The fair process effect in taxation: the roles of procedural fairness, outcome favorability and outcome fairness in the acceptance of tax authority decisions

This article presents two studies examining the impact of procedural fairness, outcome favorability and outcome fairness on the acceptance of tax authority decisions. Twofactor experimental studies were conducted using a sample of 198 people: 2 x procedure (fair vs. unfair) and 2 x outcome favorability or outcome fairness. Study 1 found no procedural fairness effect, level of tax decision acceptance being dependent only on outcome favorability. Respondents accepted personally favorable decisions more even when they resulted from an unfair procedure. The fair process effect was only related to outcome fairness, higher acceptance of tax decisions resulting from fair procedures regardless of their consequences. Results suggest outcome favorability has a stronger influence than outcome fairness.


Introduction
Justice is one of the most crucial features underlying tax systems and laws. The issue of how to fairly distribute tax burdens between citizens is discussed widely by economists, philosophers and politicians. The issue is also prominent in social discourses surrounding the state, citizens, and their duties. As Frey (2002, 2007) emphasize, taxpayers are motivated not only by a concern to maximize their own well-being, but also by concerns for the state of their country. A citizen's sense of duty underlies their voluntary cooperation with state authorities and compliance with tax laws. Trust is the basis of civic commitment (Kirchler 2007). If people believe that an authority is respectful they trust its motives and work toward a common goal. Fairness in taxation, especially procedural fairness, helps to build and maintain trust. It also has a significant influence on taxpayer morale (Niesiobędzka 2014).
Where fairness is perceived, two types of motivational posture towards taxation might develop: commitment posture and capitulation posture (Braithwaite 2003). The former is based on a sense of moral obligation and the perception of taxpaying as an act of goodwill, the latter is based on the perception of tax authorities as representing a legitimate authority. Procedural fairness provides authorities with psychological legitimacy (Tyler 1997).
The present studies focused on the fair process effect and aimed to increase understanding of relationships between procedural fairness, outcome fairness and outcome favorability in the context of acceptance of tax authority decisions. The fair process effect is a commonly replicated finding and concerns the issue of how perceived procedural fairness affects people's reactions to decisions. Thibaut and Walker (1978) demonstrated that people are more willing to accept a judge's decision if the trial procedure is fair, even if the outcome is unfavorable. The acceptance of negative outcomes when a procedure is perceived as fair is known as the fair process effect. The current studies examined the fair process effect in the context of taxation. More specifically, we attempted to demonstrate the effects of perceived procedural fairness on taxpayers' acceptance of tax authority decisions.
The studies also tried to clarify the issues of whether the constructs of outcome favorability and outcome fairness can be used interchangeably, as is done in much research, and whether the fair process effect is equally related to outcome fairness and outcome favorability in the context of taxation.

Tax Fairness and Tax Compliance
Distributive tax fairness refers to the distribution of tax burdens across citizens and comparison of one's own tax burden with that of similar others, as well as to the relationship between the amount of tax paid and the amount of money received back in benefits from the state. Horizontal tax fairness concerns taxpayers' evaluation of their tax burden compared to the amount of tax paid by taxpayers in the same revenue bracket. Vertical fairness refers to comparisons between one's own tax burdens and the amount of tax paid by taxpayers in different revenue brackets. Finally, exchange tax fairness refers to the extent of government services received by a taxpayer relative to the tax they have paid: taxpayers compare their personal contributions to the state's finances (paid taxes) with the benefits they receive from the state (e.g., access to various public goods and services; Wenzel 2003). Studies have demonstrated a significant impact of subjective distributive tax fairness on tax compliance. The greater the horizontal and vertical tax fairness, the higher the satisfaction with the balance between tax burdens and the public goods provided by the state, and the lesser willingness to evade tax (Alm et al. 1992;Bosco and Mittone 1997;Braithwaite 2003;Cowell 1992;Falkinger 1995;Levi and Saks 2009;Moser et al. 1995;Niesiobędzka 2014;Pocarno 1988;Pommerehne et al. 1994;Spicer and Becker 1980;Verboon and Van Dijke 2011).
Procedural tax fairness refers to the decision making process (the degree of taxpayer participation and the perceived possibility of controlling the tax decision process) and to interpersonal relations between taxpayers and tax authorities. Furthermore, it also includes assessment of the quality of information provided by legislators and tax authorities. Research shows a significant influence of procedural tax fairness on taxpayers' decisions, with perceived fairness of the formal procedures involved in decision-making, perceived quality of treatment by tax authorities, and tax authorities' informational practices all encouraging tax compliance. The more decisions are free from bias, stable, neutral, respectful, sympathetic and honest, and the greater the clarity and comprehensibility of the information about tax regulations, the less the propensity for tax evasion (Alm et al. 1993(Alm et al. , 1999Alm and Torgler 2006;Cuccia and Carnes 2001;Maroney et al. 1998;Murphy 2003Murphy , 2004Niesiobędzka 2014;Pommerehne et al. 1994;Wartick 1994).
The term procedural fairness has both instrumental and relational meanings. Fair practice guarantees fair outcomes in the long-term and sends the symbolic message that one is valued and respected as a member of the collective (Lind and Tyler 1988;Tyler 1997;Tyler and Lind 1992). Bies and Moag (1986) emphasized the importance of interpersonal justice as a construct of respectful, unbiased, confident and friendly interactions with authorities. Bies (2005) considered relational fairness as primary because it connects with self-image, violations of which are particularly painful to people. The manner in which taxpayers receive notification explicitly reveals the character of mutual relationships between tax authorities and taxpayers. A formal announcement with strict sanctions classifies the individual as a suspect and casts them automatically into the role of a fraudster. The sense of being under constant control with the implied lack of trust negatively influences intrinsic motivation to pay taxes. Also, reminder letters based upon principles of procedural justice yield greater tax compliance than standard letters emphasizing penalties (Wenzel 2003), andTorgler (2004) demonstrated that procedural fairness determines the effectiveness of moral persuasion aimed at encouraging tax compliance. Thus, taxpayers are sensitive to moral persuasion only when tax authorities treat them with respect, dignity and in an unbiased manner.

The Fair Process Effect, Outcome Favorability and Outcome Fairness
The important role of procedural fairness was shown by Thibaut and Walker's (1978) study using simulated court decisions, which demonstrated that, irrespective of the outcome, participants were more willing to accept a judge's decision if the trial procedure was fair. The acceptance of negative outcomes when a procedure is perceived as fair is known as the fair process effect, and has been replicated in legal, social, political and organizational settings (Brockner et al. 2003;Folger et al. 1979Folger et al. , 1983Greenberg 1987Greenberg , 1993Lind et al. 1990;Tyler 1994;Tyler and Degoey 1996;Van den Bos et al. 1997. Features such as respect, dignity, trust, neutrality, accuracy of information and participation in decision-making, which are all embodied in the notion that procedures are fair and which are responsible for the acceptance of disadvantageous decisions, contradict the assumption that people are fundamentally selfish and concerned with maximizing outcomes in their self-interest. Lind and Tyler (1988) claimed that overall fairness judgments depend more strongly on procedures than outcomes. Van den Bos (2015) argued that procedural fairness has informational value for people, leading to many positives reactions, such as decision acceptance, conflict resolution, organizational commitment. Folger and Cropnzano (1998) assume that unfair procedures lead people to believe that outcomes would be more favorable if procedures were fairer. Unfair procedures also diminish internalization of responsibility and lead others to be blamed for unfavorable outcomes. Conversely, fair procedures make such beliefs difficult to maintain and enhance internalization of responsibility. As Brockner (2002) demonstrated, the tendency to make such internal attributions depends more on structural aspects of procedures than interactional aspects.
Inputs to the decision-making process (the opportunity to voice one's position and accuracy of information) affect attribution of responsibility more than being treated with dignity and respect (Brockner 2002).
The fair process effect has been examined by manipulating both outcome favorability and outcome fairness. Most research has used the first approach, participants receiving positive or negative outcomes without any relevant social comparison information. Outcome favorability concerns whether or not outcomes are personally beneficial. Outcome fairness refers to the rules underlying allocation processes and information about outcomes for others in a social exchange (Skitka et al. 2003). The equity theory of distributive justice proposes that fairness judgments rely on comparisons of a person's outcome/input ratio with those of others (Adams 1965;Walster et al. 1978). Brockner and Wiesenfeld (1996) claim that the constructs of outcome favorability and outcome fairness are different but closely related, and thus may be used interchangeably as is done in much research. Nevertheless, meta-analysis of studies examining relationships between distributive fairness, outcome favorability and procedural fairness supports the distinction between outcome fairness and outcome favorability, and has demonstrated that procedural fairness has a greater effect on perceptions of outcome favorability than on perceptions of outcome fairness. Skitka et al. (2003) concluded that the fair process effect is weaker when a manipulation concerns outcome fairness rather than outcome favorability. However, it should be noted that they only studied the fair process effect in terms of whether a higher level of procedural fairness led to increases in perceived outcome fairness or outcome favorability. The issue of how procedural fairness interacts with outcome fairness and outcome favorability to affect decision acceptance was not examined.

The Present Studies
The purpose of the studies was to investigate the effect of perceived procedural fairness on taxpayer decisions. The studies examined whether the fair process effect is present in the context of taxation. It was expected that unfavorable tax authority decisions would be more acceptable if they were the result of fair, rather than unfair, procedures. In addition, we also attempted to establish whether the fair process effect is present to the same extent under conditions of outcome fairness and outcome favorability.
We conducted two experiments, both manipulated procedural fairness, but one manipulated outcome favorability (Study 1) and the other manipulated outcome fairness (Study 2). Contrary to most studies, which mainly use voice/nonvoice procedures for fairness manipulation, we included interpersonal tax fairness and informational tax fairness in the studies. In Study 1 participants read a scenario about a tax office visit with a description of the tax authority's behavior (kind, respectful, a lack of unpleasant comments, care about the taxpayer's position, comprehensive information about the possibility of appealing against the decision and cancellation procedures, vs. unkind, disrespectful, suspicious, with unpleasant comments, and no information about cancellation procedures). The Study 2 scenario concerned governmental implementation of planned changes in tax law, to be introduced as a result of social consultation either via different channels of communication or without consultation with citizens.
In Study 1 outcome favorability was manipulated by varying information about the results of a visit to a tax office, which were either favorable or unfavorable to a taxpayer (tax exemption vs. the necessity to pay the full amount of tax demanded). In Study 2 outcome fairness was manipulated by varying information about the results of planned changes in tax law. These changes were presented as beneficial for the taxpayer individually but unprofitable for their tax group overall or unprofitable for the taxpayer individually but beneficial for the tax group to which they belonged.
The dependent variables were degree of acceptance of the decision made by the tax authority (Study 1), and acceptance of the tax changes proposed by the government (Study 2).

Study 1 -Materials and Method
The purpose of Study 1 was to determine the impact of outcome favorability on the acceptance of a tax authority's decision under conditions of fair and unfair procedures. Based on previous studies of the impact of procedural fairness on decision evaluations we assumed that unfavorable tax authority decisions would be more accepted if they resulted from fair rather than unfair procedures. A 2 (procedure: fair vs. unfair) × 2 (result: favorable vs. unfavorable) independent groups experimental design was used.

Design and Participants
The study was conducted using a group of 97 participants (30 of them males) in the age range 19 to 49 years (M = 26.93, SD = 7.43). Respondents were assigned randomly to one of four experimental groups. Three groups consisted of 24 and fourth groupof 25 participants. Data showed that 57% of respondents completed their yearly income tax form themselves, 25% had the form completed by their employer and 18% did not pay income tax.

Materials and Procedure
Participants read a tax scenario and were asked to put themselves in the shoes of the taxpayer who experienced the situation. In the scenario, respondents were informed that the tax authorities had discovered an undeclared donation from a member of their family in the amount of 11,000 PLN (approximately 3000$). Initially, respondents were given brief information about tax legislation regarding tax exemptions in respect of donations and the maximum permissible amount of non-taxable donations from family members. Then respondents read a letter from the tax office informing them about detection of the undeclared donation during inspection of their tax return. According to Polish tax law, in cases of donations by family members, taxpayers should complete and send a tax form to their tax office within 14 days. In the event of failure to do this the tax office may impose a penalty of up to 33,600 PLN (approximately 9200$).
Procedural Fairness The second part of the scenario described a fair or unfair procedure related to the behavior of an official during a visit to the tax office. In the fair procedure condition respondents read that an official explained the situation in detail, outlined tax law in an understandable way, was friendly and treated the person with respect, demonstrated an interest in the taxpayer and gave full information about the appeal procedure. In the unfair procedure condition the official was described as disrespectful, rude, hostile and suspicious, and as not mentioning the possibility of appealing against the decision made.
Outcome Favorability In the third part of the scenario, respondents were acquainted with the tax final authority's decision. The favorable decision amounted to tax exemption while the unfavorable decision resulted in the necessity of paying the tax due, in the amount 344.40 PLN (approximately 94$), within 14 days.
Decision Acceptance In the last part of the scenario, respondents were asked to indicate the extent to which they would accept the tax authority's final decision on a 5-point Likert scale from 1 (BI accept it completely^) to 5 (BI do not accept it at all^). In statistical analyses these data were recoded.
Procedural Fairness Evaluation Participants also indicated the extent to which they thought the taxpayer was treated fairly by the tax official on a 5-point Likert scale from 1 (Bcompletely unfair^) to 5 (Bcompletely fair^).

Manipulation Check
A manipulation check confirmed the effectiveness of the procedural fairness manipulation, t(95) = 16.65, p < .001. Participants who read the scenario describing the tax official as respectful, trustful, informative and supportive considered their behavior as more fair (M = 3.82, SD = .62) than participants who read the scenario in which the official was presented as being disrespectful, hostile and suspicious (M = 1.69, SD = .66).

The Effect of Procedural Fairness
Analysis of variance was used to determine the effects of procedural fairness and outcome favorability on decision acceptance. The results of the analysis only revealed a main effect of outcome favorability, F(1,93) = 63.73, p < .001, η p 2 = .41). Participants accepted the decision with favorable outcomes to a greater extent (M = 4.35, SD = .81) than the decision with unfavorable outcomes (M = 2.73, SD = 1.14). Neither the main effect of procedural fairness nor the interaction effect was significant (see Fig. 1). Thus, contrary to expectation, the fair process effect was not dependent on outcome favorability: decision acceptance depended only on outcome favorability.

Study 2 -Materials and Method
Study 2 investigated the effect of perceived procedural fairness in implementing governmental tax law changes (either positive or negative for the taxpayer) on evaluations of these changes in the context of outcome fairness. In Study 1, the fair process effect was not related to outcome favorability. We therefore wanted to examine the strength of any procedural fairness effect under conditions of different individual and group tax law change consequences (outcome fairness). Again, a 2 (procedure: fair vs. unfair) × 2 (result: beneficial for the taxpayer individually but unprofitable for the tax group to which they belonged vs. unprofitable for the taxpayer individually but beneficial for the tax group to which they belonged) independent group experimental design was used.

Design and Participants
The Study 2 sample consisted of 101 people (63 male) aged between 20 and 51 years (M = 27.89, SD = 7.28). As in Study 1, respondents were assigned randomly to one of four experimental groups. Two groups consisted of 26 participants, and two other groups consisted of 25 and 24 participants each. Fifty-nine percent of participants completed their yearly income tax form themselves, 38% had the form completed by their employer, and only 3% did not pay income tax.

Materials and Procedure
Respondents were presented with a scenario about substantial tax law changes planned by the government. These were described as being particularly relevant to the group of taxpayers to which participants belonged, and referred to changes in methods of calculating tax deductible expenses as well as a new directory of tax deductible items. These regulations were said to affect the amount of tax to be paid by the respondents.
Procedural Fairness The second part of the scenario manipulated procedural fairness. In the fair procedure condition it was stated that the changes would be preceded by public consultation aimed at understanding the opinions of taxpayers, which form the basis of future legislation. Furthermore, it was said that the government intended to launch a nationwide toll-free number, where operators would answer calls from citizens and call taxpayers to obtain their opinions. Finally, the government was said to be planning both a series of meetings in many towns and media information campaigns informing people of the possible changes in tax law. In the unfair procedure condition, respondents were informed that changes would be designed on the basis of the conclusions of tax inspections aimed at detecting existing irregularities in tax returns. It was said that the government did not plan public consultation with taxpayers during the preparatory period.
Outcome Fairness In the third part of the scenario, respondents were informed about the consequences of the planned changes that corresponded to the two variants of outcome fairness. In the individually beneficial condition, respondents read that the changes would lead to an increase in the total amount of tax paid in the group of taxpayers to which they belonged. However, due to the specifics of the respondent's employment situation, despite the increased tax burden for their group, these changes would result in a reduction of the amount of tax that they would pay personally. In the individually unprofitable condition, respondents were told that changes would result in a reduction of the overall tax paid by the group of taxpayers to which they belonged. However, due to the specifics of the respondent's employment situation, despite a reduction in the tax burden for their group, the changes would increase the amount of tax they would personally have to pay.
Decision Acceptance As in Study 1, respondents' task was to register their degree of acceptance of the changes on a 5-point Likert scale from BI accept them completely^to BI do not accept them at all^). Again, in order to achieve greater transparency of results, in the statistical analyses presented below, acceptance data were recoded.
Procedural Fairness Participants also evaluated the fairness of the method of implementing the tax law changes on a 5-point Likert scale from 1 (Bcompletely unfair^) to 5 (Bcompletely fair^).

Manipulation Check
Manipulation check results demonstrated that, regardless of the outcome for the taxpayer, the fair tax law change implementation method was considered fairer than the unfair implementation method, t(99) = 6.22, p < .001. The average fairness evaluation in the group of respondents with the unfair procedure was 2.10 (SD = .65), which was significantly lower than the average for the group given the fair procedure (M = 3.06, SD = .88).

The Effects of Procedural and Outcome Fairness
Analysis of variance was used to examine the effects of procedural fairness and outcome fairness on tax law change acceptance. For the tax law change acceptance dependent variable, results confirmed the significant impact of both fairness Fig. 1 Level of decision acceptance as a function of procedural fairness and outcome favorability in the four experimental groups of the tax law change implementation procedure, F (1101) = 31.74, p < .001, η p 2 = .247, and outcome fairness, F (1101) = 8.65, p = .004, η p 2 = .082. The interaction between the two independent variables was nonsignificant. Tax law changes introduced in a procedurally fair way were more accepted (M = 3.25, SD = .80) than changes introduced in an unfair way (M = 2.46, SD = .71). Also, there was greater acceptance of changes that were personally beneficial but not beneficial for the tax group to which participants belonged (M = 3.06, SD = .86) than of changes unprofitable for the taxpayer individually but beneficial to their tax group (M = 2.66, SD = .80). As can be seen in Fig. 2, there was higher acceptance of the tax law change in the group with individually beneficial changes and with the fair procedure (M = 3.60, SD = .71) than in the group with individually beneficial changes but with the unfair procedure (M = 2.54, SD = .65). Mean acceptance of changes unprofitable for the taxpayer but profitable for their own group was significantly higher in the fair procedure condition (M = 2.92, SD = .74) than in the unfair procedure condition (M = 2.37, SD = .77).
The above results showed a significant impact of both procedural fairness and outcome fairness on the evaluation of tax law changes. Changes involving fair implementation procedures were more acceptable regardless of their individual or group consequences. However, of the four scenarios presented, individually beneficial changes that were planned to be implemented in a procedurally fair way were the most acceptable (regardless of the adverse consequences for the group).

Discussion
The aim of the studies was twofold: to examine whether the procedural fairness effect exists in the area of taxation and to assess whether the effect is present to the same extent in conditions which vary outcome fairness and outcome favorability.
The first experiment concerned outcome favorability. Here, a taxpayer received a favorable or unfavorable tax decision issued using a fair or unfair procedure. The results suggested that we found no procedural fairness effect because the extent of decision acceptance depended only on outcome favorability. A decision favorable to the taxpayer (meaning no extra tax liability), was more acceptable even if this resulted from an unfair procedure. This suggests that acceptance of tax decisions is primarily influenced by the final result for the taxpayer, not by the fairness of tax authority procedures. Furthermore, the finding indicates that existing research results demonstrating the role of procedural fairness in willingness to accept a decision in the legal, social, political and organizational spheres cannot be generalized to the area of taxation.
The findings of the present study are in line with Cohen (1986), who argued that where the rules of procedural fairness are violated but an outcome is favorable, people notice injustice but do not take action that could adversely alter their privileged position. This explains the observed difference in levels of acceptance of a binding decision for an individually favorable outcome, despite respondents' prior accurate distinctions between fair and unfair procedures.
In the second experiment, procedural fairness was manipulated by varying the method of implementation of governmental tax law changes. Moreover, outcome fairness was also manipulated in the scenarios, varying in their beneficial effects for the taxpayer individually or for their tax group. The Study 2 results indicated that willingness to accept the proposed tax law changes was influenced both by procedural and outcome fairness. As expected, respondents were more accepting of changes introduced in a fair manner than in an unfair manner, even if the changes resulted in higher tax burdens. In line with Folger and Martin's (1986) study, these results suggest that procedural fairness in the field of taxation might be particularly important in the case of systemic solutions introducing permanent tax decision rules that could counteract unequal treatment of both sides of this interaction. Consequently, in situations involving proper, fair procedures taxpayers might be more willing to accept a particular decision, even if this is unfavorable for them individually.
Furthermore, the disparity in the procedural fairness effect observed in both studies strengthens Brockner and Wiesenfeld's (1996) conclusions that outcome favorability Fig. 2 Level of decision acceptance as a function of procedural fairness and outcome fairness in the four experimental groups and outcome fairness are different constructs. However, Skitka et al. (2003)'s argument that the fair process effect is less strongly associated with outcome fairness than with outcome favorability was not confirmed in the present studies.
Studies also aimed to compare the impact of outcome fairness and outcome favorability under conditions of varying procedural fairness in the area of taxation. Outcome fairness was manipulated by distinguishing between tax law changes favorable to the individual taxpayer and to their tax group, introducing a social comparison, while outcome favorability was manipulated by varying the personal result of a tax decision (favorable or unfavorable). Results showed greater strength of an outcome favorability effect in comparison to an outcome fairness effect, which again may suggest that the personal profitability of a decision is the main factor determining its acceptance in the context of taxation.
As discussed above, the present results complement existing knowledge regarding tax-related behavior, however the studies had some limitations. First, it was assumed that the two studies focused on levels of tax decision acceptance only. However, another important consequence of varying levels of procedural fairness is that procedures, in addition to outcomes, arouse emotions (Krehbiel and Cropanzano 2000;Weiss et al. 1999). In particular, as far as the discrepancy between outcomes profitable for the individual taxpayer and for the group is concerned, a sense of guilt is often observed. A lack of procedural fairness may also produce other negative emotions (e.g., anger), and it can be assumed that there are relationships between outcome favorability, outcome fairness and procedural fairness with respect to the extent to which emotions are aroused and subsequent willingness to accept tax authority decisions.
It should be also mentioned that the unfair tax decision making procedure presented in the Study 1 scenario described the individual behavior of a tax authority official, while in the second study unfairness was connected with systemic issues regarding tax law change implementation. However, both studies analyzed the relationship between a tax authority and an individual taxpayer since the tax official was a representative of the tax authority and behaved according to its rules.

Conclusions
The experimental results demonstrated that, in comparison with outcome and procedural fairness, outcome favorability has a predominant role in citizens' acceptance of tax authority decisions. Despite the existence of the fair process effect in many other fields, the fairness of tax authority processes did not influence acceptance of decisions.
Another important finding concerned the differential impact of outcome favorability and outcome fairness on tax decision acceptance. Results suggested that these two constructs cannot be treated synonymously in the domain of taxation. It can be assumed that outcome favorability was associated with the personal benefits connected with a particular decision: that an individual considered it in terms of profits or losses. On the other hand, outcome fairness may be more connected with social comparisons, individuals assessing a particular decision through the prism of social fairness. Thus, as shown in the present results, an alternative point of reference in these two comparisons might lead to differing levels of decision acceptance.
Compliance with Ethical Standards All procedures were conducted in accordance with the ethical standards of the institutional and/or national research committees, and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.