Exploring the Pathologies of One-Party-Dominance on Third Sector Public Policy Engagement in Liberal Democracies: Evidence from Meso-Government in the UK

In liberal systems governing-party-turnover and third sector organisations’ engagement in public policy-making are seen as key factors maintaining the health of democracy. However, a significant lacuna in current understanding is the effect on engagement when governing-party-turnover is absent. Accordingly, drawing on qualitative interview data, this study examines the effects of one-party-dominance (OPD) in Wales; a regional polity in the UK where the Left-of-centre Labour Party has held uninterrupted government office since a new meso-legislature was created in 1999. The findings reveal OPD introduces a range of pathologies related to party institutionalisation, path-dependency and cognitive locks. These affect third sector organisations’ resource dependency and strategic bridging to elected representatives. The resulting democratic ills are self-sustaining and include diminution of NGOs’ autonomy, trust and criticality. This study’s wider significance lies in underlining the importance of governing-party-turnover- not only to effective third sector public policy engagement, but also the health of contemporary liberal democracies.

Resumen En los sistemas liberales, la rotación del partido que gobierna y el compromiso de las organizaciones del sector terciario en la formulación de políticas públicas son vistos como factores claves que mantienen la salud de la democracia. Sin embargo, una laguna significativa en la comprensión actual es el efecto sobre el compromiso cuando la rotación del partido que gobierna está ausente. Por consiguiente, recurriendo a datos cualitativos de entrevistas, el presente estudio examina los efectos del dominio de un partido en Gales; una política regional en el Reino Unido en la que el Partido Laborista de Centro Izquierda ha ocupado los cargos gubernamentales de manera ininterrumpida desde que se creó una nueva mesolegislatura en 1999. Los hallazgos revelan que el dominio de un partido introduce una serie de patologías relacionadas con la institucionalización de partido, dependencia de la trayectoria y bloqueos cognitivos. Todo esto afecta a la dependencia de recursos de las organizaciones del sector terciario y al nexo estratégico con los representantes elegidos. Los males democráticos resultantes se autosustentan e incluyen la disminución de la autonomía, confianza y capacidad de crítica de las

Introduction
This study addresses a key lacuna by exploring the effects one-party-dominance (OPD) on third sector engagement in public policy-making. This matters because, as the scholarly literature attests, over recent decades there has been renewed political emphasis on the democratic benefits of third sector engagement in the work of government. These include the bolstering of trust, accountability and participation (Pateman 1970;Putnam 1993Putnam , 2000Hirst 1994;Giddens 2000;Bullock et al. 2001;Kendall 2009;Chaney 2012Chaney , 2014. In definitional terms, this paper follows existing research practice (Casey 2004) using the umbrella term 'third sector' to refer to the principal collective signifiers associated with non-government advocacy and service organisations (including, 'voluntary sector', 'civil society' and 'non-profit sector'). As a burgeoning literature confirms, power inequalities in the state-third sector nexus are amongst the principal barriers to reaping sought-after democratizing benefits (Schwartz 2001;Kendall 2009;Sinclair 2011). In this electoral politics play a pivotal, if frequently overlooked role. This is because governing parties' attitudes and practices towards the sector vary across the political spectrum; with some favouring neo-liberal, market-based approaches (Fyfe 2005) and others coproduction, mutualism and inter-sectoral working (Pestoff 2012).
Accordingly, the type of party elected to form the executive clearly matters to understanding the patterns and processes shaping NGOs' engagement in public policy-making. For this reason the turnover rate of governing parties is pivotal not only to third sector-government relations but also the vitality of a given democratic system. According to Huntington's seminal work (1991) turnover is an effective predictor that democracies have become well established and will endure. Inter alia, turnover shapes: circulation of the political elite (Mosca 1939), opportunities for new political leaders and champions of sectoral interests (Schlesinger 1966)-and the introduction of innovation and changes in public policy and administration (Bunce 1981;Brady 1988).
Indeed, as Jackson (1994, p. 270) observes, the absence of turnover presents a number of threats to democracy-or, 'a danger of the three A's setting inarrogance, apathy, and atrophy'. A key lacuna in the scholarly literature is whether such assumptions are supported by empirical data on third sector organisations' policy engagement. As Casey (2004, p. 247) cogently notes: 'weak parties and a fragmented party system are likely to provide more opportunities for independent third sector organisations' (TSO) policy participation; [in contrast,] strong parties… are likely to exclude TSOs or force them into close relationships with existing party structures'. In order to explore such matters, in the following discussion the negative effects of OPD on third sector policy-making are captured by the term 'pathologies'. This follows Hogwood and Peters' (1985) seminal work comparing the human body and the body politic. This uses the language of medical pathology to investigate the disorders and challenges some governments experience in making and implementing public policy (an approach that has subsequently been applied to other aspects of political science, see for e.g. Green and Shapiro 1994;Yap and Kim 2008). 'Pathologies' is a particularly apposite term for, as the following discussion outlines, negative effects of OPD on third sector policy-making have wider ramifications for the 'health' of liberal democratic systems-including issues of trust, representation, legitimacy and accountability.
Notwithstanding the centrality of governing-party-turnover to the state-third sector policy nexus, strikingly little progress has thus far been made in addressing the invocation made three decades ago that, in 'extending the polity model to deal with different states and regimes, including the development of neo-corporatism… the largest vacuum lies in the analysis of one-party regimes' (Jenkins 1983, p. 527, emphasis added). The following analysis heeds this call. It must be acknowledged that a number of studies have examined the role of the third sector in one-party systems. However, they have centred on authoritarian regimes (Huntington 1991;Chazan 1992;Clarke 1998, Heurlin 2010Hsu and Hasmath 2014). Insufficient attention has been given to the impact of OPD on the third sector's policy role in liberal democracies. As Greene (2010a, p. 155) observes, such 'dominant party systems are ''odd ducks'' that combine genuine electoral competition with the absence of turnover' (see also Giliomee and Simkins 1999). Prominent international examples include Japanese politics 1955(Inoguchi 2005 and South African politics 1994-present (Friedman 2003;Bénit-Gbaffou 2012). Far from being an obscure and remote eventuality, OPD is a mainstream issue; all liberal democratic systems have immanent potential for periods of OPD.
There are competing definitions of what constitutes OPD. In his seminal text Duverger (1954, pp. 308-309) alludes to a dominant party as one whose 'influence exceeds all others for a generation or more […whose] doctrines, ideas, methods-its style, so to speak, coincide with those of the epoch [… such that] even enemies of the dominant party, even citizens who refuse to give it their vote, acknowledge its superior status and its influence'. In a similar vein, Butler (2009, p. 159) notes that OPD: has been used to refer to the protracted electoral and ideological dominance of a single party in a representative democracy. It requires, but suggests more than, a series of electoral successes. OPD implies institutions that translate electoral success into political power; the capacity to attract support from substantial electorates over an extended period; the presence of a unifying historical project; and the ability to dominate the policy agenda of a country (see also Pempel 1990).
Elsewhere the literature on OPD systems reveals a number of defining characteristics (Greene 2010b, p. 823).
(1) A power threshold for the dominant party ('it means holding the premiership, [and] at least a plurality of legislative seats'); (2) A longevity threshold (this 'should capture the notion of a dominant party system as a stable pattern of interparty competition, but one not so restrictive as to make the category disappear'). Typically a four election/20 year span; (3) and finally, electoral competition must be meaningful (this entails three procedural elements (a) a legislature that cannot be dismissed by the executive and is chosen through regular popular elections, (b) opposition forces that are allowed to form independent parties and compete in elections, and (c) the incumbent does not engage in outcome-changing electoral fraud (without which dominant party rule would have ended).
As the following discussion reveals, when OPD systems emerge they have significant implications for the democratic health of the polity as a whole; not least in relation to the policy-making activities of third sector organisations. The main potential problem is that OPD skews the political context for third sector-state interaction; thereby subverting the beneficial democratizing elements of criticality and resource exchange seen in other liberal democracies. In such instances, as Giliomee and Simkins (1999, p. 340) note, 'the vital elements of democracy, namely genuine competition and uncertainty in electoral outcomes, are removed in a process that is self-sustaining'. The way that this impacts on third sector organisations' policy work is a key knowledge gap. It matters because, as predicated by pluralist theory (Dahl 1961), exogenous civil society interests perform a pivotal role through critical engagement in policy-making as part of the wider process of holding government to account.
For the latter reason OPD has the potential to introduce a range of pathologies to democratic governance. For example, as Brooks (2004, p. 18) notes, it may result in 'a blurring of the boundary between party and state, which has the effect of reducing the likely formation of independent groups from within civil society that are autonomous from the ruling party'. To examine such matters the following analysis draws upon data from 100 one-hour research interviews undertaken in 2011-2014 with third sector managers and policy officers in Wales, a regional polity in the UK. The latter is a propitious research context for the Left-of-centre Labour Party has held a hegemonic position in politics throughout the post-war period. This is aptly illustrated by the fact that the Party has continuously held government office following state decentralisation and the creation of a meso-legislature in 1999. Moreover, a singular provision in constitutional law means that meso-government is required to publish and uphold a judiciable Third Sector Partnership Scheme that is subject to annual review. As Kendall (2000, p. 542) observes, such attempts at formalised state-third sector working represent 'an unparalleled step in the positioning of the third sector in public policy… [and constitute] a calculated attempt to supplement the 'vertical' policy arrangements in specific fields that have dominated in the past'.
The remainder of this paper is structured as follows. First, the research context is outlined. Discussion then centres on relevant aspects of the literature on OPD and third sector organisations' policy role in liberal democracies. Following a summary of the methodology, attention turns to the research findings on the effects-or pathologies-of OPD on third sector policy engagement. 1 The paper concludes with a discussion of the principal findings and their significance.

Research Context
The third sector in the UK comprises 163,763 organisations, has an income of $58.9 billion, and employs 765,000 staff. In recent years peak annual funding for the sector totalled $61 billion (NCVO 2012, p. 6). 2 It is made up of four territorial policy frameworks covering the three constituent nations and province in the UK. These geographical distinctions existed during the pre-existing era of administrative devolution, 3 yet they have taken on heightened significance with the (re-)establishment of legislatures for Scotland, Wales and Northern Ireland in 1998/1999. This has accentuated divergence of the prevailing legal and policy frameworks applying to the third sector(s) in the UK and resulted in contrasting funding regimes-as well as levels and structures of engagement with government. Crucially, it has meant third sector-state relations have been shaped territorially specific patterns and processes of meso-level electoral politics.
Contrasts also exist in the size and operation of the sector in each constituent polity (reflecting the population base of each). In Wales, the present case study area, it contains 32,798 organisations, employs 34,370 people and has an income of $2.6 billion (WCVA 2012, p. 3) In contrast the third sector in Scotland comprises 45,000 organisations, has a turnover of $7 billion and employs 137,000 people (SCVO 2012, p. 2); whilst in Northern Ireland it is made up of 4,700 organisations, has an income of $915 million and a workforce of 26,737 (NICVA 2009, p. 4). 4 The present locus of enquiry is appropriate because 'devolution'-or move to quasi-federalism in the UK (Gamble 2006) is part of the wider international trend of state restructuring. Under the revised governance structures the Welsh, Scottish and Northern Irish governments have responsibility for the respective third sectors in their territories. Moreover, in each case, structural mechanisms and political discourse have emphasised co-working between government and the sector. As noted, the third sector in Wales presents a particularly propitious research context, not only because of singular structural arrangements such as a statutory partnership (established in constitutional law) that links government ministers with a series of 1 Denoting third sector organisations' involvement in a range of policy-making activities with government including lobbying, membership of expert groups and working groups, and policy consultations. 2 2007/2008. 3 Territorial government ministries oversaw aspects of public administration in Scotland and Wales. 4 The foregoing third sector data need to be viewed in the context of the population of each polity: England, 53 million; Scotland 5.25 million; Wales 3.1 million; Northern Ireland 1.8 million (Circa April 2011). state-sponsored policy networks (designated 'the Third Sector Partnership Council'), but principally because, over the past 150 years, Welsh politics has been characterised by single party dominance.
Throughout the period to the First World War the country experienced one-party domination under the Liberal Party. Thus for example, in the 1885 election, the Liberals won thirty out of a total of thirty-four Welsh seats in the UK parliament (Morgan 1981, p. 26). Liberal fortunes waned in the first decades of the twentieth Century, principally owing to decline in religious non-conformism and the Welsh language; hitherto both were integral to party support. Subsequently, following a brief inter-war hiatus, the Labour Party has been the pre-eminent political force in Wales (McAllister 1980;Hopkin et al. 2001). Having secured 58.6 % of the vote in the 1945 election, Labour has subsequently had a majority of Welsh Members of Parliament (and frequently an absolute majority of Welsh votes). Such hegemony is reflected in the Party's mean percentage share of general election votes in each decade from the 1950s through to the 1990s, respectively: 58.2, 59.3, 48.7, 41.3, and 52.1. Notably, in electoral terms the other main state-wide party in UK politics, the right-of-centre Conservative Party, has done worse in Wales than in England at every general election since the 1880s.
However, since 1997 Labour's mean share of general election votes in Wales has declined to 42.5 %. It is in large measure thanks to the 'first past the post' system that Labour has been able to hold on to the majority of Welsh Westminster constituencies, but in several seats only by a narrow margin. This has led some to question whether the era of OPD is over (Johnes 2012). Limited evidence to support this idea can be found in local government elections (where Labour has lost ground in recent years) and the 2009 European Parliament election (when the party's support level in Wales was the lowest since 1918). Yet it is argued here that such claims are precipitous. They are rebutted by the fact that in the post-1999 mesoelections Labour has always gained the largest share of the vote. In 1999 they won 28 of the 60 seats in the National Assembly for Wales (polling 37.6 % of the vote, 9.4 % ahead of their nearest rivals). In the next three elections the party won exactly half of the seats in the legislature-the same as its three main rivals combined (with 40 % of the vote in 2003, 32.2 % in 2007 and 36.9 % in 2011). As noted, this record has ensured that Labour has continuously held government office since the National Assembly for Wales was created in 1999. 5

Methodology
This study adds to the contemporary body of qualitative work on the third sector (for example, MacNeela 2008; Pick et al. 2011;Sinclair and Bolt 2013). It explores data gathered from semi-structured interviews with managers and policy officers in a purposive sample of 100 third sector organisations in Wales undertaken in 2011-2013. The interviews were based on an interview schedule consisting of core questions developed from leading texts on third sector engagement (Fyfe 2005;Kendall 2009). The use of a basic interview framework-or standardised set of research stimuli (Gubrium 2012), allowed comparison of participants' accounts. Semi-structured interviews were particularly suited to the present task because they permitted the use of probes and secondary, supplementary questions (King and Horrocks 2010). This boosted reliability of the data by enabling the exploration and clarification of the issues and experiences described by participants.
The sample of TSO interviewees was constructed to reflect the range of policy responsibilities devolved to the meso-government (including, for example, health, housing, education and the environment). 6 In terms of geographical focus, care was taken to include both regional and cross-border organisations, 7 as well as a combination of local-level and national NGOs. 8 Moreover, participants were selected so as to include the views of organisations employing a range of engagement techniques (such as lobbying, membership of expert groups, taskforces, policy networks, partnership structures with the third sector and public petitions), as evidenced by official records of policy consultations and other modes of engagement available on the government and legislature websites. 9 The interviews were conducted in either Welsh or English and lasted approximately 1 hour. 10 Data gathering and analysis adhered to established ethical practice (BSA, 2012); inter alia this extended assurances to participants regarding anonymity, confidentiality and use of research data. TSO managers and policy officers were interviewed in order to gather rich narrative accounts of individuals' experience of engaging in government policy-making. Interviews were recorded digitally and transcribed verbatim using a word processing package. Subsequently they were analysed using Nvivo 9 software. An inductive coding frame (Sayer 1992;Saldaña 2012) was employed as the most appropriate means of 'creating meaning in complex data through the development of summary themes or categories from the raw data' (Thomas 2006, p. 239). Twelve key themes (Table 1) emerged from the data, thereby informing subsequent analysis of the effects of OPD on third sector organisations' policy-engagement.

Third Sector Policy Engagement in a One-Party-Dominant System
The policy director with a campaigning organisation captured a prevailing view amongst the interviewees in this study when she stated: ''it's not helpful to have such a dominant party. It's not good for democracy that a single party is in power for such prolonged periods without a single break at all, I think it's not healthy for that party; the whole point of democracy is that you have to be accountable to the electorate. Yet that's how it is and it seems that's how it's going to be'' (Participant 18). In this section third sector policy actors' views of engaging in government policy-making in a one-party-dominant system are considered. The discussion is structured around the twelve (non-discrete) themes that emerged from inductive coding of the data (Table 1).
(i-iii). Strategic Bridging, Veto Players and Collective Incentives Strategic bridging is a concept allied to Charles Tilly's classic work (1978, pp. 125-133) highlighting individual agency and co-working between policy actors in civil society and parliamentarians (see also Gamson 1975;Brown 1991;Sharma et al. 1994). In the present case it applies to NGOs' targeted lobbying of-and engagement with, individual ministers and backbenchers. It is a conception that resonates with bridging social capital, a relational resource stemming from networks and reciprocity (Coleman 1988). Both Granovetter (1985) and Putnam (1993Putnam ( , 2000 describe how it takes bridging social capital to link associative life in civil society with the political classes. This presents to third sector organisations a strategic challenge; namely, which parliamentarians to engage with in order to advance their policy demands (Faerman et al. 2001). In this regard interviewees' comments allude to how OPD skews the political dynamic. Crucially, it reduces the value of alliance building with opposition party members (because they are viewed as less influential owing to the fact that they are unlikely to hold future government office). It also accentuates the political power of members of the dominant party because, effectively, NGOs have no choice but to deal with them in order to advance their policy claims. In this way OPD limits what Dewan and Myatt (2010, p. 270) dubthe 'talent pool' of current and prospective ministers (because, over successive election cycles, it is likely that there will be a greater number of individual parliamentarians holding ministerial office when two or more parties alternate in forming the executive than is the case under OPD, simply because the field from which they can be selected in greater). This raises the associated challenge of 'veto players'-or individuals who, by virtue of their office, may block progress on an issue (for a discussion see Tsebelis 1995). Whilst they may also be present in 'regular' liberal democratic systems, OPD makes their presence (and thus power) more significant and challenging. Their impact is captured in interviewees' comments. For example, the director of an NGO reflected: The wider significance of these findings is in providing empirical support for Bollen's (1990, p. 12) observation that governing-party turnover has profound implications for the political opportunity structures open to third sector organisations for, 'it is the relative balance of power between elites and non-elites that determines the degree of political democracy' (see also Schlager 1995, p. 248). The findings also provide an important insight into the impact of OPD on collective incentives for political engagement. This is captured by the Collective Interest Model (Finkel et al. 1989, p. 39) that outlines how groups' incentives for political participation to challenge the political agenda fall into three categories: (1) 'high levels of discontent with the current provision of public goods by the government or regime', (2) the belief 'that collective efforts can be successful in providing desired public goods'; and crucially for the present purposes, (3) the belief that third sector organisations 'own participation will enhance the likelihood of the collective effort's success'. In particular, it is the latter aspect that is transformed by OPD. As interviewees' comments reveal, OPD dis-incentivises NGOs from engagement with opposition party members because, as noted, of a prevailing view that it is unlikely they will be elected to government office over future election cycles. Thus, in interviewees' eyes they lacked potential power and influence. Significantly, a troubling feedback loop can be seen to operate here. Over time reduced engagement between NGOs and opposition parliamentarians not only weakens the opposition's influence, it also increases dis-incentives for NGOs to engage with them; in turn increasing OPD.

(iv). Trust
Trust is a core component of bridging social capital and a further predictor of policy engagement (Portes 1998;Mishler and Rose 2001). Amongst a range of institutional factors it is influenced by party balance in legislative settings (Dunn 2011, p. 396;Norris 1999), as well as perceptions of the executive party's political performance (Price and Romantan 2004); or as Lühiste (2006, p. 493) puts it, 'the regime's capacity to produce the so-called ''procedural goods'' and desired outputs such as fair treatment of its citizens, protection of civil liberties, and transparent and effective administration' (see also Aberbach and Walker 1970;Rohrschneider and Schmitt-Beck 2002).
The present study addresses a knowledge-gap by providing empirical data on how OPD affects NGOs' levels of trust in government. Specifically, individuals spoke of uncertainty and mistrust in dealing with members of the governing party. The lack of certainty centred on whether parliamentarians were genuinely concerned with advancing a given policy issue, or were more concerned with party interests. For example, the chief executive of one NGO referred to an official enquiry over maladministration by a state-sponsored policy network. This raised questions about the network's links with the governing party. 11 She said: ''it's difficult, I have to be honest… we're not political, we're here for everyone… things like that do raise questions about whether they [government] can be trusted… that has a knock-on [effect] as far as our members are concerned'' (participant 27). A further example is provided by the comments of a policy officer with an environmental organisation who said: ''there has been a loss of trust in the government… we were naïve I suppose… [then] one of our trustees said look at the politics of it! As a party their support base is mostly urban and in the south of the country… now you can hardly trust them to put rural interests ahead of that! … no matter what they promised [at the election]'' (Participant 91).

(v). Neo-corporatism
The past three decades have given rise to a burgeoning literature on the relationship between government and social capital; in other words, norms and networks of trust and reciprocity in society (Granovetter 1985;Coleman 1988;Fukuyama 1995, and Portes 1998. For the present purposes two key issues stand out in this body of work: (1) the idea that social capital is a resource for governments that is related to the efficiency of public administration and health of democracy; and (2) That social capital has been subject to decline in many democracies around the globe. Both factors have spurred governments to intervene in attempt to boost social 11 This report (WAO 2012, p. 11) concluded: 'we have found no evidence of inappropriate Ministerial influence-on party-political or other lines-in the Welsh Government's decisions about… funding'. Writing about another report on government funding of the policy network it also noted: 'The fact that the report author joined the Welsh Government's equalities unit on secondment prior to the completion of the review was being completed inevitably calls into question the extent of the Welsh Government's influence ' (p. 16). capital (Lappé and du Bois 1997). In this regard successive administrations in the present case study are no exception.
Such matters have led to a renewed international emphasis on neo-corporatism; an approach to governing concerned with fostering state engagement with nongovernmental bodies and vice versa. Its provenance is managerialism and its aim is to get the best ideas and the most reliable information into the decision-making process. Typically, as in the present research context, it takes the form of government-sponsored policy networks, financial support for NGOs, advisory groups and partnerships designed to inform public policy-making.
The conditions for neo-corporatism are described by Mansbridge (1999, p. 495): First, it values interest groups as ongoing institutional mechanisms for representing interests not easily represented in the territorial representative process. Second, it attempts to bring the laissez faire system of interest representation partly under public control… As a system of interest representation becomes more directly involved with state law-making and law-enforcing processes, it more fully deserves the name of 'corporatist', and to the degree that it recognises non-traditional interests, it more fully deserves the prefix 'neo'.
De Jager (2005, p. 56) is clear about its inherent dangers: 'civil society derives its very legitimacy from its ability to act and then to act independently… The development of more formal and regulated civil society-state relations may subvert the character of civil society and compromise its role in enhancing democracy'. Whilst this concern is germane to liberal democracies in general, it has heightened significance under OPD because of the increased power of political patronage centred on a single party.
Amongst interviewees there was widespread concern over the negative consequences of neo-corporatist government intervention. This is consonant with a burgeoning literature on how, far from boosting social capital and securing the democratizing benefits of associative life (Putnam 1993;Hirst 1994), government support for policy networks may actually have the contrary effect. A core pathology alluded to by interviewees was co-option, loss of freedom to criticise government. On one level this is nothing new. It is an issue that has received considerable scholarly attention (Boydell et al. 2008;Chapman et al. 2010). However, the original contribution made here is in showing how OPD significantly exacerbates known shortcomings.
For example, the director of an NGO reflected: ''we don't want to alienate [ministers] but also we don't want to be in the pocket of the minister and it's that balance which is difficult'' (Participant 79). Another offered candidly: ''we are afraid to say 'the emperor has no clothes' 12 … we are afraid to show when failure is happening, because you know you have several people that have to wear several hats [i.e. perform multiple roles as NGO representatives, members of government policy task and finish groups, and recipients of state funding], there are conflicts of interest…'' (Participant 32). Another manager alluded to how OPD even limited NGOs' willingness to provide information to assist opposition backbenchers in criticising government policy. This is because the specialist nature of the information meant its provenance would be apparent (and thus likely to be viewed as tantamount to an overt attack on government, their chief sponsor). One manager said, ''What is the value of being attacked by the opposition using evidence provided by the [third] sector? It puts you in a difficult position, it can misfire… having the right points advanced by the 'wrong party'-er, [irony] it doesn't always advance your cause!'' (Participant 30). Another was more direct, ''I can't say anything because the Welsh Government is funding me'' (Participant 31). Another NGO director recalled, ''one of the organisations received a [telephone] call saying 'don't you realise who's paying your bills'? [adding]-human beings don't like being criticised, why should politicians be any different?'' (Participant 36). Another interviewee reflected on the overall impact of OPD and neo-corporatist practices: in effect you now have state sponsored civil society which compounds the problem that it's hard to criticise government. So they've created a civil society but it's an anaemic one because they control the flow of blood and they don't want it [civil society] to get too strong (Participant 27).
A further democratic 'cost' that emerged from the interview data was that NGOs' need to be seen to be influential meant they were deterred from effective engagement in policy-making. This is a striking example of government intervention having the opposite effect to that intended. A public affairs manager explained: Organisations themselves haven't been prepared to be ''critical friends'' or be critical at all because it's been more important to be seen to be part of ''the centre''. If you were to look at the discourse that goes on… it's the manufacture of consent rather than the manufacture of discontent (-which is how you'd normally characterise the relationship between media, civil society and government). What's driving it? Left-learning [politics]-the need to be seen to be close to the [governing Labour] Party, to guarantee their [i.e. NGOs' own] significance… and that's determined the flavour and style of policy work (Participant 75).
As Piattoni (2001, p. 4) notes, clientelism and patronage involve 'the trade of votes and other types of partisan support in exchange for public decisions with divisible benefits'. Blakeley (2001, p. 100) describes how a strong civil society may play a pivotal role in tackling this owing 'to the strength of civil society and the ability of associations in civil society to avail themselves of the state's resources without compromising either their aims or autonomy'. However, she concludes on a cautionary note, 'clientelism is likely to remain an inevitable feature of liberal democratic polities' (p. 104). The current empirical data support this conclusion. For example, the director of an NGO said: We have a ''client state''. The government-either directly or indirectly through its various agencies, puts an awful lot of funding into third sector organisations: that shuts them up. We are very lucky that we don't receive any funding from the Welsh Government or any government, that means we are free to stand-up and to criticise and to challenge. Whereas an awful lot of organisations feel they can't do that (Participant 16).

(vi and vii). Blurred Dominant Party-Civil Society Boundaries; and 'informal' politics
As (Chatterjee 2004, p. 136) notes, existing studies often point to political parties and civil society as largely discrete entities (Evans 1996;Kopecky and Mudde 2003). Some have taken issue with this and highlight how contemporary societies are characterised by blurred boundaries (Mair 2000;Kitschelt and Wilkinson 2007). In particular, recent research on the effects of OPD in South Africa identifies issues that resonate with the present case study (Friedman 2003). For example, as Bénit-Gbaffou (2012, p. 186) explains, civil society organisations' criticism of the dominant party government may be limited for a range of reasons including citizens' loyalty to the party as well as a shared sense of identity. It may also be due to the party 'exert[ing], through its members, a subtle but efficient social control on what is said… it may be shaped by everyone's perception that their discourse and position might be reported to [party] structure' (Bénit-Gbaffou 2012, p. 191).
The latter point chimes with the present study for interviewees spoke of 'toningdown' or avoiding criticism of government for fear that some third sector managers-who were also Labour Party figures, may pass on comments to ministers; which in turn could jeopardise their government funding. According to one, ''you do have to stop yourself. Like at the [name of policy network] meeting… [named policy officer] was saying how fed-up she was with the lack of progress, backtracking and stuff… I thought hold-on [uses own name] watch yourself! 'cos I know a couple of policy people there are strong Party [people]'' (Participant 86).
A further, related issue raised by interviewees was the negative aspects (opaqueness, lack of accountability and exclusive nature) of informal networks that crosscut formal political channels and processes (Masket 2009). This corresponds with existing work such as by McClurg and Lazer (2014, p. 4 forthcoming) who allude to how 'party organizations, interest groups and others in the policy process… often form connections that are at times fluid and at other times enduring… suggest[ing] a host of new substantive questions about organised influence in politics that centred on explaining relationship formation and impact in political organisations'. Thus interviewees referred to how critical engagement with government routinely took place outside legislative channels. In other words, OPD was ''short-circuiting'' the formal practices of parliamentary representation and instead operating more as an informal, intra-party process. Accordingly, a public affairs manager stated, ''some of the civil service they are very much steeped in the old ways, the Whitehall approach… and that's where we've had to make representations, you know, through political channels…'' (Participant 76). Another said: I found where we have had some successes is in raising concerns with Labour backbenchers who have gone on to raise the concerns themselves with ministers. So we tend to shape the policy practices indirectly through the Labour Party machine itself, shaping policy internally. I think there's more chance of a Labour backbencher getting concessions out of a [Labour] minister than engaging opposition parties ''from the outside'' (Participant 23).
Another said: We have personal relationships… an awful lot is done via personal relationships-and that's how stuff gets done. It's not right, it's not proper, it means that the better your relationships the more influence that you have, I don't agree with it but it's a paternalistic, party system where, if you have those networks, you will get policy through and that's how things happen… it's a network thing. I don't agree with it but I will use it when I have to (Participant 68).

(viii). NGO Action Repertoires
Interviewees referred to the way that OPD shaped NGOs' repertoire of contention (Tilly 1995, p. 42); in other words, the varied means by which they advance their policy demands on those in power. A notable finding is interviewees' claim that this involves a deliberate element of tactical coordination and a differentiation of modes of engagement between organisations across policy sectors. For example, one manager asserted: ''I see the environmental movement as being on a bit of a continuum, [named organisation] are going to be able to do some of the 'far out there-really critical stuff', others like [named NGO] keep mainly to new media stuff, whereas we [named organisation] are more mainstream-and we are going to be more ''in the tent'', but when we do ''chuck our toys out of the pram'' [in other words, a significant intensification of campaigning against government policy]-we will do it when we really need to' (Participant 47).
A concern expressed by interviewees was that the dominant party and associated political culture lacked the maturity and experience to respect the third sector's right to vigorously challenge government policy. This was captured in the following observation from the manager of an environmental NGO: ''I don't think-we're unlike Westminster where everybody understands the rules of the game-that we've got that maturity with Labour dominant here. What I mean is that I don't feel they [Welsh government] respect our right-if you like-to 'up the ante' and be really critical… there are big political consequences for really ''pressing the button'' [i.e. meaning a significant increase, or step-change in third sector organisations' criticism of government] … the drawbridge could go up at any moment… if I've really cheesed the minister off-and that's it! Thatisit. There aren't any other people or channels'' (Participant 24).
A minority of interviewees spoke of deliberate avoidance of neo-corporatist structures. 13 This was driven by a concern with strategic positioning. A public affairs manager explained the approach of his NGO: If you're perceived to be in the 'inner circle', the 'policy bubble' by members, stakeholders, patrons, supporters then you are in a very difficult position. You are going to have a very difficult time… the last thing I want to be seen as on the inside of that group. That's deliberate, that's a cognitive step taken. As an individual that can be quite difficult because you can be in a minority of one (Participant 8).

(ix). Legitimacy versus Legitimation
As Rawls (2003, p. 171) notes, 'political legitimacy aims for a public basis of justification and appeals to free public reason, and hence to all citizens viewed as reasonable and rational'. In contrast, 'legitimation involves communicative actions aimed at managing the public's perception that government actions are effective in promoting their desired ends, whether that is in fact true' (Moore 2001, p. 712). Interviewees revealed issues and tensions between legitimization and legitimation that are compounded by OPD. For example, the director of an NGO in receipt of government funding reflected that their involvement and access to government policy-making was more about legitimation than legitimization: ''they dine out on us and whenever they [ministers]

(x). Party System Institutionalisation
Party system institutionalisation is 'the process by which a practice or organisation becomes well established and widely known, if not universally accepted. Actors develop expectations, orientations, and behaviour based on the premise that this practice or organisation will prevail into the foreseeable future' (Mainwaring and Scully 1995, p. 4). As such it is allied to the notion of 'path dependency' in (neo-)institutionalist theory (Collier and Collier 1991;Ertman 1997). As Piersen and Skocpol (2002, p. 6) explain, this refers to the dynamics of self-reinforcing or positive feedback processes in a political system… Outcomes at a ''critical juncture'' trigger feedback mechanisms that reinforce the recurrence of a particular pattern into the future…'. Party institutionalisation and path dependency (Ennis 1987;Morris 1993) emerge in interviewees' references to third sector organisations' tactical approach to government policy-making. In this regard Tarrow and Tilly (2001) explain that in regular liberal democracies new tactics arise, are modified and adopted in the context of peaks or cycles of protest activity, or 'cycles of contention'. In contrast, the present findings suggest that OPD disrupts such cycles of protest because the trigger provided by a change of government, which often sparks new modes of NGO campaigning, is removed. Potentially, this has a stifling effect on NGOs' repertoires of contention. Accordingly, a policy and media officer said: There are some really difficult decisions to be made in the [state] health service… When you don't have the fear of being voted out, why then are you avoiding those decisions? It's really difficult, you know the policies and strategies aren't going to change and you almost end up working with them [governing party], because that's the way it is-there's not likely to be any change [of governing party] in the future and you get sucked into the bubble, you wonder ''am I enabling them''? (Participant 49).
Another manager said: ''my take on it is a lot of what we do in lobbying and public affairs is ritualistic so we follow a set of routines but actually beneath that there is very little to play for in terms of high stakes political and policy gain-that's because we are likely to have the same shape and colour of government in respect of where its sits in terms of party balance'' (Participant 75). In a similar vein, another policy officer observed ''all of it is performative to provide government with a bit of press cover-to me the real dynamic that needs to change as devolution grows up is more organisations being prepared to step back and rethink and extend ways to be critical-until then we are not going to get the kind of grown-up [political] settlement that we need'' (Participant 27).

(xi). Cognitive Locks
Cognitive locks draw on the literature of historical institutionalism (Pierson 2004). They are what Forestiere and Allen (2011, p. 381) term 'intellectual path dependency in policymaking'. Once established, they become a guiding set of ideas or 'ideological mantra to be repeated and applied no matter… [what] the actual conditions of a situation' (Blyth 2002, p. 229). They arise under one-party dominant regimes because change or reversal of government policies as a result of exogenous pressure-what Hirschman (1970, p. 30) calls 'voice'-is limited or absent. In the present research context the discourse espoused by the governing Labour Party emphasises universalism, collectivism, community empowerment, coproduction and equality of outcome.
Thus interviewees referred to what might be viewed as a cognitive lock linking the governing party and much of civil society. One said: I don't think there's enough space in civil/civic society in Wales about what kinds of policies we want, what types of things do we want to achieve, the existing voices come from the Left-of-centre, they come from a socialist campaigning background… that imposes challenges in terms of tone, and positioning our messages, you can't be seen as being market-driven by tooth and claw, you've got to try and package your message in a particular way (Participant 86).
Another interviewee referred to the governing Labour Party's resistance to neoliberalism and espousal of mutualism as a predominant policy frame or cognitive lock. She said: ''it presents us with a bit of a dilemma. We can't take the confrontational route because we have a government which we are ideologically supportive of… we agree with the underlying principles, not always with the implementation… it's quite difficult then to be adversarial'' (Participant 23). Another said: ''we wanted regulating the market of social care on the face of the Bill, like in England. That was heresy to the minister! … instead it was all on mutualism and co-production…. Finally the market was buried in a sub-section'' (Participant 75). According to another NGO director, ''we learned early on its futile… we were naive… rather than present a direct challenge [to government] you need to wrap things up in their language… [their] kind of orthodoxy'' (Participant 51).

(xii). Party Politicization of Third Sector Engagement
One strand of the policy literature emphasises public policy-making as a rational, bureaucratic exercise with an emphasis on the role of professional knowledge and the power of ideas (e.g. Sabatier 1987;Sabatier and Jenkins-Smith 1993); in contrast another underlines how policy work has much more of a political provenance and emerges from conflict between parties. It is argued that the latter view allows policies to be understood in terms of consequences for different social interests in a manner that is influenced by decision-making contexts and party competition (Lowi 1964). Thus far work within this tradition has generally failed to give sufficient attention to the effects of OPD. In response, the present findings reveal how NGO policy engagement may become party-politicised as a result of dominant party institutionalisation.
Thus, for example, interviewees underlined how policy engagement was not always viewed as a politically neutral process because third sector organisations were obliged to negotiate the challenging and singular positional politics that characterise a one-party-dominant system. This can be problematic because NGOs can be drawn into party politics and erroneously perceived as being partisan purely by dint of taking a critical stance in relation to government policy. It is a viewpoint captured by the public affairs manager of a leading NGO: My personal view is the fact that it is a one party state actually does create quite a sterile environment in terms of policy-making. It very much feels that at times to challenge a policy is very much to challenge the Party, and to challenge the hegemony of the Welsh Labour Government (Participant 16).
Another policy officer said, ''it can get a bit 'political' at times… someone from the [policy network] said to me 'you were giving the minister a hard time, they're not Tories' [official opposition]-it was clear he meant 'why are you giving Labour a hard time'? [i.e. the viewpoint they are on our side]-but that's not the point, is it? It's the government you're dealing with' (Participant 49).

Discussion
The foregoing interview data show how one-party-dominant systems have considerable implications for third sector interest representation, co-working, criticality and 'voice'. Twelve pathologies related to the third sector-state nexus emerge from the accounts of NGO policy actors; each with a potentially significant impact on the democratic health of the polity. Crucially for the body politic they may collectively be seen as self-sustaining. This is because, over time, diminishing engagement and criticality from third sector organisations may further strengthen the dominant party. In turn, this makes a change of governing party even more unlikely.
The present findings evidence how OPD skews the political dynamic. For example, it reduces the value of NGOs' strategic alliance building with opposition party members. At the same time it accentuates the political power of members of the dominant party; by, for example, extending the influence of 'veto players' who may block third sector policy demands. Such factors illustrate how governing-party turnover has profound implications for the political opportunity structures open to the third sector. It also distorts the collective incentives for engagement with opposition parliamentarians, and increases the executive's power of political patronage. Moreover, the blurred boundaries between dominant party and civil society may result in a 'toning-down' (even cessation) of NGOs' criticism of government. This is due to the fear that some third sector managers who are also influential in dominant party circles will report back to ministers; thereby threatening NGOs' state funding and political access. 'Informal' or extraparliamentary politics was a further pathology identified in this study. It arises when third sector organisations' engagement with government routinely takes place outside formal political channels. Because such interaction is vectored through internal party contacts and networks it is neither transparent nor accountable.
Further pathologies associated with OPD revealed in the data analysis include limited NGO repertoires of contention (because OPD systems lack the usual 'trigger' presented by a change of government); government's over-emphasis on neo-corporatism (a legitimation response to their hold on power); and the rise of 'cognitive locks' in policymaking (whereby NGOs are constrained by the 'intellectual path dependency' of a one party regime). The present findings also reveal how, in undertaking policy work, NGOs have to negotiate the singular and challenging positional politics of a one-party-dominant system (whereby criticism of government may become conflated with criticism of party qua party).
Given this plethora of democratic ills, it should be restated that OPD is a mainstream issue; all parliamentary systems have immanent potential for periods of OPD. With such a caveat in mind, the wider contribution of this study to understanding the health of contemporary liberal democratic systems is twofold: (1) It reveals the types of pathologies that may emerge when a party retains power over several electoral cycles; and (2) It underlines the need for contemporary academic analysis to be cognizant of how the third sector's public policy role is contingent on electoral competition and governing-party turnover.