‘There is so much good data out there that each topic could be its own book’

Stuart Newman’s new book, Britain by Numbers, is an ingenious portrayal of a nation through data and statistics. It takes a broad look at many aspects of British life from religion to recreation, health to employment, and most places in between.  The Southville-based author and Economic Advisor at the Office for National Statistics (ONS) tells Bristol 24/7 what data can reveal about who we are.

When did you first get become interested in statistics and data?

I think we all have an inherent interest in statistics, even those who don’t like numbers or maths. The key is a salient topic.

For example, after hearing that approximately 31,000 gay people in the UK are married to someone of the opposite sex, we can immediately picture someone in that group. We may even judge or empathise with them and their partners.

But it’s only when you work with data that you really delve into data quality. That’s the point where I go into super-nerd mode. I love to think about how the data are put together – how was the data collected (if it’s a survey then what questions were asked), who compiled the stats and why. I consider those to be key considerations and the answers are often as fascinating as the headline numbers, yet few of us look into the data we encounter to this kind of depth.

………………………………………..

Bristol24/7 relies on your support to fund our independent journalism and social impact projects. Become a member and enjoy exclusive perks from just £5 per month.

………………………………………….

What can we learn from statistics and data that we can’t from anything else?

Raw data provides invaluable insights that are impossible to achieve in any other way. And these data are important for making decisions – for government, corporations and in our own lives.

However, data has inherent problems. Firstly, any aggregation, computation or data analysis requires statistical judgements that can be challenged. A recent example that made the news relates to immigration data from the Office for National Statistics (ONS). The ONS underestimated the number of EU citizens who stayed longer than they intended and overestimated the equivalent for non-EU citizens. The ONS have to account for people who answer survey questions incorrectly, maybe because they changed their plans, but these estimates can be incorrect so the data have to be revised when we receive better or updated information.

The second problem is that data are difficult to communicate well. If you look at wage data, one of the key indicators is Average Weekly Earnings (AWE – again, published by the ONS). Yet it’s incorrect to assume that movements to AWE are due to pay increases for existing employees. They are just as likely to be due to the type of jobs becoming more prevalent – more software engineers and fewer retail posts will probably lead to higher AWE figures. That kind of nuance is impossible to fit into headlines.

Thirdly, we tend to consider all the data we see as the number (or maybe we disregard it entirely). But most data are estimates. They’re usually the best estimates we have, but they’re still estimates. This is poorly understood, which is a shame.

Stewart Newman’s Britain by Numbers will be launched at Foyles in Cabot Circus on November 7

How much material did you leave out of the book?

Tons! There is so much good data out there that each topic could be its own book. It would be great to do something on gender in the UK. And it would be fascinating to do something on ethnicities. Although there’s an issue with the latter topic: as soon as you break ethnicities down into different groups the samples sizes get smaller and smaller, so unless we have data for the whole group, the data become less reliable. I had to be conscious of this when using ethnicities data for the book.

You say in the introduction that “It has been an eye-opening journey”. What surprised you the most when putting the book together?

There are so many topics in the book that I investigated for the first time. I was surprised how detailed our house price data is and disappointed by our lack of equivalent rental data. I was shocked by the cancer statistics (grouping cancers together, they account for more UK deaths than I imagined) and there are differences between ethnicities that were unexpected. For example, Bangladeshi women are less likely to be either in work or looking for work than other women in the population. I’ve since heard the Cabinet Office is doing some research to understand why this is the case.

On the downside, I found it difficult to compare some data across the UK. Data for England and Wales are often collected by different departments to data for Scotland and that of Northern Ireland (look out for this in newspapers where data often refers to England or England and Wales, as opposed to the UK). The collection offices sometimes use different methodologies and collection methods, although there are efforts to harmonise these data. For example, there are efforts to create a UK-wide view of homelessness. These are important projects that we should all support.

What do you hope readers will take from the book?

The book has two elements: the images and the text. In the text I’ve tried to pull out interesting elements of the data, including how it was collected and compiled. I’ve aimed to impart some knowledge about how to interpret statistics, so I hope that readers will come away with a greater respect for the data they see around them and additional skills in thinking about what those data are actually saying. But I recognise that some people will just want to flick through the pretty images and look for interesting titbits. And that’s OK with me.

Stuart Newman’s Britain by Numbers will have a publication day launch at Foyles in Cabot Circus on November 7. For more details, visit https://www.foyles.co.uk/Public/Events/Detail.aspx?eventId=4071

Read more: Jane Duffus publishes a second volume of The Women Who Built Bristol