Tapping into new data sources

Jonathan Athow explains how hard data from government and business is helping ONS to produce better economic statistics.

The ONS produces a wide variety of statistics describing the UK’s economy and society.  Most are based on large-scale surveys, often the biggest of their kind regularly undertaken in the UK.  For example, the Labour Force Survey takes data from around 100,000 individuals in 40,000 households each quarter to provide estimates of unemployment, part-time working and average earnings for the population as a whole.

Increasingly there are new sources of data that can help us understand our economy. Every day, huge amounts of new data are collected by businesses and the public sector as part of their normal activities. In the public sector, the Government regularly collects information from individuals and firms, for example to administer the tax system. At the ONS we are not interested in individuals or firms. Rather we want to aggregate that data together to provide fresh insight into the UK’s economy. In some cases they could even replace some of our surveys, reducing the form-filling burden we place on businesses and people across the country.

…The prize is more timely , detailed and reliable statistics, and the better decisions that will be based on them…

To exploit these new data sources, one of the first projects we have undertaken is turnover data that companies report to HM Revenue and Customs as part of VAT system.  This gives huge amounts of information that can be used to provide a measure of economic output.

On 4 October we published aggregate VAT turnover data for the first time. This is a major step on the road to including VAT data in our regular statistics on economic output, which we intend to do next year.

One advantage of VAT data is the sheer volume of data available. Our usual Monthly Business Survey covers around 45,000 firms, while VAT data will give us over 1.75 million firms. This means we will be able to give a much more detailed breakdown either by industry or geography by using VAT data. For example, our Monthly Business Survey includes around 250 companies involved in publishing, too small to allow for a full regional analysis. On the other hand, we have around 10,000 firms in the VAT data, which should allow for a fuller breakdown of the publishing industry by country and region.

While new data sources like VAT data are very welcome, it is important to invest time and effort in understanding the data. How well does it track the underlying economic concepts we are interested in? How timely is it? What is the underlying data quality? What data cleaning or imputation is needed? Only then will you know how effective the new data source will be.

It is also vitally important that all data provided to the ONS is properly protected. The ONS and its predecessors have a long history of guarding personal information. Much of the survey data we currently collect from individuals and businesses is sensitive. We have carefully designed systems and processes to help ensure that all data – survey data and new data sources such as VAT – are fully safeguarded.

These protections are also reinforced in the Digital Economy Bill, currently before Parliament, that will give ONS much better access to a wide range of information already provided to other government departments.

The revolution in data and digital technology has hitherto been driven largely by the private sector.  Now public organisations like ONS are becoming better equipped to take advantage. The prize is more timely , detailed and reliable statistics, and the better decisions that will be based on them.

Jonathan Athow is Director General for Economic Statistics at ONS   

Feedback to the editor.