views before one makes a decision The provided information of the first source was not clear or was incomplete. Out of interest To already gain more information while I have to wait for an answer provided by my first consulted source For other reasons

n

%

301 206

50.8 34.7

161

27.1

152

25.6

44 29

7.5 4.8

6

1.0

In summary, these results imply that channels are selected in various stages of the information-seeking process. The website and telephone channels are regularly used as primary channels to obtain information. Different paths are possible; for instance, the website provides preliminary information, and interpretation occurs via another channel such as the telephone or face-to-face. An information seeker may also start by using the telephone, followed by contact via e-mail to exchange further information, which serves as preparation for face-to-face contact. Together, the above results indicate that sources and channels fulfill various roles in information-seeking processes because they are selected in various stages of the search process. The following section provides more insight into the particular source-channel combinations selected throughout the information-seeking process and studies whether there is an association between the source and channel. 4.4. Source-channel combinations during the information-seeking process To study whether there is an interrelation between source and channel choices and how source-channel combinations are selected over time in the search process, two chi-square tests were executed. One test focuses on the first selected source in combination with the first selected channel, whereas the second provides insight into the first selected source with second chosen channel. All types of source choices and channel choices were included in both analyses. The results of the first chi-square test are shown in Appendix A, and the results of the second test are shown in Appendix B. Both the first and second test show a significant moderate association between the selected source and the channel (χ2(18) = 536.59, p b .001 with Cramer's V = .384, p b .001 and χ2(18) = 206.73, p b .001 with Cramer's V = .310, p b .001, respectively). We use the values of the standardized residuals to interpret further the results of both analyses. A value greater than z = 3 or less than z = − 3 reflects a true effect of over- and underrepresentation at the p b .001 level (Agresti & Finlay, 2009). First, as shown in Appendix A, fewer respondents selected the combinations NTCA with face-to-face (z = −5.8) and NTCA with e-mail (z = −6.3) than would be expected. Contrasting findings were found concerning the combination NTCAwebsite, which is a very common selected combination (z = 11.6), much more than expected. Nevertheless, this same channel in combination with the advisor is a much less selected combination (z = −11.0) than expected. For this source, the telephone channel (z = 3.4) and email channel (z = 6.5) are commonly used. Concerning personal sources, the face-to-face channel is very often selected (z = 8.8). These results reflect which combinations tend to be selected during the first stage of information-seeking processes. These combinations were selected more often than expected. The above results are related to the first selected source-channel combination. When comparing these results with the second selected

Please cite this article as: van den Boer, Y., et al., In search of information: Investigating source and channel choices in business-to-government service interactions, Government Information Quarterly (2015), http://dx.doi.org/10.1016/j.giq.2015.11.010

8

Y. van den Boer et al. / Government Information Quarterly xxx (2015) xxx–xxx

Fig. 6. Chosen channels for the first consulted source.

source-channel combination (first source with the second channel) (see Appendix B), we observe a shift. First, the written channel is now commonly selected to obtain information from the NTCA (z = 4.1) more often than expected. The overrepresentation of the NTCA-website combination has decreased somewhat, although it remains an oftenselected combination (z = 5.6). Although fewer respondents selected the telephone as a first channel to contact the NTCA, more respondents selected this channel in the second instance than would be expected (z = 2.3). To contact the advisor, more respondents selected the faceto-face (z = 3.8) or e-mail channel (z = 2.5) than would be expected. Therefore, the results show that the second step (i.e., selecting a second source-channel combination) in the information-seeking process differs from the first step, because information seekers are inclined to select different source-channel combinations in a second instance. This implies that source-channel combinations fulfill different roles in informationseeking processes. Which channel is chosen to obtain information depends on the selected source and vice versa. 4.5. The position of the information seeker in the organization We also performed various analyses to determine whether the position of the information seeker in an organization influences (a) the average number of sources and channels chosen or (b) which sources and channels are selected throughout the information-seeking process. First, we ran an ANOVA test with a Bonferonni post hoc test to study the effect of position on the number of chosen sources and channels. There was no significant effect observed for the position in the organization on the number of selected sources (F (6, 1171) = .405, p = .876) or the number of selected channels (F (6, 1199) = 1.511, p = .171). Second, we ran two chi-square tests to determine whether information seekers in different positions in organizations select various sources and channels (only the first choices were analyzed). Concerning source choice, there is a significant but weak association observed (χ2(8) = 54.65, p b .001 with Cramer's V = .150, p b .001). The results for this test can be found in Appendix C. There is no significant over- or Table 4 Reasons for consulting more than one channel (more than one answer allowed).

Not all information was available through the first channel. To be sure that the information obtained earlier was interpreted correctly To confirm exactly what was agreed or discussed via the other channel To send extra information The source was not available via the first chosen channel. To ensure that the information arrived correctly To further discuss the topic For other reasons

n

%

408 295 173 136 80 73 15 13

56.7 41.0 24.1 18.9 11.2 10.1 2.1 1.8

underrepresentation found concerning managing directors without employees. The results indicate that fewer managing directors with employees select the NTCA source (z = − 3.1) than would be expected. The exact opposite is observed for financial specialists. Furthermore, managing directors consult the advisor (z = 3.5) more often than expected, whereas there are fewer financial specialists than expected that consult the advisor (z = −2.8). Therefore, managing directors with employees have a tendency to consult an advisor as a first source from whom to obtain information more often than expected. Financial specialists are inclined to obtain information from the NTCA more often than expected. Additionally, for channel choice, we found a significant but weak association with position (χ2(12) = 58.76, p b .001 with Cramer's V = .156, p b .001). The results for this test can be found in Appendix D. Again, there are no significant results observed for managing directors without employees. The results indicate that fewer managing directors with employees select the website channel (z = −3.8) than would be expected. In contrast, many more financial specialists obtain information from a source via the website channel (z = 4.3) than expected. Furthermore, fewer financial specialists choose to communicate face-to-face (z = − 2.6). More managing directors with employees select the telephone channel (z = 2.5). These findings teach us that managing directors with employees and financial specialists do have a tendency to obtain information via the telephone and website, respectively, more often than would be expected. 5. Conclusions, discussion and implications This quantitative study was undertaken to investigate which sources and channels businesses select, and in what sequence, during their search for information about tax matters. By doing so, the underlying roles of sources and channels in information-seeking processes were determined. We believe that these insights help governments address the available numerous sources and channels to improve the efficiency of their service delivery. Our study occurred in the context of B2Gservice interactions. This section answers the following research question based on the results of our study: Which source and channel choices do businesses make during information-seeking processes? In addition, answers to each sub-question will be presented. 5.1. How many sources and channels are selected in a single informationseeking process? And what are the underlying reasons to select a source and channel? RQa and RQb address the number of sources and channels information seekers use in a single information-seeking process, and RQc addresses the underlying reasons for using more than one source-channel combination. The results suggest that many information seekers do not end their

Please cite this article as: van den Boer, Y., et al., In search of information: Investigating source and channel choices in business-to-government service interactions, Government Information Quarterly (2015), http://dx.doi.org/10.1016/j.giq.2015.11.010

Y. van den Boer et al. / Government Information Quarterly xxx (2015) xxx–xxx

information-seeking processes after having selected and used a single source-channel combination. Instead, they often continue their searches and select one or more additional source-channel combination(s). Such new combinations might consist of another channel and another source or another channel and the same source. Our study shows that, on average, slightly fewer than two sources are consulted in a single information-seeking process. In addition, concerning the first selected source, more than two channels on average are used to obtain information from that particular source. Furthermore, our findings on specific source-channel combinations in the first and second instance suggest that for all sources, a considerable number of channels are frequently used to contact a source. Although some combinations are chosen more frequently than are others, the results suggest that for all types of sources, switches in source-channel combinations occur. These results, in combination with the previously mentioned average of two channels per source, imply that a single information seeker switches between channels to contact a particular source. Although previous studies (e.g., Pieterson, 2009; Stephens, Sornes, Rice, Browning, & Saetre, 2008; Van den Boer, 2014) have found that people use different sources or channels in one information-seeking activity, our study shows that people use a combination of multiple channels and sources to obtain the information they need. All the above leads to the initial conclusion that solving tasks concerning tax issues is often not that simple – that only a single source-channel combination is not sufficient. Our results indicate that feelings of uncertainty about the accuracy of the obtained information, insufficiency of information and the desire to compare different views or answers often lead to the selection of a second source-channel combination. The feelings may arise from task characteristics, because these affect the number of selected source-channel combinations (Van den Boer, 2014; Byström & Järvelin, 1995; Byström & Hansen, 2005). This possibility seems in line with Chowdhury, Gibb, and Landoni (2014), who argue that such feelings of uncertainty can be related to topicality and the information needed. However, our results do not support the effect of task characteristics on the number of selected channels. We did found that task complexity and the consequences an answer or solution has for the business influences the number of selected sources, although the effect is not very strong. Furthermore, although other studies found that relationship characteristics influence the number of selected channels (Van den Boer, 2014), the results of this work do not support this particular notion. Lack of awareness or unfamiliarity with the various potential sources and their channels may increase uncertainty and lead to unsuitable source and channel choices (Chowdhury et al., 2014). In addition, there may simply be a mismatch between the functionalities of the sources and channels on the one hand and the expectations of the information seekers on the other hand. These findings elucidate that it is important for both information seekers (i.e., businesses) and information sources (e.g., the government) that the appropriate combination be selected for each stage of the seeking process. 5.2. Are there differences in source and channel choices related to the information seeker's position in the business? The final sub-question (RQd) addresses whether differences in the position of information seekers in the organization lead to various source and channel choices. Our findings indicate that managing directors (with employees) and financial specialists differ in their search behavior. For instance, managing directors with employees consult the advisor via the telephone channel in the first instance, whereas financial specialists visit the website of the NTCA. Possible explanations for these findings are that they differ in knowledge about the topic (Byström & Hansen, 2005), in the time available for them to solve their tax-related tasks (Van den Boer et al., 2015) or in their interests (e.g., a director and major shareholder in a business also has his own funds). Although we found only a weak association between position and source and channel choices, Van den Boer (2014) found a strong effect of position on these choices. The effect of position appears to be stronger than

9

aspects such as social influences and situational factors (e.g., available time, point in time and place), which have minimal effect (Van den Boer, 2014). This supports the notion that, in small organizations, a small team or single person is responsible for addressing obligatory public matters (Bergers, 2003). It is a secondary task outside the core business. Therefore, searching for information about public matters (e.g., tax matters) is predominantly based on individual choices. 5.3. Which source and channel choices do businesses make during information-seeking processes? The primary research question of this study focuses on which sources and channels are selected by businesses during informationseeking processes and in what sequence. Table 5 provides an overview of our findings. The descriptive results concerning source choices indicate that the NTCA and the advisor are the most popular sources to consult in the first instance. For a second source choice, the advisor is by far the most consulted, and in the third instance, the information seekers seem to switch to the more personal sources. With respect to channel choices, our findings suggest that the telephone and the website are oftenchosen channels for a first round of obtaining information. The website channel was also the first used channel in the study of Agarwal et al. (2011). In the second instance, the telephone and e-mail are the two most used channels. For a third channel choice, e-mail and face-to-face are often selected. Thus, the telephone channel seems relevant to various types of stages and is found to be very popular for all types of sources. Therefore, the advisor as a source and the telephone channel play key roles in information-seeking processes of businesses addressing tax matters. That the advisor seems to be the primary source in the search process is not surprising. It is perceived as the source with the highest expertise, is the most accessible and is trusted most, which are all aspects that strongly influence source choices (e.g., Hertzum, 2014; Agarwal et al., 2011; Cross et al., 2001). It is also the source with a prominent place (i.e., influence on business processes) in the environment of small organizations (Bergers, 2003). The findings and line of argumentation in the study of Cross et al. (2001) that sources/channels provide various benefits seem compatible with our results. However, there is difficulty in comparing the findings because they studied internal source choices of employees working in a global organization, whereas we addressed mostly external source choices. Nevertheless, there are some indications in our data of the existence of various benefits. The channel-related results initially suggest that channels fulfill various roles regardless of which source is consulted. However, the results of our more-detailed analyses concerning the selection of specific source-channel combinations indicate that the selection depends on the type of sources and may change during the search process. The results indicate that some channels often are only selected to consult a particular source (e.g., the website channel for obtaining information from the NTCA). Moreover, specific sourcechannel combinations fulfill different roles in information-seeking processes, which means that some are more often consulted in the first step

Table 5 Overview of selected sources and channels, including specific combinations throughout the information-seeking process. Beginning of the information-seeking process

Later in the information-seeking process

Sources

NTCA, Advisor

Channels

Website, Telephone

Source-channel combinations

NTCA-website, Advisor-telephone and (to a lesser extent) advisor-e-mail

Advisor, personal sources (e.g., colleagues, friends/family) Telephone, e-mail, face-to-face contact NTCA-telephone, Advisor-face-to-face contact

Please cite this article as: van den Boer, Y., et al., In search of information: Investigating source and channel choices in business-to-government service interactions, Government Information Quarterly (2015), http://dx.doi.org/10.1016/j.giq.2015.11.010

10

Y. van den Boer et al. / Government Information Quarterly xxx (2015) xxx–xxx

(e.g., NTCA-website and advisor-telephone). However, others are often selected later in the process (e.g., advisor face-to-face and NTCA telephone). The results of our previously conducted qualitative study (Van den Boer, 2014; Van den Boer et al., 2015) showed that the NTCA website is often consulted to orient in preparation to contact (a) the primary source, who is often the advisor who helps with the interpretation, or (b) the primary channel of the same source, which can be the telephone channel, to gain more certainty about the information previously obtained from the website. 5.4. Limitations This study has several limitations. First, the decision to focus on source and channel choices in the context of public service delivery – or more specifically the context of the NTCA, with its obligatory regulations – hinders the generalizability of the results to commercial services in the private sector. Second, our choice to study source and channel choice behavior in the business context hinders generalizability of the results to the citizen context. Therefore, replication of our findings is necessary. Furthermore, it is questionable whether our findings apply to larger businesses. We argue that seeking tax-related information is primarily an individual activity in small organizations. However, in larger organizations, aspects such as social norms may influence the behavior of an employee. In particular, when there are several employees responsible for tax matters, it seems plausible that information seeking becomes a lesser individual activity. Therefore, further research is needed to determine whether the individual-behavior approach is appropriate for larger businesses. Another point of discussion relates to whether respondents correctly interpreted the concepts of source and channel. During the pre-tests, respondents were consistently asked to respond to the provided explanations of the source and channel, and adjustments were performed when necessary. However, it is questionable whether all respondents correctly interpreted the concepts because some provided incorrect answers to the question ‘Which sources did you consult?’ Some respondents answered that they selected the Internet. The respondents who indicated the Internet as one of their first consulted sources were removed (b 5%). Despite these various drawbacks of our study, we believe that we provide useful implications for both practitioners and theorists, along with ideas for future research. 5.5. Theoretical implications This paper contributes to current theories by elucidating the importance of broadening our view and of studying source and channel choices in more real-life settings, such as information-seeking activities, instead of focusing solely on the moment of choice. We wonder whether media choice and information-seeking theories fully address the details of source and channel choices throughout information-seeking processes. Because media choice theories primarily focus on choices in an isolated context, it is doubtful whether they apply to all steps (i.e., each source or channel choice) of the search process. In addition, because information-seeking theories explain aspects of sequence, they fail to provide insight into the factors that determine the flow of information-seeking processes. These findings highlight the need for a tighter integration between theories and research rooted in both domains to investigate further and extend our knowledge of source and channel choices in information-seeking processes. Future research should focus on (a) which factors hold influence at which stages of the information-seeking process and (b) the aspects that trigger information seekers to move on to the next stage. 5.6. Practical implications This paper brings to light the notions that (a) businesses use several source-channel combinations to obtain information that answers a

single question and (b) sources and channels fulfill various roles throughout the seeking process. Therefore, whereas current service delivery strategies of governments focus solely on the management of their own channels, our findings suggest that this view is too narrow because other information sources also play a crucial role in B2G service interactions. Thus, it is advisable for governments (and other types of organizations) to anticipate the roles of the relevant key information sources in the network and to determine their own role as one of those sources. In addition, it is important to link the role of an information source to the role of its channels. Governments and other information sources in particular should be aware of and must focus on those channels that are first selected. The use of these channels reflects the first moment of contact between a seeker and the source. For sources, this is the chance to fulfill the information need of seekers or to guide them directly to the appropriate channel. This prevents unnecessary use of channels and services. However, what if the observed role of an information source and its channels is not the desired one? For instance, what if it is observed that the NTCA website is often consulted to solve specific tasks, but the NTCA website primarily offers general information? In such situations, it is important to start the debate about what role the government should fulfill. What expectations should SMEs have for public service delivery? What are the responsibilities of SMEs themselves? Because governments impose tasks (e.g., tax declarations) on SMEs, those governments should provide a minimum level of services and cannot act as commercial service providers. Only for the more specific tasks, for instance, would it be appropriate to expect SMEs to solve their tasks with the help of advisors (i.e., intermediaries). Debating these aspects would help governments to formulate service-delivery strategies that incorporate both their own role and the advisor's role in a networked setting. Therefore, we suggest that governments should implement strategies that guide information seekers to desired source-channel combinations, which we call multisource and multichannel management. Crossreferrals should guide seekers to appropriate source-channel combinations, which include referrals to another channel or other sourcechannel combinations. For example, the NTCA can guide information seekers for specific information to the advisor by adding the following sentences on their webpage: ‘If you want more specific information concerning whether changing the legal form would be appropriate for your business, please call or visit a financial advisor. They can help you with the interpretation of the rules and finally making a decision’. It would further be advisable for governments to ensure that advisors have the correct information to help information seekers. This can be done simply by providing information and services that are customized to the needs and knowledge of advisors. Advisors (e.g., accountants, administration offices, and bookkeepers) obviously play a key role in B2G service interactions. Respondents in our study consider this intermediary the most knowledgeable, accessible and trusted information source. These advisors not only provide support and advice concerning tax-related issues but also assist businesses in applying for a subsidy or addressing pension schemes and paying salaries, for instance. For subsidy activities, advisors may also be involved in B2G service interactions on the levels of local government and provincial authorities. It seems plausible that our results are also applicable to these interactions about public matters. Addressing such matters seems similar because they are all outside of the core business of small organizations (Bergers, 2003). Such organizations lack the knowledge and time to analyze and interpret different rules and regulations. Currently, the NTCA develops service-delivery strategies based on general characteristics of businesses. Examples of such characteristics are businesses size, ‘age’ of businesses and their life events (e.g., start up, expansion, and take-over). However, the findings of this study indicate that choice making in B2G service interactions is predominantly determined by individual choices of employees. Information seekers in various positions in the organization select various sources and

Please cite this article as: van den Boer, Y., et al., In search of information: Investigating source and channel choices in business-to-government service interactions, Government Information Quarterly (2015), http://dx.doi.org/10.1016/j.giq.2015.11.010

Y. van den Boer et al. / Government Information Quarterly xxx (2015) xxx–xxx

channels. Van den Boer (2014) further found that ‘age’ and size of businesses are weak predictors of source and channel choices. Therefore, knowing who the organization's representative government contact is for tax matters, for instance, seems quite important to the design of the public-service delivery process. To anticipate the position of the information seeker and therefore to be able to provide customized information in ‘adapted language,’ we suggest that screening questions are implemented. Such an approach would help to determine the position of the information seeker when contacted. Although some implications are more evident than others are, we believe that they all can significantly contribute to the design of future studies that increase our knowledge on the selection process of sources and channels in real-life situations such as seeking for sometimescomplex information. The study also provides new directions for more efficient and effective public service delivery strategies. However,

11

more in-depth (quantitative) research is needed to gain insight into the various factors that influence source and channel choices of individuals in different positions in organizations in various steps of the information-seeking process. Such insights help governments to (a) cope with the numerous sources and channels available in the network, (b) provide an efficient and effective interpretation of their own role as a source, and (c) address information seekers in various positions who differ in background knowledge and available time to solve their tasks.

Acknowledgements The research was funded by the Netherlands Tax and Customs Administration.

Appendix A. Results of linear regression analyses: The effect of different task characteristics on the length of the seeking process

Number of Consulted Sources

Number of Selected Channels

B(SE) Constant Complexity Specificity Consequences Importance

.996 (.09) .045 (.01) .032 (.02) .050 (.02) .016 (.01)

β

B(SE)

β

.109⁎⁎ .063 .114⁎⁎ .043

2.103 (.19) .067 (.03) −.019 (.04) .031 (.04) −.026 (.03)

.075⁎ −.018 .033 −.032

Note. R2 = .056 for Source Choice and R2 = .003 for Channel Choice. ⁎ Significant at p b .05. ⁎⁎ Significant at p b .01.

Appendix B. Results of the chi-square test for first Source and first Channel Choice

Type of channel

Type of Source

NTCA

Advisor

Personal Source1

Other2

Observed frequency Expected frequency Std. residual Observed frequency Expected frequency Std. residual Observed frequency Expected frequency Std. residual Observed frequency Expected frequency Std. residual

Telephone

Face-to-face

E-mail

Written

Website

SMS, Whatsapp

217.0 258.4 −2.6** 354.0 295.2 3.4*** 19.0 23.7 −1.0 21.0 33.7 −2.2*

4.0 41.4 −5.8*** 65.0 47.3 2.6** 21.0 3.8 8.8*** 8.0 5.4 1.1

22.0 77.0 −6.3*** 149.0 87.9 6.5*** 6.0 7.1 −0.4 5.0 10.1 −1.6

9.0 4.7 2.0* 1.0 5.3 −1.9 0.0 0.4 −0.7 1.0 0.6 0.5

260.0 128.6 11.6*** 14.0 146.9 −11.0*** 0.0 11.8 −3.4*** 30.0 16.8 3.2**

0.0 0.8 −0.9 1.0 1.0 0.0 1.0 0.1 3.3 0.0 0.1 −0.3

Note. χ2(18) = 536.59, p b .001, Cramer's V = .384, p b .001. * p b .05, ** p b .01, *** p b .001. 1 Friends, family, colleagues, and external contacts are grouped together into personal sources. 2 Examples of ‘others sources’ are unions, associations.

Appendix C. Results of the chi-square test for first Source and second Channel Choice

Type of Channel

Type of Source

NTCA

Observed frequency Expected frequency Std. residual

Telephone

Face-to-face

E-mail

Written

Website

Social Media

SMS, Whatsapp

108.0 86.3 2.3*

13.0 50.9 −5.3***

66.0 90.0 −2.5*

27.0 12.5 4.1***

49.0 22.5 5.6***

0.0 1.5 −1.2

1.0 0.4 1.0 (continued on next page)

Please cite this article as: van den Boer, Y., et al., In search of information: Investigating source and channel choices in business-to-government service interactions, Government Information Quarterly (2015), http://dx.doi.org/10.1016/j.giq.2015.11.010

12

Y. van den Boer et al. / Government Information Quarterly xxx (2015) xxx–xxx

Appendix C (continued) (continued) Type of Channel

Type of Source

Advisor

Personal source1

Other2

Observed frequency Expected frequency Std. residual Observed frequency Expected frequency Std. residual Observed frequency Expected frequency Std. residual

Telephone

Face-to-face

E-mail

Written

Website

Social Media

SMS, Whatsapp

108.0 128.8 −1.8 11.0 9.5 −0.5 7.0 9.5 −0.8

109.0 75.9 3.8** 9.0 5.6 1.4 7.0 5.6 0.6

163.0 134.3 2.5* 9.0 9.9 −0.3 6.0 9.9 −1.2

6.0 18.7 −2.9** 0.0 1.4 −1.2 1.0 1.4 −0.3

7.0 33.6 −4.6*** 0.0 2.5 −1.6 5.0 2.5 1.6

1.0 2.2 −0.8 0.0 0.2 −0.4 3.0 0.2 7.1

−0.7 0.0 0.0 0.0 0.0 −0.2 0.0 0.0 −0.2

Note. χ 2(18) = 206.73, p b .001, Cramer's V = .310, p b .001. * p b .05, ** p b .01, *** p b .001. 1 Friends, family, colleagues, and external contacts are grouped together into personal sources. 2 Examples of ‘others sources’ are unions, associations.

Appendix D. Results of the chi-square test for position in the organization and first source choice.

First selected source

Position

Managing director without employees

Managing director with employees

Specialist1

Observed frequency Expected frequency Std. residual Observed frequency Expected frequency Std. residual Observed frequency Expected frequency Std. residual

NTCA

Advisor

Personal source

Other

232 220.6 .8 160 204.5 −3.1** 120 86.9 3.6***

226 251.6 −1.6 287 233.3 3.5*** 71 99.1 −2.8**

26 19.8 1.4 15 18.4 −.8 5 7.8 −1.0

35 28.9 1.1 22 26.8 −.9 10 11.4 −.4

Note. χ2(8) = 54.65, p b .001, Cramer's V = .150, p b .001. * p b .05, ** p b .01, *** p b .001. 1 Examples of specialists are controllers, financial employees, and administrative employees.

Appendix E. Results of the chi-square test for position in the organization and first channel choice.

First chosen channel

Position

Managing director without employees

Managing director with employees

Specialist

Observed frequency Expected frequency Std. residual Observed frequency Expected frequency Std. residual Observed frequency Expected frequency Std. residual

Telephone

Face-to-face

E-mail

Written

Website

SMS, Whatsapp

237 263.2 −1.6 283 244.1 2.5* 91 103.7 −1.2

47 41.8 .8 44 38.8 .8 6 16.5 −2.6**

85 78.4 .7 71 72.7 −.2 26 30.9 −.9

5 4.7 .1 5 4.4 .3 1 1.9 −.6

142 130.5 1.0 79 121.1 −3.8*** 82 51.4 4.3***

2 .9 1.2 0 .8 −.9 0 .3 −.6

Note. (χ2(12) = 58.76, p b .001, Cramer's V = .156, p b .001. * p b .05, ** p b .01, *** p b .001. 1 Examples of specialists are controllers, financial employees, and administrative employees.

References Agarwal, N. K., Xu, Y. C., & Poo, D. C. C. (2011). A context-based investigation into source use by information seekers. Journal of the American Society for Information Science and Technology, 62(6), 1087–1104. http://dx.doi.org/10.1002/asi.21513. Agresti, A., & Finlay, B. (2009). Statistical methods for the social sciences (4th ed.). Upper Saddle River: Pearson Prentice Hall. Arendsen, R., van Engers, T., & Velde Te, R. (2006). An empirical study on business-togovernment data exchange strategies to reduce the administrative costs for businesses. Project E-Society: Building Bricks, 311–323. Bergers, A. M. (2003). Communication with SME entrepreneurs: assistance for communication advisors of the federal government. (Retrieved from http://communicatieplein. nl/dsc?c=getobject&s=obj&objectid=). Boase, J., Horrigan, J. B., Wellman, B., & Rainie, L. (2006). The strength of internet ties: the internet and email aid users in maintaining their social networks and provide pathways to help when people face big decisions. (Retrieved from http://www.learningtimes. trainingshare.granatogroup.pewinternet.com/~/media/Files/Reports/2006/PIP_ Internet_ties.pdf.pdf).

Byström, K., & Hansen, P. (2005). Conceptual framework for tasks in information studies. Journal of the American Society for Information Science and Technology, 56(10), 1050–1061. http://dx.doi.org/10.1002/asi.20197. Byström, K., & Järvelin, K. (1995). Task complexity affects information seeking and use. Information Processing and Management, 31(2), 191–213. Case, D. O. (2012). Looking for information: a survey of research on information seeking, needs, and behavior (3rd ed.). UK: Emerald Group Publishing Limited. CBS Statistics Netherlands (2015). Vestigingen van bedrijven; grootte, rechtsvorm, bedrijfstak, regio's. (Retrieved from http://statline.cbs.nl/StatWeb/publication/?DM= SLNL&PA=81644NED). Chowdhury, S., Gibb, F., & Landoni, M. (2014). A model of uncertainty and its relation to information seeking and retrieval (IS&R). Journal of Documentation, 70(4), 575–604. http://dx.doi.org/10.1108/JD-05-2013-0060. Christensen, E. W., & Bailey, J. R. (1997). A source accessibility effect on media selection. Management Communication Quarterly, 10(3), 373–387. http://dx.doi.org/10.1177/ 0893318997010003005. Contractor, N. S., & Eisenberg, E. M. (1990). Communication networks and new media in organizations. Organizations and Communication Technology, 143, 172.

Please cite this article as: van den Boer, Y., et al., In search of information: Investigating source and channel choices in business-to-government service interactions, Government Information Quarterly (2015), http://dx.doi.org/10.1016/j.giq.2015.11.010

Y. van den Boer et al. / Government Information Quarterly xxx (2015) xxx–xxx Creswell, J. W. (2014). Research design-qualitative, quantitative & mixed methods approaches. SAGE Publications. Cross, R., & Sproull, L. (2004). More than an answer: information relationships for actionable knowledge. Organization Science, 15(4), 446–462. http://dx.doi.org/10.1287/orsc. 1040.0075. Cross, R., Rice, R. E., & Parker, A. (2001). Information seeking in social context: structural influences and receipt of information benefits. Systems, Man, and Cybernetics, Part C: Applications and Reviews, IEEE Transactions on, 31(4), 438–448. Culnan, M. J. (1985). The dimensions of perceived accessibility to information: implications for the delivery of information systems and services. Journal of the American Society for Information Science, 36(5), 302–308. Daft, R. L., & Lengel, R. H. (1984). Information richness: a new approach to managerial behavior and organizational design. In L. L. Cummings, & B. M. Staw (Eds.), Research in organizational behavior, 6. (pp. 191–233). Homewood, IL: JAI Press. Daft, R. L., & Lengel, R. H. (1986). Organizational information requirements, media richness and structural design. Management Science, 554–571. Davis, K. (1981). Human behavior at work: organizational behavior. New York: McGrawHill. de Vos, H. (2008). Kanaalkeuze: voorkeuren van ondernemers. Enschede: Novay. Ebbers, W. E., Pieterson, W. J., & Noordman, H. N. (2008). Electronic government: rethinking channel management strategies. Government Information Quarterly, 25(2), 181–201. http://dx.doi.org/10.1016/j.giq.2006.11.003. Fidel, R., & Green, M. (2004). The many faces of accessibility: engineers' perception of information sources. Information Processing and Management, 40(3), 563–581. http:// dx.doi.org/10.1016/S0306-4573(03)00003-7. Foreman, E. K. (1991). Survey sampling principles. New York: Marcel Dekker. Fulk, J., Schmitz, J., & Steinfield, C. W. (1990). A social influence model of technology use. In J. Fulk, & C. W. Steinfield (Eds.), Organizations and communication technology. London: Sage Publications. Gerstberger, P. G., & Allen, T. J. (1968). Criteria used by research and development engineers in the selection of an information source. Journal of Applied Psychology, 52(4), 272–279. Giddens, A. (1979). Central problems in social theory: action, structure and contradiction in social analysis. Berkeley: University of California Press, 50. Hertzum, M. (2014). Information processing and management. Information Processing and Management, 50(5), 775–795. http://dx.doi.org/10.1016/j.ipm.2014.04.003. Ingwersen, P., & Järvelin, K. (2005). The turn: integration of information seeking and retrieval in context. Springer. Jaffee, D. (2001). Organization theory: tension and change. NY: McGraw-Hill. Jansen, J., van de Wijngaert, L., & Pieterson, W. (2010). Channel choice and source choice of entrepreneurs in a public organizational context: the Dutch case. Electronic Government, 144–155. Julien, H., & Michels, D. (2000). Source selection among information seekers: ideals and realities. Canadian Journal of Library and Information Science, 25, 1–18. Kuhlthau, C. C. (1991). Inside the search process: information seeking from the user's perspective, 1–11. Leckie, G. J., Pettigrew, K. E., & Sylvain, C. (1996). Modeling the information seeking of professionals: a general model derived from research on engineers, health care professionals, and lawyers. The Library Quarterly, 161–193. Lu, L., & Yuan, Y. C. (2011). Shall I Google it or ask the competent villain down the hall? The moderating role of information need in information source selection. Journal of the American Society for Information Science and Technology, 62(1), 133–145. Mulgan, G. (2004). Connexity revisited. Demos Collection, 49–62.

13

OECD (2012). Forum on tax administration: working smarter in revenue administration—using demand management strategies to meet service delivery goals. Paris: Organization for Economic Co-operation and Development. OECD (2014). Forum on tax administration: increasing taxpayers' use of self-service channels. Paris: Organization for Economic Co-operation and Development. O'Reilly, C. A., III (1982). Variations in decision makers' use of information sources: the impact of quality and accessibility of information. Academy of Management Journal, 756–771. Pieterson, W. (2009). Channel choice: citizens' channel behavior and public service channel strategy. Enschede: Gildeprint B.V. Pieterson, W., & Ebbers, W. (2008). The use of service channels by citizens in The Netherlands: implications for multi-channel management. International Review of Administrative Sciences, 74(1), 95–110. Poole, M. S., & DeSanctis, G. (1990). Understanding the use of group decision support systems: the use of adaptive structuration. In J. Fulk, & C. Steinfeld (Eds.), Organizations and communication technology (pp. 173–191). Newbury Park, CA: Sage. Reddick, R., & Anthopoulos, L. (2014). Interaction with e-government, new digital media and traditional channel choices: citizen-initiated factors. Transforming Government: People, Process and Policy, 8(3), 398–419. Saunders, C., & Jones, J. W. (1990). Temporal sequences in information acquisition for decision making: a focus on source and medium. Academy of Management Review, 29–46. Sitkin, S. B., Sutcliffe, K. M., & Barrios-Choplin, J. R. (1992). A dual capacity model of communication media choice in organizations. Human Communication Research, 18(4), 563–598. Stephens, K. K., Sornes, J. O., Rice, R. E., Browning, L. D., & Saetre, A. S. (2008). Discrete, sequential, and follow-up use of information and communication technology by experienced ICT users. Management Communication Quarterly, 22(2), 197–231. van den Boer, Y. J. P. (2014). What's your favorite blend? Analyzing source and channel choices in business-to-government service interactions. Enschede: Gildeprint B.V. van den Boer, Y., Pieterson, W., & Arendsen, R. (2011). Channel choice in the G2B context: towards a research agenda. In T. Skersys, R. Butleris, L. Nemuraite, & R. Suomi (Eds.), Building the e-world ecosystem. Ifip aict, Vol. 353. (pp. 14–27). Heidelberg: Springer. van den Boer, Y., Wijngaert, L., van de Pieterson, W., & Arendsen, R. (2012). On the interaction of source and channel choice in the government-to-business context. In H. J. Scholl (Eds.), Electronic government. Kristiansand, Norway, September (pp. 27–39). Heidelberg: Springer. van den Boer, Y., Pieterson, W., Arendsen, R., & van Dijk, J. (2015). Exploring informationseeking processes by businesses: analyzing source and channel choices in businessto-government service interactions. International Review of Administrative Sciences. http://dx.doi.org/10.1177/0020852314564309 (Online first). van den Woudstra, L., Hooff, B., & Schouten, A. P. (2012). Dimensions of quality and accessibility: selection of human information sources from a social capital perspective. Information Processing and Management, 48(4), 618–630. http://dx.doi.org/10.1016/ j.ipm.2012.02.002. van den Woudstra, L., Hooff, B., & Schouten, A. (2015). The quality versus accessibility debate revisited: a contingency perspective on human information source selection. Journal of the Association for Information Science and Technology. http://dx.doi.org/ 10.1002/asi.23536 (n/a–n/a). Zimmer, J. C., Henry, R. M., & Butler, B. S. (2007). Determinants of the use of relational and nonrelational information sources. Journal of Management.

Please cite this article as: van den Boer, Y., et al., In search of information: Investigating source and channel choices in business-to-government service interactions, Government Information Quarterly (2015), http://dx.doi.org/10.1016/j.giq.2015.11.010