The webs of belief around 'evidence' in legislatures

A wide-ranging literature has explored the relationship between research, knowledge and policy. However, legislatures have often been overlooked in this research. While some studies have looked at ‘who has access’, the literature on how parliaments seek to engage with knowledge claims is particularly scarce. This article addresses this gap through a case study of UK select committees. By adopting an interpretive lens, the article explores how MPs and officials make sense of evidence in committee settings. It finds that legalistic definitions around ‘evidence’ shape wider beliefs in how to engage with knowledge claims and the practices of undertaking inquiries, and are underpinned by a distinctly political function of knowledge use in politics. Beliefs around evidence have significant repercussions and highlight tensions (i) around the authority of committee reports, (ii) between epistemic and democratic claims, and (iii) with respect to who is included and excluded.


Introduction
The relationship between science, knowledge and policy continues to be a fertile area of debate and scholarly research. Issues including the distrust of elites and expansion of nationalist movements have questioned the established relationship of expertise and public administration (Nichols, 2017), while most recently the global health pandemic of Coronavirus/Covid-19 has shone a bright spotlight on policy-makers' use of scientific advice and knowledge claims. It is in this timely context that this article explores the role of 'evidence' in the principal democratic arena of most polities: legislatures.
Parliaments are uniquely placed in the debate about expertise, public policy and democracy because they regularly juxtapose diverse political claims by MPs and the wider public with those of experts through parliamentary debates and committee inquiries. In this arena, representatives make use of a diverse set of knowledge claims, including lived experiences of constituents, reports from interest groups, and research published by universities, as well as information from in-house research organisations such as the Parliamentary Office for Science and Technology (POST) in the UK, the Wissenschaftliche Dienste (research services) in Germany, or the Congressional Research Service in the US. Knowledge claims are pervasive, and MPs make use of it for a range of different reasons, whether instrumentally for policy learning or symbolically to bolster arguments (Boswell, 2009(Boswell, , 2018. Furthermore, the legislative arena is important in discussions about expertise because they are almost always the single-most important representative political institution in a polity. So, legislatures are key to law-making and policy processes, who scrutinise legislation and give legitimacy to political decisions. However, and with some exceptions discussed below, research on the interaction between knowledge and legislatures is sparse when compared with a much larger literature on evidence use in public administration, government and state bureaucracies. And while executives are important given their role in decision-making roles, legislatures' unique position in the political system means they face distinctive issues and challenges with respect to knowledge claims, including the representative and accountability functions that do not preoccupy executives in the same way.
In this article, I examine the webs of belief that surround 'evidence' in parliamentary contexts through a case study of select committees in the UK House of Commons. Specifically, I argue that the beliefs around political balance and breadth of evidence are privileged over ideas of their quality or weight. This is a consequence from a legalistic definition around 'evidence' in the legislature, which constrains how evidence might be recorded. The beliefs and practices around evidence use are important for three reasons: first, for the authority of a committee and its policy claims; second, for opening a tension between democratic and epistemic claims to make policy; and third, for ongoing debates about who is included and excluded. This article is structured in the following sections: first, I summarise what we know about the use of evidence in public policy and parliaments; second and third, I explain the theoretical framework that underpins this study, followed by a summary of methods; fourth, I examine the role of evidence use through my case study of UK select committees; and finally, in the conclusion, I discuss the broader relevance of this article for understanding the relationship between science, knowledge and policy in legislatures and beyond.

The relationship between science, knowledge and parliaments
Studying the relationship between science, knowledge and politics is not new, with philosophical debates stretching back centuries. Today, the virtues of making policy based on evidence are extolled by a range of national governments and international institutions (Lavis and Panisset, 2010;OECD, 2015), whereby 'evidence' can be as broadly defined as an argument backed by information, but usually has connotations of scientific-based knowledge (i.e. underpinned by research and a recognised, rigorous methodology). However, while a significant literature on the topic exists (see Boaz et al., 2019;Oliver and Boaz, 2019), the dominant focus of research has focused on governments and state actors, such as the civil service, arm's-length bodies or streetlevel bureaucrats (e.g. LSE GV314 Group, 2018;Boswell, 2009). Meanwhile, legislatures have been somewhat neglected. The oversight may be partly explained because, traditionally, parliaments have not been seen as key to decision-making, with many noting a shift towards executive power since the mid-twentieth century (e.g. Andersen and Burns, 1996;Richardson and Jordan, 1979). However, these arguments underplay the continuing centrality of parliaments in the political process. Notably, they are crucial legitimating institutions for policy decisions, in which legislatures not only serve to debate political issues but also where laws are passed and policies are scrutinised. Most of the executive governance and public administration literature has arguably underestimated the influence of such processes on decision-making through the often informal yet persistent scrutiny by MPs through a range of parliamentary processes, committees and groups (e.g. Blidook, 2010;Russell and Gover, 2017). Furthermore, as explained in the introduction, legislatures regularly invoke a diverse number of claims to knowledge and are supported by a range of research organisations both within and outside of the institution. In this context, legislatures deserve greater attention in order to understand how their use of evidence shapes policy and political debate.
One important area of research regarding legislatures focuses on committees, which are often seen as fundamental to improving the efficiency and effectiveness of legislative activity (Martin, 2014) and act as information-gathering tools for parliamentarians (Geddes, 2020). So, understanding the link between committee work and knowledge claims is a central way by which legislative power can be explained. There are some pools of literature that contribute to answering such questions.
First, some studies have looked at the institutional factors that shape interest group access to parliamentary committees, finding that better-resourced groups with economic power (e.g. business groups) play a disproportionate role compared to others (Binderkrantz et al., 2015;Chaqués-Bonafont and Márquez, 2016;Eising and Spohr, 2017;Pedersen et al., 2015). Meanwhile, a related pool of research has focused on social diversity of committee participants, often finding that men dominate legislative arenas (Bochel and Berthier, 2018;Geddes, 2018;Rumbul, 2016).
While there is a growth of academic interest in who has access to the legislative arena, research has been less forthcoming about how the knowledge claims are mobilised or what function they play in decision-making. There are some exceptions, especially from the UK. For example, John Turnpenny et al. (2012) find that chairs and advisors of the Environmental Audit Committee in the House draw sharp boundaries between preconceived notions of 'science' and 'non-science', privileging the former as more credible, while Christina Boswell (2018, pp.98-120) finds that committee scrutiny focuses on transparency and publication of reliable information in order to validate trust in government. Elsewhere, one study explores the perceptions of witnesses in committee inquiries, where the authors found that witnesses are largely positive of the process and accusations of grandstanding and bias by parliamentarians are exaggerated (LSE GV314 Group, 2020). Finally, and beyond committees, others have looked at the role of evidence in the legislative process (Crewe, 2017) and the use of specifically academic evidence in Parliament (Rose et al., 2020).
What all this tells us is that, on the whole, there are interesting emerging findings on evidence use in parliaments, particularly in the UK (for a comparative example, see Hendriks et al. (2017)).
However, there are many fruitful opportunities to further develop research in this area. This is especially important given the democratic challenge of declining trust and satisfaction with democracies and the epistemic challenge of the entwined relationship between science and policy.
Legislatures remain at the epicentre of these challenges.
This article contributes to the above literature on how different groups interact with legislatures by asking: what does 'evidence' mean to MPs and officials, and what factors affect the way that evidence is used? So, rather than looking at which groups knock on the door of parliaments, this article looks at what happens behind that door before, during and after it is opened. The focus of the article is around select committees of the UK House of Commons. This is a valuable case study for three reasons. First, the Westminster parliamentary tradition has been exported around the world, which has shaped the legislatures of Commonwealth countries and beyond (Kumarasingham, 2013). Today, the UK Parliament is often compared to other legislatures from this tradition, with many parliaments leaning on each other for best practice innovations (Rhodes et al., 2009). Second, there is a lively debate in UK parliamentary studies about Parliament's role in shaping policy decisions, including the impact of committees (Benton and Russell, 2013;Russell and Gover, 2017). While this article does not directly shed light on the effectiveness of the legislature, it does give a greater understanding of committees' ways of working -something which has been understudied. And third, select committees (as explained below), are often perceived to be influential based on their evidence-led processes. So, this case study has wide appeal in helping us to understand the interaction of expertise and knowledge. I now explain the theory and methods, before outlining the political logic and institutional context of select committees and the main empirical findings.
3. Theoretical framework I focus on how parliamentary actors interpret the role of 'evidence' in their work and, consequently, the practices to which those beliefs lead; i.e. the 'local knowledge' of elected representatives and parliamentary officials (Yanow, 2004). I draw on a rich and varied literature of anti-naturalist approaches to the social sciences (Bevir and Rhodes, 2003;Bevir and Blakely, 2018). While this has led to many epistemological and methodological debates, these arguably go beyond this paper's core contribution. I briefly identify the merits of interpretive approaches and its capacity to answer the research questions before explaining the methods used.
The central starting point for my approach is that social reality is constructed through human activity, a point which has been articulated by many constructivist and interpretive scholars (Wagenaar, 2011). This means that a central task for social scientists is to explain the meanings of social life and, therefore, examine the beliefs, values, passions and interests that shape individuals' behaviour. To give an example: elected representatives' attitudes to the concept of 'accountability' necessarily affects how they go about in holding government to account. This might include detailed policy scrutiny in committees or through media appearances; or it might focus on either shedding light on government policy, i.e. making it explanatory, or focusing on changing government policy, i.e. making it outcomes-focused. Depending on representatives' interpretations, this affects accountability relationships in legislatures and evaluations of success.
So, ideas can have an explanatory function in political analysis.
No single belief can be understood in isolation. They are part of wider 'webs' of values, ideas and interpretations that relate to one another to make sense of the world. People draw on these as wider organising perspectives, to make cognitive shortcuts or to act as guides for interpreting events (Bevir and Blakeley, 2018). To return to the above example of accountability, it is often put into conversation with other beliefs about transparency, good or responsible governance, and as part of a healthy democracy, all of which have their own definitions, meanings and challenges (Bovens, 2010;Page, 2010). In order to understand accountability, we need to understand other concepts in these 'webs of belief'.
Beliefs play out through people's actions. Some of these become regular and may exhibit a pattern, which are commonly labelled as practices. They are the ways in which beliefs manifest themselves in everyday life, and may include diverse activities, such as representatives' weekly routines, voting rituals, or processes over how information is collected and shared. Practices often rely on 'local knowledge', which Dvora Yanow (2004, S12) explains as 'the very mundane, yet expert understanding of and practical reasoning about local conditions derived from lived experiences'.
In this way, then, when we talk about and analyse beliefs, we must put them in connection with lived experiences (the relationship between beliefs and practices is complex, as discussed by Cook and Wagenaar (2012) and Wagenaar (2012)).
Of course, MPs and officials do not distinguish between beliefs and practices in their day-to-day lives; they are analytical concepts. Nonetheless, bringing attention to them as part of a wider interpretive approach to studying social and political life has benefits. It highlights the importance of different ways that actors behave in parliamentary institutions, and the effect of their behaviour.
It brings new perspectives to understanding political institutions, which sits alongside a growing range of research that has applied similar insights in different legislatures, including India (Rai and Spary, 2019) and the UK (Crewe, 2005(Crewe, , 2015. The interpretive approach is particularly apt for this research article because it pushes our analyses of knowledge use from who accesses parliamentary arenas to how forms of knowledge interact with MPs and officials (as discussed above). I now explain my application of interpretive concepts and ideas.

Methods
In this study, I have used three qualitative methods. First, I undertook participant and nonparticipant observation by working as a full-time research assistant to a select committee over a 14-week period in the House of Commons Committee Office (amounting to over 600 hours of observations accumulated on a daily basis from 8.30am until 7.00pm, and occasionally earlier/later). Through deep immersion, I was able to participate in supporting 'my' committee in private meetings with MPs, through staff meetings and team discussions, by writing briefing papers and drafting reports, and in offering other support as required. Furthermore, I was a non-participant observer in the sense that I witnessed private and public meetings of a range of committees over three months, proceedings of the chamber, and a range of interactions between MPs and officials throughout the parliamentary estate. Being inside the heart of Parliament gave me unprecedented access and allowed me to see, first-hand, how actors use 'evidence'. I was in a privileged position to see how established beliefs and values met with and clashed against the limits of procedure and formal interpretations of evidence. This inherent value of observation would be difficult to capture in other ways. In order to draw on my experiences, I kept a fieldwork diary; a personal, private and confidential journal. I draw on this diary throughout the empirical sections (cited as 'FWD', with a paragraph number). This research received ethics approval from the University of Sheffield, and both the Head of the Scrutiny Unit of the Committee Office and I agreed to the terms of the placement verbally and in writing (Geddes, 2020, pp.149-55). The identity of all participants is protected through anonymity, which restricts some details, including the timing of the placement and the policy area of 'my' committee.
While observation has many benefits, it also has challenges. Each researcher must consider the specific types of access that they receive, and more broadly their positionality. As Mike Crang and Ian Cook (2007, p. 9) have pointed out, research embodies observers and their identities, and researchers must be sensitive to this. In my own research, I have sought to remain critically aware of power dynamics and my privileged position as a white man in a legislative arena (Geddes, 2020, pp. 149-155). I ensured that I spent time in different academic as well as political environments in order not to get 'lost' in the world within which I was immersed. Importantly, I have complemented observation with other methods. I undertook 46 semi-structured interviews with committee members, chairs and officials (that were recorded with their consent), which took place alongside informal conversations with parliamentary actors during the research placement. Interviews focused on select committee work broadly, in which I sought to get a better understanding of participants' interpretations of scrutiny. Finally, a third method on which I draw are the findings of a focus group with eight parliamentary officials (from across Parliament) that took place in June 2016, and centred on barriers and facilitators to academic engagement with the legislature .
These methods created a wealth of data. Observations from my fieldwork diary formed the basis of inductive coding, in which I drew out key themes of my research. In interviews, I was interested in how committee work is structured, and so these interviews encompassed themes including (i) ideas about scrutiny, (ii) views on inquiry processes and (iii) relationships between key actors.
From this, I drew out understandings of how MPs and officials interpreted ideas around 'evidence'.
They are the hooks around which the empirical sections are based, to which I now turn.

Interpreting 'evidence' in UK select committees
In order to undertake an interpretive analysis of evidence use in select committees, this -the main empirical section -is broken down into the following: first, I give an overview of the political logic of select committees in the House of Commons and their epistemic function; second, I examine the beliefs around evidence by MPs and officials; third, I explore the practices around evidence along the committee corridor in the House of Commons; and fourth, I discuss the main implications.

The political logic and epistemic function of select committees
Select committees in the UK House of Commons are the principal accountability mechanism for Parliament vis-à-vis the executive, and have been widely praised (e.g. Fisher, 2015). They are made up of a small cross-party group of usually between nine and 18 MPs. The party balance reflects that of the House generally. Chairs are allocated across the committee system based on party balance (informally agreed by parties' business managers, or whips) and, since 2010, directly elected to their positions by the whole House (and by secret ballot). There is one committee for each ministerial department (e.g. the Defence Committee shadows the Ministry of Defence), as well as further cross-cutting policy ones (e.g. the Environmental Audit Committee looks at environmental policy across government). They are tasked to examine the expenditure, administration and policies of government, which takes place in the commonly assumed adversarial political system of the UK. This political context is important, because it reminds us that MPs have a range of motivations for scrutiny which is not limited to improving policy-making, and may invite grandstanding and bias (LSE GV314 Group, 2020).
Committees are a core part of the political process, enjoying widespread media coverage and a positive reputation (Gaines et al., 2019). This is despite the few formal powers of committees, restricted to 'send for persons, papers and records' (House of Commons Standing Orders).
Committees cannot force government to change course and must rely on cross-party agreement through inquiries. Ordinarily, committees hold an inquiry into a subject by publishing terms of reference, calling for written evidence and holding committee hearings (oral evidence), followed by a report that includes policy recommendations. Committees usually agree these reports unanimously to strengthen their appeal and persuasive power (White, 2015). Indeed, it is now largely expected that a 'good' committee is one that avoids divisions on party lines. The level and range of recommendations varies, and while the government is expected to respond within two months of the report's publication, adopting recommendations is optional. For this reason, many academics have dismissed these committees as comparatively weak arenas for policy influence (e.g. André, Depauw, and Martin, 2016). However, this view is contested, with a number of studies finding that the UK Parliament's scrutiny committees play an active role in shaping policy -even if not easily observable (Benton and Russell, 2013;Hawes, 1993;Hindmoor, Larkin, and Kennon, 2009;Russell and Gover, 2017, pp. 205-233).
Alongside their policy scrutiny role, select committees can and do play a much bigger role for MPs in the House of Commons by acting as sites for professional development, socialisation and institutionalisation (Geddes, 2020;Norton, 1998;Rush and Giddings, 2011). Committees are spaces that allow for policy learning because they act as information-gathering tools of a range of different types of information. This allows MPs to gain expertise around key policy areas, whilst also putting MPs, officials and stakeholders in contact with each other through committee work.
As such, the role of different claims to knowledge and 'evidence' in this setting is especially important. In order to explore this issue, and as noted in the theory section above, I focus on (i) the beliefs around 'evidence' in committee work and (ii) the consequent practices of how MPs and officials approach evidence.

Webs of belief around 'evidence'
As a starting point, it is worth noting the formal meanings that surround 'evidence' in a parliamentary context as set out in Erskine May, the most authoritative reference book on parliamentary procedure for the UK Parliament. As noted above, committees take part in an evidence-gathering process, by which they collect 'written evidence', which may be submitted by any individual or organisation, and 'oral evidence', where individuals or organisations are invited to appear in front of committees and answer questions as 'witnesses' (usually in public and, if necessary, on oath). In both cases, it is required and expected that evidence is 'truthful' (Erskine  Both quotes are revealing in that they use the metaphor of the courtroom to explain their beliefs of the meaning and significance of evidence. They are illustrative of a wider web of beliefs in which select committee work is viewed as a court that is reliant on only evidence that is directly presented to the committee. The legalistic definition makes evidence-gathering flexible because minimal, usually procedural, criteria need to be fulfilled. As one official remarked, the concept means 'testimony' in practice: When we say evidence, what we mean is testimony. … Someone's told us something … somebody's written a letter, it says, you know, "I was waiting over an hour, I've seen my GP earlier this morning" -we would call that evidence (Focus Group, HC committee clerk 3).
The clerk went on to say that evidence is, therefore, 'something you can publish' and 'see', but it does not necessarily explain its role. For this, another official explained that evidence is information with a narrative (Focus Group, HC Librarian 2).
In taking a legalistic yet flexible view of evidence, MPs and officials subsequently echo a courtroom notion of 'for' and 'against'. This is especially important given the balance of political opinion on the committee, which includes representatives from across parties. So, political balance receives a crucial priority. This was borne out in focus group discussions: This excerpt is illustrative in showing a 'balancing' dilemma that staff continually negotiate as a result of a legalistic belief in evidence, which contrasts with an ideal for the ostensible quality or truthfulness of evidence. So, evidence must meet the needs of politicians. One interviewee explained: We did some work on charities and … [two committee members] decided, at the end of an evidence session with [a moderate third sector organisation] -hardly the most left-wing body in the country -that they wanted more … right-wing witnesses and that there's been too many charity sector people in it. I'm like, so okay, inquiry on charity sector … talking to charities … kind of makes sense? So we ended up getting in some … particularly right-wing think tank and got them in and, to try and appease that (Interview with Official 2). This quote is an example of how beliefs affect the role of evidence in committee inquiries and is therefore closely interwoven with committee practices around evidence use to which I return in the next subsection.
The role of evidence changes at a later stage of a committee's inquiry, most notably when the committee seeks to write its report, at which point: So, evidence becomes more explicitly political with a legitimising function to increase the authority of the report's claims. This has clear implications for the practices around evidence use but also about the basis and credibility of a report. As one committee chair explained: … Members on that committee know that the conclusions will be evidence-led, that they will have had every opportunity to identify the evidence themselves and come to their own conclusions and that we will collectively agree a formal wording which might accommodate their concerns and won't be used just for party political [purposes] (Interview with Chair 9). This quote is insightful because it links beliefs about evidence to those about consensus. Consensus is often seen as the ideal outcome for committee scrutiny because it suggests that multiple political parties are able to agree on a policy solution. In this context, evidence can be used to substantiate policy conclusions and recommendations and a mechanism to prevent partisanship. Once again, therefore, evidence has a key legitimising function. This theme of authority is something to which I return in the discussion.
This subsection reveals to us a specific yet flexible belief in evidence for select committee work, which is often subjected to political stress-tests. Evidence is constrained by formal beliefs around legal notions of evidence, such that information collected by committees become their property and protected by parliamentary privilege. The echo of the courtroom seeps into committee processes: political balance is important throughout, but towards the end it is used to substantiate one side over others. These beliefs are, of course, closely interwoven with practices, to which this subsection has alluded in a number of places and to which our discussion now turns.

Evidence practices in committees
While evidence is pushed into a particular legalistic direction, these beliefs rub against the everyday practices of MPs and officials along the committee corridors. Here, I focus on three ways where beliefs and practices intersect.
Select committees in the UK work to tight timeframes and often in response to wider political issues, so they are, as one official put it, 'essentially reactive' (Focus Group, HC Librarian 2). I noticed this during fieldwork. On one occasion, my committee's chair asked for a one-off 'newsy' evidence session to raise the committee profile (FWD 39.9.16). So, a committee is not necessarily led by evidence; rather, it is led by a political cycle and set of processes (and thus the first way in which beliefs and practices interact). Given the pace of the news cycle, this makes it more difficult for committees to have fully thought-out inquiries and less time with which to undertake those inquiries. As a result, the pool from which witnesses might be drawn is reduced. During fieldwork, it was sometimes a struggle for officials to ensure that witnesses could attend certain sessions because they had not been planned sufficiently in advance. As a consequence, some committees are disproportionately reliant on those that have the resources to be in London at the right time (Geddes, 2018). This point is related to a bigger issue about time. MPs and officials are time-poor, which has direct knock-on effects on who is able to engage with evidence in committee work in two ways. Not only does it skew those available to participate but, it also means that politicians are often reliant on committee staff. MPs readily acknowledge this: The role of officials is a second factor that I want to explore, specifically their epistemic networks.
The evidence-gathering process relies heavily on officials to be able to analyse written evidence as well as make useful suggestions for oral evidence. The basis from where these suggestions come can vary significantly, but usually includes written evidence, suggestions from colleagues across Parliament (e.g. the House of Commons Library, POST), specialist advisers and/or informal policy networks of officials -I was reliant on all of these groups during my placement. To ensure breadth, one senior official argues that it is part of the clerks' role to get 'stuck into the wider policy community' and encourage them to find relevant expertise (Interview with Official 12). One MP, for example, explained that he would 'not very often' suggest names for oral evidence so long as the committee hearings explored 'a range of views' (Interview with Committee Member 8), which was a common response in interviews (behaviour which arises out of the beliefs around evidence needing to be diverse and balanced). During fieldwork, it would be rare to receive many suggestions other than from the chair of my committee (FWD 2.1.23).
Officials' networks are often established in two informal ways. First, it is by consulting colleagues within the institution 'who were in the know' (Focus Group HC committee clerk 2). This was a general finding in the focus group, in which committee reports, POSTnotes (short, impartial and accessible evidence syntheses/briefing papers produced by POST and peer reviewed externally), and Library briefing papers were regarded as not only valuable in the first instance, but also authoritative to inform parliamentary work. Later in the focus group session, officials acknowledged that this is 'inward-looking' while MPs themselves are far more outward-looking through their involvement with All-Party Parliamentary Groups or 'various lobbying and campaigning bits of work' (Focus Group, HC committee clerk 2). The second way by which officials' networks are sustained is through word of mouth with academics: … you can go along to the UCL lot and say, "Who do you rate?", "Oh we think Leeds are great", "Leeds, who do you think are great", "Oh, we like Southampton", "Southampton, who do you think are great?" … it's a bit like … Facebook for academics. Everybody's part of the groups and liking each other's work and you're starting to build up that network and blogging is a, you know, great thing. Stuff that gets tweeted out. You can quite quickly see who's active within a particular sphere (Focus Group, HC committee clerk 2).
This stresses an exceptionally network-based approach, but one that is common due to the space and time constraints that many officials seek to manage.
So far, this discussion has focused principally around the effect of everyday practices on who participates in evidence-gathering processes. What about the effect of those practices on evidence itself? For this, I want to look at committee hearings. They are the 'front stage' of committee scrutiny and arguably the most important stage of the evidence-gathering process for three reasons: first, it is the most high-profile form of participation in committee inquiries; second, they bring knowledge claims and scientific evidence to life, with which written evidence rarely compares; and third, oral evidence depends on invitations, therefore revealing to us the priorities of committees and to whom they are listening (Geddes, 2020, pp. 104-105).
An important element of oral evidence is its 'performative' nature. This was noticeable throughout fieldwork and in interviews. As one clerk noted, committee hearings are a 'theatrical performance', which meant that 'getting the right witnesses is quite tricky because you want good evidence but you also want people who are going to have something to say … in a way that's accessible and understandable to members' (Interview with Official 10). Similarly, a clerk explained that academics (though this could be extrapolated to witnesses generally) needed to be 'sharp', 'on the ball' and 'talking something relevant' (Focus Group, HC committee clerk 1). This stresses the importance of building authority and credibility as a witness in a select committee context, which is, of course, mediated through identities such as gender, ethnicity or education (Puwar, 2004), as well as non-verbal cues (Schonhardt-Bailey, 2017).
The theatricality or performance of evidence is inextricably linked to the political nature of committee hearings. As one official explained: Yes, it depends entirely on who's giving it out. Members are very aware [of] all the political nuances of the sources as much as they are of the political nuances of the content. And so who presents it to you can have a huge difference (Focus Group, HC committee clerk 1).
This underscores the political nature of committee hearings, and returns us to the point about the role of balance in evidence sessions. This has implications for how scrutiny is enacted. To give an example, I give a 'vignette' from a committee hearing that took place during fieldwork. 1 The session in question was the first of many for an inquiry about government decision-making. The hearing was organised to 'set the scene' for the committee and, as such, had three researchers to provide evidence about the policy landscape and context. While some MPs' questions were factual to find out underlying dynamics of the policy issue, other MPs wanted to raise specific political issues related to it. For example, one MP was adamant to explore how one particular policy actor had misused public funds and who phrased leading questions in an attempt to get independent researchers to agree with his point. Another MP also raised stories from the media and comments from policy actors, asking the researchers to assess the validity of those stories and comments. In most cases, the researchers answered cautiously and framed their replies in abstract ways to focus on broad lessons for government decision-making but, in more than one instance, they refused to comment. When issues under discussion focused on technical matters, researchers were confident and eloquent; as soon as it turned to the political interests of committee members, there was caution and silence -including in one case where two MPs argued with each other over an unrelated matter on membership of the European Union. After the session, one academic witness noted (to her surprise) that the hearing was 'very intense' and 'very political' despite the fact that she was trying to give the committee information in her role as expert (FWD 15.4.6). Although this example has been ripped from its rich context and the heat of the debates in the committee room, it serves as a useful illustration in demonstrating that scrutiny and evidence can be used for primarily political functions (even if what some MPs regard as common sense are contested political points). Some MPs will treat committee hearings as a way to analyse information to contribute to holding the government to account, but others may use the committee arena to push their own, their constituents' or their party's interests further, while yet others still might use the opportunity to learn about policy areas rather than scrutinise. This is an important point because it suggests that the role of evidence can be pushed in different directions for the benefit of those MPs and the styles of scrutiny they seek to enact. In the case of the committee hearing above, MPs were using the evidence presented for their own particular interests, clashing with other MPs and witnesses. It meant that one of the academics felt the session was not particularly useful; she was unable to impart her expertise, despite the fact that her expertise had the potential to significantly impact the committee's inquiry.

Discussion
The committee arena -or the stage on which scrutiny and different forms of knowledge play outis a complex space that is not entirely 'led' by evidence. First, MPs and officials define 'evidence' in a legalistic way, which has cast a wide yet simultaneously constraining web for gathering evidence in that minimal procedural criteria must be met at the same time as meeting a test of political balance. Second, ideational constraints are further affected by everyday practices around evidence use, including the reactive nature of scrutiny, the network-based approach of officials, and the performative element of committee work. All of these arguably deserve further research attention but, taken together, already offer broader themes. In this discussion, I want to draw out some wider cross-cutting themes, based on the empirical observations and interviews above: first, that evidence is often used for the mobilisation of authoritative reports, and arguably less so for instrumental purposes; second, that the role of evidence in committees is not only political but also mediated by a particular democratic lens; and third, that who is included or excluded is affected by beliefs and practices inside the legislative arena.
Throughout the empirical discussions, an implicit theme has focused around the need for committees to publish authoritative reports in order to influence government policy. Committees derive a lot of authority from their ability to agree to reports on a cross-party basis, which is seen to show that politicians from different backgrounds can agree on a set of policy proposals (White, 2015). However, and importantly arising out of the above, the role of evidence in this is crucial: committees rely on a reputation that those reports are evidence-led and that MPs are able to agree on a report on the basis of evidence that supports their claims. For example, during a private meeting of a committee, where MPs were discussing a draft report, one committee member's proposed amendments were rejected because they did not reflect the evidence of what the committee had received (FWD 53.12.5). So, evidence plays an important role in driving committee work forward and links closely to consensus. It is used to ensure that a party politics is at the margin of committee scrutiny (and could potentially be threatened through greater party polarisation). That said, the credibility of evidence itself is often dependent on the way it is performed, and by the identity of the performer. Evidence therefore plays a complex and significant role -and it does so distinctively in a legislative arena as opposed to bureaucracies or governments because, even if evidence has symbolic dimensions inside the bureaucracy (Boswell, 2009), it does not need to be publicly performed against the backdrop of diverse political viewpoints -as it does in legislatures.
A second wider theme from the empirical discussion is that evidence remains only one of multiple considerations in committee work. Although committees pride themselves on the extent to which they are evidence-led, they are also framed by political processes. This chimes with findings from the wider literature on research utilisation (Cairney, 2015) and contradicts the assumption amongst many within Parliament that select committee inquiries are 'evidence-based'. As Boaz et al. (2019) note with respect to other arenas, it makes more sense to talk of 'evidence-informed' processes.
Moreover, while select committees are therefore not only conducting scrutiny through a political lens, they are also doing so through a representative lens. So, in the first instance, evidence must be politically balanced in order to ensure that the viewpoints of different MPs are respected. But additionally, there is a strong priority for committees to also represent the wider public in the political process, in part because of the very public nature of the process. For this reason, there is a strong desire to ensure that committees hear from a socially diverse group of people (Beswick and Elstub, 2019;Bochel and Berthier, 2018). As Hendriks et al. (2017) have shown in Australia, too, this has created some tensions between the epistemic and democratic functions of committees. This is unique to legislative arenas in the sense that governments usually rely only on those forms of expertise that support their policy goals.
Finally, and related to the previous point, there is a growing debate about who is included and excluded from parliamentary processes. Indeed, the question of 'who has access?' is one of the main focal points of the academic literature on this topic. In the UK, this has led to a debate about the social diversity of witnesses select committees (House of Commons Liaison Committee, 2018; Beswick and Elstub, 2019). However, this issue also brings attention to the wider ecosystem of evidence in parliamentary settings and the way by which different groups are able to participate.
So, there are increasing questions about democratic innovations, such as the use of mini-publics or citizens' assemblies, introduction of virtual evidence-gathering, or, giving more space in reports to citizens' lived experiences. This article has brought to light some of the traditional practices of how evidence is gathered and the extent to which this contributes to skewing the evidence base in the House of Commons, namely, a tight timetable and reliance on pre-existing networks. It raises questions about how democratic innovations can be integrated within the webs of belief of parliamentary actors and established practices in evidence use.

Concluding remarks
This article has focused on how MPs and officials interpret the idea of evidence and how their interpretations are enacted day-to-day. It chimes well with other studies that have demonstrated the political nature of knowledge claims and evidence use. In shifting focus from government and the bureaucracy to the legislature, it demonstrates distinctive characteristics and challenges for parliaments, including the public nature of performing evidence; the need for political balance; and appealing to representative as well as epistemic ideals.
This study is important for at least three key reasons. First, committees across legislatures are often seen as particularly important in influencing government policy, yet we will know little about the basis on which their claims are made. In the UK select committee context, this could amount to more than 40% of policy recommendations being accepted and implemented by government (Benton and Russell, 2013). Clearly, understanding from where those recommendations come is crucial to further explain the influence of select committees and of different knowledge claims.
Second, and now turning to a wider point, knowledge claims are used pervasively in legislatures, which informs public debate and the law-making process. To overlook how politicians in the legislative arena engage with different forms of knowledge, or with what sorts of organisations they do so, is a blind spot for scholarly understanding of policy and political processes. In the UK, this has become ever-more critical at a time of prevalent political discord through, for example, claims made about the UK's departure from the European Union. Such debates in the chamber of the House or through inquiries by committees shape the wider narratives of policy-making.
Furthermore, committees can act as a policy learning tool for MPs, meaning that their epistemic foundations may have considerable effects.
A third and final reason for why studying knowledge claims in parliaments matter is that, at a time of significant policy shifts and political debates in the western world as a result of Coronavirus/Covid-19, it has revealed deeper tensions in society about the role of knowledge in making political and policy decisions. Parliaments are uniquely placed within those tensions because they juxtapose diverse political claims made by political parties, social movements and citizens with those of experts, scientists and other stakeholders. Committee work is one such arena where different knowledge claims can and do clash, and therefore makes research into this area all the more important. In understanding these issues better, we can be better informed about the effects of parliaments on policy-making processes. This article has addressed these issues using an interpretive framework. Its strengths lie in adding depth and nuance, but it has limitations in that it required anonymising rich data and focusing on a single case study (for further reflections, see Geddes (2020, pp.149-55)). This article has offered a distinctive angle, but also raises more questions, from how public voices are integrated into parliamentary decision-making to whether certain forms of knowledge are more or less credible and authoritative than others.