Exploring the Descriptive Representation of Women: The Case of the National Assembly of Wales

With the devolved institutions nearing their 20th anniversary, the Scottish Parliament, National Assembly for Wales and the Northern Ireland Assembly have changed the face of politics within the United Kingdom. The National Assembly for Wales, in particular, has made extensive advances in the field of female representation, notably being the first devolved legislature to achieve perfect gender parity. However, the question remains as to whether this parity was achieved through luck or design. This dissertation draws on the work of Pitkin to identify the importance of the relationship between both substantive and descriptive representation and then explores the concepts and contemporary academia to identify the leading theories within this field. The National Assembly will then be analysed in order to establish the measures that have been put in place by the institution in order to protect and promote a commitment to the equality of representation. Four political parties within the National Assembly for Wales will be assessed including the Welsh Labour Party, Welsh Liberal Democrats, Plaid Cymru and the Welsh Conservatives drawing on evidence from 1999–2016. Both chapters will draw upon semi-structured interviews conducted with the Diversity and Inclusion Team and female Assembly Members from across political parties in order to allow for conclusions to be drawn as to what the most influential factors have been in the success of the National Assembly’s female representation, and whether the Assembly has been successful due to luck or design.


Introduction
In the United Kingdom, women 'make up 51% of the population' (Jegede, 2009, p. 12), and yet, despite this, women are still under-represented within political institutions in comparison to their male counterparts. The World Health Organisation (2008, p. 145), 'formally recognised gender inequality as one of the major hurdles to development' with, 'the position of women [having] changed dramatically over the last century in many countries' including the UK where progress has been made, albeit slowly. Breitenbach (2006, p. 1) identified, 'the election of a Labour government in 1997, committed to a programme of constitutional change and modernisation, has provided new opportunities and new structural space in which politics of gender can be pursued'.
All parties have since increased their female representation and yet, there has been a limtied change to the 'masculinist political culture' (Khoury, 2015, p. 119) and ' old-boys club' (Coote, 2000, p.  The graph illustrates some areas of success with the percentage of female MPs increasing steadily. However, by comparison to other state-level institutions, progress is limited with the UK currently 48 th in the world parity league table ('Women in Parliaments: World Classification', 2016).
The discrepancies within Westminster are clear with the number of women representatives limited at the statelevel institution. However, 'the modest levels of female representation at Westminster stand in sharp contrast to the Nordic levels of representation' (Kenny and MacKay, 2013) achieved in the devolved institutions. The National Assembly for Wales (NAfW), the Scottish Parliament (SP) and, to a lesser extent, the Northern Ireland Assembly (NIA) have all achieved high levels of female representation with ' one apparent advantage of devolution [being] the opportunity for lesson-learning across jurisdictions' (Acton, 2013, p. 231). Across all devolved institutions a variety of measures have been introduced in order to increase female representation with the instittuions themselves unconstrained by the tradition of practices as seen within Westminster. The NAfW, inparticular, have achieved greater success than both the NIA and the SP and therefore will be the focus of this research. During the creation of the NAfW ' campaigners … successfully … lobbied for the inclusion of an equality clause in the 1998 Government of Wales Act' (Bagihole, 2009, p. 84). The clause is said to go 'beyond statutory requirements placed upon other UK legislatures' (Bagihole, 2009, p. 84) and requires the NAfW to: '… [m]ake arrangements with a view to securing that the functions of office-holder or authority are carried out with due regard to the need to met the equal opportunity requirements' (Hepple et al., 2000, p. 8).
The relationship between the descriptive representation of women, identifying how representatives reflect those represented, and the substantive representation of women, relating to the activities of these representatives, have been studied extensively within the relevant literature (Abels, 2015;Celis et al., 2014). Due to this extensive research, the main focus of this dissertation will be understanding the descriptive landscape within the NAfW with a view to possible further research to be undertaken concerning substantive representation. Analysis will be conducted of both the NAfW institutionally and the political parties of Wales to allow understanding of the success of female representation thus far, and how this was achieved.
In order to analyse the question in hand qualitative methods were used with Mason (2002, p. 3) identifying qualititve methods are: i) 'Grounded in philosophical position which is broadly 'interpretivist' ii) 'Based on methods of data generataion which are both flexible and sensitive' iii) Hold recognition ' of complexity, detail and context'. Therefore, by using qualitative methods through combining analysis of documentary data and a small number of semi-structured interviews this will allow for identification as to the prominent factors that contributed to the success of the NAfW in regard to female representation. Semi-structured interviews were conducted with both the Diveristy and Inlcudion Team (DIT) and female Assembly Members (AMs) from across political parties in order to underpin further analysis although, the Welsh Conservative Party (WCP) and Welsh Liberal Democrat (WLD) members all declined when asked for an interview.
Section one will be dedicated to the analysis of the contemporary literature initially focusing on Pitkin's typology, specifically in reference to the relationship between substantive and descriptive representation, before analysing both 'Critical Acts' and 'Critical Mass'. With this in mind, chapters two and three will then analyse the NAfW as an institution and the political parties in question. Chapter two will identify what the NAfW has put forward, since its creation in 1997, to increase female representation whilst analysing leading works by the likes of Chaney, Childs and Krook. Chapter three will then continue by scrutinising the political parties within the NAfW; investigating the period from 1999-2016 in reference to the four major political parties of this era. The discussion will examine the evidence in order to recognise the successes concerning female representation and the impacts of achieving parity within the institution. This will illustrate both areas in need of improvement, areas of success, and whether it is justified to assume that the NAfW will continue on this path into the future.
Section One: Pitkin's four-part typology, female representation and policy change The expanse of literature relating to the representation of women within political institutions has allowed for detailed analysis to be carried out relating to contemporary feminist theory. This chapter will focus on Pitkin's representation typology in order to understand the relationship between substantive and descriptive representation, before then analysing the impact on current institutions such as the NAfW through the use of contemporary research. Analysis will then be conducted on the debate surrounding the theories of 'Critical Mass' and 'Critical Acts' ultimately leading to the understanding of the correlation seen between female representation and policy change.
'Pitkin's well-established four-part typology of representations, which makes distinctions between formal, descriptive, substantive and symbolic representation' (Childs et al., 2010, p. 172), offers a comprehensive grounding in relation to the concept of the representation of women. Each element of the typology embodies a form of representation with formalistic relating to 'the formal structures and arrangements' (Ryden, 1996, p. 15), symbolic referring to 'standing for the represented' (Katz and Crotty, 2006, p. 42), descriptive identifying the ' extent to which a representative resembles those being represented' (Ethridge and Handelman, 2014, p. 121), and finally substantive which relates to the ' activities of the individual legislator' (Mezey, 2008, p. 52), and legislators as a group. This typology holds a prominent place throughout the literature with focus placed on the relationship between substantive and descriptive representation. Abels (2015, p. 9), identifies this relationship to be 'the most prominent concern in gender studies on political participation' with focus therefore frequently applied to these forms of representation. The debate that stems in reference to Pitkin is that: ' a relatively high descriptive representation does not automatically ensure better and effective substantive representation' (Abels, 2015, p. 9).
Due to the somewhat contested correlation between both substantive and descriptive representation, analysis of the leading debate is crucial in order to develop understanding of their importance. Phillips (1995), quoted in Abels (2015, p. 9), argues that a greater percentage of women in legislatures 'is considered to be a favourable condition for the promotion of gender equality in politics'. As identified by Mansbridge, quoted in Krook (2010, p. 201), ' disadvantaged groups may want to be represented by descriptive representatives' for four main reasons: (1) 'Adequate communication in contexts of mistrust (2) Innovative thinking in contexts of uncrystallised, not fully articulated interest (3) Creating a social meaning of "ability to rule" for members of a group in historical contexts where the ability has been seriously questioned and (4) Increasing polity's de facto legitimacy in contexts of past discrimination' (Krook, 2010, p. 201).
Krook and Childs (2010, p. 201) continue, detailing, ' descriptive representation enhances the substantive representation of interests by improving the quality of deliberation'. This illustrates one argument put forth detailing the relationship between substantive and descriptive representation; formally known as 'Critical Mass Theory' (Osmond, 2010, p. 18). Critical Mass Theory is argued by Krook and Childs (2008, p. 725) to be, 'frequently invoked to explain why women do not always appear to represent women once they are in political office'. Critical Mass details that ' only as [female] numbers increase will women be able to work more effectively together to promote women-friendly policy change' (Krook and Childs, 2008, p. 725). It is argued, by the likes of Schnellecke (2013, p. 5), 'that as soon as women constitute a particular proportion of parliament, political behaviour, institutions and public policy will be transformed'. Abels (2015, p. 2) further supports this claim recognising that 'the number of female parliamentarians has been higher in the European Parliament than in the national parliaments of member states' finding that this has led to 'female members of the European Parliament [being] able and willing to introduce more innovative ideas and take innovative action to promote the gender cause'.
Nevertheless, despite the development of strong support among political theorists in reference to Critical Mass Theory, Annesley (2010, p. 50) identifies that 'the scholarly focus has shifted away from counting the number of women in politics … to theorizing and empirically demonstrating the link between their presence and policy change'. Dahlerup, as quoted in Skjelsboek and Smith (2001, p. 113), argues that Critical Mass should be replaced 'with the new concept of critical act[s]' which is 'better suited to the study of human behaviour'. The Critical Act Effect is identified to be a process 'which will change the position of the minority considerably and lead to further changes' (Barry et al., 2003, p. 29) with the most comprehensive definition being 'the willingness and ability of the minority to mobilize the resources of the organisation or institution to improve the situation for themselves and the whole minority group' (Childs and Krook, 2010, p. 230). Lovenduski identifies that 'the critical point is whether women politicians develop some common ideas they want to fight for' (2000( , p. 128) with Annesley (2010 identifying ' an excellent example is … backbench MP Christine McCafferty [who] pushed for the removal of indirect taxation (VAT) on sanitary products in the United Kingdom'. Chaney, McKay and McAllister (2007, p. 53) have also done research on the relationship between substantive and descriptive representation looking at 'specifically the institutionalisation of women's voice, needs and interests through mechanisms and statutory duties'. They recognise the importance of the NAfW when interpreting the institutions current record in terms of parity arguing: 'the outcome of the 1999 Assembly election raised hopes that the direct or "descriptive", representation of women, who [then] comprised 42 per cent of the Assembly's total membership, would, in turn, translate into "substantive" representation, whereby women's concern and interests would be heard and taken into account in the policy-making process' (Chaney et al., 2007, p. 117). Chaney, McKay and McAllister (2007, p. 131), focusing specifically on the election of the first Assembly, argue that there is a clear relationship between both substantive and descriptive representation identifying '47.7 per cent of all plenary debates included discussion of "women's issues", and just over half of all plenary debates included references to equality'. However, Childs and Lovenduski (2010, p. 172) take a stance of a differing nature arguing, 'it has proved difficult to demonstrate that the representations on women's interests necessarily follows from the presence of women representatives'. They continue arguing the relationship between substantive and descriptive representation, is often used to explain away the problems of female representation.
In addition to this debate, a number of studies have 'found that party affiliation explains more than sex' (Waylen, 2013, p. 500) and, that 'very often women members of left-leaning parties, rather than all women representatives from the rest of the political spectrum, who make the most effort to raise women's issues' (Waylen, 2013, p. 500). Fundamentally, as argued by Mansbridge, quoted in Childs and Lovenduski (2010, p. 172), 'the sex of a candidate does not matter so long as the interests of both men and women are represented in national parliaments' which, ideally, would be true. Men should be as capable as women to represent female-only issues and, despite not having first-hand experience, this should not be a handicap to raising women's issues. However, given the concern among feminist scholars, and the clear under-representation of femaleonly issues, it appears that the relationship between both descriptive and substantive representation is potentially one of great importance. Acker (1999, p. 28) argues that 'representation is more than participation' and claims that 'the presumption that all interests can be articulated and satisfied when equal numbers of women and men participate in politics is hopeful but unconvincing'. Research by the Hansard Society, quoted in Branch, examining the NAfW also further illustrates the need for female representation identifying 'that 62% of AMs who spoke on childcare were women' (Branch, 2014, p. 10) with a further 'two thirds on equal pay' (Branch, 2014, p. 10).
What is ultimately clear is that, ' despite reservations about a researchable concept of women's interests and the complexities of constructing a "women's" policy agenda … most feminist political scientists are attracted to the potential for descriptive representation to deliver at least a measure of substantive representation' (Waylen et al., 2013, p. 494). However, what is further identified is that this also provides a simplistic and sweeping generalisation of elected officials and how they will affect the legislation concerning women's issues. Ultimately, 'political representation is best conceptualized as an active, multiface, and contingent process, driven by a broad swathe of actors with various views on group issues and interests' (Celis et al., 2014, p. 152) which will be further analysed through the following Chapters.

Section Two: The National Assembly for Wales
The NAfW, as noted by Beynon and Bennett (2012, p. 3), 'led the way with a world-first perfect gender balance of 30 men and 30 women' however, whether this was by luck or design is contested. Kirsty Williams, leader of the WLDs, spoke on the Assembly's success detailing: '… an equal split of male and female AMs has been one of the proudest achievements of the National Assembly' with 'Wales's new democratic institution [having] marked a different era in women's political participation ' (quoted in: Chaney et al., 2007, XI).
A core philosophy for the NAfW, pioneered by Ron Davies -leader of the campaign to create the NAfW -was 'inclusiveness' (Chaney et al., 2001, p. 48) detailed to be ' a concept that links the language of political rhetoric with the process of social change' (Chaney et al., 2001, p. 48). Chaney, Hall and Pithouse (2011, p. 48) argue that this 'rhetoric' is unique as it has 'structures and mechanisms for furthering equality outcomes not found in other devolved assemblies'. The NAfW has an 'institutional blueprint' (Chaney et al., 2007, p. 53), lobbied for by women activists who capitalised on a 'short and intensive period of opportunity' in order achieve a new "gender settlement" (Chaney et al., 2007, p. 53). This, coupled with the newness of the institution and lack of traditional constraints, has contributed to the NAfW's ability to champion female representation within the UK.
The National Assembly Advisory Group (NAAG), publishing their recommendations for the devolved legislature in late 1988, put forward extensive proposals promoting equality in order to ensure a gender neutral blueprint for the institution. The NAAG stated the Assembly ought to be, ' democratic and inclusive, flexible and responsive, efficient and effective and transparent and accountable' (Chaney et al., 2007, p. 61). Chaney and Fevre (2002, p. 4) identify that: '… this legal duty goes beyond the statutory requirements placed upon other UK legislatures for it effectively requires the government to … "make appropriate arrangements with a view to securing … equality of opportunity for all people"'.
With this legal duty in place, the NAfW has had extensive success concerning female representation linked to a number of factors including the architecture of the building, the institutional arrangements and the familyfriendly ethos of the institution. However, despite this success, further areas of improvement have also been noted such as an increase in the size of the Assembly. These areas are assessed in greater depth throughout this Chapter.

Architectural Design
The NAfW illustrates both inclusivity and consensual politics, not only through its equality blueprint, but through the physical architecture of the institution. The NAfW, designed by Richard Rogers, is identified by Boutkhill and Touaf (2009, p. 21) to reflect the ' core philosophy of openness and transparency' further recognising that '[t]he debating chamber is the physical and metaphorical centre of design and is surrounded by public space so that … it is open and accessible to all' (Touaf and Boutkhill, 2009, p. 21 In an interview conducted with the DIT they identified that it was possible that women could be more encouraged to run as AMs due to the:

'… fact that it's a new modern institution with a chamber that is designed in a circular configuration. This could help to augur for a less adversarial atmosphere than a non-circular configuration might generate' (DIT1).
This suggestion by the DIT of the success of the architectural aim of inclusivity illustrates how the creation of a new Assembly, moving away from the traditions of Westminster, may well have encouraged female participation particularly in the form of running for elected office.

Institutional Arrangements
In addition to the architecture, the institutional arrangements in place have also proved pivotal concerning the achievement of gender parity. It is important to note that the NAfW currently functions under a conferred powers model and therefore has limited jurisdiction in certain policy areas; one being equal opportunities. However, despite this, gender parity has been placed 'higher up the political agenda in Wales than ever before' (Fevre and Chaney, 2002, p. 10). Fevre and Chaney (2002, p. 21) identify that 'equality of opportunity is beginning to be addressed in a systematic way', but nevertheless, Chaney et al., note that there is 'a need for continuing work to overcome prevailing discrimination and inequality ' (2007, p. 206), illustrated in Figure 2 below depicting the limited number of women in certain areas of public life. Fevre and Chaney (2002, p. 21) argue the introduction of a 'statutory equality imperative' and the physical architectural design, has resulted in, 'significant progress' in relation to ' achieving gender balance amongst elected politicians'. However, in an interview conducted with a WLP AM she identified that: '… the Assembly does not provide enough provisions and there are issues. It has done better than Westminster in terms of equality measures such as family friendly working hours so they think they have done well and although it is not awful it could certainly be improved'. (Labour AM 2) Chaney (2003, p. 5) further supports this identifying that, 'despite such undoubted progress, much further work remains to be done in developing appropriate institutional arrangements' such as improved familyfriendly hours. Bennett (2014, pp. 3-5) argues that the achievement of parity, 'set a benchmark for fairness and lit a beacon of hope', and that 'since then, overall progress has been unacceptably slow', with 'fewer women [AMs and] fewer women in the Cabinet'. This is illustrated in Figure 3 below demonstrating that, despite the success of 2003, the number of female AMs has declined since 2005.
Please note that the 2003-2007 term is split into two sections due to the by-election of Patricia Law during this period.
However, despite the decline in female AMs in recent years, it is important to note that there have been a number of institutional arrangements to encourage equality to become a cultural norm; one such reform being the introduction of family-friendly hours. Specific hours were introduced in order to allow members in caring roles to fulfil their duties and also to allow members to spend more time with their families. However, a WLP AM who was interviewed argued that: '… the Assembly says is has family friendly working hours with, for example, the living expenses arrangements having been adapted but I commute to the Assembly and am only able to stay here one night a week … I leave the house at around seven and then get home around nine so if you have small children you couldn't possibly do that' (Labour AM 1).
Furthermore, in interviews conducted by Chaney MacKay and McAllister (2007, XI) in a prior study, a WLD further supported the claim heard in this research detailing: '… the adoption of "family friendly" hours … has been a step forward, although whether they are very friendly to AMs who live outside of Cardiff is another matter'.
Jones and Osmond (2001, p. 7) identify that there was a 'rigid rule that meetings of the Assembly as a whole could not extend beyond 5.30pm'. However, due to the recent legislative changes there have been examples of AMs sitting until much later in the evening which will have brought with it a disproportionate impact on AMs living in rural areas. Current sitting hours, as prescribed by Standing Orders, detail that timetabling should not schedule work before 9:00am or after 6:00pm in order to protect rurally seated AMs. Yet, if you live in a rural area of Wales finishing at 6:00pm could still result in anti-social working hours. Therefore, although a positive reform, family-friendly hours are limited concerning more rurally situated AMs within the NAfW. However, it is argued by Trench (2004, p. 42) that family-friendly hours have only been implemented in their current form as 'women members … added more weight' to their enactment illustrating the impact that a higher number of women within the institution has had.
In addition to family-friendly hours, there are also other potential areas of development within the NAfW that could benefit the ethos and working parents in particular. Currently there is not an on-site crèche although a salary sacrifice system is available. However, it is important to note that despite the absence of an on-site crèche, and the feedback concerning family friendly working hours, the salary sacrifice system is seen to be very effective. Furthermore, when combined with the statutory imperatives initially introduced, this has produced a range of successful institutional arrangements within the NAfW which can be seen to promote the gender equality ethos.

The Size of the Assembly
Indirectly related to this, and a factor that has been raised repeatedly both in interviews and academic research, is the work load of AMs due the relatively small size of the institution. McAllister, Cole and Stirbu (2013, p. 7) identify that 'the Assembly is less than half the size of the Scottish Parliament (120 MSPs) and substantially smaller than the Northern Ireland Assembly (108 MLAs)' with the size of the Assembly having caused controversy since it was first established. A WLP AM identified that: '… the size of the Assembly holds great barriers for women due to the fact that the work-load for every AM is so high'. (Labour AM 1).
McAllister, Cole and Stirbu (2013, pp. 6-17) identify that, 'the small number of AMs lies at the heart of many of the capacity restrictions in the Assembly. In particular, it [is] associated with the intense workload of the third Assembly and the reduction in scale of committee scrutiny in the fourth Assembly'.  Therefore, although not directly related, the limited number of AMs can be seen to cause barriers to areas such as scrutiny, which have been exacerbated due to the shift to full law-making powers. However, few women would cite the capacity of the NAfW as a barrier to their desire to become an AM so the effect it would have on female representation remains questionable.

Discussion
In analysing the NAfW it has become apparent that there are a number of successful initiatives in place in order to protect and encourage female representatives whilst there also being a need for further reform to policies such as family-friendly hours. However, as identified by the DIT:

'… provisions to encourage equality can be in place throughout the organisation but there is a limit on what an organisation can do; in terms of increasing female representation, that is a matter for political parties'.
Therefore, although institutionally important, the statement by the DIT illustrates that, arguably, the institution has less of an impact on parity than that of the political parties.

Section Three: The Political Parties
Political parties are instrumental in the achievement of gender parity with it having 'long been recognised that party selection procedures … are of central importance in political recruitment' (Mitchell and Bradbury, 2004, p. 288) and, ' ahead of the first election to the … [NAfW] in 1999 … were emphasised as a potential opportunity for innovation ' (Mitchell and Bradbury, 2004, p. 288). This innovation came, in part, as the NAfW was 'being elected afresh, and [was] utilising a novel additional member electoral system, combining constituency and regional list members' (Mitchell and Bradbury, 2004, p. 288). Within the NAfW there are five parties as of the 2016 election being: the Welsh Labour Party (WLP), Plaid Cymru (PC), the Welsh Liberal Democrats (WLD), the Welsh Conservatives (WCP) and the UK Independence Party (UKIP). However, the analysis in this chapter will span from 1999-2016 and therefore will exclude UKIP.
As well as the current term of the NAfW, past results will also be important to assess with the table below showing female representation in the NAfW from 1999-2016. Table 1 illustrates the notable strides taken regarding female representation in some parties, with perfect parity being achieved in 2003. However, the decline in female representation among certain parties is troubling. It is, of course, important to identify there are certain variables causing anomalies in these results, such as seat number, however, the table below demonstrates, as identified by Richardson, 'the difficulties in maintaining a gender balance of political representatives' (Robinson and Richardson, 2015, p. 54). In addition to this, the 100% female representation by the WLD must be identified an as anomaly due to their retention of only one seat in the 2016 election held by Kirsty Williams.

The Welsh Labour Party
The WLP have seen the highest levels of female representation since 1999 which is illustrated in Figure  4 below demonstrating the WLP's success in terms of female representation.
The WLP, as identified by Charles and Aull-Davies (2010, p. 4), ' committed itself to a gender balance in the devolved legislature', which, 'led to the policy of twinning constituencies' for the 1999 NAfW election. Twinning entails 'that for each pair of constituencies a woman and a man [are] selected as candidates' (Charles and Aull-Davies, 2010, p. 4) which proved highly successful resulting in 'fifteen women [being] elected out of 28'. During interview, a WLP AM detailed: Osmond (2010, p. 8) further supports this identifying that ' although the policy was effective in bringing female representatives to the Assembly there was a good deal of unhappiness about the policy's imposition'. Krook (2010, p. 145) identifies that, on the introduction of Twinning, ' opponents sought to undermine the policy by threatening legal actions and refusing to cooperate' however, ' despite these struggles, legal challenges never materialized, although opponents expressed some hostility toward the women who were selected through the twinning policy'. Furthermore, there were other limitations of Twinning, namely that it could not be as successfully implemented once there were incumbents within the NAfW. Therefore, the WLP had to adopt a new form of positive action for the 2003 election, with Osmond (2010, p. 8) identifying that, 'perhaps surprisingly, given the anguish generated by the first policy' six of the sixteen 'seats seeking new candidates' were 'required to be all-women-shortlists' (AWS). However, much like Twinning, the quota caused numerous objections across Wales with Cynon Valley being on such example. Ann Clwyd, the AM from 1999-2003, decided to step down with the WLP implementing an AWS to replace her. However, Alun Williams, secretary of the constituency, stated 'the constituency needed the best candidate, regardless of their gender' (List, 2014) which is a leading concern with the implementation of quotas. A WLP AM responded to this concern identifying that:

'… I wouldn't be here without Twinning so it has worked but it did leave a nasty taste in people's mouths. I don't disagree with the policy but the way it was implemented was damaging'. (Labour AM 2)
'… unless you take the point of view that we're a meritocracy and that white males in their 50s get elected because they're the best candidate … then clearly the previous system was biased towards white males' (Labour AM 1). Furthermore, the WLP success due to AWS cannot be denied. Due to their sustained positive action, and ' as the largest party by some distance, [the WLP] has been the most important factor in achieving gender parity' (Electoral Reform Society, 2011, p. 9) with quotas having been successful in maintaining at least a gender balance in every election. However, it is important to note that, 'women [were] in their most vulnerable seats' (Electoral Reform Society, 2011, p. 11)  This is further supported by Brooks and Gareth (2016, p. 19) arguing that, of the ten seats that they predicted to have been 'vulnerable seats held by Labour in 2011, eight [were] being defended by women'. This illustrates that, despite the positive action measures that the WLP have deployed, the fact that women were in more vulnerable seats means that the 2016 election could have resulted in damaging loses for female representation. Ultimately, Labour women successfully defended the majority of these seats and yet, it is important to note it was both by small margins and by no means the expectation.
Furthermore, although there has been notable success, the lack of organisation within the WLP is a leading critique about positive action measures which was identified by a WLP AM stating: '… my bug bare is that they make it up as they go along with there being no over-arching strategy. I think you need to look at the whole of Wales … as to who should have [AWS] and develop a strategy' (Labour AM 2).
Furthermore, the WLP candidate selection process has also been heavily criticised for not encouraging female representation. A WLP AM detailed that: '… my party feels that people who have attended a lot of meetings and does evening work for the party is more entitled to get endorsed to become an AM.

However, the meetings are held at half-six and seven o'clock so if you have family responsibilities it is very difficult to come out'. (Labour AM 1)
A further WLP AM also raised similar concerns when interviewed stating: '… the political parties are very isolated. The odd thing is that things that go on in the party would never happen in the work place. It is still very oldfashioned and traditional for example, when they short list people the criteria aren't published as that's not allowed'. (Labour AM 2) Mitchell and Bradbury (2004, p. 297) argue that, despite quota success, there has not been ' a general cultural shift towards gender parity in party recruitment' due to the fact that the party selection process shows little to no signs of change. The WLP have also had the advantage of incumbency having been in either a majority or Coalition Government since 1999 and yet, as identified by Beynon (2014, p. 5) there are 'fewer women in the Cabinet' than ten years ago. However, despite the claim by Beynon, it is important to note that certain Cabinet positions hold a heightened importance with both Edwina Hart and Jane Hunt holding top positions, including Minister for Health. This therefore illustrates that a gender balanced Cabinet is not crucial as it is dependent on the positions in question. Furthermore, despite certain critiques of the WLP an AM stated: '… I think the party has done well to have arrangements of this sort in place. I would certainly rather have them than not have them' (Labour AM 1).
Ultimately, the WLP have made clear and notable strides towards the achievement of gender parity, having done so through a number of positive action measures such as Twinning and AWS. However, what becomes clear from analysis of the WLP is that their policies have often been disorganised and, despite their success, there are areas in need of improvement such as their methods of implementation.

Plaid Cymru
PC, currently the second largest party, saw devolution as '[providing] an impetus to continuing consideration of party change' (Bradbury et al., 2000, p. 62) and, as identified by Charles and Aull-Davies (2010, p. 4), 'took specific measures to ensure a gender balance'. PC have illustrated a commitment to a gender balance electing their first female leader, Leanne Wood, in 2012 with a PC AM detailing the impact of having a female leader saying: '… in terms of encouraging women into politics it speaks volumes as she is so high profile. She embodies the ambition that people can have and if you see that at the front of the party it does really encourage people to become more involved in politics' (Plaid Cymru AM 1).
PC have also implemented a range of quotas which have increased female representation, achieving a perfect gender balance in 2003. Initially, as identified by Bradbury et al., PC conducted ' a … formal effort to attract more women … by including in the person specification criteria agreed with the party's gender balance commission and also holding sessions designed to encourage women' (2000, p. 59). However, due to the limited number of candidate applications all reached interview stage illustrating the somewhat limited effect this had concerning female representation.
'Ultimately … the best chance of election for [PC] … came from selection high up on the regional lists' (Bradbury et al., 2000, p. 169) and therefore PC were far more innovative here. As identified by Osmond (2010 p. 8), 'to a large extent the combination of first-past-the-post for the 40 constituencies and the Additional Member to-up proportional vote for the 20 list seats defined how … [PC] had to implement policies to increase women [AMs]'. Therefore, as further recognised by Osmond (2010, p. 8), PC's 'interventions have focussed on Regional Lists, where it has won many of its seats. PC 'adopted the "zipping principle" for its party list, with women in first and third positions' (Sawer and Tremblay, 2006, p. 15) with a PC AM detailing: '… regional lists are our most important change -I wouldn't be here without zipping and these kind of policies'. (Plaid Cymru AM 1) Bradbury et al., identifies that zipping 'involved separate votes for male and female shortlisted candidates … and rank-ordering only the first three candidates, starting with the top woman, followed by the top man, and then the second woman ' (2000, p. 169), ultimately resulting in 'five of Plaid Cymru's seven regional list members being women' (Mitchell and Bradbury, 2004, p. 297) in 2003. However, there was 'some unhappiness' (Bradbury et al., 2000, p. 169) among the party members and the public although, 'the under-representation [of] women for Plaid in the constituency contests confirmed to moderate opinion in the party that zipping was necessary' (Bradbury et al., 2000, p. 169).
However, PC, as identified by Charles and Aull-Davies, have 'recently abandoned the policy of putting women at the top of regional lists although if a man heads the list a woman has to be in second place ' (2010, p. 5). This was due to controversy ' caused by the fact that the popular ex-leader, Dafydd Wigley, was second on the North Wales list due to the policy' (Electoral Reform Society, 2011, p. 9) and therefore the policy was abandoned in 2007. Furthermore, PC have also fluctuated in terms of seats since 1999 and therefore, with both the variation in seats and removal of positive action, the percentage of women has varied as illustrated in Figure 5 above.
A PC AM detailed in interview:

'… I don't think that I would be here without zipping but it has now changed with there being if there is a man at the top of the list a woman has to be second and that has meant our female numbers have declined at times' (Plaid Cymru AM 1).
Furthermore, PC also faced a similar scenario to the WLP with concerns of damaging election losses in 2016 predicting as low as 6 seats for the party. However, once again results were less harmful to female representation than expected with PC ultimately increasing their seat number to eleven. However, among these 11 seats only four were held by women with this being their lowest female return since 2003. As identified by Osmond (2010, p. 4) the 'gender balance in the [NAfW] did not come about by accident. During the first decade of devolution … Welsh Labour and Plaid Cymru used measures to ensure that female candidates were given priority or encouraged to stand for office' and the impact of these quotas implemented by PC are clear from the achievement of perfect parity in 2003. However, 'for [PC], a poor electoral performance coupled with their retreat on positive actions makes regaining traction on the issue more difficult' (Electoral Reform Society, 2011, p. 11). It is also now problematic to reintroduce the policy as this would 'be, effectively, an attack on the male incumbents now in place' (Electoral Reform Society, 2011, p. 9) and therefore, although influential in the first decade concerning female representation PC have faltered in attempting to continue their high rates of female representation.

The Welsh Liberal Democrats
The WLD, despite an electoral disaster in the 2016 election, had prior to this point not ' escape[d] demands for an increase in female representation' (Bradbury et al., 2000, p. 170). The WLD party leader, Kirsty Williams, detailed on her election to leader in 2008 that 'the party had broken the mould by electing a woman' (Williamson, 2008) and yet, despite the election of a female leader, it is debated as to whether the WLD policies reflect a desire for gender equality. Bradbury et al., identifies that for the 1999 election the WLD ' decided to adopt an equal-opportunities approach to gender balance, laying down that half the shortlists should be women ' (2000, p. 171) which, as recognised by Lovenduski (2005, p. 128), allowed the WLD to have an ' equal number of women and men representatives without quotas'. However, Bradbury continues arguing, ' as a small party, simply getting enough candidates was an issue' (Bradbury et al., 2000, p. 170) with there being ' a lack of women applicants for many constituencies' (Bradbury et al., 2000, p. 172). This subsequently meant, as identified by Bradbury et al., 'the creation of genderbalanced shortlists [was] impossible in more than one of the Welsh regions ' (2000, p. 173).
Furthermore, Bradbury et al., identifies that, 'when it came to list selection, the party experienced controversial debate as to whether to introduce zipping (2000, p. 172) and, following extensive internal division, opting 'instead … to extend the equal opportunities approach of genderbalanced shortlists ' (2000, p. 172) illustrating a lack of positive action by the party. In addition, there was also 'some resentment at the perceived pressures on members to place' (Bradbury et al., 2000, p. 173) 2000, p. 173). However, it is important to identify that 'the party […] did reasonably well in female representation' (Bradbury et al., 2000, p. 73) with equal representatives in 1999.
WLD success is however, questionable with it argued that gender parity was 'the result of accident rather than design' (Hassan, 2009, p. 48). Mitchell (2004 identifies that 'in defending all six seats … in 2003' (Mitchell and Bradbury, 2004, p. 297) the WLD maintained gender parity however, 'this chance outcome of gender balance may actually distort future debates about the need for formal procedural changes' (Bradbury et al., 2000, p. 173) due to the fact that the WLD success has been argued to be largely due to chance circumstances. The Electoral Reform Society (2011, p. 10) detailed that, on the election of AM Parrott due to Mr. Dixon's disqualification in 2011, 'if previously, parity had been achieved more by luck than design -in this case the increase in the proportion of women was through a miracle of chance and circumstance'. Figure 6 above illustrates both the result that would have occurred if Mr. Dixon had not been disqualified (WLD 2) and demonstrates the chance circumstance that was the election result in 2011 (WLD 1).
Furthermore, and unfortunately for the WLD, they have experienced near obliteration in the 2016 elections returning just one AM. Their female contingent may now amount to 100% with their leader, Kirsty Williams, the only returning WLD however, in a party of one this can hardly be hailed as an achievement for female representation.
The WLDs provide an anomaly to analyse female representation due to the fact that they have maintained parity despite their limited use of quotas which, at first glance, implies their success. However, in analysing their procedures in both regional and list selection the success of the WLD in terms of female representation is seen to be the result of luck and chance rather than formulated design.

The Welsh Conservative Party
The WCP have consistently returned the fewest number of women to the NAfW, despite having been the only party to increase their seats in every election up to 2016, therefore making their candidate selection process important to analyse. It is identified by Bradbury et al., that 'the Conservatives … felt no pressure to respond to devolution with their own "new politics" (2000, p. 174) choosing to ' oppose quotas in favour of a policy of persuading constituencies to select women to winnable seats' (Dahlerup and Krook, 2013, p. 204). The Conservatives, although having begun with a modest nine seats, are still behind concerning female representation achieving the smallest percentage number of female AMs in all assembly elections with their results demonstrated in Figure 7.
The WCP decided on 'the simple adaption of previous UK-wide procedures' (Bradbury et al., 2000, p. 174) and therefore 'the selection procedure was devised by the party chair and vice-chair' (Bradbury et al., 2000, p. 174) which 'prevented any debate' (Bradbury et al., 2000, p. 174) and was concerning regarding the representation of women. In terms of the constituency selection, claims of central party control arose which was argued by the likes of Bradbury et al., to make the process 'both undemocratic and unfair' (Bradbury et al., 2000, p. 176) although, in terms of gender equality, the regional list system produced far heavier criticism. The WCP made ' candidacy … closed to aspirants who had not been selected for a constituency contest' (Bradbury et al., 2000, p. 176) resulting in the gender balance being compounded and However, this is argued to be only due to Contagion Effect. Contagion Effect is identified to have encouraged the Conservatives in terms of female representation due to feeling 'pressure to move towards the mainstream' (Lucas, 2006, p. 43). Furthermore, the WCP failed to adopt any other quotas as they resulted in ' considerable and hostile coverage in the [UK]' (Dahlerup and Krook, 2013, p. 54). Bradbury and Mitchell (2004, p. 289) identify that, 'had [the WCP] repeated their nine seats of 1999 it would have once again been an all-male group [however], in increasing their seats to 11 in 2003 they absorbed two lower placed women'. This can also be seen regarding the 2011 election where the WCP again increased their seats and increased their number of female AMs. However, the 2016 election, where the WCP only managed to return eleven seats saw the end of this trend with only three female members.
Ultimately, the WCP have been the party that have introduced the least in terms of positive action. Their party strategy, although in part encouraging to women, has had little to no effect in increasing female representation with election results being heavily influential concerning their number of female representatives. Furthermore, the fact that their female AMs were defending particularly vulnerable seats in 2016 demonstrates little promise for the future with it once again argued that their return of female members was due to lack rather than design.

Results
Having analysed each party individually a recurring theme throughout has been, not only the success of quotas, but continued concerns relating to the candidate selection process. The lack of party regulation in place has meant that the parties have become out-dated and discrimination has been allowed to take-root through party functions. The analysis further demonstrates the success that has been experienced concerning female representation by parties who implement quotas. The Electoral Reform Society (2011, p. 9) identified that 'Labour's record in Wales is … an illustration to all parties of what can be achieved through political will and positive action' and this can be further be attributed to the work of PC. The 2016 election was argued to be the most concerning election in terms of female representation in the NAfW's history and, despite the fact that the number of women did not 'flat line', as predicted by the likes of Brooks and Gareth (2016, p. 7), this was partly due to luck rather than design with women defending a number of vulnerable seats.

Conclusion
Throughout this dissertation, discussion about the NAfW and its female representation has been undertaken through analysis of the contemporary literature, the devolved institution and the political parties. This has led to the conclusion that the political parties can be considered the most influential when affecting the gender parity of the institution. Both the WLP and PC have led the way in terms of positive action and have had extremely encouraging results proving that success can be achieved when quotas are introduced and combined with a progressive organisation. However, the WLD and the WCP have had less of an impact concerning gender equality, introducing little to no positive action and yet still experiencing the benefits embodied by having a gender equal institution.
In addition to the parties, it is important to note that the NAfW, institutionally, has also played a vital role in perpetuating an ethos of equality. The institution has achieved this through measures such as family friendly hours and salary sacrifice child care and, despite the need for some improvements, the policies have facilitated a more gender equal environment. This is partly, as previously cited, due to the youth of the institution and the ability to work with a blank slate creating a gender neutral blueprint, which has allowed for the organisation to become far more feminised. Statutory Amendments, such as the clause initially introduced detailing by law that equality of opportunity must be upheld, can be seen to have been influential in terms of the running and functioning of the NAfW. Furthermore, this can also been see through the architecture of the building which has contributed to the desire for transparent politics.
Ultimately, the combined research into the literature surrounding this issue, the NAfW as an institution and the political parties has allowed for analysis to be drawn as to why the NAfW has experienced such success. Through analysis of Pitkin's typology, specifically the relationship between substantive and descriptive representation, an insight into contemporary feminist theory was able to frame the analysis performed in this dissertation and provide crucial understanding relating to different forms of representation. This dissertation has specifically focussed on the descriptive representation of women due to the limited ability to discuss, in depth, substantive representation. Therefore, in terms of further research, understanding the substantive representation of women within the NAfW would allow for a more exhaustive analysis of what impact these quotas and positive action measures have had, and the longevity of their influence.