ISET Economist Blog

Tax Inspection With A Human Face?
Saturday, 14 November, 2015

Apostle Matthew was a tax collector in Galilee – perhaps the most hated occupation in the ancient world. By quitting his job and deciding to follow Jesus, Mathew accomplished one of the greatest transformations possible in a person’s life. Modern tax inspectors are certainly not expected to follow Jesus. However, by following insights from behavioral economics, they may gain respect and social status while also improving tax collection, in Georgia and elsewhere around the world.

As discussed in our previous article, Georgia’s tax administration has improved quite significantly following a massive crackdown on corruption and a series of reforms implemented after the Rose Revolution of 2003. As a result, the Georgian shadow economy has indeed shrunk, but not by as much as one might have expected. According to a World Bank study, Georgia’s shadow economy declined from 68.3% in 1999 (at the pinnacle of Shevardnadze's rule) to 62.1% in 2007 (four years after the glorious Rose Revolution). Six percent in 8 years is certainly an improvement, but…

Anyone puzzled by the resilience of shadow economic activities should understand that being in the “shadow” is not necessarily about plain tax evasion. Equally important is tax avoidance (a variety of tactics used by corporations to shift profits to tax havens or otherwise legally reduce their tax liability). Being legal, tax avoidance is very difficult to fight in the absence of motivation for voluntary compliance.


One way to increase voluntary compliance might be to invest in the government’s own image and quality of public services. Corporations (and individuals) might be more willing to pay taxes knowing that their contributions are used for good purposes by a benevolent, lean, and efficient administration. 

It is worth noting that voluntary compliance may be stronger in a decentralized fiscal system in which at least some share of locally collected taxed is visibly contributing to the quality of local infrastructure, public schooling, etc. Rather than hiring expensive tax consultants, a business might just as well contribute a bit more to the local budget, if it sees any tangible benefits from doing so. 

[To avoid any misunderstanding, fiscal decentralization (or fiscal federalism) is not about breaking a centralized tax payment system of which Georgia may be proud; it is about linking the amount of taxes collected in a region to the amount of public funding it is able to spend on local needs]. 

Unfortunately, Georgia’s fiscal system is overly centralized, killing the incentives of local government to cultivate local businesses, and those of local businesses to reciprocate by contributing resources, ideas, and know-how. For example, the Rooms hotel is by far the largest taxpayer in the Kazbegi municipality and the only paying client of the government-controlled United Water Supply Company (UWSC). Yet, none of its taxes remain in Kazbegi, which helps explain the poor quality of the road leading to the hotel and the fact that, until 2014, UWSC did nothing to prevent chronic water shortages that poisoned the hotel’s relationship with the community. 

This is clearly not the way to strengthen voluntary compliance. 


Corporations do care about their reputation. The fact that Starbucks, Google, and Amazon came under fire and were publically vilified for tax avoidance in the UK may have reduced their appetite for tax minimization. Thus, when in May 2013 grassroots movements, such as UK Uncut, staged sit-ins at 40 Starbucks branches, what followed was a series of lengthy explanations by the company’s global CFO, and a decision to pay corporate tax in the UK (for the first time since 2008). We do not manipulate anything, anywhere, insisted the company, but we listen to our customers.

A more elegant method to exploit reputational considerations of businesses is presented by Fair Tax UK. Instead of naming and shaming companies for tax avoidance, Fair Trade provides positive stimuli for socially responsible and ethical behavior. Companies are awarded Fair Tax Mark if they cooperate with Fair Tax experts and meet basic criteria of tax reporting, tax payment, and transparency. The incentives for applying for Fair Tax Mark are reputational. Ed Mayo, Secretary General of Co-operatives UK puts it well: “All it takes is for consumers, people who are taxpayers themselves, to back the companies that pay what they owe”. Fair Tax found widespread support in the UK as over 50 MPs endorsed the idea. SSE, a British supplier of energy, phone, and broadband services, became the first FTSE100 company to apply and qualify for Fair Tax Mark in 2015.

In Georgia, FINANCIAL has a practice of recognizing businesses for success in areas such as corporate social responsibility (CSR). It might as well be time to recognize those Georgian corporations that contribute their fair share to society and thus put consumer pressure on those who practice dodging. The Fair Tax UK team will be happy to help us get started.


A recent randomized field experiment conducted by Kristina Bott, Alexander Cappelen, and Erik Sørensenin from the Norwegian School of Economics demonstrates yet another possibility to increase the citizens’ tax morale. Based on reports from foreign tax authorities, the Norwegian Tax Authorities (NTA) established that at least 51% of income earned outside of Norway in 2011 never reappeared on Norwegian tax returns, and around 20,000 tax residents were classified as tax evaders in this fiscal year. In collaboration with NTA, the NHH team conducted a nationwide randomized field experiment to study the reasons for tax evasion and the relative importance of the probability of detection and moral motivation. 

Shortly after sending the pre-populated tax returns for 2012, the NTA mailed eight different treatment letters randomly assigned to people in the pool of tax evaders. A key finding was that self-declared incomes for 2012 were significantly increased by letters containing a moral argument for paying taxes (including visual aids highlighting the government sponsorship of public services, health, education, and scientific research) or a reference to the fact that the tax authorities know that this person has previously misreported. Further, the team found that moral motivation mainly works on the intensive margin (increasing the amounts of self-reported income), whereas the probability of detection works also on the extensive margin, increasing the share of individuals self-reporting any positive amount of income.


These and other results have been recently presented at the workshop “How to Increase Tax Compliance – New Tools and Insights from Behavioral Economics” organized by the Joint Research Centre of the European Commission in Kyiv. The workshop was attended by Giorgi Tabuashvili, head of the Georgian Revenue Service (RS), and ISET Policy Institute’s Lasha Lanchava. Inspired by what he heard, Mr. Tabuashvili came up with the idea to engage international experts and Dr. Lanchava in piloting and testing the effectiveness of different communication strategies targeting potential tax avoiders in Georgia. 

Two communication treatments were tried and evaluated: 

  • A moral appeal letter that highlights the importance of tax compliance for funding vital public projects such as police, healthcare, education, pensions, etc. (moral treatment); 
  • Telephone communication by an RS agent informing taxpayers that they had been selected for monitoring and assistance with filing their tax returns properly (telephone treatment).

In line with the literature on tax evasion, these treatments were applied to two different categories of taxpayers: those who did not pay taxes at all (Group A), and those who paid but are likely to have filed their taxes incorrectly (Group B). 

To estimate the causal impact of the intervention on voluntary tax payments it was necessary to make sure that those contacted either by mail or telephone (treatment group) are more or less similar to those who did not receive any communication (control group). In principle, the best way to assure such similarity is to randomly assign people to treatment and control groups. Since randomization could not be performed in this case, to establish proper controls for each type of treatment, Lanchava established control groups using the so-called Propensity Score Matching (PSM) procedure. This statistical technique uses observable characteristics (such as age, number of family members, the average income of family members, and value of taxable wealth) to match taxpayers that have not been treated to similar ones among the treated. The results of the experiment are shown it Table 1 which summarizes the average effects of the two treatments on the treated (ATT). 

 Table 1. Average Treatment Effect on the Treated (ATT) Estimates
 Treatment All Group A Group B
  #obs ATT #obs ATT #obs #ATT
Moral treatment 103 44.71* 77 64.48* 27 -18.10
Telephone treatment 136 58.20* 100 67.24* 28 14.80

Notes: * indicate significance level at 1% level

As can be seen, the moral and telephone treatments had a positive and significant impact on tax payments by individuals in group A (those who have not been paying taxes) – by about 64 and 67 GEL, respectively. Interestingly, neither treatment had any significant effect on those in group B. The immediate takeaway from this is that the moral and telephone treatments work on an extensive margin, affecting those who do not pay taxes at all. One additional observation is that the effect of moral treatment is as good as telephone communication – the two treatment effect estimates are statistically indistinguishable. 

These results provide support for a national strategy aimed at increasing voluntary tax compliance in Georgia. Communication interventions of the kind evaluated in our pilot are cost-effective and not overly intrusive. Given that communication is costly, our policy recommendation for the Revenue Service would be to focus on Group A taxpayers only.

Finally, there might be scope for the Georgian Orthodox Church – by far the most respected public institution in the country – to play a role in elevating the tax morale of the Georgian churchgoer population by more forcefully preaching the message “Render unto Caesar the Things that Are Caesar's”. This message may go down particularly well if Caesar – the Georgian tax inspector, in our case, – learns behavioral economics and acquires a human face.

The views and analysis in this article belong solely to the author(s) and do not necessarily reflect the views of the international School of Economics at TSU (ISET) or ISET Policty Institute.