Today is my last day as Chief Statistician after just under ten years in the job (depending on how you define my starting date, in true admin-data style).
It’s been an absolute privilege to have held this post for so long through some fairly momentous times in recent Welsh history. It is fundamental that statistics put out by government are accurate, useful and trustworthy – to ensure the right decisions are being made, and government can be scrutinised effectively. As Chief Statistician I’ve tried my best to ensure that statistics produced in Wales meet those high standards. I have aimed to be as open and transparent as possible, working closely with the people who use our statistics and keeping people informed – including through the introduction of this blog a few years ago.
It’s also been a privilege leading such a fantastic group of intelligent, enthusiastic and helpful people. Each of our small teams of statisticians work tirelessly to balance a huge number of conflicting demands. Not only do they need to hit publication deadlines week in week out (after all if we don’t meet those deadlines we have to publicly explain why!), but also collect the data, provide support to policy officials, and answer hundreds of queries from Ministers, the Senedd and the public all year round. I wanted to take the opportunity to thank them publicly for all their toil and hard work over the years. The work we do – be it the production of national well-being indicators, the collection of the first trade survey for Wales or the annual collection of vast amounts of data from schools, farms, people, helps shine a light on what is going on in Wales.
COVID-19 has highlighted the strengths of good practice in presenting statistics
Never has data been so important and we’ve all seen throughout the period of the pandemic the challenges that have arisen due to the demand for almost real-time operational data on what is happening in the country. This has simply reinforced the importance of the Code of Practice for Statistics and the values of the Government Statistical Service (GSS) in providing trustworthy data, transparently, and with data quality issues clearly explained. On a number of occasions during this period Ministers and officials have asked for my advice on how we can ensure data related to the pandemic are available in an authoritative and transparent manner – which is a recognition of the strength of what we do in the GSS.
The teams have risen to this challenge by working in a virtual and agile way, learning swiftly about new datasets which needed to be published in days rather than weeks or months. As a result we’ve done the best we can to try and help paint the picture of what is happening in Wales, and to help those producing the operational data in using it and publishing it.
The pandemic has also emphasised some of the statistical challenges that my successors will now have to try and make progress on – so I wanted to just leave a few reflections on these challenges.
The need for inclusivity in our data is stronger than ever
The report of the BAME Advisory Group on Covid-19 and the Gender Equality Review published last year have highlighted the importance of collecting and publishing data which allow us to better understand outcomes for different groups of the population. We are already trying to do this – we need to continue with this work as part of our wider response to those reviews. This includes ensuring we are capturing good quality equality data wherever possible and considering how we can link datasets together to re-use data already held by government.
But this can’t be done by statisticians alone. We’re very reliant on data captured through interaction with public services so we will need to drive up the quality of data throughout the system. It’s important to understand why this doesn’t always happen. What are the barriers for public services in ensuring they have high quality data? What might stop citizens – and sometimes particular groups of citizens – providing their equality data to public services? We will need to work with public services and different groups of the population to communicate the long-term value of capturing comprehensive data. The prize is clear – with better data, we will understand far more about the experiences and outcomes of all sections of our population.
Data on Welsh citizens should be available for Welsh research
Our work on using administrative data for research here in Wales has been UK-leading. Through the work of Swansea University on the SAIL databank and the Administrative Data Research – Wales partnership we are linking together, securely and ethically, data from different topics to better understand the outcomes and impact of interventions. For example through the course of the pandemic we have undertaken research to understand how many children are living in shielded households, and how many teachers are shielded. As a result of our efforts in this area we have drawn significant amounts of research funding into Wales.
But we could so much more. We are working with colleagues across the UK to ensure that researchers in Wales are able to access data about people in Wales, even where it is not devolved, so as well as understanding health and education outcomes we can use the wealth of data on society and economy held by UK departments to help us develop policy. We have seen progress on this in the last couple of years with positive commitments from the Ministry of Justice and the provision by ONS of 2011 Census data to researchers in Wales to support our response to COVID-19.
But getting access to data for research is still slower than we would like, and this has been noted in reviews over the years by ESRC and the Office for Statistics Regulation. I want to encourage all departments to realise the benefit of access to data in a secure, robust and ethnical manner that will allow us all to produce high quality research that will benefit Wales and the rest of the UK.
Where is the user need for “England and Wales” data?
Another issue the COVID-19 pandemic has highlighted is the historical challenge of getting statistical producers, and the media, to present data about Wales separately and not as part of “England and Wales”. There are historical reasons – because of some functions not being devolved, or the way the data landscape has evolved. I have regularly pressed for my fellow statistical producers in government to make available data about Wales easily to users in Wales and, more importantly, to tell the story about Wales separately and not as part of “England and Wales”. Regardless of whether or not a function is devolved, users in Wales want to understand what is happening in terms of society, the economy and crime and justice in Wales.
Even if the data is buried in a spreadsheet somewhere, if the key messages of a statistical report talks about “England and Wales”, this is what the media will report.
Our Welsh Parliament, stakeholders and civic society need to have clear messages from our statistics about what is happening in Wales to scrutinise the actions of Ministers, and to understand what interventions can be taken. Our public services need to plan on the basis of trends in Wales, not an “England and Wales” statistic.
This is not simply a challenge for our fellow producers of statistics. The media are also too often happy to present data on “England and Wales” even where data is easily available broken down into individual countries – the presentation of mortality data by some media outlets during the crisis has been a good example of this.
Whilst this is my last Chief Statistician blog, since I’m moving to become Chief Digital Officer for Welsh Government this certainly won’t be the last “Digital and Data” blog you’ll hear from me! As CDO I will do my best to be open and transparent about the work we are doing to ensure that Welsh Government embraces digital and on the role it will play in the COVID-19 recovery work. I would very much welcome hearing any feedback on how we are doing, and what priorities should be for Welsh Government and the wider public services.