The impact of annual summaries, texts and apps on behaviour

• Context
• The study
• Key findings
• Points to consider

Context

There are approximately 65 million active personal current accounts in the UK. These accounts generated £8.1 billion revenue in 2013; an average of £125 per account. In 2014 it was estimated that 94% of UK adults have at least one account, with 40% having two or more. While the importance and penetration of the current account market is irrefutable, there have been major public concerns about how well it works. Two of the main concerns are the low transparency of overdraft charges and the low levels of switching between personal current accounts. In 2008 the Government recommended annual summaries including charges to consumers, while in 2012 text alerts were available at all major UK banks. Over the same period, mobile banking apps were also rolled out.

The study

This 2015 report from the Financial Conduct Authority, in conjunction with the Office of Fair Trading, looks to understand the impact of annual summaries, text alerts and mobile banking apps on customers. To infer the impact of the initiatives, monthly data was analysed (using advanced econometric methods) from customers of two major UK banks in the UK between 2011 and 2014. Account level data on a representative sample of 500,000 customers over 30 months from one bank (Bank A) was obtained, along with aggregated results over 36 months from another bank (Bank B), allowing the authors to cross-check and validate their results.

Because annual summaries are sent out on a rolling basis, some people receive them before others. This allows for a natural experiment that makes it possible to analyse their impact by comparing the behaviour of those who receive them early to those yet to receive them, while controlling for other factors.

To infer the impact of text messages and apps from customers of Bank A, the authors compare the behaviour of those who do sign up for these services to their own behaviour before doing so, along with the behaviour of those who do not sign up.

For Bank B, the authors use a spike in sign-up rates after the release of the mobile banking app as a natural experiment to estimate the impact of it. The switching behaviour or the effect of text alerts could not be analysed for customers from Bank B due to limitations of the data.

Key findings

• The annual summaries have no effect on unarranged and arranged overdraft charges, balance levels or switching to other current account providers (though there is a small effect on internal switching of a 1% increase per year).
• Text alerts reduce monthly unarranged overdraft charges by 6% (£0.22) for Bank A.
• Mobile banking apps reduce monthly unarranged overdraft charges by 8% (£0.33) for Bank A and by 5% (£0.23) for Bank B.
• Signing up to both services reduces monthly charges by 24% (£0.93) for Bank A, with an additive, combined effect greater than the sum of the individual effects of each service.
• Those on higher incomes and in the 40-59 age group see the largest reduction in unarranged overdraft charges.
• The mobile app at Bank A increases monthly easy access savings balances by 7%, though the text alerts have much less of an effect.
• The findings also showed that switching rates systematically decrease with age and income.
• The authors state that the strong combined impact of alerts and apps shows the benefit of receiving timely information through automatic triggers.
• They also suggest that future regulatory work might consider the possibilities of:
  • Targeting annual summaries at people with financial issues;
  • Rules to help those with difficulty understanding their finances such as making alerts ‘opt-out’;
  • Understanding which aspects of technological innovation are most helpful;
  • Thinking how to ensure that incentives for innovation are sufficiently strong.

Points to consider

• Methodological strengths/weaknesses: The methodology used appears sound and the findings are tested for statistical significance, so we can have a fair degree of confidence in the robustness of the findings.
• The authors acknowledge that it is not possible to control for the fact that those who sign up to alerts might do so as part of a deliberate effort to change their behaviour in relation to managing their account. However, they state it is unlikely that this is the case for a large number of the sign-ups during the study.
• Generalisability/ transferability: This report is of significant interest to stakeholders and policymakers in the field of consumer financial behaviour, and of particular relevance to those studying the effects of nudges and interventions in relation to consumers’ current account activity and the charges they incur.
• Relevance: The findings are based on consumer data from two large UK banks. While not necessarily representative of the entire consumer market, it is quite likely that the general trends are typical of other large banks in the UK.