Career Models in the New Tripolar Order. Political Profiles of the Italian MPs after the 2022 General Elections

,


Introduction
The Italian party system has undergone several changes in the last twenty years, transitioning from an imperfect bipolarism (D'Alimonte & Chiaramonte, 2010) -also triggered by the changes of the electoral law towards a majoritarian perspective (both with the so-called Mattarellum and Porcellum) -to a new tripolar phase (Chiaramonte & De Sio, 2015). In particular, after the 2013 general elections, mainstream parties of the centre-right, Go Italy 1 (FI-PDL-FI) and of the centre-left (Democratici di Sinistra, DS before and Partito Democratico, PD later)-that dominated the political scene from late 1990s to late 2000s-started losing votes whereas new populist parties such as the Five Star Movements (Movimento Cinque Stelle, M5s) (Chiapponi, 2017) and radical parties such as the League under the leadership of Matteo Salvini (Passarelli & Tuorto, 2018) and Brothers of Italy (Fratelli d'Italia, FDI) under the leadership of Giorgia Meloni (Baldini, Tronconi & Angelucci, 2023) increased their vote share. Additionally, referring to the party organisational literature that draws a distinction between thick and thin parties (Gunther & Diamond, 2003), we can expect that parties possessing a well-defined territorial structure are more inclined to present similar political figures, particularly when it comes to their political careers. Conversely, parties with lower levels of institutionalization tend to offer candidates of diverse backgrounds, resulting in less predictable and stable career paths.
By considering the last 2022 general election, we provide an original dataset which collects socioeconomic data and political career experiences of 392 deputies and 196 senators, for a total of 588 MPs.
Our article proceeds as follows: section 1 provides our theoretical framework and exposes our main hypotheses; section 2 explains our methodological approach and choices; section 3 provides our descriptive analysis and section 4 shows our findings. Section 5 concludes.

Theoretical framework and Research Design
The study of parliamentarians holds a significant place in the field of political science and has long been associated with the concept of political professionalization (Weber, 1919). In fact, a substantial body of contemporary European and American literature on political professionalization primarily revolves around parliamentarians, encompassing both country-specific analyses and comparative studies. (e.g. Squire 1993;Searing, 1994;Norris 1997;Saalfeld 1997;Shabad & Slomczynski, 2002;Borchert, 2003;Cairney, 2007;Koop & Bittner, 2011;Siavelis & Morgenstern, 2012, Heuwieser, 2018Allen 2018;Ohmura et al., 2018). However, so far in Europe the most important comparative work with a longitudinal perspective is that of Best and Cotta (2000), which has analysed the socioeconomic background, the political experience and turnover rates of the European parliamentary representatives from 1948 to 2000.
From a theoretical point of view, the research on political careers can be distinguished between actororiented and context-oriented approaches (Jahr & Edinger, 2015), which focus respectively on individual characteristics and on contextual features as explaining factors. Vercesi (2018) provides a systematic framework for reviewing both approaches in the study of political careers. Notably, among the former, the personality approach suggests that individual personality traits have an impact on career trajectories (Lasswell, 1948;Caprara & Silvester, 2018).
The ambition theory approach (Schlesinger, 1966;Nicholls, 1991;Lawless, 2012) posits that politicians are office seeking and pursue different career paths based on their ambitions. The social background and socialization approach (Putnam, 1976;Pilet et al., 2014) asserts that specific offices can only be attained by politicians with particular social characteristics and expertise. Lastly, the selection and deselection approach (Searing, 1994;Dowding & Dumont, 2015) operates within a rational choice framework, suggesting that politicians (as agents) are chosen based on their previous experiences.
In contrast, context-oriented approaches in the study of political careers encompass the following perspectives. The opportunity structure approach (Botella et al., 2010;Stolz, 2003Stolz, , 2015Stolz & Fisher, 2014;Grimaldi & Vercesi, 2018) asserts that career paths are influenced by the availability, accessibility, and desirability of political positions. The intra-organizational approach (Thurber, 1976;Hazan & Rahat, 2010;Peters, 2010) suggests that career choices and opportunities within an organization are shaped by formal and informal organizational rules and procedures. The supply and demand approach (Norris & Lovenduski, 1995;Krook, 2010;Carnes, 2016) focuses on individuals representing specific social sectors (e.g., women, workers) whose resources and motivations contribute to the supply, while the demand is determined by elite preferences that act as gatekeeping factors. In summary, this approach posits that political careers evolve through a step-by-step process influenced by the interplay of supply and demand dynamics. Our paper builds on the opportunity structure approach which proved to be the most suitable when it comes to consider the broad contextual factors that characterised multi-level democracies. In fact, in Italy, as in all multi-level democracies, politicians have tended to exploit all possible offices from local to regional and national level as opportunities for their political career (e.g. Grimaldi & Vercesi, 2018;Di Capua et al. 2022).
By relying on Borchert' work (2011) which distinguished three career models based on the hierarchy of the political offices and the direction of career paths, the article investigates what career model prevail among MPs of the XIX legislature in Italy and to what extent such models are affected by the politicians' social background and by the structure of their party organisation. To recall, according to Borchert (2011) in multi-level democracies three career models are likely to emerge, the unidirectional model refers to a political system in which a clear hierarchy of political offices exists, and it implies a movement from local to regional to national or European level. The alternative model implies a less clear hierarchy of offices with a high degree of autonomy between the levels of government and scarce possibility of "multi-level jumps". Consequently, national and subnational careers are clearly separate as there is no movement from regional to national or European level.
Finally, the integrated model implies that there is no hierarchy among territorial levels and, thus, movements from the European or the national level to the regional and local level are more likely.
Party experience is excluded from this analysis since the career model framework only concerns the occupation of previous institutional positions.
In order to explain the prevalence of one of the three career models mentioned above, we identified specific drivers that draw from the social background and socialization approach, the intraorganizational approach, and the influence of party organizational structures on career paths. In Italy, previous research (Grimaldi & Vercesi, 2018) have proved that, as a result of institutional changes towards greater regionalization, all such career models may coexist at the same time..On the one hand, MPs' socio-economic characteristics seem to be quite similar in comparative perspective (Best & Cotta, 2000) and thus we can assume that certain specific features matter in their selection process.
In particular, several studies pointed out that there is a systematic under-representation of women in the legislative assemblies (e.g. Philipps, 1995;Caul, 1999;Matland, 2005;Matland & Montgomery 2003;Norris, 1996;Tremblay, 2008). Despite the introduction of rules to favour gender balance (Dahlerup, 2007;Krook ,2007;Franceschet et al, 2012;Hughes et al,. 2019) in many European contexts, women continue to represent a minority and in addition, they also have less stable career On the other hand, the existence of a direct connection between the organisational dimension of political parties and the career model has been emphasized repeatedly in the literature (Janda, 1980;Panebianco, 1982;Gunther & Diamond;Deschouwer, 2006;Thorlakson, 2013). In a nutshell, a strong party organisation at the territorial level is correlated with the presence of more linear, stable and easily predictable career paths, while more fluid organisations may be more susceptible to "horizontal" inflows of political personnel (Panebianco, 1982). However, with a progressive party institutionalisation, the specialisation of internal roles and a greater vertical integration between the political elites of the different levels, career paths become considerably long and stable (Tronconi & Verzichelli, 2019). The reasons for such influence must be sought precisely in the intermediary role that political parties play in recruiting the political class.
In line with these premises, our research questions are the following: To answer these RQs, we rely on different streams of literature related to political class studies with a gender perspective and to organisational party literature. Notwithstanding the introduction of a strong system of quota provision since 2017 (i.e. art. 1.10, e law 165/2017), recent studies pointed out that parties seem to have learned how to circumvent such formal requirements by exploiting the electoral rules and in particular the possibility of multiplecandidacies, so much so that most of the female candidates have not been elected (Regalia 2021;Pansardi & Pedrazzani, 2022). Furthermore, many parties filled in the electoral lists with female political outsiders to comply with the norms (Pansardi & Pinto 2020) and therefore in the XVIII legislature more than 70% of the female newcomers have never held a political office before (Sampugnaro & Montemagno, 2020).
Furthermore, as Verzichelli et al. (2022) have illustrated, following the peak in 2013, approximately 40% of newly elected female politicians were re-elected in subsequent elections. Consequently, these women exclusively gained political experience within the parliamentary arena, thereby exemplifying an alternative career model. Similarly, prior to the implementation of quota systems, studies by Sampugnaro and Montemagno (2020) revealed that a significant number of incumbent women MPs had previously held solely national offices, further highlighting an alternative career trajectory. If these observations hold true for the XIX legislature, it implies that women not only continue to constitute a minority in Parliament, but regardless of their political affiliation, they seldom exhibit a "planned political career" characterized by ascending from lower-level territorial offices to higherprestige positions.
Consequently, it is more likely for women MPs being picked up without previous experience to fill in party lists -also to comply with the formal regulations on gender balance-and later they gain political experience limited to the national level in comparison to their male counterparts. Therefore, our first hypothesis is that: H1: Women MPs are more likely to develop an alternative career model in comparison to men MPs.
A second set of hypotheses is related to the territorial organisation of political parties. Party organisation represents one of the most investigated concepts in political science (Krouwel, 2006) and it is one of the most important criteria to classify political parties from the very beginning. Despite numerous scientific research having highlighted the diminishing importance of the organisational dimension (Kircheimer, 1966;Katz & Mair, 1995;Carty, 2004;Bolleyer, 2009 In our paper we explicitly rely on Gunther and Diamond' s (2003) classification of parties according to their organisational strength. They have distinguished between "organisational thick parties" as parties characterised by a strong territorial structure, a large mass membership base and a strong collateral associative network and "organisational thin parties" as parties mainly centred on networks of personal relationships and with a weak (if not absent) territorial organisational structure. Therefore, according to this classification, parties of cadres or notables, purely personal parties Calise, 2015;Pedersen & Rahat, 2021), but also movement parties (Kitschelt, 2006;Della Porta et al., 2017), would be classified as "thin" parties as they are all characterised by a low level of organisational articulation; on the contrary, traditional mass parties and their heirs, endowed with a strong organisational structure, would be classified as "thick" parties.
In particular, in line with such distinction it is likely that thick parties-due to their strong territorial organisation in sections-may facilitate the movement of party personnel between the different levels of government, thus favouring the affirmation of the unidirectional career model and (to a lesser extent) of the integrated model, whereas thin parties with scarce territorial organisation would favour the affirmation of the alternative career model (Pilet et al. 2014;Tronconi, 2018, p. 618) by allowing people to gain a political office and maintain it for a while to achieve a certain degree of specialisation Therefore, parties would constitute the instrument through which politicians could make the most of the career opportunities offered by the multi-level democracy. We expect that: H2: MPs belonging to thick parties with strong territorial organisation are more likely to develop a unidirectional career model in comparison to MPs affiliated with thin parties.
H2 bis: MPs belonging to thick parties are more likely to develop an integrated career model in comparison to MPs affiliated with thin parties.

Data & Methods
The To investigate the MPs' profiles and how gender and party characteristics affect their model of career, we deemed a quantitative analysis to be the most appropriate choice.
Bivariate analysis and descriptive statistics were used to check similarities and differences among the MPs' profiles related to gender, age cohort, education, profession, the career model and the party membership. Then, a multinomial logistic regression was used to test our initial hypotheses, namely, to check to what extent gender and party territorial organisation affect the development of a specific career model.
Variables were operationally defined in the following ways. On the one hand, the career model, which constitutes the dependent variable of the logit regression model, is operationalised as a three-way polytomous variable related to the three different types of careers: "unidirectional", "integrated" and "alternative".
A career was labelled as "unidirectional" whether it occurred incrementally among different levels of government (municipal, provincial, regional, and national/European), regardless of the initial level and of the length of the career. For example, in the unidirectional career type fall MPs such as Chiara Appendino (M5s), municipal councillor (2011)(2012)(2013)(2014)(2015)(2016)  Finally, a career type was considered as "alternative" if it was characterised only by experience at the national or at the European level (thus devoid of previous local or regional offices) Subjects such as Pierferdinando Casini (PD), MP continuously since 1983 and currently senator and Giuseppe Conte (M5S), Prime Minister between 2018 and 2021 and currently deputy, belong to this category. It is worth mentioning that we drop from the analysis the newcomers without any previous political experience such as Ilaria Cucchi who never held institutional positions and now she is senator among the ranks of Green -Italian Left Alliance. As a consequence, our analysis was run on a total of 516 observations. The alternative model represents the reference category of the multivariate analysis 2 . In Italy, it is noteworthy that the position of a parliamentarian can be held concurrently with that of a municipal councillor, municipal executive, and mayor of municipalities with a population under 15,000.. Since it is not possible for the dates of assuming office to coincide in two different positions at different governance levels. we can always distinguish between the unidirectional model (when the entry in local office occurred before) and the integrated model (when the entry in the national office occurred after).
On the other hand, party organisation and MPs' gender were considered as independent variables.
Both variables were operationalized in a dichotomous way.
Indeed, in accordance with Gunther and Diamond's (2003) classification, we opted to categorize parties as thin (assigned a value of 0) or thick (assigned a value of 1) based on their organizational structure. The choice of a dichotomous representation of such characteristic, although it constitutes a simplification of reality, was dictated both by reasons of practicality and comparability. We are aware that starting from the important studies of Janda (1980) and Panebianco (1986), scientific literature proposed a series of indicators to measure the territorial organisational strength of parties (e.g. Katz et al., 1992;Harmel & Janda, 1996;Tavits, 2012;Poguntke et al., 2016). However, some authors have pointed out that some of such indicators cannot be easily applied to new personal or populist parties (Vercesi, 2015, Musella & Vercesi, 2019, making their use in contemporary political systems more complex or scarcely reliable. Moreover, the collection with relative certainty of some of these indicators for small, newly founded parties or for electoral lists that include several different parties may be complex. For these reasons, we deemed the choice of a dichotomous, albeit simple, operationalization as the most appropriate. Consequently, basing our evaluation on the rich scholarly literature on Italian party organisations (Diamanti, 2009;Ignazi, 2013, Ignazi & Pizzimenti, 2014Ignazi & Bordandini, 2018, Pizzimenti, 2020Vittori;2020;Bordandini et al., 2023), the Democratic Party (PD), the League and Brothers of Italy (FdI) were defined as "thick" parties, whereas Go Italy (FI), the Five Star Movement (M5S), Noi Moderati (NM), + Europe (+E), Azione -Italia Viva (AZ-IV) and the Green -Italian Left Alliance (AVS) were defined as "thin" parties 3 . Table 1 summarises the modalities of such a variable. Lastly, we operationalized gender as a dichotomous variable with value 0 if the MP is male and 1 if is female.
Specific control variables were also used to increase the robustness of the analysis.
The first set of control variables are linked to the MPs' socio biographical profiles. First, we consider age in years from the year of birth to 2022. Scientific literature (Verzichelli, 2010) underlined that most politicians reach the greatest availability of political capital and resources in the middle-aged cohorts 45-55. To avoid a bias related to the non-linear relationship between age and political career, in the multivariate analysis we chose to operationalize age as the square distance between the age of the MP and the apex of the career (50 years) 4 .
We incorporated also these other control variables: education (due to the small number of subjects with less than a bachelor's degree, operationalized as a dichotomous variable with value 1 assigned to bachelor's degree) and original profession, operationalised as polytomous variable divided in six modalities: professional politicians; entrepreneurs, managers and directors; professionals (lawyers, doctors, accountants, etc.); public employees; private employees; other professions.
We also included some context related control variables, such as the House in which the MPs are elected (value 0 for the Chamber of Deputies and 1 for the Senate), whether MPs are elected in a single-member district (SMD) or not and the geographical area of election of each MP, operationalised as a variable with three modalities, North, Red-Belt and South 5 .
Data were extracted from the National Registry of the Local and Regional Administrators available on the website of the Ministry of the Interior 6 , from the Chamber website and from the Senate website 7 .
3. Descriptive Analysis: the socio-economic characteristics and the career models of the

MPs of the XIX legislature
In this section we focus on the socio-economic characteristics and career models of the Italian MPs of the XIX legislature by pinpointing differences and similarities among parliamentary parties.
The over-representation of men in the Italian Parliament compared to women comes as no surprise.
However, this time the share of women in the Chamber (Fig. 1) is less in comparison with that in the Senate (Fig. 2)   Looking at the original profession both in the Chamber and in the Senate the three most frequent categories are: liberal professionals (respectively 40.3% and 48%), managers or entrepreneurs (respectively 16.1% and 19.9%) and employees in the private sector (13.8% and 11.7%). However, in the Chamber of deputies there is also a consistent number of political professionals (12.5%).
For almost all parties, most deputies are liberal professionals and then managers or entrepreneurs.
The only exceptions are, the AVS whose majority of deputies are employed in the public sector (25%) and the PD whose majority of deputies (27.7%) are political professionals. Other peculiarities are the following. For AZ-IV it is worth mentioning that the second most frequent category is that of employees in the private sector (28.6%) rather than manager-entrepreneurs and that there is a high share of political professionals (19.1%). For NM the most frequent category is that of managers entrepreneurs (37.5%) rather than liberal professionals which is the second category along with employees in the private sector (both 16.9%). For the League, the second cohort is that of managersentrepreneurs and employees in the private sector (both 16.9%) whereas the third is that of political professionals (12.3%).
Most of the Senators, irrespective of their party affiliation, are liberal professionals. The only exception in this respect is that of Greens and the Left whose majority of senators are union officials (50%) and employees in the public sector (25%). The second most frequent category is that of manager-entrepreneurs for all parties with the exceptions of the PD and the M5s. In the former case, the second most frequent category is that of union officials (14.3%), in the latter case that of artisans and shopkeepers (17.9%). Finally, it is worth mentioning that 13.8% of the League senators are political professionals. All in all, deputies and senators seem to be quite a homogeneous class with regards to the original profession. There are only a few differences among parties, and these are typical of leftist parties and the M5S. On the one hand, for the leftist parties (AVS and PD) being union officials 8 is still an important background to serve as MPs. On the other hand, especially in relation to the M5S senators the profiles are the most dissimilar with those of the other parties since both artisans and shopkeepers and public employees are highly present.
When it comes to the career models of MPs, the unidirectional model is the most frequent (respectively 44.1% of the deputies and 41.8% of the senators), followed by the integrated model (23.2% of the deputies and 25% of the senators) whereas the alternative model is the least widespread (respectively 20.2% of the deputies and 21.9% of the senators). In addition, the share of newcomers is quite similar in both chambers (respectively, 12.5% and 11.2%).
However, when analyzing gender-based data, it was found that despite the unidirectional model being the most common for both men and women, there is a notable proportion of female deputies (30.5%) and female senators (26.5%) who follow an alternative career model, which ranks as the second most frequent in both cases. Furthermore, a significant percentage of female MPs (15.6% in the Chamber and 17.7% in the Senate) are newcomers without prior political experience. Consequently, considering both categories, it becomes apparent that women face challenges in fully harnessing the opportunities provided by multilevel democracy, as the integrated model remains marginal, and the unidirectional model falls well below the 50% mark for both female deputies and senators. Moreover, there are certain interesting differences in the career paths of deputies and senators according to their party affiliation.
To begin with the Chambers of Deputies, for most deputies, irrespective of their political affiliation, the unidirectional model is the most frequent followed by the integrated model. This means that generally deputies are more likely to obtain political offices by exploiting at the best the possibilities of a multilevel democracy by going up and down territorial levels when political offices are available avoiding concentrating on the national posts only. In particular, this holds true especially for AZ-IV and the League where the integrated model is typical of 33.3% and of 38.5% of their deputies.
Differently, the M5S is the party whose around 65% of deputies displays an alternative model, namely almost all M5S deputies only have experienced national political posts or are newcomers without previous experience (17.6%) and just a minority (15.7%) had some experience at subnational level before (Fig. 3). (Source: Authors' elaboration) When looking at the senators (Fig. 4)

Findings
As stated before, we tested our hypothesis with a multinomial logistic regression with the alternative career type as a base reference. The results of the logit regression are displayed in Table 2; the first part of the table shows the effect on the unidirectional versus the alternative career type whereas the following shows the results on the integrated versus the alternative career type. In the second column is also presented the relative risk ratio (RRR) 10 .
As table 2 shows, both gender and the party organisational feature are statistically significant in the two different models. As far as gender is concerned, the correlation is negative which implies that women MPs are less likely to develop a unidirectional or integrated career pattern as they are more likely to develop an alternative one. The result confirms our first expectation, namely women struggle to develop a linear career, which ascends through the various territorial levels and finally reaches parliamentary posts. Conversely, women are often elected directly to national bodies without other types of political background. Thus, hypothesis H1 is confirmed.
10 When a multinomial logistic regression is used, the relative risk ratio is used more frequently (Kleinbaum & Klein, 2010). It is calculated exponentiating the multinomial logit coefficients and it shows the risk of the outcome falling in the comparison group compared to the risk of the outcome falling in the referent group changes with the variable in question. probably due to the greater ease of access to office in the political system. In other words, politicians who are in the central age group are probably those with the longer political and institutional curricula and therefore they should have a better chance of exploiting the incentives of the institutional system, by moving more easily between different levels of government, than younger or older politicians who are more likely to be elected directly for national institutions.
Moreover, the regression highlights a significant negative effect of being elected in the South both in Model 1 and Model 2 (coefficient -0.839 and -0.783, RRR 0.43 and 0.46). This is a partially surprising result since, given the literature on the topic, it could be expected that southern politicians have career patterns with long local experiences before reaching national ones (Emanuele & Marino, 2016).
However, we believe that this result can be traced to the stronger affirmation of the M5S in the South.
In other words, the high electoral strength of the M5S in that area of the country, led to a higher share of Southern MPs without local experience.
However, such speculations should require further empirical investigations.

Conclusion
This article examined the composition of the Italian parliamentary class that emerged from the last general elections in 2022. More precisely, through a quantitative statistical analysis, the research highlights the similarities and differences of the socio-political characteristics among parliamentarians of different party groups both in the Chamber and in the Senate and it also provides an exploratory analysis on those factors that can influence their career model.
The results of the research have a double relevance, both on empirical and theoretical grounds.
From a theoretical point of view, the research proved that personal and organisational factors such as gender and the nature of the party to which MPs belong, influence their career models.
Indeed, the analysis has highlighted how, at least in the Italian context, women are not only fewer in number than men, but also have less structured and concentrated career models.
In other words, our findings show that, despite the amount of time women invest in politics, their career model often diverges from that of men. Specifically, our data indicates that women face challenges in fully leveraging the range of political positions available in a multi-level democracy when compared to men. This observation remains consistent regardless of the level of institutionalization within their respective parties.
Furthermore, the research highlighted how the organisational dimension of the party influences the emergence of a specific type of career pattern. In fact, highly organised parties seem to better exploit the opportunities offered by the multilevel institutional system, better plan career trajectories and thus favour unidirectional or integrated models.
Consequently, our research contributed to the debate on gender disparities in politics by looking at women predominant political career models and on intra-organizational approaches by focusing on party structure as drivers of specific career models.
From this point of view, therefore, the research partially fills the gap in the literature, examining whether party institutionalization influences career models.
From an empirical point of view, the analysis underlined that socio-demographic characteristics do not differ particularly from those found for MPs elected in previous electoral rounds. Parliamentarians are in fact generally men, with a high level of education, performing highly qualified professions and with previous political experience at the local level. Specifically, by looking at the gender dimension, the Parliament elected in 2022 has a lower number of women than in 2018, halting the path towards an increase of gender balance that seemed to have started in the two previous legislatures. Some further considerations with reference to Brothers of Italy and the M5S are explored below. Giorgia Meloni's party, the actual winner of 2022 elections, due to its "novelty" within the Italian political system, might have presented certain peculiarities in regard to the recruitment patterns of its parliamentary class. However, as data show, the party seems to largely follow the recruitment methods of traditional parties. Thus, despite the party getting the first female party leader who became PM, such a novelty did not affect the party recruitment of political personnel. In fact, its parliamentary class is largely made up of male politicians already active in politics during the so-called Second Republic (and in some cases even during the First Republic). Furthermore, women -when presentare newcomers or show a less structured career model. Thus, Meloni's leadership not only did not affect political agenda in terms of increasing gender-related issues (De Giorgi et al. 2023), but also did not affect the recruitment of political personnel in a gender balanced way, so far. The M5S, on the contrary, seems to maintain its specificity. Differently from what commonly expected, there was not a normalisation of the party, which maintained a recruitment mostly centred on political outsiders.
The new leadership of the former PM Giuseppe Conte, also favoured by the departure of numerous historical exponents (including the former Minister Luigi Di Maio), managed to maintain such a distinctive trait of the M5S. We will see in the future whether the M5s will be able to maintain this specificity, escaping traditional political and institutional socialisation patterns or not.
Our findings can open interesting future research paths especially by expanding the analysis by adopting a longitudinal and comparative perspective or through a more in-depth and multidimensional operationalization of the party organisation dimension.