Chapter 1 Introducing well-being data
Why the book is written in this order?
This book is a game of two halves, with a post-match pint to digest what we have just watched: the performance of the players and those calls which are on the edge of the rules of the game. The first half is about how different kinds of well-being data (data about well-being) came about. It begins with the historical traditions of philosophy, governance and social science that led to ‘well-being data’ becoming a thing that is useful and looks at the methods, innovations, contexts and limitations of these.
The second half looks at how well-being data are relied on as evidence in social and cultural policy, also how they are used to answer questions beyond the contexts they were collected in. Ideas of a cultural society as a good society have long-shaped social policy and informed future philosophy. We look at how this enabled cultural policy to become an aspect of social policy, before presenting a number of case studies on the relationship between well-being and culture that I have elsewhere  called the culture–well-being relationship.
The conclusion aims to be a sort of post-match pint down the pub. It reflects on moments of tension, recapping on what has happened and reflecting on how these might be understood from a different position. We end with trying to understand ‘understanding’ in a number of ways. First, as the ways we understand the world, through data, information, knowledge and wisdom.6 Second, as a reflection on the work that needs to be done towards a shared understanding of data. Third, how in using well-being data, we may become more understanding of each other.
The First Half
We start by setting up some of the background story to well-being data. Chapter 2, ‘Knowing Well-being: A History of Data’, puts the concerns of this book into context, these contexts being historical, political and technical. There are different theories of well-being from different times and places, and how these are understood today by researchers, national statisticians and policy-makers affect what data are collected to understand well-being.
We look at the project of measuring well-being as one that wanted to understand how to improve human welfare. We also consider well-being as a tool of policy, as the very idea of it is used to make arguments for one policy decision over another. Or in more real terms, to fund one social project over another. This is deeply connected with developments in national politics and governance, which changed and increased the role of economics in auditing, efficiency and valuation. We consider how these processes led to not only more well-being data, but more well-being data practices. In other words, more uses of more data. This chapter will help the reader think more critically about why and how well-being became such a default ‘good idea’—and some of the issues at play here. It will also help think about how striving for a good society became inextricably linked with well-being data.
Chapter 3, ‘Looking at Well-being Data in Context’, moves more specifically into thinking about the uses of data and measurement in policy, practice and research. The previous chapter’s historical focus on measurement as an expression of objectivity and governance is extended here. This chapter is a more focussed appraisal of contexts in which data are collected and used. We think about the role of methods and methodology (explaining what this word means). We look at specific examples of how well-being is measured and how that maps onto philosophical accounts of well-being. This is not a methods textbook, as there are plenty out there that do this job. Instead, this chapter’s focus on context, difference and limitations across mundane, critical and authoritative contexts aims to help us think about how we might understand well-being better, or differently.
Therefore, we think about the implications of different kinds of data, starting with how they are collected. Well-being data can be collected in various ways: through administrative processes, such as the recording of births, marriages and deaths, or crime-rates. These data will be used as quantitative data, to understand and develop measures we see in the press, like ‘mortality rate’. Quantitative data can also be collected using surveys that allow understanding of more complex aspects of people’s lives. Asking people questions means you can know how long it is since they visited their GP (general practitioner), for example, or how far they have to walk to their nearest children’s play area. These data are easily turned into numbers to give a picture of how people’s lives compare, or how we are doing overall, and can help governments make decisions about how to allocate resources.
Data collected in questionnaires and online surveys can also be qualitative, as can interviews, diaries and observations. Qualitative data are most generally text-based, and so are good to understand how people have described their experiences or opinions; although can also involve image or sound, for that matter. Using qualitative data can allow researchers to understand the complexities of a situation and the specificities of people’s personal lives. While quantitative and qualitative approaches tend to be discussed separately, some data collection methods, such as surveys and questionnaires, collect both quantitative data (by ticking a box) and qualitative data (by a free text field), so surveys are able to gather data that offer a bigger picture and more detail at the same time.
Qualitative data often have lots of rich detail about few people in a specific context that have to be interpreted by the person analysing it.
Quantitative data will have been collected so they can be quantified, removing contextual detail for analysis using numbers and comparison across a population. Somewhat confusingly, if you have enough qualitative data, you can quantify them, but this is less common and we look at how and why that can be useful sometimes. While quantitative data also require interpretation, there are standardised mathematical approaches, usually drawing on statistical methods to support these decisions and analyses. This means quantitative approaches are considered to be more neutral and objective. But as we shall see, lots of decisions are needed, and this poses key questions about the idea of objectivity in the data used to make statements about what is good for society and to make arguments that one thing over another will improve well-being.
Chapter 3 is the first chapter where we start to look under the bonnet of well-being data. At some points we get up-close to specific research examples and ideas, including quotes from focus groups and examples of well-being survey questions in an imagined context of evaluating a local community event. We also look at so-called objective well-being indicators (e.g. mortality rate) that feature in well-being metrics, like the OECD’s Better Life Index. We ‘follow these data’ using qualitative data in reports to think about how objective these measures really are. We will reflect on the distinction between objective well-being, as something experts decide is important to well-being, such as an aspect of health, and subjective well-being measures which involved asking people how they feel. All data and ways of using them have pros and cons, which is why context is important. Understanding how different data work in different contexts is key to well-being data and key to data for well-being.
‘Discovering “the New Science of Happiness” and Subjective Well-being’ is the title of Chap. 4. Here we consider the formation of happiness as something that can be measured. Happiness is part of a broader academic concept called ‘subjective well-being’—as an idea of how well-being is felt. Subjective well-being becomes extremely influential in the well-being agenda and we look at the role that these new measures hold. The chapter begins by describing how ‘happiness’ became a ‘new science’ including the different academics, politicians and fields of study involved. It describes the evolution of positive psychology and happiness economics and their influence in the realm of policy-making.
Disciplines like psychology and economics often group subjective well-being data into different types. They refer to evaluation, experience and eudaimonic measures. This chapter does the same to explore what these mean in practice, and how they are used or useful to understand specific aspects of the human experience, which is then used in evidence for policy-making. Again, specific examples of the contexts in which these sorts of data are collected and used reveal their limits, as well as contradictions in their use. We then focus on subjective well-being measures in the UK and the Office for National Statistics’ (ONS’) Measuring National Well-being Programme.
Looking at the invention of subjective well-being measures in the UK offers context behind the ubiquity of well-being measurement practices. Understanding the recent history behind, and breaking down the different ways of measuring a particular idea of well-being, is vital to appreciate the limitations of such projects. While the innovations and limitations of well-being data remain unaddressed, their positive contribution for society can never be fully realised. This chapter’s comprehensive survey and critical lens aim to offer tools to promote better understanding of subjective well-being and happiness data, their capacity to change culture and society, and the limits of their application in areas of social and cultural policy and practice.
Chapter 5 looks at Big Data, which is an enormous topic to try and cover in one chapter. ‘Getting a Sense of Big Data and Well-being’ asks many questions, beginning with: what do we even mean by the term?— how are data big? The amount of data on individuals that is now collected is quite simply mind-boggling. The International Data Corporation (IDC) predicts that by 2025, the total amount of digital data created worldwide will rise to 163 zettabytes . That is 1021 (1,000,000,000,000,000,000,000 bytes) or one trillion gigabytes. The European Commission forecasted the European ‘data market’ to be worth as much as €106.8 billion by 2020 . We can therefore see that not only have the amounts of data increased, but their economic value has as well. It is, therefore, even harder to maintain that all uses of well-being data enable neutral decisions about how society is managed, when it is being called ‘the new oil’ .
We begin by asking the question: ‘What even is Big Data?’ We look at what the term means, as well as what Big Data are and what they can do, including how as soon as someone tries to define it, somehow that definition is not quite right. Emergent technologies from all walks of life are producing and collecting and analysing data about us as we move about the online and offline world. This means that more can be known about people—which we discover means that data are a double-edged sword for well-being.
Big Data are often attributed with much power—by those in favour of their use, and those who actively work to limit the negative possibilities of these new data and how they are used. The chapter demystifies Big Data by putting them into historical and a number of practical contexts. For example, smaller organisations, in the arts and social sector, use data mining in small, mundane and often unobtrusive ways . It is possible to use data in research like this in a way that is ethical and without much software skill or financial resource. We revisit a practical example of a manageable project I undertook to reanalyse Twitter data using a hashtag that was started by a Mass Observation project to understand what makes people happy. As a spoiler, there are many cats.
Mass Observation was a project originally established by an anthropologist, a poet and a filmmaker in 1937 who wanted to record everyday life in Britain. The project emerged at a time where there was a desire for more detail in data, and around the same time as social surveys were becoming more complex to understand more detail about people’s everyday lives, particularly around World War II. More data were wanted to understand quality of life and manage populations beyond the administrative data collected on mass-scale, like the census.
Most countries now undertake a census of sorts, and in the UK, the ONS have collected its census data every ten years since 1801. The new ‘enthusiasm for numbers’ in the early to mid-nineteenth century  coincided with a growing infrastructure to collect and analyse data. This desire for numbers, and the data processes that were required to provide them, led to the ‘great explosion of numbers that made the term statistics’ . In this ‘avalanche of numbers’, ‘nation-states classified, counted and tabulated their subjects anew’ . Censuses date back far farther, of course, and the ONS’ website offers an interesting history of censuses in the UK, back to the Domesday book ordered by the Norman (French) King, William the Conqueror in 1086 . Again, censuses precede these European data moments by some 4000 years in both Egypt and China, who recorded who lived where how wealthy they were. The Romans held regular censuses to keep track of their expanding—and then contracting—empire. Further back still, the clay tablets of Sumerian script  might be considered a dataset of Big Data from 6000 years ago. The promise of Big Data is therefore not new.
We look at the promise of Big Data to predict a pandemic, reflecting on the obvious failings of Big Data to forecast COVID-19’s impact in a way that could have averted international crisis. We also look at a company that claims to have predicted the pandemic, yet failed to stop it: is it possible that the commercial value of the intelligence they had was a barrier to more effective global prevention? We start some years before that, in 2009 with the failings of Google Flu Trends (GFT), which promised to beat the slow infrastructures of health services and testing in the US. GFT analysed what people searched for on Google, analysing what, where and when people typed symptoms into the search. Yet, this did not work for a number of reasons tied to a lack of capacity to understand context.
Back in the UK, I took part in a home testing programme that the media said would ‘clear up [the] “Wild West” of Covid-19 estimates’ . In what has been called the ‘largest testing study for Coronavirus’ , tests were posted to you, using the UK’s traditional Royal Mail postal system. That all worked fine for me, but there were a series of steps registering different barcodes and I found myself wondering how accessible this was for everyone (when I say everyone, I often think of my once tech-savvy Dad, who’d have been bewildered at this whole process). As a result of these steps, a courier was ordered to collect the test, but failed after three attempts (that I describe in more detail in the chapter). A neighbour told me in passing that this particular courier company was infamous for not bothering to try and collect from my high-rise flats, probably because the buzzer has never worked and it can take too long for a resident to come down. This looks bad for the drivers’ performance data, which are meant to encourage them to make as many deliveries and pick-ups as possible.
In my case, while some aspects of the traditional data infrastructure (the post) worked fine for this COVID-19 data collection research, they didn’t necessarily all work together as they might. This meant that my test remained uncollected; therefore my data became ‘missing data’. Thinking about the contexts in which data are collected (or not) can be both extraordinary and mundane, and we often don’t hear of these stories— when they work, and the odd occasion when they don’t, and what that might mean for the data.
We follow other case studies of data from mobile phone usage, social media data and tracking apps, for example. We, again, ‘follow the data’ and how they are used to interpret whether these data projects are primarily concerned with improving human well-being, or with refining data practice. It is crucial to problematise the ethics of Big Data for well-being, particularly their commercial aspects, rooting these in the larger questions of what data can do more generally and the limits of data for understanding well-being or improving well-being.
The data we look at in the first half of this book are either all collected to better understand people or society, or have been analysed to do so to enable a government or a company to make better decisions. There is a sense that these data are all neutral—they are not affected by bias and can all be treated as fact. These chapters reveal the fragilities in the assumptions behind these kinds of data. When you consider the hypothetical and real-world examples, you can see lots of humans mainly doing their best to work with data. We can also see mistakes in the systems and analysis, and therefore, some of the data-driven decisions we live with are not the best decisions they are assumed to be.
The fact that data have real-world impacts and implications is not something that is often made clear by those who use data, or advocate data-driven decision-making. The impact of Big Data has seen an increase in those considering their social effects. Consequently, the negative aspects of data are an issue of government agendas with new emphases . However, the ways that data about people make the problems of society legible are not necessarily new, and neither are the problems. Data on residents, together with a map produced by the City Office of Statistics of Amsterdam, enabled the rounding up of the city’s Jewish population under Nazi occupation in 1941 . Yet, the same techniques of mapping people and personal data about them also led epidemiologists to identify how the AIDS pandemic was spreading and of course the current COVID-19 crisis.
We need context to understand data practices and the possible ramifications of their social effects. They have their own ‘social life’ , meaning they might be thought of as living in that they act on the world as much as humans do. Data and numbers ‘make up’ people  and tese later theorists enable us to think. Decisions are made about our lives without asking us, but looking at how we are represented by data. Data decide whether you will get a commercial loan or access to financial support by the state. Postcode data in the UK will decide how you will receive medical treatment and what drugs you are entitled to. Data hold much power through metrics  and algorithms . But also, the very idea of data is powerful; it affects our day-to-day behaviour. Crucially, however, it is also in the desire for data where its power lies.
The Second Half
We ‘switch ends’ in the second half. The goal instead is thinking more about how society has increasingly required well-being data. So, while we do not entirely leave thinking about contexts of data collection, we think more about the contexts in which they are used. We continue to focus on how society works, its relationship to governance and decision-making, and the role of data in this. Given that data are social and cultural, we will, therefore, look at areas of social policy, focussing on cultural policy in particular to make comparisons more readily across some simple arguments about well-being that use data. To be truthful, it is also in looking at data in the cultural sector and in cultural policy that I came to understand data, and is my natural data habitat.
Chapter 6, ‘Well-being, Values, Culture and Society’, provides an overview of how cultural policy became a form of social policy, specifically looking at the role of well-being. The chapter historicises the idea that particular aspects of culture have a social role and are good for well-being using accessible interpretations of key philosophers from Aristotle to Kant. We reflect on the fact that much like population data, the arts have an honourable and dishonourable history , as both have been co-opted for political projects, such as fascism: that didn’t just damage well-being, but were almost indescribably catastrophic for people and society. The chapter brings these empirical accounts of uses of culture into play with social theory from cultural studies scholars, including Raymond Williams . These later theorists enable us to think through some assumptions around the role of culture, even what gets to be called culture, and why that is a problem for cultural and social policy. In turn, we are in a position to contextualise how the institutions and historical assumptions that decide what is good culture, and manage cultural policy, are not so different from thinking about the institutions that manage data and the way we work with and understand data. These overlaps are rarely acknowledged.
We reflect on a genealogy of the idea that culture (broadly defined) is good for well-being (broadly defined); how that has been naturalised over time and then popularised. By this I mean, there is a generally accepted view that culture is good for well-being, and we look at the lineage of this idea as something that began with philosophers and is now common sense. We will then investigate how this relationship has been instrumentalised as a form of social policy. This involves looking at how culture is used as a means or ‘instrument’ for attaining goals in other areas of society. Examples of this can be found in policy documents, research agendas and in practitioner movements including ‘arts in health’  or the use of culture in urban regeneration projects . The idea that the arts can be used to directly address societal problems has led to arguments that culture is—in fact—instrumental to these social policy areas.
The idea that arts are instrumental in delivering broader social projects and improving social infrastructure has been operationalised to advocate for funds for the arts. We have, therefore, witnessed changes in the value of culture from something belonging to everyone , to how much social impact it can demonstrate, or indeed financial estimates of the creative industries . In return for advocating the value of culture, the sector is increasingly required to evaluate how much of this value it has generated in response to funding, or to argue for more funds.
This has also seen the slippery nature of culture and its definitions be instrumentalised in arguments, where one meaning of culture is used to justify another aspect of it. The benefits of culture as something more everyday  are used to justify the funding of art-forms which are considered the opposite of commonplace in that they are elitist, with often small numbers of people interested in participating (opera being the default perpetrator in this argument). This slippery effect is also used when it comes to ‘creativity’ and arguments surrounding the economic impact of the arts, where ‘the arts’ become ‘the creative industries’, including some professions in IT, which in many cases do not seem to be very creative at all—in the way we would normally use the word.
We have, therefore, seen a process in which the culture–well-being relationship is theorised (through philosophers) and become naturalised in people’s day-to-day thinking: making it common sense. Figure 1.1 shows the full journey of processes described in the chapter. The common-sense nature of the relationship is operationalised in policy and instrumentalised to argue the value of the arts and culture to other areas of social policy. This process, however, has led to the cultural sector finding itself in a bind to the burden of proof. It has to evidence the social impact of the work it does, which is a costly exercise of data production and analysis.
These shifts in the culture–well-being relationship have seen the value of data increase and become capitalised on . The increase in funding saw an upturn in evaluations required to report back to funders. With this came demand for data and data practices that are often outside of the skills and confidence of many working in the cultural sector, and broader areas of social policy. These skills therefore often need buying in from elsewhere. With the newer forms of well-being data introduced in the first half of this book, come new metrics and valuation tools, which are presented as a solution to issues of advocacy and proof in the sector. They also perpetuate this cycle of funding and evaluation, which preserve this process of instrumentalising, operationalising and capitalising on the culture–well-being relationship. We will therefore look at some examples of how well-being data are used to make arguments about culture—and we will follow the data in different ways to see how they work.
Chapters 7 and 8 draw from the framing in Chap. 6 to look at how the culture–well-being relationship has been operationalised in research to provide proof. Chapter 7 is called ‘Evidencing Culture for Policy’. It takes three fundamental arguments about the culture–well-being relationship- that are used in advocacy and looks at them more closely. The first is that culture warrants funding, because it is good for well-being. We look at a number of different examples of data to establish if a relationship between public funding and well-being can be found. Again, through investigating the contexts of data collection and analyses, we are able to think about the limits of what can be known using these data.
Why are well-being data in demand to understand some relationships and not others? Despite the naturalised belief that we should invest in culture for its well-being benefits? There is little research which explores whether a pattern can be established between increased funding and well-being. Why are some questions repeatedly asked and not others? Is this a matter of the data (what can be known) or the limits of what people want to know?
We look at the question of ‘how much is culture good for well-being’ in more detail. The chapter considers two pieces of research which investigate the well-being of cultural practitioners and creative professionals who are often presented as similar, even the same, population. The two studies ostensibly use the same approach to analyse survey data to understand this culture–well-being relationship. In comparing these two cases, we unpack differing findings and look at limitations of data, in categories, populations and analyses, and question how they help us understand well-being in this instance. Crucially, this is not necessarily a case of comparing studies to see if one is better than the other. Instead, we look at how asking (at least superficially) the same question using similar data about similar people at comparable points in time does not present the same results. So what does this mean for ideas of evidence?
The final section looks at a piece of research that is found in important and high-profile reports as evidence that culture is good for well-being. The article uses what it calls ‘data mining’ to understand ‘cultural access’. We look under the bonnet of this idea of cultural access and the data that have been used to measure it. We also follow the authors’ data mining practices and analyses to find combined variables which change the meaning of the category ‘cultural access’, resulting in an inflated outcome.
Unpacking the different ways that culture has been packaged as something that is good for people and society is important. In this chapter we discover how particular findings become popularised as ‘common knowledge’ and how they then become operationalised in reports, the media and policy documents. This is crucial to grasping the idea that the relationship between data and evidence is cultural, and relies on practices, understandings and meanings.
Once we begin to question the social value of generating evidence in this way, the economic value of contracting in well-being data and research practices warrants investigation. In Chap. 8, ‘Talking Different Languages of Value’, we follow a piece of research that was commissioned to help with advocacy for the arts. The commissioners were an organisation called the Happy Museum, and the research was funded by Arts Council England. Building on the work we have done in previous chapters to understand how data work in contexts , we look at how culture and well-being are operationalised in this study, and walk through the processes, step by step.
The chapter opens with this idea that this book seeks to challenge: that the arts and data speak different languages. Breaking down what is happening, we follow the data in various ways. There is a description of how the data were collected in a national-level survey. We look at the questions, as they appear in a survey, because it can be hard to imagine the mundane contexts that data originate from, when you are looking at the complex results. We follow the data forward, to see how key findings are interpreted by the world. This allows us to ask questions like: what does research do? How does it affect the world or change things?
We follow the conceptual work behind what is being measured before reflecting on some of the steps in the analysis. There was another way that these data were followed, as I was part of a research project to reproduce findings, using details on the processes and the data available. Crucially, the second piece of research arrived at different conclusions from the first. What does that mean for the very idea of ‘evidence’?
How does commissioning well-being data analysis to support the arguments people want to make change the nature and role of evidence in different social policy areas? How does this affect overall knowledge of ‘what works for well-being’ in terms of social policy? Importantly, how does ‘capitalising’ on well-being data affect their capacity to do social good or to be good data? Do the economic value of data and their analysis change the relationship between well-being data and a good society? We have found indications that this is the case with COVID-19, but is this more generalisable?
Chapters 7 and 8 break down various aspects of how data are used in cultural policy to communicate quantitative expressions of well-being to people who lack confidence in these areas. Crucially, this will enable readers to think about how something that is described as culture or cultural is said to impact on well-being, whilst also looking at the limits of the data we have to make such claims. These chapters aim to encourage you to make your own mind up (with a little help) as to whether everything adds up (not just the numbers). Do the arguments make logical sense based on the evidence we actually have, rather than what we are told we have? How can considering the contexts of data help those working in data and working in social policy do more good with data? History tells us the dangers of ignoring the good and the bad that can be done with data, and that how it is used is a matter of culture.
The final chapter is simply called ‘Understanding’. Here we will reflect on different ways of understanding well-being and different ways of interpreting data. We will look back on how well-being and data are related by way of policy and politics. We consider the relationship between evidence and policy, and the politics of data. How do these conflicting ideas work together when the aim of the game is well-being?
We reflect on how understanding contexts of data helps us better understand the politics of data and evidence for policy. We look at the limitations of well-being data that we have explored in terms of claims that can be made and we look at their limitations when it comes to calling data objective. The huge amounts of decisions involved in establishing the well-being measures in Chaps. 3 and 4 show these are not neutral decisions. Furthermore, Chaps. 7 and 8 reveal the decisions made in modelling: what data to clean, weights and adaptations to valuation techniques when well-being data are used to make arguments about value.
We think about what understanding means. It means understanding as knowledge, shared understanding of how something works and being understanding, or having empathy. Well-being data promise information that leads to knowledge and wisdom, but these do not currently lead to a shared understanding. Research is commissioned for the cultural sector and presented in ways preoccupied with proof, rather than communicating findings with those who work in the sector.
The concluding chapter presents a case study of how people crave understanding of why they are being asked certain questions on equality monitoring forms, what will happen to and with the data they offer. Yet, it is not common practice to share understanding of how and why different data are valuable. There is much room for understanding and empathy in approaches to inequality and well-being data, and this is currently overlooked in most projects that work with these data in the name of social justice.
The ‘social life of methods’ is a body of research proposing that methods are not neutral ways of capturing an objective reality, but have their own social effects; in fact, changing the reality they claim to capture. Data: how it is collected, shared, analysed and where the results are published are a fundamental part of this. We have looked at how data are cultural, in that they change culture, making new cultures, and we look at the implications of these social effects. Those who are campaigning for data rights are very focussed on what can be known about people from data. However, this is often framed as an issue of privacy as an abstract human right or as an issue of social justice, as the effects of data-driven decision-making disproportionately affect marginalised groups. This, of course, is an important ethical question.
A broader question, however, is what can these data actually tell us about people? There are limitations to most data when it comes to what we can actually understand about society that are not always taken into consideration. Crucially, the question we must ask ourselves at this moment is how can we also rethink questions of what can be known about people from data to incorporate data’s limits, as well as their power? How might well-being data improve well-being? Can we be better at moving from understanding people as units of analysis to becoming more understanding in the way we collect and use data?
These are the provocations this book leaves us with and I hope to continue to do work that not only tries to answer these questions, but which goes about changing things. This book is set up so that we can look at the work that well-being does in policy and practice contexts for social and cultural policy, for third sector organisations and arts managers, for charities. Most of all this book is meant to help us all have a better grasp of ideas of well-being and ideas of data, how they work in different contexts and how they are used and manipulated for different ends. Neither are neutral. They are imposed by historical traditions which say what works and what doesn’t. They are imbued with values—and I hope this book will help you value your own judgement to decide what they mean for you.
- Oman 2015a, 2015b[↩]
- We look at the idea of eudaimonia in greater detail in Chaps. 2 and 4. Most simply, eudaimonia means feeling purpose, or flourishing.[↩]
- Coughlin 2018[↩]
- Ram and Murgia 2019[↩]
- The Economist 2017[↩]
- Kennedy 2016; Oman 2013[↩]
- Mass Observation is a project that has long aimed to record everyday life in Britain. More detail can be found on the different phases of the overall project and its smaller projects, here: http://www.massobs.org.uk, and in Chap. 6.[↩]
- There were a number of iterations of Mass Observation (n.d.), with different people initiating them, but the original founding members were anthropologist Tom Harrisson, poet Charles Madge and filmmaker Humphrey Jennings.[↩]
- Hacking 1991, 186; Porter 1986, 1996[↩]
- Porter 1986, 11[↩]
- Hacking 1990, 2; 1991, 186[↩]
- ONS 2016[↩]
- Harford 2017[↩]
- Devlin 2020[↩]
- Ipsos Mori 2020[↩]
- DCMS 2020[↩]
- Scott 1998, 77[↩]
- Beer and Burrows 2013; Oman n.d.[↩]
- Hacking  2002[↩]
- Beer 2016[↩]
- Kennedy 2015[↩]
- Belfiore and Bennett 2008[↩]
-  1971, 1977,  1989a,  1989b[↩]
- ACE 2007; AHRC n.d.; AHSW 2019[↩]
- DCMS 2004; LGA 2020; UNESCO 2018[↩]
- Hall 1977; Keynes 1945[↩]
- Campbell 2019; DCMS 2011[↩]
- Williams  1989a[↩]
- Oman and Taylor 2018[↩]
- see also Oman n.d.[↩]