Deterrence Of Tax Evasion Research Paper ⋆ Research Paper Examples ⋆ EssayEmpire

This sample Deterrence Of Tax Evasion Research Paper is published for educational and informational purposes only. If you need help writing your assignment, please use our research paper writing service and buy a paper on any topic at affordable price. Also check our tips on how to write a research paper, see the lists of criminal justice research paper topics, and browse research paper examples.

Overview

Building on the logic of Gary Becker’s rational choice model of crime, Allingham and Sandmo (1972) developed the theory of income tax evasion. Their model studies the choice of a taxpayer who rationally trades off the benefits from evading taxes and the risky costs from detection and incurring fines. By increasing auditing frequencies or the magnitude of fines and penalties, tax authorities can deter evasion. This survey discusses three selected aspects of the literature that studies the general deterrence hypothesis in the context of tax evasion. First, the paper will summarize several empirically studies that convincingly solved the key identification problem. The majority of these studies provide evidence that strongly support the deterrence hypothesis. In a second step, the present survey will address the role of taxpayers’ imperfect information about the actual enforcement policy. This point, which is closely related to the field of perceptual deterrence research, has recently gained attention in empirical work on deterrence in the context of evasion. The third and last aspect covered in this survey concerns the role of standard deterrence incentives as compared to the tax morale and alternative noneconomic factors that shape tax compliance. The paper concludes with a brief summary and remarks on future directions for research on tax enforcement.

Introduction

Four years after Becker’s seminal work on crime and punishment, Michael Allingham and Agnar Sandmo published their analysis of income tax evasion (Allingham and Sandmo 1972). They considered a risk-averse taxpayer who decides how much of his income to report to tax authorities and how much to conceal. Taxes on concealed incomes are evaded. With a certain probability, however, the taxpayer is audited and all the concealed income is detected. In this case, the true taxes plus a fine – which is increasing in the concealed income – have to be paid. Rational taxpayers choose the optimal level of underreported income, trading off the benefits (saving on taxes) with the costs (higher fines) from evasion. Similarly as in Becker’s (1968) work, there is scope for general deterrence: by increasing the probability of detection and the magnitude of fines (and other penalties), evasion becomes less attractive. A rational taxpayer will therefore conceal less income. Along these lines, tax authorities can enforce a certain level of compliance with taxes.

The work by Allingham and Sandmo (1972), one of the most cited articles in public economics, triggered a wave of theoretical studies that quickly extend the original analysis. (For detailed surveys of this literature, see Cowell (1990), Andreoni et al. (1998), and Slemrod and Yitzhaki (2002).) The empirical assessment of the model’s implications, however, evolved at a slower rate. In particular, the test of the deterrence hypothesis in the context of tax evasion turned out to be quite complicated. First of all, evasion and tax enforcement – i.e., the level of auditing determining the detection risk as well as the size of the fines – are endogenously determined: evasion responds to enforcement, but enforcement policies also react to evasion behavior (exactly as with the policecrime circularity). By now, only a handful studies managed to break this endogeneity. Section 2 (Causal Evidence on Deterrence) discusses several of these contributions. Except for one study, these contributions provide evidence suggesting that deterrence works.

A second issue in empirical studies on deterrence in the context of tax evasion emerges from concerns about the limited knowledge of taxpayers about the actual enforcement policies. While the imperfect information of decision makers about the likelihood and magnitude of sanctions is of general interest for the research on deterrence (Bebchuk and Kaplow 1992), this point appears particularly relevant for the case of tax enforcement. In contrast to visible indicators for enforcement such as police on the streets, there are few cues that taxpayers can use to assess tax enforcement policies. In addition, tax enforcement agencies do not reveal details about their auditing practice and often – at least in many continental European countries –operate quietly in the background. Section 3 (Tax Enforcement and Perceptual Deterrence) discusses a very recent strand of literature on general deterrence in the context of evasion that explicitly addresses this point. This literature, which closely related to the field of perceptual deterrence research (see the survey in Nagin 1998), points to the importance of the channels through which information on detection and punishment is spread.

A further complication in the empirical analysis of deterrence emerges in the operationalization of the theoretical concept of “the detection risk.” The problem is best illustrated with the group of (not self-employed) labor income receivers. In modern economies, the evasion rate in this subpopulation is almost zero (Slemrod 2007). At the same time, the effective auditing frequency in this group is close to zero, too. The observation of this evasion and enforcement pattern has motivated many researchers to question the validity of the model of rational taxpayers: given the low auditing rate, the standard model would predict more evasion. To explain this puzzle, the importance of noneconomic motives for compliance – often subsumed among the label “tax morale” – has been put forward (Torgler 2007). Section 4 (Tax Morale and Third-Party Information) will briefly summarize this literature. The section will also highlight that the modern literature classifies the zero-evasion, zero-auditing observation quite differently. In practice, an individual’s auditing risk is not a fixed parameter (as assumed in the simple benchmark model), but rather a function of her evasion choice: the more income she conceals (given a certain level of true income), the higher is her auditing risk. Observing an effective auditing frequency that is close to zero corresponds to just one point of this auditing function (see, e.g., Kleven et al. 2011). Hence, full compliance can be a rational response to standard incentives from deterrence if small levels of underreporting would be detected with a high probability.

Causal Evidence On Deterrence

The identification of a deterrent effect from auditing and the severity of sanctions is complicated by several problems. First of all, it is obviously difficult to measure evasion (Slemrod and Weber 2012). Secondly, enforcement strategies are no exogenous: it is not only evasion, which should respond to enforcement; the choices of enforcement authorities are also determined by the (past, present, or expected future) level of evasion.

Early studies on the determinants of tax evasion were mainly concerned with the first issue (e.g., Clotfelter 1983; Feinstein 1991). Based on random auditing data from the US tax authority’s Taxpayer Compliance Measurement Program (TCMP), these contributions derived individual level measures for evasion. In some studies, this dependent variable was regressed on regional measures of auditing frequencies or income type specific proxies (e.g., Dubin et al. 1990; Beron et al. 1992). In doing so, however, they did not cope with the second issue, the endogeneity of enforcement. Cross-sectional and intertemporal variation in regional auditing rates, for instance, might reflect variation in the unobserved determinants of the taxpayers’ noncompliance that also drive evasion.

A more recent wave of contributions that studies the determinants of evasion behavior is based on randomized field experiments. These experiments provided truly exogenous variation in enforcement policies and therefore contributed to what Slemrod and Weber (2012) called the “credibility revolution in the empirical analysis of tax evasion.” One of the first studies is based on the Minnesota Income Tax Compliance Experiment (Slemrod et al. 2001). In cooperation with a team of researchers, the Internal Revenue Service (IRS, the US tax authority) approached a randomly selected set of taxpayers from a predefined group of income earners with different types of letters. Next to a baseline letter that served as a control treatment, one treatment included a letter with the information that the tax return filed by the taxpayer was selected to be “closely examined.” The comparison of the first-differences in reported incomes provided mixed results.

In line with the deterrence hypothesis, lowand middle-income taxpayers responded to the audit treatment with an increase in reported income. This effect was considerably stronger for those with more opportunities to evade income (e.g., self-employed income). For highincome taxpayers, however, the audit threat backfired: relative to the control group, it induced lower reported tax liabilities. One possible interpretation of this observation is that high-income taxpayers who know that they will be audited with certainty behaved strategically. By entering the bargaining-like interaction with the IRS with a lower reported income, these taxpayers might expect to end up with lower tax payments.

More clear evidence on deterrence is offered by a recent field experiment in Denmark. In cooperation with the Danish tax authorities, Kleven et al. (2011) gained access to a stratified and representative sample of more than 40,000 income tax filers. Half of the individuals were randomly assigned to similar audit threat treatments as in Slemrod et al. (2001), communicating either a 100 % or a 50 % chance of an audit. Relative to a control treatment that did not receive any letter, the audit threats triggered a significant increase in reported incomes, with a larger effect under the 100 % auditing risk (Kleven et al. 2011).

In contrast to the other contributions, the study by Fellner et al. (2013) does not rely on a random draw from a certain population but rather a selected sample of 50,000 potential evaders. They experimentally study the effect of various mailing types on the evasion of TV license fees. Relative to a neutral mailing, an audit threat treatment that also stressed the severity of sanctions had a significantly positive impact on compliance. The evidence in Fellner et al. (2013) therefore shows that deterrence also works for the most important group – those that self-selected into evasion.

All these contributions focused on independent individual or household level choices of evasion. The field experiment by Pomeranz (2011) documents an interesting spillover effect from deterrence in the context of joint, interdependent evasion decisions. Based on a large-scale field experiment with more than 400,000 small Chilean firms, she finds that audit threat letters not only reduces the evasion of Value Added Taxes (VAT) of the targeted firms (i.e., the firm that receives the letter), but also for untreated suppliers, who are also involved in the evasion decision. The logic behind this observation is straightforward: the potential detection of the VAT evasion of the treated firm increases the detection risk of the untreated firm. Hence, in the context of interdependent evasion decisions – as it is the case with VAT evasion – the deterrence effect from tax enforcement is multiplied.

Beyond this strand of research that experimentally varied the threat from an audit, there is still little evidence on the specific impact of actual audits. A rare piece of causal evidence on specific deterrence in the domain of tax enforcement is again found in the Danish study from above (for an earlier studies, see Erard 1992). In addition to and independently of the letter treatments, Kleven et al. (2011) randomly selected a group of taxpayers that was actually audited. Experiencing an audit turned out to have a substantial positive effect on the level of reported incomes in the years following the audit. Quantitatively, this effect was much stronger than the impact of the threat letters (Kleven et al. 2011). While the mechanisms driving this observation are not yet identified, the evidence suggests that actual audits have a much stronger and more long-lasting effect than simple audit threats (on the longevity of deterrence effects, see also Fellner et al. 2013).

Tax Enforcement And Perceptual Deterrence

As Becker’s model of crime, the rational model of evasion assumes that decision makers know the precise detection probability set by the enforcement authority. Perceptual deterrence research provides some evidence that “there presumably is a positive relationship between actual and perceived levels of enforcement”; however, it is considered “implausible that individuals’ probability estimates are generally accurate, particularly when the probability is extremely low” (Bebchuk and Kaplow 1992, p. 366). While this observation is of general interest for research on deterrence, it appears particularly relevant for the case of tax enforcement. In contrast to visible cues such as the number of policemen on the streets, there is hardly any publicly accessible information that allows potential tax evaders to accurately predict their true detection risk. This raises the question, how changes in the enforcement policy translate into changes in compliance behavior.

The question was addressed in the field experiment in Fellner et al. (2013). The researchers confronted subjects with the different letter treatments, in particular, letters with and without an audit threat (see above). Subjects were then asked to indicate their perceptions about the detection risk and potential fines, etc. The results from the perception survey neatly complemented the evidence on the behavioral responses observed in the field experiment: the audit threat considerably increased the perceived detection risk. The evidence suggests that the threats shaped perceptions and that individuals rationally adjust their compliance behavior to these perceptions.

While Fellner et al. (2013) provide evidence on the impact of a targeted policy on individual perceptions and compliance behavior, the impact of “hidden” policies – i.e., generally unobservable enforcement activities – in a society remains unclear. A theoretical analysis of this issue is provided by Sah (1991), who models agents that update their perceived detection risk based on information obtained from neighbors and acquaintances. A higher number of detections among this sample ceteris paribus result in an increased risk perception. In turn, the inclination to comply with the law increases. Implicitly, the model therefore highlights the crucial role of communication for the dispersion of information on enforcement activities.

A recent study that addressing behavioral responses to interpersonal communication in the context of tax enforcement is Alm et al. (2009), who study tax compliance in a laboratory experiment. They implemented a design in which there was no “official” information on auditing risks available; however, they allowed the experimental subjects to communicate. Alm et al. find that income reporting is sensitive to information on auditing frequencies that is spread via communication. More closely related to the model in Sah (1991) is the contribution by Rincke and Traxler (2011), who provide field evidence on a substantial deterrent effect that is mediated by word of mouth. They analyze the enforcement activities of field inspectors that approach potential evaders of TV license fees (compare Fellner et al. 2013) at their homes. These field inspectors unregularly visit different neighborhoods in different months. The presence of inspectors in an area is not announced; inspectors are not uniformed nor do they use any police-like cars. Hence, it is basically impossible to observe if inspectors are visiting an area.

Making use of monthly panel data on the activities of field inspectors in more than 1,000 small municipalities, Rincke and Traxler (2011) identify a substantial impact of these “hidden” enforcement activities. Following an instrumental variable approach, they find that three additional detected evaders induce one further evader (that was not approached by field inspectors) to switch to compliance. Making use of microlevel data, they provide further evidence that strongly suggests that the effect is driven by communication about detections among neighbors. In addition, they show that the enforcement spillover is concentrated in the close spatial proximity to the detected household.

The findings in Alm et al. (2009) and Rincke and Traxler (2011) provide two important insights. First and most importantly, the fact that enforcement activities are often “hidden” and that auditing rules are typically not announced by tax authorities does not imply that an increase in enforcement activities fails to deter evasion. Communication on experiences with enforcement authorities seems to be sufficient to convey the information about an increased detection risk. This implies – and that is the second point to take away – that patterns of communication and information dispersion become decisive for general deterrence.

Tax Morale And Third-Party Information

Third-Party Information

The rational model of evasion predicts that a sufficiently high detection probability should deter evasion. The comparison of evasion levels for different sources of incomes (Slemrod 2007), however, seems to provide a picture that first appears at odds with this prediction. In modern economies, the evasion of taxes on nonbusiness labor income is nearly zero. At the same time, the auditing rate of labor income receivers is close to zero, too. As noted earlier, this pattern has motivated many researchers to question the rational model of evasion: it has been argued that the standard model would predict “too much evasion.” As an explanation for this “puzzle,” noneconomic motives for compliance, often subsumed among the label tax morale, have been put forward (e.g., Torgler 2007). The modern taxation literature, however, provides a different explanation to the zero-evasion, zero-auditing observation: the presence of third-party income reporting, which makes many taxpayers unable to evade.

In its most simple version, the standard model of evasion captures the risk of detection in one single parameter, the auditing probability p. In practice, however, an audit does not necessarily result in the detection of all evasion. In addition, and more importantly, an individual’s auditing risk is not given by a constant parameter, but rather a function of her underreporting and her true income. It appears plausible to assume that, for a given level of true income and a given income source, an individual’s risk of detection increases with a higher level of underreporting (several extensions of the basic model follow this approach; see the surveys in Andreoni et al. 1998; Cowell 1990). If one thinks about the observation of zero-auditing and zero-evasion for labor incomes along the lines of this more realistic model, one arrives at a straightforward explanation.

In economies with a modern tax enforcement system, labor income is typically “third-party reported.” This means that not only the labor income receiving taxpayer declares her incomes to the tax authority, but also the firm who pays these incomes (and often withholds taxes) reports the payments. Tax authorities can therefore easily compare the taxpayer’s declared income with the income that is reported by her firm. For a taxpayer who exclusively receives labor income, this implies that she is basically unable to conceal income (see, e.g., Kleven et al. 2011). While there is a zero-auditing risk under full compliance (which is observed on average), any noncompliance will get detected basically with certainty. Given this nonlinear detection risk, full compliance is a rational response to standard economic incentives – and perfectly in line with a more realistic extension of the baseline model of rational evaders.

This point appears trivial. In fact, it was well understood among policy-oriented researchers for a long time (see, e.g., Long and Swingen 1990). For the theoretically motivated research, however, it took more time until it was acknowledged that the model by Allingham and Sandmo (1972) is a simplified model of a taxpayer that fully self-reports her, e.g., self-employed income, rather than a model that describes the evasion of third-party reported income. (For a theoretical analysis of the case where employer and employee jointly evade taxes on third-party reported incomes, see Kleven et al. 2009.) In line with the baseline model, fully self-reported incomes – in contrast to third-party reported incomes – are actually concealed to a significant amount (Slemrod 2007).

The availability of third-party information provides a convincing explanation for the high compliance and the low auditing risk that is observed for labor incomes. Hence, the argument that this pattern can only be explained by tax morale or social norms for tax compliance is not justified. This does not mean, however, that tax morale – or any other behavioral incentive beyond those captured in the standard model – is irrelevant for tax compliance. It is therefore interesting to consider possible noneconomic incentives and their relevance for tax enforcement. The remainder of this section therefore discusses the role of tax morale. Models of bounded rational taxpayers, probability weighting, and loss aversion are beyond the scope of this survey (see, e.g., Dhami and al-Nowaihi 2007, and references therein).

Tax Morale

The modern literature uses the term tax morale quite broadly, typically subsuming the idea of a social norm for tax compliance, perceptions about civic duties as “honest taxpayer,” but also the notion of stigmatization (for a survey, see Torgler 2007). Theoretical approaches typically model tax morale in form of an additional, noneconomic cost of evasion (an overview of different modeling strategies is provided by Traxler 2010). In these frameworks, concealing income triggers economic consequences (in case of detection) as well as psychic (e.g., shame) or social cost (e.g., from social exclusion) that are motivated by tax morale. These costs might differ between different individuals – some agents might not be affected by tax morale, others might even perceive a moral obligation to evade taxes. In addition, it is typically assumed that the noneconomic costs of evasion depend on the compliance behavior of “relevant others” (Traxler 2010). In a society, where everybody complies with taxes, the noneconomic costs of evasion might be large. Hence, individuals experience strong incentives to comply with taxes. However, if evasion would be widespread in this society, the same individuals might perceive hardly any noneconomic costs of cheating and might evade taxes. A tax morale that depends on the behavior of others therefore has an important implication: it gives scope for a multiplicity of equilibria (or, more generally, social interaction in tax compliance; see Galbiati and Zanella 2012). For a given level of tax enforcement, a society might coordinate on a state where evasion is widespread or a state where there is a strong norm to comply with taxes (and, potentially, further equilibria between these extremes; see Traxler 2010).

These ideas on tax morale have several important implications for the effectiveness of deterrence. The first two are mainly of theoretical interest: changing the level of tax enforcement might eliminate the existence of one (out of several) possible equilibrium. Hence, minor changes in the enforcement policy could, in principle, produce large shifts in compliance behavior. Related to this point, one can think more generally about the interaction of the formal, legal and informal, private enforcement of compliance: if stricter tax enforcement is an expression of the societies’ weight attached to compliance with this rule (an idea from expressive law), it might also crowd-in stronger psychic or social cost from tax evasion. Similar as with the private versus public provision of public goods, however, there might also be scope for crowding-out (Kube and Traxler 2011).

A third implication relates to the notion of the “relevant others.” If the most relevant others (in terms of shaping perceptions about tax morale) are, for instance, politicians or high-profile members of the business world, the models highlight potential benefits from focusing enforcement policies on these groups: Enforcing a high level of compliance in a narrow group might contribute to a high tax morale among a broader population (Traxler 2010). Hence, there might be a positive spillover from deterrence. One should note, however, that stricter tax enforcement among these specific groups could also backfire. This is the case, since media reporting on detected evaders from moral reference groups could erode the tax morale in the general population.

A further interesting implication for deterrence is contained in models that consider tax morale and stigmatization (e.g., Kim 2003). These studies assume that stigma-related costs are only incurred in case of detection. Given that taxpayers want to avoid stigmatization and any further negative social consequence from being known a detected evader (e.g., ostracism, social exclusion), the fear of stigmatization will increase the overall level of (economic and noneconomic) punishment and therefore renders an increase in the auditing frequency more effective. Relating the idea of stigmatization to the previous point, one might argue that the costs from stigmatization crucially depend on how many other taxpayers are detected evading. If stigmatization is more severe in a society where evasion is a rare phenomenon, this means that the deterrent effect from an increase in auditing is larger as compared to a society where evasion is widespread.

The discussion from above shows that tax morale has nontrivial implications for tax enforcement and deterrence. It is therefore worth asking, how important tax morale really is. Unfortunately, the empirical literature provides little causal evidence that allows assessing the role of tax morale. Beyond a huge number of correlational studies, there is only one very recent contribution that documents a causal impact of (survey measures of) tax morale on actual evasion behavior (Halla 2012). In addition, evidence from field experiments that tested approaches which rely on tax morale suggests that it is difficult to effectively apply this type of alternative enforcement strategies. The Minnesota Income Tax Compliance Experiment discussed above, tested – next to an audit threat letter – also different moral appeal letters. Relative to a control treatment, these moral letters turned out to be ineffective (Blumenthal et al. 2001). Similar evidence is provided in the field experiments of Fellner et al. (2013) and Pomeranz (2011). However, Fellner et al. (2013) found some evidence that an information letter, that emphasized the high level of compliance in the population, had a positive effect on compliance (relative to a neutral treatment which did not include this information).

Concluding Discussion And Future Research

This research paper provided an overview of different strands of research on deterrence within the domain of tax evasion. Section 2 (Causal Evidence on Deterrence) summarized field studies that randomly varied the auditing risk that is communicated to potential evaders. While some of these studies point to the particularities that distinguish tax evasion from the deterrence of noneconomic crimes (e.g., the evidence on responses from high-income group in Slemrod et al. 2001), the overall picture provides clear support for the deterrence hypothesis: a higher auditing risk reduces evasion. These randomized studies made a major contribution to the empirical analysis of deterrence in the domain of tax evasion, as they established credible, causal evidence on a conceptually important issue (Slemrod and Weber 2012). Nevertheless, there is still a lot left for future research.

First of all, there is little causal evidence on the effectiveness of actual audits. Existing results suggest that there is a substantial effect of specific deterrence (Kleven et al. 2011). Due to the lack of data, however, the robustness and intertemporal persistence of this observation is still unclear. To analyze these points and to better understand the impact of audits, more studies with exogenous variation in auditing are needed.

Second, the empirical literature offers only little evidence on the impact of different sanction levels. Moreover, there is basically no field evidence on the optimal mix of the two policy tools for deterrence – the detection risk and the magnitude of the punishment. One theoretical solution – an enforcement system with infinitely high sanctions and an infinitesimally small auditing risk (“hang taxpayers with zero probability,” as suggested in an early contribution), which is cheap in terms of auditing resources – will most likely not be tested empirically. However, one can easily imagine a randomized study that varies both dimensions of the enforcement system – the detection risk and, e.g., the size of fines. Such an experiment would make a huge contribution to the important policy debate on the composition of an optimal enforcement system as well as to the theoretical literature on the trade-off between certainty and severity of sanctions.

Section 3 (Tax Enforcement and Perceptual Deterrence) of this research paper discussed a recent literature which links perceptual deterrence research with the empirical analysis of tax enforcement. The first contributions in this domain suggest that there is a clear link between policies and perceptions, as well as perceptions and behavior (Fellner et al. 2013). In addition, the crucial role of communication about own experiences with auditing and detection is highlighted (Rincke and Traxler 2011). In the context of imperfect perceptions about the objective auditing risk, word of mouth about detections crucially shapes the effectiveness of enforcement strategies.

Given the difficulty to obtain precise microdata, this young literature has mainly focused on communication within neighborhoods and spatial distances. It would be important for future research to bring this analysis to further layers of social interaction and to future measures of “proximity” (such as friends in a social network). This would allow assessing, for instance, the role of communication at the workplace or among business partners for spreading information about auditing policies and auditing technologies (“what can easily be detected”).

Naturally, this research agenda seems to be linked with the growing empirical literature on networks. Given that communication between taxpayers is crucial for their compliance behavior, there emerges an interesting research question: how should an optimal auditing policy allocate resources between different taxpayers within a (neighborhood, social, workplace) network? Who is the key evader to target? Empirical studies that address these questions would provide important insights for the design of optimal enforcement strategies and would make major contributions to the applied work on networks.

Section 4 (Tax Morale and Third-Party Information) addressed the role of third-party information reporting and tax morale for tax compliance. It is now widely acknowledged that the benchmark model of evasion is tailored to the case of fully self-reported incomes and that the availability of third-party information explains the observed auditing and evasion pattern for different occupational groups and income sources – in particular, self-employed and business versus nonbusiness, labor incomes. A straightforward policy implication is that an extension of the application of third-party information reporting – or more generally: policies that give tax authorities broader access to matchable data – will render a tax enforcement system more efficient. In order to increase compliance, investing in clever information systems therefore appears to be an alternative to investments in stricter auditing.

Finally, several possible implications of tax morale for deterrence were discussed. Field experiments that tested moral appeals or other approaches to enforcement that are rooted in the notion of tax morale are generally found to be ineffective in turning evaders into compliant taxpayers. It would be misleading, however, to interpret this observation as an indicator for tax morale being irrelevant. Based on the available evidence, the question about the importance of tax morale for actual behavior remains unanswered. Again, this is an important issue to address in future research.

Bibliography:

Allingham M, Sandmo A (1972) Income tax evasion: a theoretical analysis. J Public Econ 1:323–338
Alm J, Jackson B, McKee M (2009) Getting the word out: enforcement information dissemination and compliance behavior. J Public Econ 93:392–402
Andreoni J, Erard B, Feinstein J (1998) Tax compliance. J Econ Lit 36:818–860
Bebchuk L, Kaplow L (1992) Optimal sanctions when individuals are imperfectly informed about the probability of apprehension. J Legal Stud 21:365–370
Becker G (1968) Crime and punishment: an economic analysis. J Polit Econ 78:169–217
Beron KJ, Tauchen HV, Witte AD (1992) The effect of audits and socioeconomic variables on compliance. In: Slemrod J (ed) Why people pay taxes. University of Michigan Press, Ann Arbor, pp 67–89
Blumenthal M, Christian C, Slemrod J (2001) Do normative appeals affect tax compliance? Evidence from a controlled experiment in Minnesota. Natl Tax J 54:125–138
Clotfelter C (1983) Tax evasion and tax rates: an analysis of individual returns. Rev Econ Stat 65(3):363–373
Cowell FA (1990) Cheating the government – the economics of evasion. MIT Press, Cambridge, MA
Dhami S, al-Nowaihi A (2007) Why do people pay taxes: expected utility versus prospect theory. J Econ Behav Organ 64:171–192
Dubin J, Graetz M, Wilde L (1990) The effect of audit rates on the federal individual income tax 1977–1986. Natl Tax J 43(4):395–409
Erard B (1992) The influence of tax audits on reporting behavior. In: Slemrod J (ed) Why people pay taxes. University of Michigan Press, Ann Arbor
Feinstein J (1991) An econometric analysis of income tax evasion and its detection. Rand J Econ 22(1):14–35
Fellner G, Sausgruber R, Traxler C (2013) Testing enforcement strategies in the field: threat, moral appeal and social information. J Eur Econ Assoc 11(3): 634–660
Galbiati R, Zanella G (2012) The tax evasion social multiplier: evidence from Italy. J Public Econ 96:484–495
Halla M (2012) Tax morale and compliance behavior: first evidence on a causal link, B.E. J Econ Anal Policy 12(1)
Kim Y (2003) Income distribution and equilibrium multiplicity in a stigma-based model of tax evasion. J Public Econ 87:1591–1616
Kleven H, Claus K, Emanuel S (2009) Why can modern governments tax so much? An agency model of firms as fiscal intermediaries. NBER Working Paper 15218
Kleven H, Knudsen M, Kreiner CT, Pedersen S, Saez E (2011) Unwilling or unable to cheat? evidence from a tax audit experiment in Denmark. Econometrica 79:651–692
Kube S, Traxler C (2011) The interaction of legal and social norm enforcement. J Public Econ Theory 13:639–660
Long S, Judyth S (1990) Third-party information reporting and compliance. In: Research conference report, how do we affect taxpayer behavior? Internal revenue service document 7302, Washington, DC, pp 3–91
Nagin D (1998) Criminal deterrence research at the outset of the 21st century. Crime Justice 23:1–42
Pomeranz D (2011) No taxation without information: deterrence and self-enforcement in the value added tax. Working Paper, Harvard Business School
Rincke J, Traxler C (2011) Enforcement spillovers. Review of economics and statistics 93:1224–1234
Sah R (1991) Social osmosis and patterns of crime. J Polit Econ 99(6):1272–1295
Slemrod J (2007) Cheating ourselves: the economics of tax evasion. J Econ Perspect 21:25–48
Slemrod J, Shlomo Y (2002) Tax avoidance, evasion, and administration. In: Auerbach AJ, Feldstein M (eds) Handbook of public economics, vol 3, Chapter 22, pp 1423–1470, Elsevier: Amsterdam
Slemrod J, Weber C (2012) Evidence of the invisible: toward a credibility revolution in the empirical analysis of tax evasion and the informal economy. Int Tax Public Finance 19(1):25–53
Slemrod J, Blumenthal M, Christian C (2001) Taxpayer response to an increased probability of audit: evidence from a controlled experiment in Minnesota. J Public Econ 79(3):455–483
Torgler B (2007) Tax compliance and tax morale: a theoretical and empirical analysis. Edward Elgar Publishing, Cheltenham
Traxler C (2010) Social norms and conditional cooperative taxpayers. Eur J Polit Econ 26:89–103