![Sample our Area Studies journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5922/TOC-Subject-Sample-2021-Area-Studies.jpg"})
Abstract
Tax non-compliance and perceptions of corruption are key challenges to state-building in developing countries. Using a social psychology approach, we develop a theoretical model in which different forms of perceived corruption can adversely influence the way individual taxpayers behave. We then apply this model to Indonesia, placing our empirical findings in the context of compliance risk management, identifying strategies to improve tax compliance, and exploring how to implement these strategies effectively. We shed light on the applicability of the traditional responsive regulatory approach (used by revenue authorities to deal with intentionally non-compliant taxpayers), which combines measures in attempting to achieve either voluntary or enforced compliance. While the empirical evidence is based on the Indonesian experience, we suggest that our model is sufficiently general and robust to be applicable to other developing countries in the Asia-Pacific region.
Ketidakpatuhan terhadap pajak dan persepsi korupsi adalah dua tantangan utama bagi penguatan institusi negara di negara-negara berkembang. Menggunakan pendekatan sosio-psikologi, kami membangun model teoretis di mana berbagai bentuk persepsi korupsi dapat secara negatif mempengaruhi perilaku individual dalam membayar pajak. Kami menerapkan model ini pada kasus Indonesia, dan menganalisis temuan-temuan empiris pada konteks manajemen risiko kepatuhan, mengidentifikasi strategi-strategi untuk meningkatkan kepatuhan pajak, serta menelaah bagaimana strategi-strategi ini dapat diimplementasikan secara efektif. Lebih jauh, kami menjelaskan bagaimana pendekatan peraturan yang tradisional responsive (yang digunakan oleh otoritas pajak dalam menghadapi pembayar pajak yang tak patuh) dapat berguna, dengan mengombinasikan berbagai peraturan yang dapat mencapai kepatuhan, baik secara sukarela maupun dipaksa. Meskipun bukti-bukti empiris ini hanya bersandar kepada pengalaman di Indonesia, kami percaya bahwa model yang dibangun bersifat umum dan relevan untuk diaplikasikan terhadap negara-negara berkembang lainnya di wilayah Asia-Pasifik.
Notes
1 In the literature on corruption in general, there are two competing views on the impact of corruption on economic efficiency: the ‘sand the wheels’ and ‘grease the wheels’ hypotheses (see Nguyen et al. Citation2016). In the context of tax corruption, we are inclined to support the ‘sand the wheels’ hypothesis—that is, that tax corruption is harmful.
2 Data from http://pajak.go.id/statistik-amnesti (accessed 27 Oct. 2016).
3 This definition deliberately excludes the ‘registration’, ‘filing’, and ‘payment’ criteria.
4 Tax morale is frequently described as the intrinsic motivation to pay tax (Frey and Torgler Citation2007).
5 We estimated the study's accessible population to be around 360,000, having taken the number of registered PITs obliged to lodge annual tax returns (as of December 2012) and multiplied it by the filing ratio of each tax office under study. Our structural equation modelling stipulated that our minimum sample of taxpayers would be 384.
6 Perception can be defined as ‘a belief or opinion, often held by many people and based on how things seem’ (https://dictionary.cambridge.org/dictionary/english/perception). As is true when defining corruption, defining the perception of corruption depends on social and cultural factors (Melgar, Rosi, and Smith 2010). However, in this context, perception does not necessarily represent belief in the TPB model. Belief in this model refers to ‘subjective probabilities’ (Fishbein and Ajzen Citation2010, 221) instead of a sense of certainty about the existence or truth of something. Smith (Citation2001) offers a fuller discussion on the relation between perception and belief.
7 This is in stark contrast to what the neoclassical economic model predicts: that most rational taxpayers would become evaders, since they are unlikely to be caught and penalised (Ritsatos Citation2014).
8 The causal relationships between elements in the TPB have been adopted to explain a variety of psychological and social factors in several studies of tax compliance behaviour. These include, for instance, the impact of moral obligation (Bobek and Hatfield Citation2003), the influence of ethics as a representation of moral reasoning (Trivedi, Shehata, and Mestelman Citation2005), and the impact of the intention to comply on predicting compliance behaviour (Langham, Paulsen, and Hartel 2012).
9 For instance, ‘Why should I pay tax if it is just financing corruption?’ was a frequently expressed view from participants―this sentiment was expressed 18 times during the interviews. Also, as several participants described, to compensate for a ‘non-deductible payment’ (that is, a bribe) for tax calculation purposes, some taxpayers might have no choice but to underreport their income. These taxpayers may otherwise have to pay double taxes, weakening their financial circumstances.
10 Considering that the survey collected data from similar portions of self-employed and employed PITs in each of the surveyed tax offices, the variability of these portions should not be evident. For this reason, the variability of types of annual return lodged by employed PITs—who may use either 1770 S or 1770 SS tax returns—was used.
11 The representativeness test of the sample frame should ideally be performed on various demographic variables, such as age, gender, level of education, and race. Unfortunately, owing to the nature and availability of data provided by the DGT, only these two variables could be used for the test.
12 Full models include the ‘level of reported income’ (TCB) variable. Partial models exclude this variable.
13 These three findings draw heavily from Rosid, Evans, and Tran-Nam (Citation2016). Mainly due to space considerations, the detailed process of empirical investigation is not discussed further in this article.
14 This result seems counterintuitive. Nevertheless, recent data on the Indonesian government's tax amnesty program might partly reflect this phenomenon. The internal data provided by the DGT on 12 October 2016 suggest that, surprisingly, more than 53% of the total tax payments collected from phase one of the 2016 tax amnesty program (July−September 2016) was paid by employed PITs. As our findings suggest, one possible explanation for this is that the employed PITs tend to conceal much of their additional income other than income from employment.
15 In this sense, the DGT describes a situation where many of its tax officers are constantly monitoring the same taxpayers without prioritising compliance, and terms this business-as-usual phenomenon asberburu di kebun binatang, or ‘hunting at the zoo’ (DGT 2015b, 36).
16 Data from http://www.ilo.org/surveydata/index.php/catalog/history, accessed 5 May 2017. By way of comparison, our findings suggest that 17% of employed PITs earn income other than from employment.
17 Consequence is evaluated in the terms of its impact on organisational objectives. It can be measured qualitatively or quantitatively or by both methods (OECD 2004). Likelihood is assessed in terms of its probability of occurring. It can also be measured qualitatively or quantitatively or by both methods (OECD 2004). In other words, it is somewhat arbitrary.
18 A subjective definition of ‘almost certain’ would be ‘expected to occur in most circumstances’, while an objective definition would be ‘likely to occur this year or at frequent intervals’ (OECD 2004, 28).
19 Three basic questions need to be addressed in understanding taxpayer compliance behaviour (OECD 2004, 42): What is occurring (for example, the underreporting of income)? Who is doing it (what are their characteristics)? And why are they doing it (for example, because of perceived inequity)?
20 The fundamental principle of this approach is that policymakers should be responsive to the culture, conduct, and context of those they seek to regulate. Thus, it requires a diversity of regulatory strategies to be context appropriate (Wood et al. Citation2010).
21 In this sense, while descriptive norms tend to guide people in uncertain situations, injunctive norms are likely to help people avoid social disapproval (Onu and Oats Citation2015).
22 See the DGT's official letter SE-15/PJ/2014 on 21 March 2014 regarding Audit Plan and Strategies for 2014. The audit coverage ratio for PITs represents the number of audited PITs against the number of PITs’ annual tax returns received by the DGT. The actual ratio is probably much lower. In 2011, for example, our calculation suggests that the random-based audit rate for PITs was less than 0.02%.
23 Although audits are crucial in enforcing compliance among non-compliant taxpayers, coercive audits might have undesired outcomes (Lisi Citation2014) because tax compliance results from a combination of effective enforcement and mutual trust between taxpayers and the authorities. Once both trust and power are at moderate levels, destructive multiplier effects will occur, resulting in a downward spiral of decreasing compliance (Kirchler, Hoelzl, and Wahl Citation2008).
24 Alternatively, this categorisation could be based on types of profession, such as lawyer, accountant, architect, artist, and other prominent jobs.
25 Focused enforcement creates greater perceived uncertainty among the population (Osofsky Citation2014). Taxpayers may know there is some chance of getting audited but not know the increased probability. Research suggests that people tend to avoid uncertainty.
26 Cognitive bias suggests that individuals tend to rely more heavily on information that is more readily accessible when evaluating the likelihood, or frequency, of an event.
27 The Indonesian government has recently (on 8 May 2017) issued Government Regulation in Lieu of Law 1/2017 on Financial Information Access for Taxation Purposes, which gives the DGT the power to scrutinise banking data.