Impact case study database

The Tax Gap is the difference between the amount of tax that should, in theory, be paid, and what is actually paid. Measurement of the Tax Gap is an important tool in helping governments to understand the relative size and nature of non-compliance, and then to develop effective strategies aimed at reducing the gap. Better understanding and measurement of the Tax Gap is a top priority for all revenue agencies. The UK’s HMRC estimates the UK income Tax Gap to be in the region of £12 billion annually and in the EU-28 it is estimated to be over €50 billion a year. Corresponding annual UK Tax Gaps relating to all forms of tax are estimated to be £31 billion (UK; 4.7% of total theoretical tax liabilities) and €1 trillion (EU-28).

The major piece of research [3.1] underpinning this impact case study is the outcome of a longstanding close collaboration between the UK’s HMRC, and Professors Kotsogiannis and Myles of the University of Exeter Business School Tax Administration Research Centre (TARC). HMRC representatives sit on TARC’s Advisory Board and a PhD was delivered thorough this research (Ana Cinta G. Cabral).

The TARC/HMRC collaboration was consolidated through a research project led by Kotsogiannis (2013-2017) on Tax Gap measurement, which specifically focused on the UK’s Self-employment Income Tax Gap – estimated to be £1 to £2 billion annually [3.1]. The project was commissioned directly and co-developed with the HMRC. Its aim was to contribute to improving understanding around some of the more challenging areas of tax gap estimation. The research [3.1] builds on the existing literature, which utilises the expenditure-based methodology to measure income tax non-compliance, and exploits the differing opportunity for under-reporting of the two employment groups, the self-employed and employed.

As income from employment is subject to ‘withholding taxes’, there are minimal possibilities for under-reporting, and thus the income of the employed is assumed to be accurately reported. Through observing the expenditure pattern of the employed, the level of income necessary to sustain a certain level of expenditure can be ascertained. Therefore, any discrepancies between the income reported by the self-employed and the employed, at a certain level of expenditure, can be used to elicit the extent of hidden income by the former group. A key research finding was that the discrepancy observed is due to the taxpayer’s decision to under-report their tax liabilities, as opposed to being due to reasons related to the heterogeneous behaviour of occupational groups and the differences in their preferences, or by consumers saving and consuming differently, or error in measuring the variables of interest [3.1]. This is an important finding and innovation because, when events are studied that do not lend themselves to measurement (as with the under-reporting of income), there might be several other explanations that challenge the underlying assumptions of the analysis, and could potentially bias the results, in this case, leading to a false attribution to income under-reporting.

Following several robustness tests, the research [3.1], confirms that the results are not caused by different preferences between the self-employed and the employed. It shows that, on average, the self-employed report only 80.4% of their true income (or the self-employed income-gap is 19.6%). Given that self-employment income in the UK for the period of study (2010–2012) represented 5.5% of GDP, this translated into an estimate of unreported taxable income of 1.6% of GDP.

Further analysis [3.2], carried out in collaboration with HMRC, identified the economic sectors which are contributing to the UK’s Hidden Economy (defined as the economy operating outside formal channels) with the highest risk of income under-reporting.

The initial major piece of research [3.1] subsequently led to cooperation with a number of other national revenue authorities, including the Bulgarian National Revenue Agency (BNRA). With the support of the BNRA, TARC was awarded £100,000 by the Alan Turing Institute (2018-2020) for the project ‘Detecting Anomalies in Networks: The Case of VAT’, which investigated Value Added Tax (VAT) fraud, being a component part of the Tax Gap. The project was delivered in collaboration with University College London. Insights obtained from the work with HMRC [3.1] were combined with Bulgarian administrative tax data, and resulted in a report and scientific contribution analysing how the VAT Tax Gap, resulting from fraud, can be minimised in Bulgaria [3.3]. Using an algorithm, the research identified 8,000 registered traders from the Bulgarian tax base, previously categorised as low risk, that were in fact non-compliant.

The initial research [3.1] was also developed to estimate the income Tax Gap in Greece in 2018 in collaboration with the Hellenic Republic’s Independent Authority for Public Revenue (IAPR).

Kotsogiannis’ continuing collaboration with UK’s HMRC’s Tax Gap team, and the new work with the BNRA [3.3], subsequently led to the development and presentation of research which considered how revenue agencies can optimally choose the sample of tax payers to audit, with the objective of narrowing the Tax Gap [3.4].

The TARC research portfolio also led to cooperation with the Rwanda Revenue Authority on an original policy report evaluating its audit strategy [3.5]. This delivered preliminary and provisional estimates of the effect of Rwanda Revenue Authority (RRA) audits on reporting behaviour after one and two years following the end of the whole audit process for Corporate Income Taxes (CIT).