Briefing: HMRC’s analysis and research work

What is HMRC’s tax research agenda?

In simple terms, we use primary research to improve the evidence base on tax policy and administration. For our administrative work, key themes might be understanding our customers, maximising voluntary compliance and dealing with non-compliance. For instance, we are undertaking research into the deterrent effect of the pilots for our evasion publicity campaign. On the tax policy side, we do modelling and empirical work to appraise policy options and evaluate the impact of potential policies.

As the department evolves, I think that a lot of the boundaries between these traditional research areas will begin to blur. We’re already seeing this in customer service and compliance, and I think we’ll also see it more generally in policy and administration. Good policy has to take into account its impact on customer service. I’m not sure that the traditional distinction between policy and administrative research works anymore.

What is the underlying rationale for this work? And how does it serve both HM Treasury and HMRC?

It’s really important that we understand how what we do affects the behaviour of our customers. This has always been the case, but it becomes all the more important as we move towards a more digital future. We must look at the best way to design our digital services to maximise compliance, improve customer service and reduce our costs.

Research and analysis is not just important for tax administration. If tax policy isn’t based on robust evidence about behavioural effects, we can get our forecasts wrong or distort people’s behaviour in undesirable ways. Research has an important role to play in improving both the design and the operation of the tax system.

How is the work organised? How does HMRC work with third parties?

We do a lot of in-house research and analysis. This covers behavioural trials, policy costings and everything in between.

We also commission external research partners to independently collect information. This helps us to understand our customers' behaviour, to monitor how it changes over time and to assess how our programmes to collect tax in the most efficient way work in practice. This external research covers a very broad spectrum, from theoretical modelling by academics to large scale surveys of taxpayers and smaller qualitative studies.

We commission this work in two ways. For research to answer specific questions, we go through the competitive tendering process, using our research procurement framework. We also facilitate independent academic research, and there are two main routes for this: the Tax Administration Research Centre (TARC), a joint partnership between the University of Exeter and the Institute for Fiscal Studies, through which we co-fund research on tax alongside HM Treasury and the Economic and Social Research Council; and the HMRC Datalab, which accredited researchers can use to gain carefully controlled access to anonymised HMRC customer data when their research requires it.  

One of the new things that TARC has allowed us to do is to undertake lab experiments to test different ways of administering the tax system in a controlled setting. It gives us invaluable information on the design of our forms and on different strategies to encourage compliance, which we can then take to the next stage of piloting with our customers.

What data is made publicly available?

For obvious reasons, we only make anonymised taxpayer level data publicly available where it is legal and where it protects our core business while maintaining the public’s confidence.

Where researchers do make use of taxpayer data, we insist on this being done in the Datalab. All outputs are checked to make sure that they don’t disclose taxpayer information.

All of the research we fund is published. Directly commissioned pieces of research are posted on GOV.UK; and all TARC and Datalab work is published in academic journals and on university websites as part of the scientific process. Like other researchers, we might not always agree with the views expressed in the TARC or Datalab research, but our knowledge and evidence base is enhanced and that can only be a good thing.

Are there any implications which tax advisers should be aware of?

A number of our individual pieces of research will be of interest to tax advisers. Last July, we published research that we’d carried out with the smallest agents to inform the development of a new online self-service tax system for these agents. In December, we published research segmenting the agent population according to their digital readiness.

There’s more research in the pipeline specifically focused on the tax agent population, including some descriptive work looking at trends in the market for tax agents and how small and medium-size business customers choose a tax agent. I’d encourage people to keep an eye on GOV.UK.

I imagine that our more general publications will also be of interest to tax advisers, insofar as our research informs HMRC’s direction of travel. A lot of our work will feed into improvements in the way the tax system is run. Some of this will obviously directly affect tax advisers, but any changes would be consulted upon as usual.

Of course, the biggest game in town for us is the design of the new digital tax system. A lot of our research will inform this work and will help to refine and evaluate alternative ways of delivering digital services. The ultimate aim is to improve the service that HMRC offers to individuals, businesses and agents alike.

Where are we headed? What plans does HMRC have for developing its analytical capacity?

We need to plug the biggest gaps in our evidence base. This means that we need to get better at working with external academics and bringing them in at the right stage. We need to be more open to external viewpoints that might challenge how we’ve traditionally done things.

I think lab experiments will become increasingly crucial as the first step in testing different digital designs and compliance processes. That’s already an important part of the TARC programme and I can only see this growing in the future.

The next stage is to move into more ‘test and learn’ practices, where ideas are piloted with groups of taxpayers to predict their potential impact. Our behavioural change team is involved with several trials of this type. For example, it has looked at how the design of debt recovery processes affects repayment rates. Of course, we don’t intend to test everything, but we aim to understand the most important design decisions using live trials.

The other thing to think about is whether we can join things up across government departments to design and test different ways of delivering public services. More consistent data sharing by departments like HMRC, the Department for Work and Pensions and the Ministry of Justice would open up huge avenues for researchers and would transform how we design public services. It’s no secret that HMRC is supportive of proposals to amend the existing legislation to make this possible and I think this will continue to be our view.

Research and analysis will only become more important to our work, as we design our new digital services. If we are to move towards a win/win situation – where we are increasing voluntary compliance, while reducing our costs and improving the service for our customers – a solid evidence base will be absolutely crucial.