The use of automated algorithms is disrupting industries across the world and tax administration is certainly no exception.
We at Inland Revenue have been using versions of this technology for years to make sense of large volumes of data and to pick out trends in information such as taxpayers’ filing and payment behaviour.
The big difference these days is that the technology is performing these tasks much faster and is organising the data in such a way that allows us to make better decisions more efficiently. It’s a lot more effective than the ‘Ctrl F’ keyword search that many of our compliance officers will remember using in the past.
What hasn’t changed is that real human beings control the process and are central to the technology being a success. So let’s quickly dissolve the myth about robot tax investigators busting tradies for not declaring cash jobs – it just doesn’t happen and never will.
Our people will continue to make all the decisions about when to open a case for investigation and how to handle the data and insights that are produced.
Tony Morris, customer segment leader at Inland Revenue
Here’s an insight into how it works in practice: a smart data search might tell us that a group of small businesses may be underreporting income, compared to other businesses of their kind. Our first assumption will be that we have a group of businesses that need some help to get their taxes right. It might be that they don’t know how to comply or perhaps they need some advice on keeping good business records.
The next step is to try to help them out. It might not provide a great plot line for an action movie but this is the approach we prefer – we just want to help customers get their taxes right from the start.
Of course we’ll take action if a customer doesn’t want to comply or it becomes evident that they haven’t but this only happens after we exhaust all forms of assistance.
Our analytics technology helps ensure the data is telling us what we need to know, which is how we can help a customer stay on top of their tax obligations and receive the social policy payments to which they are entitled. Data is only ever used for these purposes and there are strict rules surrounding how we treat it, especially when it’s passed on from other agencies.
Algorithms are a part of any data analytics operation but at Inland Revenue they are used to help construct that story about the customer. Algorithms don’t make any decisions, only inform them. The ones IR uses are predictive in nature, programmed and checked by real people and the privacy of taxpayer information is always paramount.
The information we glean from analytics technology helps us make targeted interventions, which are designed to prevent the customer from running into further problems.
A big part of modern day tax administration is about turning the huge volume of data that comes to us digitally every day into useful information. Analytics technology gives us the ability to sort through that quickly and solve problems.
As it turns out, we are using technology to help customers far more often than we use it to make a case for auditing them.
Tony Morris is a customer segment leader at Inland Revenue
Media contact: Baden Campbell 029 8901674