Trust in authorities as a boundary condition to procedural fairness effects on tax compliance

Abstract

We explored the moderating role of trust in authorities in the positive effect of procedural fairness of the tax office on voluntary compliance with tax authorities. Building on fairness heuristic theory, we predicted that particularly low trust in authorities makes people carefully attend to the fairness with which the tax office enacts procedures. This should result in positive procedural fairness effects on endorsement of norms prescribing taxpaying and, consequently, in voluntary tax compliance, particularly among citizens with low trust in authorities. Results from an experiment and a field study revealed converging support for these predictions. We conclude that high trust in authorities forms an important boundary condition to the effectiveness of procedural fairness as a tool to enhance tax compliance.

Introduction

When authorities enact procedures in a fair manner, members of the social collective they represent are more likely to voluntarily comply with their decisions (see e.g., Cropanzano et al., 2001, De Cremer and Tyler, 2005, Tyler, 2006, for overviews). The present research explores the role of procedural fairness in enhancing voluntary compliance with decisions of tax authorities. Modern societies rely heavily on voluntary compliance with tax laws and regulations. The study of tax compliance has thus burgeoned over the last fifty years and contributions from scientific disciplines as diverse as psychology, sociology, and economics have resulted in significant progress in this area (see Kirchler, 2007, for an overview).

In focusing on procedural fairness, the present research fits into a stream of work that increasingly recognizes the role of social variables in tax compliance (e.g., Alm et al., 1992, Murphy, 2004, Wenzel, 2002, Wenzel, 2004a, Wenzel, 2004b, Wenzel, 2005). This recognition results from a growing awareness that traditional deterrence based analyses alone cannot account for the observed high levels of tax compliance (e.g., Bordignon, 1993, Braithwaite, 2003; see Andreoni, Erard, & Feinstein, 1998, for an overview of deterrence based tax compliance research). In fact, although deterrence based approaches have dominated economic analyses of tax compliance for decades (Allingham and Sandmo, 1972, Andreoni et al., 1998), studies generally reveal only small positive deterrence effects on tax compliance and sometimes even negative effects (Frey, 1999, Kirchler, 2007). Recent theoretical developments in the field of tax compliance thus consider social variables equally worthy of study as deterrence variables (see e.g., the “Slippery slope framework”, Kirchler, Hoelzl, & Wahl, 2008).

We are not the first to study the role of procedural fairness in tax compliance. However, prior work has not always revealed consistent results, sometimes reporting positive procedural fairness effects (Alm et al., 1993, Murphy, 2004, Murphy and Tyler, 2008, Wenzel, 2002), and sometimes failing to reveal such effects (e.g., Porcano, 1988; Worsham, 1996). The principal aim of the present research is to clarify when (and thus why) procedural fairness positively influences voluntary tax compliance. We build on well developed procedural fairness theory and identify trust in authorities as an important boundary condition to the effectiveness of procedural fairness. Trust in authorities is arguably highly relevant to consider in this context: It may be the core reason why people care about procedural fairness (Colquitt, Greenberg, & Scott, 2005) and it has a central place in recent theorizing about voluntary tax compliance (Kirchler et al., 2008).

Procedural fairness refers to the perceived fairness of procedures used to make allocation decisions (Tyler, 1988). Procedures are, for instance, perceived as more fair when followers are allowed to voice their opinion in authorities’ decisions (e.g., Brockner et al., 2001, Thibaut and Walker, 1975) and when authorities take decisions accurately and without regard for self-interest (de Cremer, 2004; see Leventhal, 1980, for an overview of procedural fairness characteristics). Much evidence for positive effects of specific characteristics of decision making procedures on procedural fairness perceptions has been obtained in organizational contexts (see Cohen-Charash and Spector, 2001, Colquitt et al., 2001, for meta-analyses). In addition, some studies reveal that when tax authorities follow procedural fairness rules, such as refraining from self-interest and giving voice in the decision process citizens judge such procedures as more fair (Magner et al., 2000, Stalans and Lind, 1997).

As noted, a host of studies have revealed that fair procedures stimulate followers to voluntarily comply with decisions made by the enacting authority. This effect has been explained in reference to the idea that people expect fair procedures to guarantee fair outcomes in the long term, increasing their willingness to invest in the social collective (Shapiro and Brett, 2005, Thibaut and Walker, 1975). In addition, there is also abundant evidence that being treated fairly by the authority representing the social collective communicates that one is a valued and respected member of the collective. This stimulates internalization of the collective’s norms and, consequently, voluntary compliance with the authority’s decisions (e.g., Tyler, Degoey, & Smith, 1996; see Wenzel, 2002, for evidence in the context of tax compliance).

Trust plays an important role in understanding why procedural fairness stimulates voluntary compliance with authorities representing social collectives. This is best explained in reference to what has been called the fundamental social dilemma (Lind, 2001; see also Kramer, 1996): members of groups, organizations, and societies face a dilemma when deciding whether to invest in social collectives. This is because such memberships present opportunities for improved outcomes and a sense of identity and belongingness. At the same time however, such memberships also include possible exploitation and identity damage from authorities who abuse their power.¹

The relevance of this idea for procedural fairness is further developed in fairness heuristic theory (Lind, 2001). This theory proposes that people are often unsure whether authorities can be trusted not to abuse their power. In such situations, people use their judgments of procedural fairness as a heuristic guide to decide whether enacting authorities will abuse their power, and, consequently, to decide about appropriate levels of personal investment in social collectives (Lind, 2001, Lind et al., 2001). In support of this idea, research shows that voice influences procedural fairness perceptions strongest when it is unclear whether an authority is trustworthy. When it an authority is either clearly trustworthy or untrustworthy, voice is less effective (van den Bos, Wilke, & Lind, 1998; see also van den Bos, van Schie, & Colenberg, 2002). Further, procedural fairness has also been shown to increase trust in the enacting authority (Konovsky and Cropanzano, 1991, Korsgaard et al., 1995). And trust explains (mediates) procedural fairness effects on important outcome variables, such as the authority’s legitimacy (Tyler, 1989) and variables that reflect involvement in social collectives, such as organization citizenship behavior (Konovsky & Pugh, 1994), turnover intentions, and organizational commitment (Aryee, Budhwar, & Chen, 2002). In the field of tax compliance, Murphy (2004) showed that trust in the tax office also mediates procedural fairness effects on tax compliance.

Recent theoretical accounts of the procedural fairness – trust relationship have recognized that procedural fairness can be an antecedent of trust, but that trust can also moderate fairness effects (Lewicki, Wiethoff, & Tomlinson, 2005). This is possible because trust in authorities is influenced by a multitude of factors in addition to procedural fairness information (see e.g., Dirks & Ferrin, 2002). In fact, people show some trust for others with whom they have no history of interaction (Meyerson, Weiss, & Kramer, 1996) and thus have no idea about whether these others will enact procedures in a fair manner (see Konovsky & Pugh, 1994). In the only study that we know of that has explicitly considered trust as a moderator of procedural fairness effects, De Cremer and Tyler (2007) showed that high (rather than low) trust in the authority makes the effect of voice on cooperation with the authority stronger. This is presumably because high trust makes opportunities to voice one’s opinion meaningful whereas low trust makes people fear that the authority will simply dismiss their input.

The present study addresses trust in authorities as a moderator of the effect of procedural fairness on compliance with tax authorities. However, our predictions diverge sharply from the work by De Cremer and Tyler (2007). These authors focused specifically on the procedural fairness rule of voice in the decisions of the enacting authority. Interestingly, voice is the only procedural fairness rule that specifically refers to active input from followers (Brockner et al., 2001, van Dijke & De Cremer, in press), making it understandable that low trust makes followers believe that the authority will dismiss their input. However, other fairness rules that have been empirically validated (e.g., accuracy, bias suppression, and consistency) specifically refer to leader behavior reflecting that the leader cares about the social collective (Brockner et al., 2001, van Dijke & De Cremer, in press). Focusing on these rules thus also suggests a different moderating role for trust in authorities.

More specifically, low trust in interaction partners (such as authorities) indicates uncertainty concerning these others’ intentions and behavior (Mayer et al., 1995, Rousseau et al., 1998). Hence, particularly members of social collectives with low trust in authorities can be expected to fear exploitation and power abuse in their interaction with authorities representing the social collective. Building on fairness heuristic theory, we therefore argue that particularly citizens with low trust in authorities will observe closely whether the tax office acts in a procedurally fair manner (e.g., unbiased), in order to assess whether the tax office will abuse their power or, conversely, cares about the interest of the social collective. This should make especially low trust citizens susceptible to information regarding how fairly the tax office enacts decision making procedures in their decision whether to voluntarily comply or not. Conversely, people with high trust in authorities, who are less likely to fear exploitation and power abuse from authorities, should be less susceptible to the fairness with which the tax office enacts decision making procedures. In sum, we expect relatively strong procedural fairness effects on voluntary tax compliance among citizens with low trust in authorities (Hypothesis 1).

Importantly, the present research focuses on voluntary compliance with the tax authority’s decisions because citizens personally endorse norms prescribing that one should pay one’s taxes. We therefore also study endorsement of norms regarding tax paying. More specifically, because citizens with low trust in authorities should be particularly susceptible to procedural fairness, we expect procedural fairness of the tax office to stimulate endorsement of norms prescribing that one should pay taxes particularly among citizens with low trust in authorities. Moreover, because endorsement of norms prescribing that one should pay taxes is known to be a strong predictor of actual compliance with tax authorities (Wenzel, 2004a), we expect that the positive effect of procedural fairness on compliance with tax authorities (among citizens with low trust in authorities) is mediated by endorsement of norms that prescribe that one should pay taxes (Hypothesis 2).

Our hypotheses were tested in two studies. Study 1 is an experimental study, in which we, as an initial test of our ideas, manipulate the procedural fairness of the tax office and trust in authorities orthogonally. For trust, we focused on general trust in authorities using a priming procedure. A criticism of experimental research is that it sacrifices external validity to obtain results high in internal validity. Therefore, Study 2 was a field study in which we measured our variables of interest. In this study, we again focused on general trust in authorities, using an established measure. Combining field research with experimental research allows one to benefit from the strengths of each method and to compensate for the weaknesses of each method with the strength of the other method (Dipboye, 1990).

In order to further increase our confidence that procedural fairness stimulates voluntary compliance with the decisions of the tax office, rather than compliance out of fear of punishment upon non-compliance, we also explicitly controlled for the role of deterrence variables in our field study. This is in line with recent theoretical developments (Kirchler et al., 2008), which have stressed that deterrence and social variables both influence compliance with tax authorities and should thus be studied simultaneously. We build on the often used operationalization of deterrence as the severity of sanctions upon getting caught cheating, multiplied by the detection probability (e.g., Wenzel, 2002).