No half measures: using tax data to monitor jobs and pay better

Access to new sources of data is helping the ONS to produce quicker and more detailed economic indicators. Now, working with HM Revenue and Customs, we have for the first time jointly published estimates of employees and earnings based on monthly real-time tax information. David Freeman explains what these new data give us and how they fit into the ongoing transformation of UK economics statistics.

In January last year I blogged about the initial work that HMRC had done to produce quarterly estimates of employees and earnings from real-time tax data. Today sees the first monthly joint publication, using robust methods to provide new insights into jobs and pay.

The story so far

Since 2014, all employers with a Pay As You Earn (PAYE) tax scheme have been required to report to HMRC every time an employee is paid. The large number of records and the detail the system holds means the data can be analysed by a number of dimensions, such as sex, age and area of residence to give much more detail about changing employment and incomes.

In January 2018, HMRC started publishing aggregate data each quarter using this PAYE Real-Time Information (RTI) data. These releases provided statistics on the number of employees and average earnings for the constituent countries and regions of the UK.

What’s new this time?

This month we are moving from quarters to a monthly publication model, as well as removing the effects of seasonality in the data.. While this may not sound like a complicated change, in fact moving from quarterly to monthly data posed two tricky problems given that weekly paid people will have four payments in some months and five in others, and that some payments are made after the period when the work is done.

To tackle these issues, HMRC have worked in collaboration with ONS specialists to produce methods that allow us to better allocate pay and employment to the month where work was carried out and a also to impute the details of earnings that aren’t available when the data are prepared.

As well as these improvements to methods, HMRC have also worked with the ONS to adjust the data for seasonality, which takes out some of the regular peaks and troughs in the data caused by regular events such as large bonus payments in the early part of the year. This means that the published figures can be used to calculate short term changes as well as changes on an annual basis.

What’s next?

Today’s release has been a genuine collaboration between HMRC and the ONS, using expertise from both Departments to produce new data that provide further insight on the labour market in the UK. The regular statistics produced by the ONS remain the headline measures of the labour market as they are consistent with international standards and have been assessed as compliant with the Code of Practice for Statistics. However, these new experimental statistics show the potential of administrative data to provide greater detail and insight into the labour market.

Over the coming months HMRC and the ONS will be working together further to improve the data and the level of detail that can be made available, such as, smaller geographic areas).

The ONS will also be investigating how RTI data could be integrated with survey data to improve the timeliness and coverage of some of the main labour market indicators. This will help us better meet our users’ needs and expectations in the future.

David Freeman is Head of Labour Market and Households at ONS