Dear Bernard,

I am writing in response to the Committee’s call for evidence for its inquiry, Governance of Statistics. I welcome this inquiry, and the opportunity it provides to support the Committee’s wider work in helping to ‘create conditions where the public can have justified confidence in public services/government.’

The following submission explains how the UK’s statistical system is mobilising the power of data to help Britain make better decisions. It starts with a brief overview of specific developments in relation to first regulation and then production, before describing a selection of the system-wide improvements that we have seen since the establishment of the UK Statistics Authority.

Ed Humpherson, John Pullinger and I look forward to speaking with the Committee to expand on these points on 19 March and indeed will follow all oral evidence sessions of the inquiry with interest.

Yours ever,
David

 

Regulating Official Statistics

In 2016, the Authority established a more visibly separate regulatory function as the Office for Statistics Regulation (OSR). OSR was built on the foundations established by its predecessor, the Monitoring and Assessment team, especially the process of assessment of official statistics. This assessment process remains the cornerstone of the regulation of statistics under OSR. OSR has supplemented this approach through transformations in three areas: the philosophy that underpins official statistics; regulatory tools and strategy; and the scope of regulation.

OSR published a revised Code of Practice in February 2018, following an extensive process of engagement and consultation. It sets public confidence in statistics as its overall aim. In an age of abundant data, it is essential that users of statistics can have confidence in statistics. The new Code describes three pillars that support public confidence: statistics must be the product of a sound and objective professional process – so an organisation’s processes, governance and systems must demonstrate trustworthiness; statistics must demonstrate appropriate quality for their intended use and not be materially misleading; and statistics must provide public value that deliver useful insights to the users of statistics.

OSR has also created a more flexible regulatory approach, including:
• A greater focus on how families and groups of statistics serve the public good, resulting in a more systemic perspective across the statistics system;
• Clearer explanation of the criteria that guide OSR’s and the Authority Chair’s public interventions on the use of statistics, and a greater willingness to make these interventions proactively as well as in response to complaints;
• Publication of a database of all extant National Statistics and a register of those National Statistics that have lost their National Statistics designation;
• Greater clarity over the range of regulatory tools at OSR’s disposal and how they are used;
• Setting out the need for greater linkage of data through the September 2018 report Joining up data for better statistics;
• More clearly separate website content, along with a more visible brand and logo.

The Authority’s regulatory regime has always focused on the ways in which statistics and data have informed public debate. As a result, the Authority has, throughout its history, been willing to comment on uses of data and analysis that are not formally designated as official statistics by the organisation publishing the statistics.

In a data rich age, OSR has built on this long-standing position. OSR has recognised the risk that the public may be presented with data from a variety of official sources – analysis, forecasts, official statistics, and management information – and that it is not helpful to confine the regulatory regime to a subset of this information. The public should be confident in all the information presented to help inform public understanding and debate.

So OSR has advocated proportionate adoption of the Code’s pillars and principles by Government and others for a wide range of analytical outputs, while continuing to expect full compliance with the Code for publications that are official statistics. This voluntary adoption approach has the potential to lead to a much more consistent set of principles covering a very wide range of analytical publications that seek to inform public debate.

Producing Official Statistics

In October 2014 the Authority launched a new strategy for official statistics, Better Statistics, Better Decisions. This provided a framework for the Authority to meet emerging demands and changes, focusing on the way we work with colleagues, the statistical community and stakeholders.

The National Statistician has led producers of official statistics, across the Office for National Statistics (ONS) and the Government Statistical Service (GSS), in delivering the ambitions set out in Better Statistics, Better Decisions. His focus has been on three priorities for change: economic statistics; contributing to public policy; and building data capability. In each area, we can look back at where we have succeeded, where we are working to deliver fundamental change, and where we would like to do more.

Economic Statistics

Professor Sir Charles Bean’s review set out a compelling vision for UK economic statistics and his recommendations have been enormously helpful as we transform as an organisation. We have made good progress against many of them, in particular establishing an Economic Statistics Centre of Excellence (ESCoE) to provide research that considers the challenges of measuring the modern economy. We have made improvements to statistics on productivity, construction and trade, such as improved flow of funds figures, productivity estimates which are more timely and comprehensive, and the inclusion of VAT data in estimates of GDP. The 2019 Blue Book round of improvements to the figures also marks a significant step forward in this work.

The establishment of our Data Science Campus illustrates our development of the long-term capacity of our workforce. Alongside this, ONS have also expanded our economic capability by increasing the number of economists in our workforce, particularly by creating a ‘London presence’ economics team to improve stakeholder engagement with key users. But responding to the Bean Review is a first step in our transformation plans. We are aware of the increase in demand of faster, more fine grained and relevant statistics arising from Britain’s exit from the European Union and other policy imperatives, and we want to improve the timeliness of our outputs, linking data where possible to provide our users with better, more innovative statistics.

Data Capability

Crucial to our transformation over this period has been the Digital Economy Act, which received Royal Assent in April 2017. This has meant that we can unlock new data sources and deliver more timely and granular data to decision makers. It has also meant we can create a permissive gateway through which accredited researchers can access de-identified data for public good research: the Secure Research Service. Of course, these improvements mean we, more than ever before, must continue to make the need to use data safely, securely and for the public good, one of our highest priorities.

We are aware that we have a role to play in discussing the public use of data, and recently published a revised framework for data handling and security. In addition, we lead within the
cross-government Data Architecture Community, bringing data leaders across government together in discussions about how to best structure and share data for the public good, and joining other cross-departmental groups in driving the data agenda for government.

In March 2017 we opened the Data Science Campus in Newport to explore new data sources and techniques, including visualising the urban forest, analysing port and shipping operations using big data and using mobile phone data to understand commuter patterns. The impact of these has led to the Campus being asked to lead on an audit of public sector data science capability.

We are also making improvements in the skills of our staff: establishing a Learning Academy at ONS, recruiting more apprentices and transforming our digital and technology estate. We have improved the desktop and mobile technology available to all staff and are moving off long-running and high-risk legacy processing systems. At the cross-government level, the National Statistician chairs the Analysis Function Board which is a federative collaboration between a number of analytical professions who deliver research, evidence and advice to a consistent, professional standard.
Staff engagement has improved, with the latest scores in the Civil Service People Survey the highest so far. The scores for the OSR are amongst the very highest in the Civil Service. Within ONS, leadership and managing change has been the most improved category and there are strong scores for being a great place to work and for diversity and inclusion. However, staff engagement in ONS is still well below the levels we aspire to reach and there is an active programme of work to make further progress.

Population and Public Policy

The Government published the White Paper Help Shape Our Future: The 2021 Census of Population and Housing in England and Wales, laying out the Authority’s proposals for the conduct and content of the 2021 Census on 14 December 2018. This details our plans for a primarily online census in 2021. The next major milestone will be the census dress rehearsal in October 2019.

We are also looking beyond 2021, consulting with users on how administrative data can come to the fore of the statistical system and ensure we are ready to make recommendations to Government in 2023 on the future of the census. The latest report on progress towards an administrative data census was published in July 2018. Alongside this, as part of the Census and Data Collection Transformation Programme, we have modernised the IT under-pinning our data collections, with approximately 360,000 businesses now supplying returns online. In addition, we have modernised the IT supporting our social survey field force.

Across the GSS there has been a transformation in the service provided, with a high demand for statistics, data scientists and other quantitative professionals to support public policy and service delivery in departments.

On our key public policy statistics, we have launched cross-GSS groups to further improve the evidence base on migration, crime, housing and income, building on work already done on student migration, violent crime, abuse, cyber-crime and affordability of rental properties; all of which will make increasing use of administrative statistics. For example, the latest update on migration and population statistics was published in January 2019. The establishment of these cross-cutting groups recognises that for most users, the split of statistics across many producers often gets in the way of clear understanding, and by joining up we can provide a better overview of these core issues. This is also true of our analysis of key public policy issues with publications on the ageing society and being 18 in 2018 setting out a much clearer picture of today’s society. All of these have been supported with improved stakeholder engagement through Population and Public Policy Forums and Select Committee appearances.

Better Statistics, Better Decisions: How the system works in practice

The Statistics and Registration Service Act 2007 established a strong, integrated system for building trust in statistics. Trustworthiness comes, however, from how the system works in
practice.

The Better Statistics, Better Decisions strategy reflects many of the themes raised by the Public Administration and Constitutional Affairs Committee in its previous inquiries on statistics. It flagged a growing demand for evidence; implications of our changing society, economy and governance for official statistics; that changes in technology meant more data was available than ever before; and that a better, more accessible online experience was expected by the public.

Together they have helped the UK statistical system drive improvements in a range of areas.

In 2016, ONS launched a new website, transforming the way in which decision-makers access official data. In July 2017, the National Statistician abolished pre-release access to ONS outputs, ensuring equality of access to official statistics published by ONS. In October 2017 the Director General of OSR published guidelines for intervention in public debate, and reviewed activity to date to provide assurance of impartiality, followed by a revised Code of Practice in February 2018.

With increased use of data comes increased responsibility. Ethics, independence and impartiality are critical to building and maintaining trust in official statistics. The National Statistician’s Data Ethics Advisory Committee is establishing principles and caselaw to guide the UK statistical system, ensuring that the access, use and sharing of public data, for research and statistical purposes, is ethical and for the public good.

And of course, perhaps most importantly, the data available to UK decision-makers are improving. The examples that follow demonstrate how the Authority has intervened to strengthen official statistics, and detail specific improvements in both the context of statistics regulation and producers of official statistics.

Crime statistics

In January 2014, as the Committee’s inquiry into police recorded crime (PRC) statistics highlighted concerns about the quality of PRC data, the Authority published an assessment of statistics on crime in England and Wales, which removed the National Statistics designation from statistics based on recorded crime data.

In the period since, ONS has been working hard to improve trust in crime statistics. In the quarterly crime publication, ONS now includes commentary on police force inspection reports conducted by Her Majesty’s Inspectorate of Constabulary and Fire & Rescue Service. This provides transparency alongside the statistics on the improvements being made and how these vary by police force.

The estimates produced from the Homicide Index were re-accredited as National Statistics in December 2016, following work done by ONS in collaboration with the Home Office to publish evidence about the quality assurance processes applied and to demonstrate the quality of this data source.

In addition, ONS is working to give the best possible overview of crime from all available sources. They expanded the survey definition of crime by adding new questions on fraud and computer misuse in October 2015, filling a knowledge gap about crimes that the general population experiences. These were confirmed as National Statistics by an OSR assessment in March 2018. ONS also added a new survey question on abuse experienced as a child, the results of which were published in September 2017.

In November 2017, OSR raised concerns about the Crime in England and Wales: year ending June 2017 statistical bulletin. ONS was encouraged to think about how to present and explain the different sources of crime statistics, and as a result, ONS introduced new commentary on which crime types are well-reported and accurately recorded, and crime types which are affected by changing recording practices.

ONS are working to provide a clear narrative across the Criminal Justice system by collaborating with the Home Office, Ministry of Justice (MoJ), the Crown Prosecution Service and support services. So far this has included publications on domestic abuse and sexual offences, to provide insight into these serious issues which will assist all users wanting to achieve better outcomes for victims. In October 2018, OSR published a blog recognising the steady improvement in the way ONS reports what is happening to crime, describing the last two publications as the clearest yet.

Migration statistics

The need for better data on international migration has been regularly noted by the Authority and the Committee for some years. A changing policy context, alongside the data sharing
powers provided to statisticians under the Digital Economy Act 2017, has offered a welltimed opportunity to reflect on the best way to deliver a population and migration statistics system in future.
OSR has responded to user concerns about these statistics, challenging ONS to develop a clear strategy of improvement. In its reports throughout 2016, OSR highlighted concerns about the differences between official migration statistics (International Passenger Survey based migration statistics) and other sources, especially National Insurance numbers registered to non-UK individuals, and encouraged ONS to bring forward the publication of analysis to explain these differences. It also published a review of the quality of student migration estimates that led to the removal of the National Statistics designation from these statistics. In 2017 OSR initiated a systemic review of how far migration statistics met user needs and concluded that the ONS’s own plans provided sufficient assurance that there was a clear improvement strategy in place. OSR continues to monitor implementation of this strategy, and in October 2018 published a compliance review of how ONS had managed and  communicated a shift from paper to tablet collection of data from travellers to and from the UK.

Working in partnership across the GSS, ONS are progressing a programme of work to put administrative data at the core of evidence on international migration for the UK and on the England and Wales population by 2020. ONS have long acknowledged that the existing approach for measuring international migration – the International Passenger Survey (IPS) – has been stretched beyond its original purpose and that all available sources must be considered to fully understand international migration in future.

Through their transformation work, ONS have already delivered a range of new research and insights into international migration using these administrative data sources. Work into international student migration, using Home Office Exit Checks data, provided new evidence on the actual long-term departure behaviour of non-EU students at the end of their study. The Migration Statistics Quarterly Reports have also been enhanced by including the best assessment of international migration, based on a wider range of sources including Department for Work and Pensions (DWP) and Home Office administrative data. OSR will advise the Authority Board on how far these developments both improve and change the nature of the main migration statistics estimates.

ONS published their latest research into how they can use administrative data to build their future system for producing population and migration statistics in January 2019. This shows the progress made towards a new approach for producing population stocks and flows using administrative data, by bringing more sources together to fill gaps in coverage.

ONS have linked immigration, education, health and income records, exploring how to use these sources to determine the usually resident population of England and Wales and immigration flows to the UK. This includes developing data driven rules, based on registrations and ‘signs of activity’ that can be identified from each data source. This research needs to go further to fully maximise the benefits of administrative data and link across a fuller range of data sources available to ONS, to continue to build an integrated system for measuring population and migration. Key to this work is close collaboration across the GSS, to address key evidence gaps identified by statistical users. ONS plan to publish their next update on this programme of work in spring 2019.

Productivity statistics

Productivity statistics in the UK are going through a period of fundamental change, mainly due to three ‘puzzles’ characterising the UK’s productivity performance. These are: a longstanding gap between UK labour productivity and other advanced economies; considerable dispersion of labour productivity across firms although economic theory suggests this should narrow over time; and lastly a sharp slowdown in the growth rate of UK labour productivity since the 2008 economic downturn, with no significant recovery since.

In its Economic Statistics and Analysis Strategy, ONS committed to improving the quality and scope of productivity statistics, to deliver a world-class set of statistics to support users attempting to address these “productivity puzzles.” Good progress has since been made: ONS now publish much more comprehensive productivity metrics, for 80 industries (up from 24 in 2015) and a new ‘flash’ estimate of labour productivity, 45 days after the end of the reference quarter, which is half the time it used to take to produce. ONS is the only National Statistical Institute to publish experimental quarterly estimates for multi-factor productivity and are therefore among the timeliest in the world.

ONS are also engaging with users to produce a series of microdata analysis pieces looking at the relationship between trade and productivity, ICT and productivity and management practices and productivity, and an interactive benchmarking tool for businesses to compare themselves with averages of their industry and size band. In March 2019 ONS will host a productivity forum to discuss developments and future priorities with users.

ONS continue to focus its efforts on the production of a wider range of statistics, analysis and research to shed light on the puzzles posed by the UK’s recent productivity performance. They are also considering the productivity gap (international comparisons of productivity) and measurement challenges, working collaboratively with OECD.

Health statistics

The UK’s health statistics landscape is complex, with a wide range of producers. NHS Digital, Public Health England (PHE), NHS England, ONS and the Care Quality Commission (CQC) publish the bulk of official statistics in England, and many other organisations publish statistics on health and care across the UK. In November 2018, 156 of the 855 National Statistics across the United Kingdom related to health and care. This also reflects the devolved nature of health policy, with England, Scotland, Wales and Northern Ireland reporting statistics separately; the availability of data from the NHS; and the public interest in the performance and outcomes from the health system. The volume of statistical outputs alone does not guarantee the public interest is well served, and the Authority recognises that users should be presented with a coherent comparable and insightful picture based on the statistics.

In July 2015, OSR’s predecessor published three assessments of statistics on patient outcomes in England: the Patient Experience Survey; the Patient Reported Outcomes Measures; and the Standardised Hospital Mortality Indices. These assessments were followed in October 2015 by an assessment of the overall Patient Outcomes Framework. These assessments fulfilled a recommendation made by the Francis Review into the problems at the Mid Staffordshire Trust in 2013.

Each assessment made some specific points relating to the statistics. But a general pattern also emerged: that there was a lack of user engagement; little insight was provided to help users interpret and use the outcomes information; and the statistics appeared to be designed for expert users within the health and care system, and did not pay sufficient attention to broader public use.

In light of these systemic concerns, OSR convened a round table in February 2016 of senior leaders from the NHS in England, which concluded that the value added by the system for health and care statistics was less than the sum of its parts, because “health statisticians often focus on servicing their immediate policy and operational users, and only within the NHS, with insufficient effort devoted to working collaboratively to address the important issues of coherent and accessible statistics to support public understanding and accountability … statisticians appear to be tentative about engaging with a broader user community.” They found that the health and care landscape is data-rich but information poor; and that a piece-meal production approach was reflected in the fact that health and care statistics can be accessed via a range of websites and portals, but there is no single source to guide researchers or the public to the most appropriate statistics to meet their needs.

Since then, OSR has undertaken strategic interventions to lead improvements, for example, intervening publicly in October 2016 to encourage the Department of Health to provide greater clarity on health funding, particularly on sources, time periods and what is being measured. This led to significant improvements to the way HM Treasury presents health spending in the Public expenditure Statistical Analyses. OSR also intervened on four occasions between 2017 and 2018 on the collection, presentation and use of accident and emergency statistics in Scotland, England, and in comparisons of accident and emergency performance between England and Wales. These interventions led to the withdrawal of new 10 guidance to Trusts in England; a recalculation of performance against the four hour waiting time in England; and revised quality assurance processes and presentation of statistics in Scotland.

ONS, with OSR’s support, has established the English Health Statistics Steering Group (EHSSG) to help ensure coherence across health statistics producers. An interactive “Health statistics landscape” has been produced to enable users of health statistics to get what they need and understand the complete coverage of published health statistics by topic. Moreover, in March 2018, OSR concluded that it should pass on responsibility for convening producers of health statistics to ONS, since this convening is more properly a production rather than a regulatory activity. OSR will continue to drive improvement using a range of regulatory interventions.

ONS health statistics have continued to progress and develop. The production of regular statistics on cause of death, including drug, alcohol and suicides is built on processing and coding of life events data, particularly death records. In addition, ONS have published statistical series on healthy life expectancy and mortality rates which have been critical in uncovering health inequality by deprivation and the levelling off of mortality rates after 100 years of consistent improvement. ONS are now collaborating across government, particularly with PHE and the Department for Health and Social Care (DHSC), to establish the causes behind these patterns.

More recently the Digital Economy Act has provided ONS with the potential to extend health analysis to new sources of data, and improved data linkage techniques and technology to produce more detailed analysis. Examples include publications on student suicides, a new statistical series on the deaths of homeless people, and ONS collaboration with DHSC using census data to analyse the family, housing and labour market participation of those with common mental health disorders and their outcomes after treatment. These outputs have resulted in significant public interest and swift government action.

The Authority is aware that there is still work to be done, and that there are many evidence gaps across the health landscape. A fundamental gap is the lack of consistent, coherent data on social care, which is a focus for the Authority in 2019. There are currently around 150 local authority providers of care, many private sector providers and a huge, hidden, informal care sector. Care and caring is going to be a key policy challenge for many years to come and ONS are working with academics and the charity sector to try and plug this gap. Another evidence gap is around disability and ONS have established a cross government group to understand and address this problem.

Trade statistics

The demand for improved and more detailed UK trade statistics has increased significantly since the EU referendum. ONS note that policy-makers, economists and users needed more
data to help provide a better understanding of the UK’s trading relationship with the rest of the world: specifically, more detail on the goods and services being traded, the sections of the UK economy engaged in this trade, and greater geographic detail of where trade is taking place.

The vital transformation of UK trade statistics is now delivering a breadth and range of trade data, commentary and insights. For example, collaboration with HM Revenue and Customs (HMRC) combined with use of more powerful technology has led to much more granular trade in goods data by country and commodity, consistent with the wider UK’s National Accounts and Balance of Payments. Following a doubling of the quarterly International Trade in Services survey sample size and optimisation of sampling by geographic region, ONS have also expanded publication of trade in services detail to quarterly (from annual) and increased the detail by country and service type. These developments, together with new experimental estimates of trade in goods by industry, have increased the number of published trade series from around 1,000 to over 100,000 in the autumn of 2018.

On the Rotterdam effect (an issue in which this Committee has expressed interest previously), ONS followed the initial analysis and estimation of its effects with further guidance and notes in their trade releases, advising users of this effect on a regular basis. While key, initial discussions with users made it clear that their priorities were for increased granularity and frequency of trade statistics. More recently, however, these users’ next set of priorities now include a better understanding of the Rotterdam effect and its possible impacts. ONS is a member of the Department for International Trade steering group considering research in this area and ESCOE are leading in this analysis.

ONS have delivered wide-reaching analysis of the UK trade asymmetries, providing context and explanation of these. The first focused on the US and the Republic of Ireland, followed by further work expanding on international collaboration and analyses to include Germany, France, The Netherlands, Luxembourg and Belgium. These analyses have highlighted cases where other countries are moving to revised international standards at a slower pace compared with the UK.

Moreover, ONS have provided innovative new tools for users to access and analyse trade data, publishing interactive maps that show 234 countries’ trading relationships with the UK, broken down by 125 types of goods and updated monthly, and further maps showing trade in services by country and type of service.

There is still a keen need for a greater understanding of trade at the local geographic level: how parts of Wales, Scotland, Northern Ireland and the regions of England are trading with the rest of the world. Recent workshops held by the Centre for Subnational Analysis within ONS highlighted even greater need for this information for local enterprise partnerships (LEPs) and combined authorities in designing their strategic economic plans and local industrial strategies. In October 2018, alongside other new trade releases, ONS published updated regional exports of services estimates.

ONS are now researching methods and investigating new data sources to continue expansion of trade in services statistics, and planning to publish experimental estimates of trade in services by industry in the first half of this year. Development work on exports of services at the regional level will continue, alongside further analysis of trade in services asymmetries, trade in value added, digital trade, modes of supply of services and further methodological reviews of data processing.

Education statistics

Statistics about children, education and schools are among the most prominently used in public debate and can also be an important influence on the choices made by families about schooling. It is essential that these statistics are presented and used in line with the Code of Practice. Education statistics are mainly produced by the Department for Education (DfE), and the respective Education departments in Wales, Scotland and Northern Ireland. The Higher Education Statistics Agency, ONS, Ofsted and Student Loans Company also publish related figures that contribute to the evidence base. OSR made a series of interventions involving DfE statistics during 2017 and 2018. These included formal letters published on the Authority’s website, and informal discussions
between OSR and the Department’s Head of Profession for Statistics during the summer of 2018. These informal discussions were based on OSR’s monitoring of statements made by
the Department which draw on official statistics or used statistical evidence. As part of these discussions, OSR queried some aspects of statistical presentation with DfE.

In October 2018, the Authority Chair wrote to the Secretary of State for Education with serious concerns about DfE’s presentation and use of statistics. The Director General for Regulation wrote at the same time to the Permanent Secretary and the Head of Profession for Statistics.

Following this intervention, both the Secretary of State and Permanent Secretary publicly reaffirmed their commitment to the principles of the Code of Practice, and DfE put in place revised procedures for the development and agreement of communication messages using statistics. OSR continues to monitor the use of statistics by DfE and to emphasise the need for appropriate communication of statistics to the Department.

This work on education statistics in England demonstrates that OSR actively monitors and provides feedback to producers of statistics; and takes reasoned judgements to consolidate cases, rather than deal with them one by one, where they judge that there is a pattern of misuse of statistics.

Homelessness and hunger statistics

Better Statistics, Better Decisions notes that ONS will provide a firm evidence base for sound decisions, and so reacting to key policy demands effectively is a priority. Examples of policy areas that require better data include homelessness and hunger. The Government set a commitment to halve rough sleeping by 2022 and eradicate it by 2027. While estimates of the numbers of homeless people are available, along with an annual estimate of rough sleepers, there is currently little systematic data available to help inform policies to prevent homelessness and monitor and evaluate initiatives to relieve homelessness. There is even less data available on rough sleeping.

The regulatory function assessed homelessness statistics for England published by the Ministry for Housing, Communities and Local Government (MHCLG) in 2015. It concluded that the presentation of the statistics, as three separate statistical reports with no coherent narrative to draw the statistics together, or to place them in context, diminished their value. MHCLG have since put in place a programme of change that is leading to more comprehensive homelessness statistics. OSR has removed the National Statistics designation from the existing homelessness statistics.
In view of the increased policy focus on homelessness, the GSS is undertaking work to understand wider data and evidence, which is often generated by charities and third sector organisations who work directly with homeless people and those who sleep rough. The aim of this work is to establish how such data could build better intelligence for local government, helping them identify risk factors and trends specific to their local populations. In addition, ONS also developed a new method to estimate the number of deaths of homeless people and published the first report in December 2018.

Furthermore, the data asset Homeless Case Level Information Classification (H-CLIC) has recently been established in England and has the potential to be linked with wider Government data sets, such as benefit receipt, income, health, crime and justice, family formation and break up, as well as characteristics such as age, ethnicity, education and nationality. Collaborative work across ONS, MHCLG and potentially DWP, MoJ and DHSC to anonymise and link this data, in order to build an understanding of underlying factors that lead to homelessness and rough sleeping, could help lead to identifying effective interventions at the local level.

There is also a growing demand on the Government to improve UK statistics on food security and hunger. Emma Lewell-Buck MP’s Food Insecurity Bill calls for ONS to address her concerns in relation to the lack of statistical data on the issue, alongside the specific requirement to measure food insecurity as part of the Sustainable Development Goals. ONS have conducted a comprehensive review of relevant existing data from both official and non-official sources and are looking at options to fill any data gaps. Following meetings with Emma Lewell-Buck MP and an appearance at the Environmental Audit Committee, the next step is chairing a round-table on food insecurity and the measurement of SDGs Goal 2 with all key third sector and government  takeholders in late February 2019 to fully understand all the data requirements before determining next steps.

Inflation statistics

The importance of, and interest in, inflation statistics is recognised by the Authority. This interest has recently increased due to the Lords Economic Affairs Committee (EAC) inquiry
considering the use of the Retail Prices Index (RPI), but it has been a key statistical issue that the Authority and this Committee have considered for many years.

A range of inflation measures are required to meet current and emerging user needs, but it is also important to ensure that the statistics present a clear and coherent picture. To this end, ONS set out three ‘uses cases’, relating these to the measures that are currently published and those that are under development. Taken together, these present the ONS approach to measuring changes in prices and costs faced by consumers and households: a comprehensive measure of inflation, based on economic principles: the Consumer Prices Index including owner occupiers’ housing costs (CPIH) which is also the ONS headline measure of inflation since March 2017; a set of measures to reflect the change in costs as experienced by households: the Household Costs Indices (HCIs); and a measure that is required under s21 of the Statistics and Registration Service Act: the RPI.

The Statistics and Registration Service Act requires that the Authority publish the RPI each month. In practice, ONS, as the Authority’s executive office, compiles, maintains and publishes the RPI. The current treatment of the RPI has its origins in the 2012 consultation on the future of the RPI. Having analysed the responses to that consultation, the National Statistician’s decision at the time was to leave the RPI unchanged.

The approach that arose from this consultation was to develop alternative, statistically robust measures and encourage their use, which was also one of the conclusions of the independent Johnson Review of consumer price statistics in 2015. The CPIH and its close relation, the CPI, are designed to capture the economic approach to the changing prices of goods and services, and HCIs are intended to measure households’ experiences of changing costs. The CPIH regained its National Statistics status in July 2017, and the first experimental HCIs were published in December 2017. Implicit in this approach was that use of the RPI would decline as alternative measures were adopted.

The Authority has long stated that the RPI is not a good measure of inflation, and that we do not encourage its use. The OSR’s predecessor also removed its National Statistics designation in 2013 because the methods used to produce the index were not consistent with recognised best practices. However, the Authority notes the EAC’s recommendations and agrees further steps are needed. We are currently considering a number of options and will keep this Committee, and the EAC, informed.