Citizens’ Communication Styles in Written Public Encounters

Abstract

Current studies of citizens’ coping behaviors in public encounters lack a direct examination of what citizens say and how they say it. Moreover, despite the ubiquity of citizens’ written communications with the state, such interactions are seldom studied. This article contributes a relational approach to studies of citizens’ coping behaviors by developing a taxonomy and detailed operationalization of citizens’ communication styles in written public encounters. This taxonomy is based on an abductive content analysis of over 1,000 citizen web contacts to an Israeli welfare agency, mostly during coronavirus disease 2019. We show that citizens’ written communications differ in their inclination to espouse demanding, pleading, civil, or neutral styles. Additionally, we tentatively show that citizens’ subjective experiences of administrative burdens and trust in government are correlated with the variation in their communication styles. The taxonomy, its operationalization, and preliminary findings open up new avenues for studying citizens’ coping behaviors in written encounters.

INTRODUCTION

Public administration scholarship has been criticized for its excessive focus on bureaucrats and its overlook of citizens’ experiences and behaviors in citizen-state interactions (Jacobsen et al. 2019). In response, a rapidly surging literature examines citizens’ experiences of administrative burdens (Herd and Moynihan 2019), that is the learning, compliance, and psychological costs citizens incur when interacting with bureaucracies (e.g. Baekgaard and Tankink 2022; Baekgaard et al. 2021a; Barnes 2021; Bell at al. 2023; Chudnovsky and Peeters 2021; Heinrich 2016, 2018; Herd and Moynihan 2019; Moynihan et al. 2015; Nisar 2018; Peeters 2020, 2023). Further advancing this citizen-focused perspective, a small number of studies analyze citizens’ “coping behaviors” (Bell et al. 2022; Masood and Nisar 2021; Nielsen et al. 2021; Peeters and Campos 2021), defined as the “behavioral efforts that citizens employ during and in preparing for interaction with public authorities in order to master the demands of the public encounter” (Nielsen et al. 2021: 3). These studies categorize the array of coping behaviors citizens employ when encountering administrative burdens or adverse bureaucratic decisions (Nielsen et al. 2021), and have made some progress toward explaining their variation (Bell et al. 2022; Gilad and Assouline 2024).

One limitation of the literature on citizens’ coping behaviors, which it shares with the bureaucracy-focused scholarship it criticizes, is its failure to directly examine the relational dynamics of citizens’ interactions with bureaucrats (cf. Bartles 2013; Boswell and Smedley 2023; Hand and Catlaw 2019; Oldenhof and Linthorst 2022; Thunman et al. 2020). Instead, studies of citizens’ coping behaviors are based on surveys and interviews (but see Masood and Nisar 2021), drawing their conclusions from interviewees’ narratives (Peeters and Campos 2021) or from respondents’ choices among structured responses (Bell et al. 2022; Nielsen et al. 2021). These methods capture a limited and distorted fraction of the variation in what citizens say and how they say it, which have been shown to influence bureaucrats’ responses (Brhun and Ekström 2017; Thunman et al. 2020). A second limitation of this emerging body of scholarship is that it takes it for granted that citizens’ coping behaviors are strategic. It thus overlooks the possibility that citizens’ behaviors, such as conveying anger, may sometimes reflect their uncalculated discharge of negative emotions in reaction to unfavorable bureaucratic decisions (Bell et al. 2022) and burdensome requirements (Hattke et al. 2021; 2020). Third, much of citizen-bureaucratic interactions are currently written and digital (Peeters 2023). This calls for extending the relational approach of studying the public encounter (Bartles 2013) to the unexplored domain of written communications. Thus, this article contributes a relational perspective to studies of citizens’ coping behaviors in written interactions with the bureaucracy. We thus extend the gaze of extant research, and the definition of coping behaviors, to what and how citizens write in order to influence bureaucratic discretion, and/or in reaction to administrative burdens. This definition clarifies our assumption that what and how citizens write is part of an interactive process wherein they both seek to influence bureaucrats and are affected by how they are being treated.

Using an abductive method (Ashworth et al. 2019; Timmermans and Tavory 2012), moving back and forth between inductive data analysis of citizens’ real-world written communications and flexible borrowing from multiple literatures, this article offers a taxonomy of citizens’ coping with bureaucracy in written public encounters. To build our taxonomy, we sampled and systematically analyzed over 1,000 citizens’ web contacts (cf. Reddick 2005; Thomas and Streib 2003) to a national Israeli welfare agency, mostly regarding unemployment benefits during the coronavirus disease 2019 (COVID-19) pandemic. Our coding and analysis unravel a variation in citizens’ demanding, pleading, civil, and neutral communication styles. These three styles involve citizens’ demeanor, by which we mean their displayed attitude towards bureaucrats and the bureaucracy, and its clustering with the type of justifications that they tend to raise.

Moreover, transcending the abductive development of the taxonomy, we employ mixed methods (Mele and Belardinelli 2019) to tentatively probe the causal mechanisms underlying the variation in citizens’ communication styles. To do so, we analyze the association between citizens’ communication styles and the type of concerns that they raise when contacting the welfare agency. We also analyze the association between burden experience and institutional trust and the communication styles of Israeli survey panel respondents when asked to write fictional letters to the welfare agency.

In what follows, we first discuss the array of theories that informed our taxonomy. We then elaborate on how we coded the data and present our conceptualization and detailed operationalization of citizens’ communication styles, which is the main goal of this article. Thereafter, we present our exploratory analysis of the mechanisms underlying the variation in citizens’ communication styles. Finally, we draw on the taxonomy and exploratory analysis to delineate directions for a research agenda regarding the variation, antecedents and consequences of citizens’ communication styles.

A THEORETICAL TOOLKIT FOR THE ANALYSIS OF CITIZENS’ COPING IN WRITTEN PUBLIC ENCOUNTERS

This article contributes a relational approach to current research regarding citizens’ coping behaviors by conceptualizing and operationalizing what citizens write and how they convey their messages when seeking favorable bureaucratic decisions or alleviation of administrative burdens. Below we discuss the diverse literature that informed our data coding, distinguishing two dimensions. The first, relating to what citizens say, involves their recourse to different justifications. The second, pertaining to how citizens communicate, regards a variation in their demeanor, which we define as their displayed attitude towards bureaucrats or bureaucracy (cf. Raaphorst and Groeneveld 2018; Raaphorst and Van de Walle 2018). Although “demeanor” can relate to various aspects of behavior, our definition, which focuses on what citizens’ behaviors convey about their attitude toward bureaucrats, is compatible with the measurement of this concept in empirical studies (e.g., Donovan et al. 2018).

Starting with the first dimension, regarding citizens’ justifications, we draw on research carried out in Israel by Danet (1971), Danet and Gurevitch (1972) during the 1960s and 1970s. That research distinctly tackled, as we do, citizens’ written communications with bureaucracy (but see Pfefferman et al. 2022, below). It involved textual analysis of citizens’ construction of “persuasive claims” when appealing decisions by the customs authority. The analysis revealed citizens’ differential inclination to appeal for tax officials’ consideration of their personal needs (e.g. “I am a sick man”), to claim their moral right for reciprocity (e.g. “I sacrificed a great deal in order to come and live in the Jewish State”), or to invoke “impersonal norms,” which we would label legal entitlement (e.g. “I understand that new immigrants are granted certain rights”) (Danet 1971: 852).

The variation Danet unraveled regarding the type of justifications citizens employ in written communications coheres with what we know about how bureaucrats judge clients. Studies of street-level bureaucrats show how their judgments of clients’ cases and requests are shaped by a blend of rules and cues for moral deservingness, relating to clients’ neediness and willingness to exert effort (e.g. Assouline and Gilad 2022; Guul et al. 2021; Jilke and Tummers 2018; Maynard-Moody and Musheno 2000, 2003; Scharm et al. 2009; Thomann and Rapp 2018). Welfare recipients’ perceived deservingness has also been shown to shape policymakers’ and bureaucrats’ willingness to impose administrative burdens on citizens (Baekgaard et al. 2021b; Bell et al. 2021). Likewise, analyses of how citizens assess others’ appeals for state support establish their inclination to distinguish between “entitlement,” which is rooted in legal rights, on the one hand, and moral deservingness, on the other hand (Feather 2003). In relation to the latter, trans-national studies by Van Oorschot (2000, 2006), and others (e.g. Heuer and Zimmermann 2020; Laenen et al. 2019; Meuleman et al. 2020; Nielsen et al. 2020) reveal that citizens’ assessments of fellow citizens’ deservingness for state support are shaped by their perceived effort and responsibility for their predicament, prior contributions to society, their belongingness to the in-group, and their perceived neediness. Since these criteria are known to shape both bureaucrats’ perceptions of clients, and citizens’ perceptions of fellow citizens’ deservingness for state support, it seems likely, as Danet and her colleagues indeed find, that when citizens seek bureaucrats’ cooperation and support, they too would allude to their legal entitlement and moral deservingness.

The second dimension that guided our coding, relating to citizens’ demeanor, is their displayed attitude towards bureaucrats, which is apparent from various studies, albeit using other labels. Masood and Nisar’s (2021) ethnography of the “coping strategies” of high-status Pakistani female doctors’ pursuit of maternity leave, finds that when facing bureaucratic hurdles, doctors are inclined to display their familiarity with administrative jargon and procedures and to insist on being noticed and heard. Conversely, Peeters and Campos’s (2021) interviews with welfare recipients in Mexico reveal their sense that they should forego overtly challenging bureaucrats on whom they depend. These studies indicate that citizens’ displayed attitude toward bureaucrats is partially associated with their internalization of their low or high social status. Still, a variation in citizens’ assertive versus docile demeanors in bureaucratic encounters has been observed even among the very poor (Mik-Meyer and Silverman 2019).

Building on the above qualitative studies, Nielsen et al. (2021) classified citizens’ coping behaviors in bureaucratic encounters, based on their clustered responses to vignettes representing prevalent bureaucratic incidents. The authors’ findings suggest that citizens’ coping behaviors differ along three axes: the first pertains to citizens’ proclivity for action versus passivity; the second to citizens’ inclination to prepare prior to bureaucratic encounters; and the third, which is the most relevant to our interest in demeanor, is the distinction between citizens who display their cooperation versus those who challenge bureaucratic authority. On the challenging side, citizens’ coping behaviors were operationalized by Nielsen et al. (2021) as involving the deployment of threats (e.g. to contact the media), questioning the quality of bureaucrats’ decisions, and “gaming.” Cooperative coping behaviors were operationalized as comprising citizens’ attempts to befriend bureaucrats by conveying an understanding of their difficulties, offering constructive solutions, or appealing for bureaucratic compassion.

Finally, the analysis of Pfefferman et al. (2022), although writing within a different research tradition (French sociology), is most relevant for our analysis, since, like Danet., it classifies citizens’ written communications in bureaucratic encounters. The authors’ analysis of archival data comprising entrepreneurs’ written applications for public support for their quest to establish new businesses during the 1930s and 1940s in pre-state Israel, found that applicants’ communications diverged in two respects. First were the justifications, or the “orders of worth,” of their claims, involving, inter alia, claims based on their role as carers for their family versus claims based on their civic contribution to society. Second, were their “formats of engagements,” involving the “rhetoric and discourse that actors employ in self-presentation to maximize their credibility” (Pfefferman et al., 2022, 5). Some, typically men, were inclined to espouse a “demanding” engagement format, asserting their right to funding and challenging bureaucrats’ actions as unfair, unjust, inefficient, or erroneous. Others, mostly women, tended to engage in “pleading,” seeking bureaucrats’ compassion, understanding, and goodwill. Finally, men were also more likely to display a “neutral” engagement format, involving their passionless justification of their pursuit of bureaucratic support as a means for achieving a distinct business goal. Additionally, Pfefferman et al. (2022) found that applicants’ justifications of their claims and their formats of engagement (which we label demeanor) tended to cluster, such that women were inclined to combine pleading and allusion to familial responsibilities and needs, whereas men either connected their past civic contribution to the community with a demanding format of engagement, or the projected financial success of their business with a neutral format of engagement.

Altogether, the above diverse literature suggests that citizens’ coping behaviors may differ in terms of their justifications, involving their allusion to their legal rights or to alternative deservingness criteria (cf. Van Oorschot 2000, 2006). Additionally, citizens may convey diversity in demeanor. Drawing on Pfefferman et al. (2022)’s notion of formats of engagements, which directly applies to written communications, the theorization and operationalization of Nielsen et al. (2021), and close reading of our own data, we sensitized our systematic coding to the possibility that citizens may adopt an oppositional or “demanding” (Pfefferman et al. 2022) approach. Alternatively, they may resort to submissive “pleading,” seeking bureaucrats’ compassion. They may also submissively convey friendliness and warmth to befriend bureaucrats (Nielsen et al. 2021), which we term as “civil.” Finally, they may adopt a “neutral” approach, focusing on the dry facts of their cases and their requests. Still, as explained below, juxtaposing our data coding against the above literature led us to adopt an integrated concept, “communication styles,” which focuses on how citizens convey their messages (i.e., their demeanor), with strong justifications, as specified below, as an integral component.

Last, before proceeding to the description of the data, its coding, and the emergence of our typology, we note that the above diverse studies are united in their assumption that what citizens write or say, and how they do so, reflects their conscious choice. That is, citizens craft their communications to influence bureaucrats’ use of discretion to impose or ease administrative burdens (e.g. Masood and Nisar 2021; Peeters and Campos 2021), or their adjudication of citizens’ applications (e.g. Danet 1971; Mik-Meyer and Silverman, 2019). It logically follows from this explanation that the variation in citizens’ justifications and demeanor reflects their divergent beliefs about bureaucratic motivations and scope of discretion (cf. Barnes and Henly 2018). For instance, citizens who display an oppositional (Nielsen et al. 2021) or demanding (Pfefferman et al. 2022) demeanor may be acting on an assumption that bureaucrats are able yet disinclined to help them unless pressured to do so; whereas those who present themselves as needy and helpless (Raaphorst and Van de Walle 2018) and plead for help (Pfefferman et al. 2022) seem to assume that bureaucrats can tailor their responses to applicants’ needs. Yet logically, citizens’ demanding or pleading demeanors may also stem from their uncontrolled emotional reactions to administrative burdens and unfavorable decisions. Namely, such communication styles may reflect citizens’ venting of emotions such as anger, anxiety, and frustration in reaction to compliance burdens Hattke et al. (2021), or unfavorable bureaucratic decisions (Bell et al. 2022). These explanations are not mutually exclusive. Citizens may display their assertiveness or helplessness because they believe that doing so is effective, and because they need to vent their negative emotions.

Drawing on the above theoretical toolkit, the following analyses advance two aims. First, to abductively conceptualize and operationalize the observed variation in what and how citizens write to the bureaucracy with an eye to what we know about types of justifications and demeanor. Second, to tentatively examine the expectation that this variation reflects citizens’ coping behavior in two respects, involving attempts to influence bureaucrats’ use of discretion, and emotional reaction to bureaucratic power and administrative burdens. In the conclusion we draw on these analyses to delineate a broader research agenda.

METHODS

The Research Context

The data for this project is mostly from the early phase of COVID-19 (March to November 2020). The web contacts we study are Israeli citizens’ written inquiries to the National Insurance Institute (hereafter: “the NII” or “the agency”), Israel’s main welfare agency, mostly relating to unemployment benefits. The Israeli government’s enforcement of lockdowns as of March 2020 generated unprecedented levels of unemployment. Moreover, due to the pandemic, the agency’s offices closed down for several months, and the web became citizens’ primary channel for contacting the organization. As a result, whereas in 2019 the agency received ~430,000 web contacts (and until September 2019 contacting the agency through the web was also technically limited), during 2020 web contacts amounted to ~2.5 million.

Web contacts are initiated by citizens via an e-form through the NII’s Web site, allowing them to freely convey their concerns in up to 1,500 characters. Citizens’ inquiries are prompted by various concerns. Some citizens seek information regarding their eligibility for benefits and the NII’s procedures. Others are facing delays in the processing of their applications or overdue payments of authorized benefits and the agency’s nonresponse to their prior messages, Still others are concerned about the rejection of their applications, or the sums of benefits received.

The Israeli context, and the COVID-19 period, likely influences the nature of the communication patterns we observe. First, the pandemic context means that the applicants who contacted the agency via the web during COVID-19 were much more diverse than would normally be the case. This diversity serves our goal of observing, conceptualizing, and operationalizing the array of citizens’ communication styles. However, future research may find that our findings are less applicable to the communication styles of a more typical welfare population of the repeatedly or permanently unemployed.

Second, due to the upsurge in applications for benefits and in web contacts, during the pandemic, applicants were subjected to objectively high burdens in terms of learning costs due to rapid change of rules, and compliance costs due to the agency’s inaccessibly, delays in application processing and in payments of benefits, and sheer mistakes in the calculation of the benefits paid to them. This exposure to burdens was discernable from the content of 85% of the web contacts (see section A of the Online Supplementary Appendix for our coding of 14 subcategories of administrative burdens).

Third, Israelis’ trust in government is generally low. According to the OECD 2020 survey data, only 39% of Israeli respondents conveyed their trust in the government, ten percentage points below the average trust across 40 countries. Thus, it is plausible that when interacting with bureaucracy, Israelis are inclined to act on the assumption that bureaucrats are disinclined to help them unless pressured or cajoled to do so. This may prompt Israelis to adopt legalistic justifications and an oppositional demeanor. In high trust contexts such coping behaviors might be less prevalent or take more nuanced forms.

Data Construction and Selection

Data access to the web contacts was provided by the agency’s then Deputy Director for Research and Planning. The selection and anonymization were carried out at the agency’s premises pre-coding in accordance with the authorization of the Hebrew University of Jerusalem’s IRB Committee. Since contacts were fully anonymized precoding, we unfortunately have no information on applicants’ demographics¹ and the precise date of their contact, nor background information regarding their case.

Our sampling, below, reflects the different channels via which citizens access the e-form and submit their contact, including: (a) the agency’s Web site, through which citizens can direct their inquiries, together with their ID and phone number, to either the (i) agency’s central contact center or (ii) to a local branch office, or (b) citizens’ password-protected individual accounts that are accessed through the agency’s Web site (hereafter: the personal account).

Data selection proceeded in two stages. We first constructed a pilot sample for the sake of inductive development of our conceptual framework and coding scheme and its gauging against multiple theories. The initial sample was manually selected by the first author, based on her reading of the texts, to reflect what appeared to be divergent communication repertoires. This purposive sample included 238 contacts, of which 77 concerned income support for the long-term unemployed and 161 related to temporary unemployment benefits, filed through the agency’s Web site between March and October 2020, containing no less than 200 characters.

Next, we randomly selected 850 contacts, submitted through the agency’s different web channels, all relating to unemployment benefits, each containing no less than 200 characters, predominantly, but not solely, from the COVID-19 period so: (a) 500 contacts initiated between 1 March 2020 and 30 November 2020 from within citizens’ personal accounts on the agency’s Web site; (b) 250 contacts initiated between1 February 2017 to 30 November 2020 through the agency’s Web site and directed to the NII’s contact center; (c) 100 contacts initiated between 1 February 2017 to 30 November 2020 through the agency’s Web site and directed to the agency’s branch offices. Including some pre-COVID-19 data was intended to enhance the diversity of the sample, but due to anonymization we cannot reliably compare pre- and post-COVID-19 contacts.

Taken together, the above sample of 1088 contacts—a pilot set of 238 and a randomly selected sample of 850—make up the data we employed to iteratively develop our conceptualization and systematic coding scheme. As evident, 350 of the 1088 contacts include some pre-COVID contacts. Since contacting the agency via the web was technically limited until September 2019, and due to the surge in web contacts during the pandemic, it is statistically given that these randomly selected 350 contacts are mostly from the COVID-19 period.

Our dataset is admittedly small compared with the population, which consists of millions of contacts, and in the absence of demographic details, we cannot weigh its representativeness. However, since our main goal is to conceptualize and operationalize the variation in citizens’ communication styles, the sample meets our need for diversity and offers a rare opportunity to directly observe what and how citizens write in public encounters with a welfare agency on a fairly large scale.

Coding Citizens’ Communication Styles

The initial pilot stage (N = 238) involved open coding, and consolidation of codes into higher-level categories (Glaser and Strauss 2017). This coding process alerted us to citizens’ frequent allusion to two broad types of justifications—legal entitlement and need—and infrequent recourse to other types, and to the variation in their demeanor. The coding scheme was then extended and fine-tuned, based on the additional set of 850 randomly selected contacts.

In relation to each contact, we systematically probed all the following aspects, each broken down into several explicitly defined codes, which were aggregated under higher-level categories: (a) citizens’ justifications, which we coded as pertaining to legal entitlement or to moral deservingness criteria. With regards to the latter, our coding followed Van Oorschot’s (2000, 2006) comprehensive categorization of deservingness criteria. If citizens alluded to multiple justifications, all were recorded; (b) citizens’ demanding, pleading, civil, or neutral demeanors, which, as mentioned above, borrowed heavily from Pfefferman et al. (2022) due to their similar analysis of citizens’ written communications. When citizens’ contacts displayed a mix of demeanors, our coding, again following Pfefferman et al. (2022), relates to their dominant approach; that is the one that manifested the most in their communication. Additionally, in view of extant research, we also coded, in relation to each contact: (c) assertions indicating citizens’ subjection to detailed categories of administrative burdens; (d) displays of administrative literacy (Döring 2021; Döring and Madsen 2022) as manifested in citizens’ allusion to specific regulations or to their change, their familiarity with specific agency forms, their usage of the agency’s online calculator, or their usage of bureaucratic jargon.

Following the first round of the full sample’s coding, we found, unlike Danet and Pfefferman et al. (2022), that in the context we study, possibly due to the diversity of citizens’ underlying concerns, citizens did not always raise a justification, that is an allusion to their legal entitlement and/or to one or more of van Oorschot’s deservingness criteria. This dimension was evident in 36% of the contacts (396 of 1088). Additionally, whilst we initially sought to formulate coding rules that would fully avoid the conflation of citizens’ justifications of their requests and their demeanor, we found that doing so was artificial. Rather, as detailed below, we found that forceful justifications, including assertions of acute need (e.g. hunger), or assertive requests based on one’s legal or moral entitlement (e.g. because I deserve/it is my right), constituted clear manifestations of citizens’ pleading or demanding demeanor, respectively. Hence, we adopted a new integrated concept, communication styles, which focuses on citizens’ demeanor, with forceful justifications as one of its manifestations. Still, we coded all justifications that appeared in citizens’ contacts and examined their clustering with their communication styles.

The following section presents and illustrates our conceptualization and operationalization of citizens’ communication styles, reflecting our elaborate coding scheme.² Section A of the Online Supplementary Appendix provides detailed information regarding other aspects of our coding, involving citizens’ justifications of their cases (20 subcategories), allegations of administrative burdens (14 subcategories), and displays of administrative literacy.

DATA-GROUNDED CONCEPTUALIZATION AND OPERATIONALIZATION OF CITIZENS’ COMMUNICATION STYLES

Our abductive coding process led us to distinguish, conceptualize, and develop a detailed operationalization to reflect four communication styles based on the content and structure of citizens’ texts. Below we illustrate the different facets of our operationalization via multiple examples, each 3-5 sentences long, with our underscoring. We do not elaborate on the content of each illustrative example, since the anonymized contacts are presented in full, and readers are not missing any background context to which we are privy. Moreover, our aim is not to explain each contact, but to illustrate the coding and the distinction between the different communication styles. Finally, although the coding guidance is detailed, directing coders’ attention to the use of specific words and phrases, punctuation and sentence structure, its application requires coders’ sensitivity to the overarching definition and meaning of each communication style.

Demanding Communication Style

First, coding the data, we detected some citizens’ “demanding” style (N = 317 of 1088 [29%]; N = 241 of the 850 random sample [28%]), which we conceptualize as a critical and/or adversarial demeanor. This demeanor conveys a sense of anger, frustration, or indignation. In terms of justifications, this communication style, as we show further below, often overlaps with citizens’ allusion to their legal entitlement (table 5). Our conceptualization and coding of this communication style draws on Pfefferman et al.’s (2022) usage of the same term and coheres with Nielsen et al.’s (2021) depiction of “resisters” and “fighters” who are mutually high on “degree of opposition” vis-à-vis bureaucrats.

More specifically, based on our inductive data analysis, we operationalized citizens’ demanding style as involving at least one of the following communication content and/or structural components. Relating to content, our coding of citizens’ demanding style comprises: (a) justifications involving strong claims of deservingness or legal entitlement such as “You owe me,” “I deserve,” “I have a right”; (b) phrases that are associated with making demands, such as “I demand,” “I assert,” (c) accusation of the NII (the agency) for unequal treatment, injustice, disregard, or immoral behavior; (d) mockery or cynicism of bureaucrats or the NII, and (e) threats to sue the NII, approach the media, the State Comptroller, and the like (cf. Nielsen et al. 2021). Additionally, we coded an applicant’s style as demanding when the contact displayed any of the following structural components: (a) use of imperatives, especially, but not exclusively, at the opening or at the end of a message, commanding agency action, provision of answers or explanations, and/or uncompromised urgent handling of one’s concerns; (b) two or more exclamation marks together or in different parts of the contact, two or more question marks combined (e.g. ??), or an exclamation and question marks combined (e.g. ?!); and (c) rhetorical questions that manifest criticism.

Table 1 provides illustrative examples for some of the above components of our coding. As is evident from the examples, contacts we coded as “demanding” typically displayed more than one relevant content and/or structural components, which together manifest citizens’ critical and adversarial attitude towards bureaucrats or the agency, and their display of negative emotions.

Pleading Communication Style

Second, also consistent with Pfefferman et al. (2022) typology of how citizens communicate, we observed some citizens’ inclination to espouse a “pleading” communication style (N = 144 of 1088 [13%]; N = 88 of 850 random sample [10%]), which we conceptualize as appeals for bureaucrats’ help, compassion and consideration, and submission to their goodwill. In terms of justifications, this style often overlaps with citizens’ allegations of need (table 5), and, in a similar vein to a demanding style, involves a display of negative emotions, yet one that conveys a sense of stress and anxiety. More specifically, we inductively operationalized pleading as involving at least one of the following components (a) justifications on the basis of an acute and immediate existential need; (b) requests for consideration or exemption from general bureaucratic requirements; (c) use of the root “beg,” or similar words; (d) combinations of “please” and “help,” and their synonyms, as well as repeated use of these terms; (e) expressions suggesting helplessness and vulnerability. In a similar vein to a demanding style, pleading sometimes involves exclamation and question marks; yet, in this case, these are employed to emphasize citizens’ helplessness and call for help. Table 2 provides an illustration of some of these components, although, as evident from the texts, citizens’ communications often displayed a combination of relevant indicators.

Civil Communication Style

Third, our coding further revealed some citizens’ inclination to engage in a “civil” communication style (N = 106 of 1088 [10%], N = 82 of the 850 random sample [10%]). By this we mean citizens’ display of their appreciation for bureaucrats’ efforts and help, and/or beyond minimum polite gestures. Drawing on Nielsen et al. (2021), the messages of these applicants are low on opposition, that is they are disinclined to challenge bureaucrats’ authority, whilst seeking to befriend them. Moreover, whereas demanding and pleading styles convey negative emotions, a civil style involves citizens’ display of warmth and friendliness. We operationalize this conceptualization as involving: (a) citizens who pose either questions or polite requests that do not signal criticism; plus: (b) at least one of the following: (i) display of beyond minimum polite gestures and greetings (that is, such that exceed one greeting at the beginning and one at the end of the message), (ii) acknowledgment of bureaucrats’ efforts and hard work. Table 3 provides illustrative examples, showing that although citizens who espouse a civil communication style allude, as others, to their experience of administrative burdens, they are inclined to counterbalance such allegations with warm politeness and recognition of bureaucrats’ workload and hard work.

Neutral Communication Style

Last, we created a residual category, which is also the largest, of citizens’ “neutral” communication style (N = 505 of 1,088 [46%], N = 424 of 850 [50%]) involving contacts that do not display a demanding, pleading, or civil styles. Whilst more diverse, these contacts typically involve non-assertive queries or requests, alongside minimal if any courtesy—a greeting at the front end and a standard closing sentence. Neutral claims, as exemplified in table 4, are pragmatic, that is, they convey the dry facts and requested action. Additionally, whereas demanding and pleading styles both involve a display of negative emotions, and a civil style is typified by a positive expression of warmth and friendliness, a neutral style is unemotional.

Clustering of Communication Styles, Justifications, and Administrative Literacy

Communication styles, as we conceptualize them, relate to citizens’ demeanors, including forceful justifications of legal entitlement or need as integral components of demanding and pleading, respectively. Still, our separate coding of citizens’ seventeen subcategories of justifications, as elaborated in section A of the Online Supplementary Appendix, is much broader than the type of powerful claims that we code as evidence of demanding and pleading communication styles.

To further unravel the patterns in citizens’ communications, table 5 summarizes the association between citizens’ communication styles and the two most prevalent types of justifications in our data, relating to legal entitlement versus need, as well as their combined usage in the same communication. The analysis relates to all contacts for which communication styles were discernable (N = 1,072). Restricting our analysis to the randomly sampled contacts yields similar patterns. The highest clustering between citizens’ communication styles and their justifications pertains to those who plead, who alluded to their needs, whether alone or alongside reference to their legal entitlement, in 71% of the contacts. The clustering between citizens’ demanding style and legal entitlement, alone or alongside allegations of need, is also substantial, involving 38% of these contacts. Finally, and partially by construction, citizens’ neutral and civil communication styles both display relatively low clustering with their allusion to either legal entitlement and/or need. These findings may suggest that citizens who plead or make assertive demands assume that stressing their neediness, or their legal entitlement are the most effective strategies. Alternatively, the clustering that we observe may indicate that citizens who actually experience high need, and those who firmly believe that are legally entitled, are more anxious or infuriated by the burdens that stand in their way.

Next, we examine the clustering between citizens’ communication styles, and their display of administrative literacy (Döring 2021; Döring and Madsen 2022), which Masood and Nisar (2021) depicted in their study of Pakistani doctors. The following two examples, with our emphasis in talics, illustrate such displays of administrative literacy, both coded as neutral (the second is admittedly assertive yet lacks clear characteristics of demanding):

“I received just 258 NIS for 9 days and this is the second time I am contacting you as I would be glad to know how you reached the calculation of such a low sum. Using your simulated calculator [on the agency’s Web site] I should be getting 111 per day so I would be glad if you please explain why I got such a small amount. Many thanks. “

“Why did I receive 5,248 gross, 4,974 net, when the average of my wages, according to what is documented on your records, is 10,927 gross, this comes out as less than 50 percent of the gross [income] whereas as far as I know, I should be getting unemployment benefits of 70 percent of my average gross income?”

As is apparent from table 6, displays of administrative literacy, the coding of which is detailed in Online Supplementary Appendix A, were low across the sample (12%), yet higher among those who espoused neutral (17%) and civil (13%) styles compared to those who displayed demanding or pleading styles (7%). This tentatively indicates that applicants who espouse a neutral style are relatively inclined to strategically seek to influence bureaucrats’ use of discretion by signaling that they are knowledgeable of their rights.

Summarizing the Taxonomy

Figure 1 summarizes the taxonomy of citizens’ communication styles in written bureaucratic encounters, which emerged from our data coding and its drawing on multiple theories. Our taxonomy maps the four styles in terms of their display of oppositional versus submissive demeanor, of positive versus negative emotions, and their clustering with justifications and/or administrative literacy. The size of the circles reflects the relative frequency of each style in our data. It illustrates that citizens’ demanding style (table 1), manifests their oppositional demeanor, which quite often clusters with their allusion to legal entitlement (table 5), and conveys negative emotions (a sense of anger and frustration). A pleading style reflects citizens’ submissive appeals for bureaucrats’ compassion and goodwill (table 2), which most often clusters with their allusion to neediness (table 5) and conveys their negative emotions (a sense of anxiety and stress). A civil style is also submissive, yet this demeanor is reflected not in allusion to one’s neediness and call for help, but via expression of warmth and friendliness towards bureaucrats (table 3). Finally, a neutral style is a residual category table 4), typified by unemotive, pragmatic messages, which correlates with citizens’ tendency to showcase their administrative competence (table 6).

A TENTATIVE EXPLORATION OF THE CAUSAL MECHANISMS UNDERLYING CITIZENS’ COMMUNICATION STYLES

The main goal of this article, as accomplished in the previous sections, is to abductively conceptualize and operationalize citizens’ communication styles. However, our theoretical contribution would not be complete without a reflection on the causal mechanisms underlying citizens’ responses. We thus further tentatively examine our assumption that the variation in citizens’ communication styles reflects their coping behaviors, involving an attempt to influence bureaucratic discretion and/or uncontrolled discharge of negative emotions in reaction to administrative burdens.

Logically, if citizens’ communication styles are purely strategic, then the intensity and cause of burdens would not systematically affect their variation. Rather, citizens’ strategic choice of response is presumably predicated on their beliefs as to which communication styles are most effective. For example, those who believe that civility is the most effective strategy would embrace it, suppressing negative emotions, even when experiencing high burdens. Conversely, if communication styles reflect citizens’ uncontrolled discharge of negative emotions, then experiences of heavy burdens, especially such that are attributable to the agency’s actions, are likely to induce writing in an emotive manner, involving pleading or demanding styles. Below we present two exploratory examinations of the association between the type and intensity of experienced administrative burdens and citizen communication styles. The conclusion draws on this exploratory analysis of the mechanisms underlying the variation in citizens’ communication styles to delineate concrete avenues for a future research agenda.

Comparison by the type of reported burdens

To examine whether burden type and intensity are associated with a variation in citizens’ communication styles we compare, based on our elaborate coding of the type of administrative burdens that citizens report (Online Supplementary Appendix A), two broad types of contacts: (a) information inquiries about entitlement to benefits, the calculation of benefits, or the NII’s procedures, and (b) allegations of delays in the agency’s processing of applications, of its failure to respond to prior contacts, and/or of its inaccessibility. We treat the latter concerns under one category since they are often bundled together in citizens’ contacts. As illustrated below, we find that information inquiries tend to be either neutral or civil. This may be because citizens who make information inquiries experience relatively low burden intensity, and/or because ciitizens do not blame the agency for their own deficient knowledge. The following contacts, the first coded as neutral and the second as civil, are typical cases:

“Greetings, I would like to inquire: 1. Until when am I entitled to unemployment benefits? 2. What is the [monthly] amount that I am being paid, and according to what calculation? 3. Why is it that in July I am entitled to just 1,204 NIS? 4. Would I need to pay social insurance at the end of the [entitlement] period? Many thanks”

“Hello, I am contacting you in order to receive information regarding extension of [the period of] unemployment [benefits]. As you of course know there is severe difficulty in finding jobs these days. As far as I know the period of my unemployment [benefits] is about to end and I would therefore be glad to understand if there is an option for an extension given the situation. Thank you in advance and good health to all”

Conversely, we find that when citizens experience the agency’s delays, failures to respond, or inaccessibility, they tend to espouse a demanding communication style. What seems to underlie this tendency is a combination of material concerns, frustration due to waiting and failing to get hold of an agent, and indignation owing to being ignored. Moreover, citizens’ demanding style seems to be associated with their attribution of these burdens to the agency’s deliberate (in)actions. The following demanding contacts are indicative:

“My file is ready for a week and a half for [benefit] calculation and you didn’t bother carrying out the calculation before the holiday or paying an advance [payment] so that we can celebrate the holiday/ I make endless contacts and I am being ignored. I request answers and most importantly that my money be deposited”

“Greetings this is the fourth time that I am writing you...you are not getting back/ [this is] simply a disgrace/I got 1,200 NIS when my gross [income] is 15,000 from two jobs/ start paying attention/ I call dozens of times every day and there is no answer/I don’t have any income and I am entitled by law”

Additionally, compared to citizens who contact the agency with information inquiries, those reporting delayed processing of their cases, failure to respond, and inaccessibility are also more inclined to plead, such as:

“Greetings/ I contacted you many many times in all possible ways and there is no response/ I didn’t get unemployment [benefits] for April... I am... begging for your response/ I have reached a state of starvation and there is no response”

Still, exposure to high burdens does not universally translate into either demanding or pleading, as illustrated by the following neutral-style message:

“Hello about two weeks ago I sent a letter from my employer about the extension of the unpaid leave. I sent you a message after that asking for confirmation [of my notification]. I still have not received a response, please confirm receiving the letter and the extension of the unpaid leave since the payment for October is supposed to be made these days and I cannot manage another month [without it]. Thank you.”

Table 7 summarizes the different rates of the four communication styles among citizens who make information inquiries (N = 179) compared to those who allege delays, non-response, and inaccessibility (N = 185). As apparent, the emotive demanding and pleading styles are much more prevalent among the latter group of contacts.

Supporting survey data analysis

To further develop a preliminary theorization of the mechanisms underlying the variation of citizens’ communication styles we sought to gauge their association with citizens’ attitudes toward bureaucracy and the administrative process. To do so, we briefly present findings from a survey we carried out among 2,160 online Israeli panel respondents (see Online Supplementary Appendix B for a detailed explanation).

The survey examined the association between respondents’ experiences of burdens as participants in our survey, which we treat as a non-ideal proxy for citizens’ experiences of administrative burdens, their trust in Israeli public authorities, and their communication styles. To examine the latter, respondents were asked to write “a persuasive letter” on behalf of a fictitious male relative who was supposedly unduly denied income support welfare benefits by the NII due to the agency’s incorrect conclusion that he fails to meet their criteria.

Our inferences are based on the following logic. If communication styles are strategic, reflecting participants’ beliefs about what makes a persuasive letter to the bureaucracy, their own experiences of burdens as survey respondents would not affect their writing of a fictional letter on behalf of another person. Hence, if we nonetheless find a correlation between respondents’ experiences of burden as survey respondents and their communication styles, it is most likely attributable to their discharge of negative emotions. However, if we find that respondents’ trust in the Israeli public authorities is correlated with their communication styles, then we could tentatively deduce that they strategically seek to influence the NII bureaucrats given their differential beliefs about bureaucrats’ motivations. As evident below, we find indications for both explanations.

At the beginning of the survey, we estimated participants’ trust in Israeli public authorities, using four items, each measured on a 5-point scale, borrowed from Grimmelikhuijsen and Knies (2017) (alpha = 0.89; mean = 2.49, SD = 0.88). Thereafter, we randomly assigned the participants to two types of objective burdens via: (a) complex versus simple presentation of the Israeli legal criteria for income support, thus generating high/low learning costs (cf. Baekgaard et al. 2023), and (b) requiring them to answer three factual questions about the legal criteria for income support and to repeatedly answer each question up to three times in case of a mistake, versus allowing them to skip these questions and providing them with the correct answers, thus simulating high/low compliance costs.

Then, to measure participants’ communication styles, we asked them, as mentioned, to write a letter to the NII on behalf of the fictional relative. The letters were classified by two research assistants,³ using the above guidance, as demanding (22.8%), pleading (35.1%), or neutral (44.5%). Due to the rarity of civil letters (0.06%), we collapsed neutral and civil messages into one category. Ambiguous letters were coded under more than one communication style. We present coded examples of respondents’ fictional letters in Online Supplementary Appendix C.

Last, we measured respondents’ subjective experiences of burdens as participants in our survey. To measure participants’ subjective learning burdens, we posed three questions (measured on a 5-point Likert-item scale), such as the following “I understand the conditions for income support as these were presented to me in this survey” (alpha = 0.85; mean = 2.17; SD = 0.8). To estimate respondents’ subjective compliance burdens, we again used three 5-point Likert-item scales, including: “Participating in this survey imposed many demands on me” (alpha = 0.86; mean = 2.3, SD = 0.93). To gauge participants’ subjective psychological burdens, we used six 5-point Likert-item scales, such as: “Filling this survey is stressful” and “Taking this survey is enjoyable” (r) (alpha = 0.84.; mean = 2.3; SD = 0.74).

As reported in Online Supplementary Appendix B, participants’ subjection to objective burdens had, to our surprise, a null effect on their communication styles. Conversely, the effects of respondents’ subjective experiences of burdens as survey respondents and their trust in Israeli public authorities were both statistically significant. Table 8 presents linear probability models of the variation in respondents’ communication styles as a function of their subjective experiences of burdens, and trust in Israeli public authorities, controlling for demographics. The outcome variables, citizens' communication styles, are measured as binary variables (e.g. demanding versus other). Trust and subjectively experienced burdens are estimated as composite indices, standardized to vary between zero and one. The effects of respondents’ compliance and psychological burdens are examined in separate models due to their strong correlation (r = 0.62, p < .001). We find respondents’ trust in Israeli public authorities decreases their proclivity to write in a demanding style (p < .01) and increases their inclination to write in a neutral manner (p < .01). This tentatively indicates that a demanding communication style reflects some individuals’ beliefs that bureaucrats cannot be trusted to rectify their mistakes unless coerced to do so. Additionally, respondents’ experiences of psychological and compliance burdens as survey respondents increase their inclination to espouse a demanding communication style (p < .1, p < .05, respectively), and decrease their inclination to write in a neutral style (p < .01). We interpret these findings as indications that individuals’ communication styles partially reflect their uncontrolled reactions to administrative burdens. Finally, women are disinclined to write demandingly (p < .01), and age is correlated with making demands and foregoing pleading.⁴

CONCLUSION AND RESEARCH AGENDA

Contributing a relational perspective (Bartles, 2013) to studies of citizens’ coping behaviors, this article examines what and how citizens write in written bureaucratic encounters to influence bureaucrats’ discretion, or in reaction to administrative burdens. To this end, we employed systematic content analysis of Israeli citizens’ real-world web contacts to a welfare agency, and flexible borrowing from multiple theoretical perspectives, to advance a novel conceptualization and operationalization of citizens’ communication styles. Our taxonomy suggests that citizens’ requests for information and alleviation of burdens vary in their demanding, pleading, civil, or neutral manner. These different styles involve citizens’ disparate demeanors (i.e. their displayed attitude towards bureaucrats or bureaucracy), and expressions of negative or positive emotions or their absence. They further cluster with citizens’ allusion to distinct types of justifications and displays of administrative literacy. In terms of operationalization, we inductively captured the variation in citizens’ communication styles via attention to their messages’ contents and structures. We created a detailed coding scheme pertaining to the occurrence of distinctive types of phrases and words (e.g. “please help me” as an indication for pleading), the extent of deployment of polite gestures, and the variation in messages’ structures and punctuation (e.g. the use of imperatives and exclamation marks as characteristics of demanding).

We acknowledge that the generalizability of the communication styles, and their operationalization is tentative and a first step in what we hope is a larger research agenda. It is based on the written communications of citizens in one country (Israel), which is characterized by relatively low levels of trust in government, in one policy domain (unemployment), and during a major crisis (COVID-19). Thus, the next logical step is to examine citizens’ written communication styles across contexts, their antecedents, and consequences. Below we offer directions for future research, which partly build upon our tentative empirical explorations in this article.

First, the relational perspective to the study of public encounters (Bartles 2013) focuses on face-to-face and voice-to-voice interactions. We advance this agenda by conceptualizing and developing operational tools to capture citizens’ communication patterns in written and digital encounters (Breit et al. 2021; Lindgren et al. 2019; Peeters 2023; Reddick and Anthopoulos 2014). Still, a significant limitation of our contribution, which future relational studies might address, is that the data on which it is based is one-sided, missing bureaucrats’ responses and their consequences for citizens’ subsequent reactions. Future studies may draw on our taxonomy alongside that of recent studies that categorize the “informational” and “relational” dimensions of administrators’ talk (Eckhard and Friedrich 2024; Eckhard et al. 2024). Uniting these perspectives, such studies may examine whether citizens’ written communication styles are affected by and influence administrators’ inclination to avoid jargon, provide an account for their actions, and signal empathy and benevolence.

Second, our article suggests the variation in citizens’ communication styles may logically reflect two types of coping behaviors, which are not mutually exclusive: calculated attempts to influence bureaucrats’ discretion, and uncontrollable discharge of negative emotions in reaction to administrative burdens (cf. Folkman et al. 1986; Lazarus and Folkman 1984). We provided preliminary empirical support for this duality. Future research may explicitly examine to what extent are citizens’ communication styles shaped by their experiences of distinct emotions, such as anger, frustration, or anxiety (Bell et al. 2022), or by their beliefs regarding the likely efficacy of alternative justifications and demeanors (e.g. their assumption that bureaucrats are responsive to expressions of need and helplessness).

Third, our exploratory analyses tentatively indicate that citizens’ demanding and pleading styles are partially attributable to their subjective experiences of higher levels of administrative burdens (cf. Hattke et al. 2021). As shown, demanding and pleading communication styles are more prevalent among citizens who allege the agency’s delays and non-response compared to those making information inquiries; and survey respondents, for whom filling the survey was subjectively burdensome, were inclined to espouse a demanding style when writing a fictitious letter on behalf of a third person. Future research may further examine these tentative findings, and whether and how the objective versus subjectively experienced intensity of administrative burdens, and their attribution to the self or to the bureaucracy, shape citizens’ communication styles. Moreover, since subjective experiences of administrative burdens vary with citizens’ human capital and competence (Christensen et al. 2020; Döring and Madsen 2022), further research may examine whether vulnerable citizens are more aggressive (i.e. demanding), or alternatively, more inclined to plead. Citizens’ innate personality traits (e.g. their agreeableness versus neuroticism) may also moderate their communicative response to similar intensity of burdens.

Fourth, another question is whether citizens’ communication styles vary with individuals’ and societies’ trust in institutions and bureaucracy. Presumably, if citizens believe that bureaucrats are trustworthy, that is competent, impartial, and benevolent (Grimmelikhuijsen and Knies 2017), then they are likely to assume that their justified requests would be met with a positive response, prompting a neutral communication style. Conversely, if citizens assume that bureaucratic services and performance are prejudiced, differently provided based on pressure or favoritism, as opposed to general rules, then they may be inclined to espouse a belligerent demanding style, or to alternatively plead on the basis of their personal need, or possibly to befriend bureaucrats via a display of civil warmth, to secure higher priority for themselves over others. Our survey data analysis tentatively confirms that individuals’ higher trust in public authorities is in fact positively correlated with their inclination to espouse a neutral style and to forego a demanding style. Future comparative research may examine these preliminary findings across countries, gauging to what extent our findings from Israel’s low-trust context travel to other low-trusting versus high-trusting contexts.

Relatedly, another research direction would be to theorize and empirically examine how citizen communication styles vary with the quality of their prior interactions with bureaucracy (cf. Soss 1999a, 1999b, 2005). Tentatively, citizens who have had negative past experiences are likely to interpret new bureaucratic encounters more negatively, and to thereby act demandingly or, alternatively, to subject themselves to bureaucratic authority through pleading or civility in fear of bureaucratic retaliation. Conversely, those who have learned to a priori trust bureaucracy may be more inclined to espouse a neutral approach.

Fifth, future research may also examine the variation of communication styles across social groups. It is plausible that citizens’ communication styles are shaped by their internalization of their social status as inculcated throughout their lives—in the family, in educational institutions, through the consumption of mass media and popular culture, and the like (Assouline and Gilad 2022). Relatedly, individuals may internalize writing conventions, such that some greetings or formalities may reflect their taken-for-granted cultural habits. Compatibly, our survey data analysis, echoing Pfefferman et al. (2021), indicates that female respondents avoided a demanding style (although writing on behalf of a fictional male applicant). Hence, research may examine whether and how social conventions, and the social status and perceived deservingness that are attached to social groups (Schneider and Ingram 1993) shape individuals’ beliefs about which communication styles are more effective in written bureaucratic encounters, as well as their actual writing.

Last is the question of how bureaucrats interpret and respond to citizens’ communication styles. Researchers may examine the extent to which our conceptualization and coding cohere with bureaucrats’ readings of citizens’ messages, and how this affects their perceptions of clients’ deservingness for their assistance and support. A priori, it seems likely that citizens’ demanding style would tend to provoke bureaucratic rejection and retaliation (Christensen et al. 2020; Raaphorst and Van de Walle, 2017; but see Davidovitz and Cohen 2022). Pleading may elicit bureaucrats’ compassion, and thereby induce their help (cf., Halling and Petersen 2024), but it may also be interpreted as manipulative (Raaphorst and Van de Walle, 2017). Civility may encourage bureaucratic cooperation, yet bureaucrats may also feel that citizens’ friendliness and warmth are incompatible with their respective roles as state agents versus clients. Ultimately, these are empirical questions that call for context-sensitive examinations.

In conclusion, this article offers an original, theoretically, and empirically grounded conceptual framework and operational tools to study the variation in citizens’ coping behaviors in written bureaucratic encounters. We believe that analyses of the variation in citizens’ communication styles, their antecedents, and, importantly, their consequences can provide a better understanding of the dynamics and outcomes of such prevalent, yet little studied, citizen-state interactions. Doing so has not only theoretical but also practical implications. Understanding the extent to which the variation in citizens’ communication styles is rooted in citizens’ uncontrolled reactions to administrative burdens versus mistrust can guide public managers in their allocation of resources, and in their supervision and training of bureaucrats’ communications with citizens.

Supplementary material

Supplementary material is available at the Perspectives on Public Management and Governance online (www.ppmg.oxfordjournals.org).

Acknowledgements

We thank three anonymous reviewers and the editor for valuable suggestions and comments. We also thank Tomer Pree-Dan and Shoam Shenav for superb research assistance, and Michaela Assouline for her advice and help. This research builds upon Gilad's former collaboration with Talia Pfefferman and Michal Frenkel.

Funding

The Israel Science Foundation grant #643/21.

REFERENCES

Footnotes