Briefing Note (BN218)

Who does and doesn’t pay taxes?

Date: 26 October 2017
Authors:
ISBN: 978-1-911102-64-9

This IFS Briefing note uses data from HMRC’s random audit programme to show which types of people are more likely to be under-reporting taxes and how their behaviour changes after a tax audit. The results are based on data from audits covering tax returns for the years 1999–2009.

Key findings

  • Every year, 10 million people file income tax self-assessments, around a third of all UK taxpayers. Using a sample of these who were randomly audited, we can learn about non-compliance among all self-assessment taxpayers.
  • More than one-third of self-assessment taxpayers (36%) who were randomly audited were found to have errors in their tax returns that led to underpayments (‘non-compliance’). The average additional tax owed by the non-compliant taxpayers is £2,320, equal to almost one-third (32%) of the initial tax amount they declared
  • HMRC has improved its targeting of non-compliant taxpayers, increasing the revenue raised per targeted audit conducted. However, the decline in the number of such audits reduced the total revenue raised from audits from a peak of £869 million in 2003, to just over a third of that by 2009.
  • Most non-compliant taxpayers (60%) owe less than £1,000. A small number of taxpayers – less than 4% – owe more than £10,000, but they account for 42% of the missing revenue.
  • Men are more likely to be non-compliant than women (40% versus 27%). However, among both men and women who are non-compliant, 32% of total tax owed was not declared.
  • Non-compliance is greater among working-age individuals, at around 40% of those below state pension age (SPA), compared with only 21% of individuals above SPA.
  • In part, this reflects the difficulty in under-reporting pension income compared with under-reporting self-employment income. 59% of taxpayers declaring only self-employment income were found to be non-compliant, compared with only 16% of those declaring only pension income. Surprisingly, 29% of those declaring only employment income were non-compliant, despite such income being also reported by employers.
  • The likelihood that a taxpayer is non-compliant does not vary substantially with income. The cash amount of tax not reported by non-compliant individuals averages around £2,200 across all but the individuals with the top 20% of incomes. For those in the top 20%, with incomes above £47,270, it averages £3,530.
  • Non-compliance is highest in the construction, transport and hospitality industries, where more than half of self-employed taxpayers were found to be non-compliant. In hospitality and transport, more than half of total tax owed was not reported.
  • Random self-assessment audits recover an initial £830 per audit on average, since one in three audits yields a return and the average return among those is £2,320. Audits also bring in income for at least five years after the audit, by changing taxpayers’ reporting behaviour: being audited leads taxpayers to declare more income subsequently. This raises another £1,230, one-and-a-half times the initial gain.
  • The longer-run gains come from the additional information HMRC acquires about audited taxpayers’ incomes. As this information becomes more dated, taxpayers again reduce their tax declarations.
  • A reasonable estimate for the cost of an audit might be around £2,500. If correct then this would be slightly more than the total revenue raised from a random audit. But targeted audits can raise significantly more than they cost. For example, audits targeting those with the top 20% of reported incomes would have an average total yield of £10,870 per audit.