What Constitutes Substantive Representation,and Where Should We Evaluate It?

Abstract

This short article introduces a novel framework for conceiving of the substantive representation of marginalized groups – in this case, racially minoritized women in the UK House of Commons. I outline a rubric of eight facets of substantive representation. These evaluate the degree to which claims that constitute a group are also substantively representative of that group. In doing so, I contribute a much-needed framework for distinguishing between representative claims which speak on behalf of a group versus those which merely speak about, or even against, that group. I argue that substantive representation must be considered intersectionally, reflecting the multiple structures positioning those represented. Furthermore, all facets of substantive representation can rarely be contributed by a single parliamentary speech, individual, or narrow group of legislators such as descriptive representatives. Therefore, I suggest that empirical studies of substantive representation should include a greater evaluation of the collective work of institutions as a whole.

Introduction

Speaking and acting on behalf of individuals and groups is fundamental to representative democracy. The positive association between what Pitkin (1967) terms ‘descriptive’ and ‘substantive’ representation – or groups’ presence in legislatures and the representation of their interests and perspectives – is well documented. This relationship is neither deterministic nor exempt from the effects of institutions (Mackay et al., 2010). It is robust, however. Conversely, the relationship between constitutive representation (describing a group) and substantive representation (speaking for or on behalf of that group) is under-explored and under-theorized (Siow, Orly, forthcoming, Severs, 2012; Squires, 2008). Likewise, there has been limited interrogation of what precisely constitutes substantive representation.

Empirical scholarship has placed great emphasis on measuring different forms of representation in debates and policy. Yet, it often makes the implicit assumption that mentioning (constitutively representing) groups such as ‘women’ or ‘minorities’ is synonymous with substantively representing those groups’ interests and perspectives. In contrast, I posit that constitutive representations of groups fall into three categories: speaking on behalf of the group, merely speaking about the group, or even speaking against the group. In doing so, I outline a novel rubric of eight facets of substantive representation. This is based on a grounded theory analysis of parliamentary speeches mentioning racially minoritized (hereafter ‘minoritized’) women in the United Kingdom over a 20-year period. Collectively, these facets of substantive representation are rarely observed in a single parliamentary speech, the contributions of a single individual or small groups of legislators such as descriptive representatives or critical actors. Therefore, I suggest that to capture all of these facets of substantive representation, empirical scholarship should include the evaluation of institutions as a whole. The specific institution considered here is the UK House of Commons, but this could extend to parliaments elsewhere, as well as parties, committees, civil society and other representative fora.

The relationship between constitutive and substantive representation is especially fragile for minoritized women in the United Kingdom. The group is descriptively under-represented and therefore afforded few opportunities to speak for themselves. Minoritized women are also stereotyped and instrumentalized by others in political debates (inter alia Joly and Wadia, 2017). This is therefore an extreme case in which to parse out the distinctions between speaking against, about, or on behalf of. In addition, the case highlights the necessity of intersectional approaches, as well as associated methodological challenges. These are detailed below, followed by a new framework for conceiving of substantive representation. Finally, I discuss the importance of evaluating institutions as well as actors.

Measuring Substantive Representation

Existing empirical studies of the substantive representation of marginalized groups have generally approached the topic from one of three starting points: descriptive representatives, critical actors or a priori issues, interests and perspectives. Studies of the contributions and experiences of descriptive representatives are essential. They have contributed immensely to our understanding of the substantive representation of marginalized groups such as minoritized women (inter alia Brown, 2014). However, studies that centre on descriptive representatives place less emphasis on whether or how these groups are also substantively represented (or not) by non-descriptive representatives. Some scholars have also argued that centring descriptive representatives risks essentializing them (Belli, 2013). This can be especially problematic for minoritized women legislators who are often presumed to be symbolic and substantive representatives of racial and gendered constituencies – even when this conflicts with their own stated intents (Bajpai, 2019).

Other scholars emphasize the responsibility of non-descriptive representatives – so-called ‘critical actors’ – to represent the interests and perspectives of marginalized groups (Childs and Krook, 2009). However, to identify critical actors on behalf of minoritized women, we need definitions of minoritized women’s interests, perspectives and issues. This presents several new challenges. First, minoritized women do not share homogeneous interests or perspectives. Second, even if we posit some degree of shared group interests, the dearth of scholarship on the politics of minoritized women in general (Alexander-Floyd, 2014) means that existing definitions of ‘women’s issues’ typically ignore intersectionality and centre the experiences of white women (inter alia Smooth, 2011). Third, the issues that minoritized women are typically mentioned in relation to within political discourse are limited to a narrow and highly stereotypical set of concerns such as Islamic dress (Joly and Wadia, 2017: 2).

Siow, Orly (forthcoming, 2023) provides a partial egress from these challenges. Starting with representations themselves, Siow, Orly employs a novel method to identify parliamentary speeches which explicitly mention (and therefore constitutively represent) minoritized women. Siow, Orly shows that descriptive representatives mention the group most often, and in relation to a wider range of issues than other MPs – including white women and minoritized men. This approach enables analysis of descriptive and non-descriptive representatives without specifying a priori ‘minoritized women’s issues’. However, it only captures limited facets of substantive representation. Here, I aim to develop a more comprehensive framework for evaluating whether and how a group is substantively represented by all members of institutions.

Method

To develop this framework, I used Siow, Orly’s (forthcoming 2023) dataset of 1630 UK House of Commons speeches constitutively representing minoritized women during five parliaments spanning 1997-2017. The detailed qualitative analysis drew on the principles of grounded theory. In particular, data collection (regarding the content of constitutive representations) and analysis (of the relationship between constitutive and substantive representation) were treated as ‘interrelated processes’ conducted ‘systematically and sequentially’ through consideration of the ‘conditions’, ‘interactions’ and ‘consequences’ of the concepts being analysed (Corbin and Strauss, 1990: 419–420). Constant comparisons of the content of constitutive representations alongside analytic memos were used to parse out eight facets of substantively representative claims, outlined below.

Speaking Against, About and on Behalf of

In Table 1, I outline the rubric of eight facets of substantive representation: eight ways in which we can evaluate whether and how constitutive representations speak on behalf of a group (and therefore substantively represent that group). These facets include representing a group without hostility; as an end in itself; in relation to relevant structural factors; as heterogeneous; on its own terms; maintaining agency, making an explicit request and constituting both problems and solutions intersectionally. In contrast, claims that speak against a group are manifestly hostile. Those who merely speak about a group are not explicitly negative but may instrumentalize, stigmatize, or homogenize the group; employ stereotypes; restrict agency; fail to make an explicit request or lack consideration of relevant structures. I provide examples of each below.

1. To be substantively representative of a group, constitutive representations must not be explicitly negative , or manifestly hostile , towards the group that is constituted. A claim or set of claims which fails to meet even this low bar can be considered speaking against. Notable examples include discussions of so-called ‘maternity tourism’ in the United Kingdom, in which minoritized women seeking maternal healthcare are constituted in explicitly negative and hostile terms. This condition is necessary but insufficient as a measure of substantive representation.

2. Next, a group must be constituted as an end in itself . The goal sought must benefit the group itself, rather than others. In contrast, claims that speak against or merely about a group instrumentalize, by constituting the group as a means to an end. This can be observed in mentions of female victims of human trafficking which are used exclusively to support arguments aimed at curbing immigration. Similarly, we might argue that minoritized women are not substantively represented when they are invoked as tools to improve the ‘diversity’ of the UK’s public institutions, with no sense of how this is beneficial to the women themselves. In both instances, the group constituted is not the group being served by the claim.

3. Third, substantively representative claims must constitute problems both within and beyond the group . Claims which speak against, or merely about a group stigmatize, by constituting the problem solely within the group. These claims fail (in the case of minoritized women) to address relevant structural racial and gendered factors beyond racialized women and their communities. Discussions of forced marriage or so-called honour crime are both typical examples. It is insufficient to denounce the perpetrators of FGM or other forms of violence against minoritized women. Neither is it sufficient to remain silent on these issues – unless doing so actively makes space for motivated descriptive representatives (see also, point 5). Substantive representation also includes articulating and addressing the structural legal and economic factors that position specific groups as vulnerable – in this case, minoritized women. For example, if a set of claims to represent women affected by FGM is silent regarding factors such as citizenship and immigration regimes, legal aid provision or the funding of specialized service providers, then the problem of FGM is constituted as solely within the affected community and is likely to stigmatize that community. However, if references to these structural factors are present then the problem is constituted as both within and beyond the affected community – reducing the risk of stigmatization.

4. Substantive representation includes representing a group as heterogeneous, and not solely in relation to stereotypical issues . A good, albeit rare, example of this is debates surrounding racial disparities in maternal health. The topic is significant for minoritized women, but beyond the stereotypical issues usually associated with the group. In contrast, claims constituting minoritized women exclusively in relation to limited set of issues, (such as Islamic dress and FGM) homogenize the group. Individual campaigning MPs focusing on these issues do contribute important facets of substantive representation. However, if within an institution as a whole, a group is only or overwhelmingly represented in relation to stereotypical issues, then this facet of substantive representation is lacking.

5. For the group to be constituted on its own terms , there must be some correspondence between the issues, interests and perspectives constituted within parliament, and those constituted beyond parliament, such as by relevant civil society. Crucially, both descriptive representatives within the chamber and responsiveness to descriptive representatives beyond the chamber are necessary for this to take place. For example, civil society led by minoritized women is not known for its focus on descriptive representation within the military. Yet, parliamentary debates include a surprising emphasis on the recruitment of minoritized women into the armed forces. Critical actors cannot substantively represent a group they do not descriptively represent without members of the group first constituting the problem and the solution. Substantive representation is ‘bottom-up’ rather than ‘top-down’. The constitutive representation of issues, interests and perspectives by descriptive representatives within and beyond parliament is therefore necessary for the substantive representation of a group.

6. Sixth, substantive representation maintains agency . Debates on issues such as Islamic dress often include contestation of the choices made by consenting, adult women. For example, the practice of veiling elicits a range of views, including among descriptive representatives. Individual representatives must of course be able to air their views within democratic fora. For Muslim women (or other groups) to be substantively represented in this context, the range of positions given floorspace must include claims which grant that group agency. Without being granted agency – as is often the case for minoritized women – the group in question is merely spoken about and not on behalf of.

7. The penultimate facet of substantive representation is making explicit requests . Merely constituting a group in glowing terms is not sufficient to substantively represent the group’s interests. For example, when parliamentarians commend minoritized women’s contributions to the NHS, this is distinct from advocacy. This form of symbolic representation makes the group ‘stand for’ something positive (Pitkin, 1967) and may complement substantive representation. Yet, it can also form a way of claiming that the group is valued while simultaneously dodging the group’s requests. Thus, explicit requests form part of the distinction between speaking about or on behalf of.

8. Finally, both the problem and the solution must be constituted intersectionally. For example, race and gender affect minoritized women’s earnings. Therefore, reference only to a ‘gender wage gap’ does not represent minoritized women’s positions. Race and gender need to be considered relative to both the problem and the solution. Claims must therefore consider the intersection of relevant structures to substantively represent intersectionally marginalized groups.

Table 1.

Facets of Substantive Representation.

Speaking on behalf of	Speaking against/about
1. Constitutes the group in a way that is not negative or hostile	Constitutes the group in a way that is negative or hostile
2. Constitutes the group as an end in itself	Constitutes the group solely as a means to an end (instrumentalizing)
3. Constitutes the group in relation to the structural factors which positions it as vulnerable	Constitutes problems as solely within racialized community (stigmatizing)
4. Constitutes the group’s heterogeneity, and in relation to a wide range of issues	Constitutes the group as homogeneous or in relation to limited range of issues (homogenizing)
5. Constitutes the group on its own terms, including relevant civil society	Constitutes the group relying on stereotypes
6. Maintains agency	Fails to maintain agency
7. Makes an explicit request	Does not make an explicit request
8. Constitutes both the problem and the solution intersectionally	Constitutes either the problem or the solution in relation to single axis or structure

Individual speeches or representatives typically contribute some of these facets of substantive representation, such as constituting a group positively or making an explicit request regarding a particular issue. These contributions are essential and important features of substantive representation; they are also overwhelmingly performed by descriptive representatives. However, it is rare and difficult for other facets – such as a group being constituted heterogeneously – to be contributed by a single speech, an individual legislator or even a small group of representatives. This is partly because institutional constraints such as legislative agendas curtail the space afforded to parliamentarians. In this context, a narrow focus is often an effective, albeit limited, strategy for campaigning MPs. Even the most powerful or highly motivated members do not consistently enjoy opportunities to articulate structural factors beyond (for example) racialized communities, or regarding a wide range of issues. This rubric therefore offers a framework to consider the degree to which multiple facets of substantive representation are present within parliaments. It may also be applied to parties, committees, other parliamentary groupings and civil society. This lays the foundations to evaluate the degree to which marginalized groups are substantively represented by institutions as a whole, and in which ways.

Conclusion

This framework makes two key contributions to existing scholarship. First, fleshing out these facets of substantive representation improves the content and construct validity of the concept while also accounting for intersectional heterogeneity among women. Second, the rubric lays the theoretical groundwork to refine empirical strategies for measuring substantive representation which (a) start with representative claims themselves and (b) evaluate the quality of representation across all the claims made within a variety of institutions. It also has promise for application to other marginalized groups and cross-national comparative analyses.

I suggest shifting focus from individual representatives towards the collective work of institutions because substantive representation is multi-faceted. My intention is neither to critique nor diminish the contributions of descriptive representatives and critical actors. The links between descriptive and substantive representation are clear. Instead, I aim to raise questions regarding the extent to which different facets of substantive representation are (a) present and (b) possible within existing institutions. The example used here is minoritized women in the United Kingdom, for whom constitutive representation often entails homogenization, instrumentalization and a failure to address race and gender as intersecting structures. What features of institutional design might allow for the group to be constituted heterogeneously, on their own terms and in ways that address the intersecting factors which position them as vulnerable? What conditions would enable and motivate representatives to achieve this, both individually and collectively? These are pressing questions for future scholars.