Public Attitudes towards Parties in Portugal: A Longitudinal Overview

This article reviews the case of the often claimed “crisis of parties” in Portugal, and argues that such controversy rests at least in part on “ambiguous” evidence. We will try to answer two fundamental research questions: 1) What motivates popular support for political parties (or lack thereof)? 2) Why does anti-party rhetoric resonate with some citizens, but not with others? The empirical data and statistical regression models used allow the following conclusion: in the eyes of Portuguese citizens, parties have become a kind of “necessary evil”, being criticized for “what they actually do” and supported for “what they are supposed to do”.


introduction
P olitical parties are universally regarded as an essential component of democratic regimes and a key vehicle in the process of political representation, voter mobilization, government organizing and public policy shaping (Ware 1987;Webb 2000).
Perhaps no political institution is as closely identifiable with representative democracy as are political parties, precisely because no one has yet shown how representative government could function without them.
Yet in established democracies today, the role of parties is often more limited than it would appear, judging from their position of formal pre-eminence, precisely because theories about the conundrum that is the relationship between citizens and parties, and its consequences in the context of contemporary representative democracies.
There are important questions in need of answers: What role do citizens consider appropriate for parties in the context of modern democracy, and how much do the parties fulfil it in their country?What do citizens really expect from the major Portuguese parties?
After a brief description of the institutional factors that help understand the position parties occupy between state and civil society in Portugal, our aim in this paper is to find out the nature and trends of popular support (or lack thereof) for parties and determine their correlates at an individual level.

Popular Support for Parties: Political Context Matters the genesis and evolution of Portuguese parties and party organization
In a democracy such as Portugal's, the roots of "party malaise" seem to be both of an institutional and an attitudinal nature, reflecting its late and very specific transition to democracy as well as the direct effects on the nature of the parties themselves For instance, in Portugal, between 2000 and 2008, the mistrust level towards parties was not just extremely high -over 80% -; it was almost unaltered throughout the whole time period.This value is bigger than the average in the established first and second wave democracies and also bigger than Spain's -70% -, which is interesting due to the obvious cultural, historical and geographical affinities between both.Again, Portugal seems closer to Eastern European countries such as Bulgaria, Slovenia, Estonia or Romania than to Western European democracies (Eurobarometer Surveys).Source: Eurobarometer Surveys (2000EB54, 2001EB56, 2002EB57, 2003EB60, 2004EB63, 2005EB65, 2006EB66, 2007EB68, 2008 EB69).
It is now important to explain the role of the democratization process in the structuring and evolution of the Portuguese party system, and also in the nature of parties and party When the first mobilization of the masses occurred, the fact that parties focused not on politicizing traditional social cleavages but on the conflict around the choice of political regime helps explain some of the main characteristics of Portuguese political parties.The first characteristic is the weakness of their social rooting, partially compensated by their proximity to the State and dependence on its resources.The second is the hybrid nature of partisan organizations (Lisi 2007;Jalali 2007;Teixeira 2009b), which despite having formally incorporated some characteristics of mass parties are, in practical terms, closer to the definition of cartel parties due to their media-boosted electoral orientation (namely through television) and occupation of the state apparatus.
Contradicting the existent literature on the origin of political parties in Western Europe, the revolutionary context at the time was the key factor in the process of party formation, heavily influenced by the choice of political regime after the April 1974 revolution.With the (partial) exception of the PCP (the only party that survived the previous authoritarian regime, and generally characterized as an orthodox Marxist-Leninist party), all other parties had to build a whole new organizational structure after 1974.This helps explain the choice to mobilize a broad and superficial social basis to the detriment of a deep and limited one.Again, with the PCP partially excluded, the main Portuguese parties did not seek to represent a single social class exclusively, opting instead for an openly "interclassist" strategy with a catch-all electoral orientation.
The need to build an organization from nothing and the weak social bases that pulled Portuguese parties away from the mass party democratic model contrast deeply with the central role they occupied at an institutional level, which has remained until today.In effect, due to the democratic transition's unstable nature, parties acquired quick access to government, even before having developed their organizational structures in the whole country (Jalali 2007;Teixeira 2009b).After that, there followed the parties' access to parliamentary representation and their monopoly, granted by the 1976 Constitution, which states that only parties can present candidates to legislative and -until 2001local elections.We can therefore state that the development of parties in Portugal was conducted by the "party in public office" at local and national levels (Biezen 2003;Bosco and Morlino 2007).
This means that the great social weakness of parties was compensated by both the occupation of the state apparatus and control of local power, extracting and distributing resources at national and local levels in an effort to maintain support networks through rewarding clienteles and supporters.Beside the state apparatus occupation and local power resource usage, the Portuguese parties' supremacy over the institutional dimension is linked to their birth in a context of public financing.By contributing to strengthening material dependence on the state, public financing caused parties to lose interest in fundraising activities as well as in massive recruitment, both mass party formation activities.
Portuguese parties were also born in an environment where the media had a deep impact on society and television quickly became the main stage of electoral campaigns.
There were two kinds of effects caused by this, the first one being the centralization of national leaders and their "charisma institutionalization", and the second one the increasing factionalizing within parties -determined more by the support given to certain national or sub-national leaders than by ideological differences (Lisi 2009).
Soon parties and their leaders became synonymic, this having been enhanced by campaigns more and more focused on the national leaders' image (as potential candidates for prime minister), as well as by being highly professionalized and media centred (Lobo 2005).

Popular Support for Parties in Portugal: Meaning and Measurement
Firstly, we need to clarify the concept of popular support.By popular support we mean the extent to which individuals evaluate political objects positively (Easton 1965).
Regarding political parties, more than theoretical distinctions between different levels of support and different objects of political support (the political community, regime and political authorities), we are interested first and foremost in Easton's dichotomy of "diffuse support", which reflects more generalized and normative orientations towards political objects, and "specific support", which is more directly tied to evaluations of a political object (Easton 1975).Sources: Sani and Ossorio (1985); Bacalhau and Bruneau (1993); Barreto (2002;2005); Freire, Viegas and Seiceira (2009).
Whilst the former represents deeper political feelings that might provide a "potential reservoir" of support in times of political stress, the latter might reflect the immediate performance of an object, namely political parties (see also Dalton 1999;Klingemann 1998;Norris 1998).
In conclusion, what stands out from the data is the existence of a contradiction in Portuguese citizens' attitudes concerning parties.If on the one hand the democratic legitimacy they give to parties is undeniable, on the other their unhappiness and

National trends of popular support for parties
Table 4 shows the results of principal components analysis on all different items or questions mentioned above.The first and second questions introduced in both tables are intended to measure the overall attitude to political parties in general, that is, the dimension of "diffuse support" for parties.The third and fourth questions seek to capture what we hypothesized to be the basic evaluative and cognitive orientations towards the traditional Portuguese party alternatives, and those are part of the "specific support" dimension of parties.Sources: Sani and Ossorio (1985); Bacalhau and Bruneau (1993); Barreto (2002;2005); Freire, Viegas and Seiceira (2009).
Note: Principal components factorial analysis (Varimax rotation with Kaiser normalization).
The results suggest some immediate conclusions.On one hand, they show us that Factor 1 is made up of items involving the recognition of the necessity of political parties for the democratic system.As we can see in Table 4, two variables related to the dimension of bpsr (2011) 5 (1)  113  105 -128 Conceição Pequito teixeira and Paulo Almeida Pereira "diffuse support" for political parties belong to this cluster; the magnitude of factor loadings, the percentage of variance explained by this factor (ranging between 33.0 % and 37.0 %).
On the other hand, a second autonomous dimension clearly emerges from the principal components analysis, now formed by a combination of feelings of lack of support for the current behaviour of parties.Also here is the magnitude of factor loadings, the percentage of variance explained by Factor 2 (ranging between 35.0 and 44.0%).
To sum up, from the results of the exploratory factor analysis we can state that popular support for Portuguese political parties has two basic dimensions.The first one is the popular acknowledgment of the representative role political parties play in the political system according to the ideology of representative democracy.The second dimension could be best described as the criticism of the current functioning of political parties, indicating "specific" popular support for parties' behaviour.
It is also important to note that in each popular support dimension three types of attitudes can be found: "pro-party", "ambiguous" and "anti-party".Therefore, regarding the "diffuse support" for parties, we can speak of "pro-party attitudes" when respondents agree with the following statements: "without parties there can be no democracy" and "thanks to parties people can participate in political life".Conversely, if respondents disagree with these statements we have "anti-party attitudes", and when they agree with one statement only, "ambiguous attitudes".
Regarding the "specific support" for parties, we are interested to know whether respondents agree or not with the following statements: "political parties criticize each other often, but in reality they are all the same" and "parties only serve to divide people".If they disagree with both, we have "pro-party attitudes", while if they agree with both statements we have "anti-party attitudes".Finally, if respondents disagree with one statement only, we then have "ambiguous attitudes".
We shall now see how these three types of attitudes are distributed with respect to each separate dimension of popular support for parties.Regarding "diffuse support", in  Sources: Sani and Ossorio (1985); Bacalhau and Bruneau (1993); Barreto (2002); Barreto (2005); Freire, Viegas and Seiceira (2009).Note: Both "Don't know/Didn't answer" and "Neither agree nor disagree" were excluded from the analysis.
With regard to "specific support", it is also surprising to observe that in 2008 "antiparty attitudes" were higher than 1985 values, exactly after the end of the democratic transition and beginning of the consolidation of the regime.This shows that the strong "critical anti-partyism" of Portuguese public opinion on parties is relatively independent of political, economic and social mid-and long-term changes in the country -a pattern more expectable from "diffuse support" for political parties.Sources: Sani and Ossorio (1985); Bacalhau and Bruneau (1993); Barreto (2002); Barreto (2005); Freire, Viegas and Seiceira (2009).Note: Both "Don't know/Didn't answer" and "Neither agree nor disagree" were excluded from the analysis.
From this data we can draw the following conclusion: the legitimacy that the majority of Portugal's population recognizes in parties as pillars of representative democracy clearly coexists with a strong and generalized feeling of discontent about the way major parties have been carrying out their roles in the Portuguese political system.

individual-level Determinants of Popular Support for Parties in Portugal
We will now examine whether this public support, understood as a multidimensional Conceição Pequito teixeira and Paulo Almeida Pereira attitudes regarding parties are not subsumable into a single concept of "popular support", and have instead two distinct and autonomous dimensions -"diffuse support" and "specific support" -, it is our objective to test some specific hypotheses resultant from the following theoretical proposal: Since "diffuse support" and "specific support" are both conceptually and empirically distinct, these dimensions of popular support for parties have substantially different individual determinants.
H1 -While "diffuse support" for parties is the result of accumulated experiences over time, we have reasons to expect its individual determinants to be more structural than cyclical, and therefore liable to be analysed through cultural-based interpretations of the sources of political support.This approach focuses mainly on the role played by socialization patterns, social or interpersonal trust and basic values orientations as explanatory variables of "diffuse support" for the political system.
H2 -While "specific support" for parties is deeply dependent on short-term performance and the main parties' outputs, it is expectable that their individual determinants are more conjuncture-related than structural, and therefore liable to be analysed through performance-based interpretations of the sources of political support.From this point of view, a lack of support for parties would be a natural consequence of the discontent shown by citizens regarding the behaviour of major parties and their short-term policy outputs.

Hypotheses and methodology
In order to test these hypotheses, in this section we will carry out a multivariate analysis of public support for parties based on the most recent survey of the Portuguese population ( 2008).Since we have three possible choices (pro-party, ambiguous and antiparty attitudes) in each popular support dimension, we will use multinomial logistic regression in this analysis.

Dependent variables
The data used in this section was drawn from the most recent ( 2008 The multinomial logistic regression models developed in this article employ a number of independent variables that result from the two main theoretical approaches regarding political support, namely the cultural-based approach and the performance-based approach.

A) Social background and socialization experiences
To the extent that social background and socialization experiences do matter to the "diffuse support" dimension, the least advantaged could be expected to be the most disaffected and/or hostile to political parties.It is these sectors of society -the women, the elderly, the less educated, the less affluent, the unemployed and also the ones who live in rural and needier parts of the country -who are presumably the least likely to feel that parties are indispensable to democracy or that the party system is responsive to their needs.
However, considering the "specific support" dimension, and given its cyclical and more fluctuant nature, it is likely that attitudes regarding parties are more resultant of their "life cycle" and that they acquire a more critical and reactive tone amongst the young.Given that younger citizens are now associated with a "new style of citizen politics" (Dalton 2008; Inglehart 1997) reflecting an even more critical and reformist view of traditional political institutions, it is expected that they will be less supportive of parties than citizens of an older age.
Regarding education, as some authors note, in the less well-established democratic regimes of Southern Europe that underwent prolonged periods of authoritarian rule -Portugal's case -a specific kind of anti-party sentiment is associated with lower education (Torcal, Montero and Gunter 2007).This is an expectation that tends to be the opposite in well-established democracies, where those with higher education tend to have greater awareness of the need for parties in a representative democracy.This occurs mainly due to a socialization process that took place under a democratic regime, and the ensuing early assimilation of all the "rules of the game" that characterize and distinguish it from non-democratic systems.
However, since a higher level of education tends to also be associated to a higher degree of political sophistication (Dalton 2008), this could explain the rising expectations of parties by the citizens of the Portuguese democracy, as well as the leaning by those who possess higher levels of education to act in a more demanding and critical fashion towards political institutions, elites and processes.
Therefore, there are reasons to expect that the more educated are more likely to become frustrated with manoeuvres performed by parties and politicians and harbour anti-party sentiments than the less educated.Also, the perceived lack of responsiveness by political institutions and authorities to the demands of less-developed regions can lead citizens to feel they are not receiving their fair share of wealth or power, hence the strong criticism of the current behaviour of the party system.
However, the fact that individuals who live in urban or semi-urban areas -and in the more developed regions near the coastline -are closer to the political decision centres and also more attentive to party performance and policy content can lead to more critical and reactive attitudes towards specific parties -typically the system's main players -or in other words, to a larger critical anti-partisanship.

B) Political sophistication
Political variables may be helpful in understanding differences in public opinion about political parties.These variables express what we can call political sophistication, conceptualized here as an interest in politics and inclination to discuss it, internal political efficiency and media -namely television -exposure in order to obtain political information.
As for this set of variables, it is expectable that a politically sophisticated public would have a greater ability to understand the central role and the important functions political parties play in modern representative democracies.As far as the "specific support" aspect is concerned, there are also reasons to expect that a politically sophisticated public would be less likely to hold unrealistic expectations of party behaviour and have anti-party sentiments (Dalton 2008; Norris 1998).
However, this interpretation is far from consensual.According to some authors, most sophisticated publics are more demanding about the functioning of political institutions, which makes them more sceptical and critical about the poor functional performance of parties.
With regard to television, it is expected that greater exposure to political news implies greater "diffuse support" for political parties as such.When analysing the "specific support" dimension, this pattern is not as expected as in "diffuse support", as some scholars argue that the increasing tendency shown by the media (television) to report on corruption and scandalous behaviour contributes to the erosion of public support for parties, making citizens much more sceptical and even cynical about the functioning of established parties.

Public Attitudes towards Parties in Portugal
On the other hand, and taking into account Ronald Inglehart's thesis, one of the alleged consequences of "cultural change" and of the increasing adoption of post-materialist values (Inglehart 1977;1997), when observed in the context of advanced industrial societies, is that individuals adopt more demanding criteria regarding the evaluation of traditional authority sources, a category in which parties are necessarily included.We can therefore admit that in that increasing demand there is a delegitimizing and criticizing anti-partyism.

C) Evaluation of economic performance
The lack of public support for parties may also have economic roots.A commonly mentioned cause for the lack of political support has been the negative public assessment of economic performance.With regard to parties, one might expect political support to reflect the ability of the major parties (the ones with a propensity to form government) to deliver what the general public expects of them in terms of economic policies.When successive governing parties fail to reduce public dissatisfaction with their poor economic performance, the public may become generally hostile to or even reject political parties.
The variables considered here are: the proximity of the respondent to the incumbent party; the retrospective evaluation of the party in the government's economic performance and, lastly, the retrospective evaluation of the country's economic situation.With "diffuse support", a kind of positive sentiments reservoir concerning political objects, we can admit that it will not be challenged by short-term phenomena or specific advantages at a given moment.This means that even when voters show deep discontent and dissatisfaction with the parties' short-term responses to the country's economic problems, they still recognize their democratic legitimacy.
Conversely, and because "specific support" for parties is related to the way parties, particularly the incumbent party, respond to the social needs of citizens and the distribution of benefits concerning their personal material well-being, we have reasons to expect the following: the greater the proximity to the incumbent party (through voting) and the more negative the assessment of both the party's economic functioning and the subjective state of the country's economy is, the greater will the frustration regarding the performance of established parties be.
What also needs to be understood is if, given the current economic and financial crisis in Portugal, public frustration might be mirrored in overall dissatisfaction not just with the party in government, but also with political parties as such.Furthermore, it is also important to know to what extent the combination of frustrations with the economic performance of the parties in government will give rise to long-term attitudes towards political parties in general.dimension, it is expected that the critical and reactive attitudes adopted by citizens as a response to the current behaviour of parties will be more expressive among those who are suspicious of the main political institutions, and also among those who consider politicians in general to be unaccountable and unresponsive to citizens' needs.

E) Civil society and social capital
Civil society and social capital theories have different intellectual origins but are still related: both argue that the stability and quality of representative democracy -therefore of its main institutions -rest on strong informal and formal social networks independent of government, and both emphasize the importance of the underlying social foundations of democracy (Putnam 1995).
In order to test the impact of social capital, we have included in our regression model another set of independent variables, namely generalized social trust and belonging, and active participation in social and civic associations.Our expectations regarding these variables are the following: attitudes towards parties will be less hostile the greater the social trust and civic engagement are, especially in the "diffuse support" dimension.

results and Discussion
The results of the multinomial logistic regression models that determine the variables that better explain "diffuse" and "specific" support are presented in Tables 7 and 8.As we can see, both models show percentages of total variance explained that are rather satisfactory, with pseudo R2 values that vary between 37% and 43%.We consider them to be rather satisfactory because democratic legitimacy and dissatisfaction regarding the performance of political parties not only coexist in Portuguese public opinion, but are also extremely diffuse and generalized phenomena among a large segment of the population, making the detection of explanatory factors at an individual level difficult.
We will now analyse Table 7, which shows the parameter estimates resulting from the model predicting citizens' "diffuse support" for political parties.
If we contrast individuals who manifest attitudes of hostility towards parties in general (anti-party attitudes) with those who recognize parties as essential to the functioning of democracy (pro-party attitudes), we verify that respondents living in rural areas, who declare themselves to be moderately partisan or non-partisan, demonstrate materialistic value priorities, and have a negative perception of the country's economy are also those who have a higher probability of adherence to an anti-partyism of the delegitimizing type.In turn, if we compare the profiles of respondents who manifest pro-party attitudes and of the ones who show a neutral or indifferent attitude concerning parties, we observe that the latter have a higher probability of differing from the former on the following essential Conceição Pequito teixeira and Paulo Almeida Pereira aspects: they live in semi-urban areas and outside the more developed region of the country (Lisbon and the Tagus Valley).In terms of political attitudes, they are non-partisan or moderately partisan, have a negative perception of the country's economic situation and have weak social capital, since they tend to be suspicious of others and participate less in social and civic associations.
When analysing the parameter estimates that result from the model that predicts citizens' "specific" support for political parties, we found that the contrast between citizens who support the concrete functioning of established parties (pro-party) and those who do not (anti-party) becomes more evident in terms of social background.This occurs because there is a higher probability that citizens who show critical and reactive attitudes towards the performance of the party-system will be younger, single, possess higher educational capital, and live in an urban habitat or the more developed region of Lisbon and the Tagus Valley.In comparison to those who manifest support for the established parties' behaviour, the probability of these citizens finding themselves unemployed, feeling distant from the party in government, and having a negative assessment of its performance and the country's economic situation in general is higher.There is also a higher likelihood of them having mistrust in political institutions or negative impressions of politicians in general than among respondents who support the main parties.
As for the comparison between respondents who show themselves to be neutral or indifferent concerning the performance of parties, and those who support their performance (reference group), what stands out is the predicted probability of the former feeling less close to the party in government and assessing negatively its economic performance.Nevertheless, the probability of them being younger and possessing higher educational capital than respondents who claim to feel satisfied with the performance of parties also stands out.
Having reached this point in our analysis, we will now contrast these results with our investigation hypotheses.Regarding the "diffuse support" dimension, we have observed that in terms of the socio-demographic background, our hypothesis claiming that diffuse anti-partyism is associated to deprivation of social, economic and educational resources is not confirmed.The hypothesis is somewhat confirmed only in the case of habitat and region.Invalidated are also the hypotheses claiming that in Portugal -more than 30 years after the April 1974 Revolution -there subsists a lack of "diffuse support" for parties among the older and less educated population segments, whose socialization occurred under the authoritarian regime, which was characterized by an anti-political and anti-party rhetoric and practice.
However, regarding the "diffuse support" dimension, the hypotheses associating the closeness to a party, adherence to materialistic values, trust and social and civic participation to a greater support for parties as institutions that play a central role in democracy are confirmed.However, it is curious to note that along with these long-term attitudinal variables there is a short-term one that stands out.We are referring to the retrospective analysis (of the last 12 months) of the country's economy.There is therefore a larger probability that anti-party feelings are greater among those who negatively assess the country's economic performance, thus seemingly dissociating it from the general performance of the party currently in government.In our understanding, this clear dissociation between the assessment of the government's performance and the country's economy allows us to state the following: the combining of short-term frustrations with the economic performance of the main parties likely to be in government (PS and PSD) may have become a political experience in itself.
In fact, as we can see in Table 5, such results far from presuppose the existence of informed and mobilized people, both critically and cognitively.As for the variables related to the social background of the respondents who showed greater criticism of the main parties' performance, such a hypothesis is widely rejected.

Conclusion
Throughout the preceding pages we attempted to shed light on the attitudinal Portuguese map regarding political parties, whose misunderstanding seems sometimes to justify some contradictions, ambiguities and perplexities.With a set of questions in four public opinion surveys (1985, 1993, 2002 and 2008) we showed that the attitudes of the Portuguese towards parties can be grouped into two dimensions -not just theoretically, but also empirically distinct.One of these translates "diffuse support" for parties and relates to the recognition and internalization of the popular role and functions that parties are often called to play in representative democracies.Another, in turn, corresponds to the satisfaction and contentment of the people with the practical operation of the established parties, and is reflected in the "specific support" to them.
The finding that these two dimensions are empirically independent allows us, therefore, to divert some more pessimistic interpretations that tend to relate the hypothetical crisis or perceived decline of parties in advanced industrial societies with the current loss (almost irreversible, according to many scholars and political commentators) of their democratic legitimacy.As we have seen, "diffuse support" for parties is not an issue in Portugal, and strong dissatisfaction with the functioning and performance of the major parties is far from producing any "contagion effect" capable of eroding the "reservoir of favourable attitudes" regarding the indispensability of parties and the functions they perform in the context of representative democracy.
If they can be considered an "evil", they are, however, as we have seen, a "necessary evil".And if there is a crisis of political parties in Portugal, it is certainly not a crisis associated with a loss of legitimacy but rather with a poor and challenged performance.socialization not only unequal (men versus women), but also dual (dictatorship versus democracy) -, the fact is that this hypothesis has not been confirmed by our analysis.
The hypothesis that criticism of the current performance of the major parties ("critical and reactive anti-partyism") would have in the younger, better educated but also the more politically sophisticated its main spokespersons in Portugal was also ruled out.If it had been confirmed, it would, in a sense, have been in accordance with the theories of Inglehart (1977;1997) andDalton (2008).But again, contrary to expectation, hostility to or rejection of the Portuguese parties as such are not fully independent of other phenomena of a more cyclical than structural nature, such as the performance of political institutions and political assessment of the current leaders, and neither are the discontent with and criticism of parties.
In short, regarding the individual determinants, the "diffuse support" for parties (or lack of it) is explained not only through a cultural-centred interpretation, as we expected, but also through performance-based interpretations of sources of political support.Exposure to TV news "How frequently do you follow political events through the media?Television." 5-categories from lower 1 (never) to higher 5 (every day) To what extent do you agree with the following statements: a) Without parties there can be no democracy; b) Thanks to parties people can participate in political life".This variable is composed by the combination of answers to these two questions, and was recoded into the following: If respondents disagree with both statements, we can say we are before "anti-partisan" attitudes.If respondents agree with both statements, then we are before "pro-partisan" attitudes.Finally, if respondents agree to one statement and disagree with the other, we are before "neutral" attitudes.After the recoding, the variable was divided in two categories: Propartisan attitudes 1, Other 0

Specific support for parties "
To what extent do you agree with the following statements: a) Parties are only good to divide people; b) Parties criticize each other often, but in reality they are all the same".This variable is composed by the combination of answers to these two questions, and was recoded into the following: If respondents disagree with both statements, we can say we are before "pro-partisan" attitudes.If respondents agree with both statements, then we are before "anti-partisan" attitudes.Finally, if respondents agree to one statement and disagree with the other, we are before "neutral" attitudes.After the recoding, the variable was divided in two categories: Pro-partisan attitudes 1, other 0 Party identification "Is there a party you feel more identified with?" Yes 1, no 0 Proximity to the ruling party "Did you vote for the party currently in government in the last legislative elections?"Yes 1, no 0 Political and economic performance Trust in political institutions "What is the degree of your trust in each of the following institutions?""Government, Parliament, courts and public administration" 4-categories from lower 1 (no trust) to higher (high trust) Opinion about politicians "To what extent do you agree with the following statement: Politicians don't care about what people like me think" 4-categories from lower 1 (completely disagree) to higher 4 (totally agree) Assessment of government performance "Thinking of the general performance of the current government, how do you assess its performance?"Categories from lower 1 (very poor) to higher 4 (very good).[Inversion of the scale] Assessment of the state of the economy "What do you think of the state of economy in Portugal?"5-categories from lower 1 (very poor) to higher 5 (very good) [Inversion of the scale]

Social capital
General trust "In general terms, do you think that most people can be trusted or that you can never be too careful".Most people can be trusted 1, You can never be too careful 0

Member of voluntary associations
Are you a member of any of the organizations or associations here mentioned: a) unions; b) employers' organizations; c) socio-professional associations; d) religious organizations; e) social organizations; f) civic action organizations; g) cultural organizations; h) sports clubs; i) recreational associations and clubs.The variable is composed by the sum of the answers to these items: Yes, I am a member 1, No, I am not a member 0 Social and civic activism "Do you actively and regularly participate in any of the organizations where you are a member: a) unions; b) employers' organizations; c) socio-professional associations; d) religious organizations; e) social organizations; f) civic action organizations; g) cultural organizations; h) sports clubs; i) recreational associations and clubs.The variable is composed by the sum of the answers to these items: Yes, I am a member 1, No, I am not a member 0 Sources: Sani and Ossorio (1985); Bacalhau and Bruneau (1993); Barreto (2002;2005); Freire, Viegas and Seiceira (2009).
organizations.Portuguese parties appeared within a revolutionary context instead of developing because of social cleavages or a pre-existent party system like most European countries.The Social Democratic Party (PSD) and CDS formed in 1974 following the April revolution, while the PS formed in 1973 by members exiled at the time.The Portuguese Communist Party (PCP) was therefore the only main party operating in Portugal at the time of the revolution, having been created in 1921.
defended the adoption of a popular-type democracy, inspired on the ex-USSR model, while the Socialist Party (PS), the Social Democratic Party (PSD) and the now Democratic and Social Centre -People's Party (CDS-PP) wanted a representative democracy of the Western kind.
way existent parties carry out their representative and governmental functions is also evident.
) national survey, based on a representative sample of the continental part of Portugal (N = 1350).It was applied to citizens of 18 years of age and over, who were also stratified by region and type of habitat.In the multinomial logistic regression models, the dependent variables are "diffuse support" and "specific support" for political parties in Portugal.These are multinomial variables, since each comprehends three categories distinguishable by the following nonnumerical characteristics: 1) pro-party attitudes; 2) ambiguous attitudes and 3) anti-party attitudes.
Lastly, in the multinomial logistic regression model we include two other long-term political variables, since they are assimilated by individuals during the first phases of the political socialization process.The first one is the ideological self-placement of citizens on the left-right scale and the second is their adherence to materialistic or post-materialistic values.We expect those who place themselves to the right of the ideological spectrum to show greater hostility towards parties, because three decades after the 1974 revolution, and as some studies show, the support for the democratic regime by voters on the left is greater than by voters on the right(Heimar et al. 1990;Magalhães 2005).
of political institutions and politicians' performanceSome indicators tell us that trust in political institutions and politicians affect what people think of political parties.On one hand, citizens are more likely to recognize the legitimacy of political parties as the level of trust in the institutions that constitute the pillars of democratic systems rises.On the other hand, and considering the "specific support" On the other hand, as we demonstrated, both popular support dimensions possess distinct determinants at an individual level.If the initial hypothesis of the lack of "diffuse support" for parties ("delegitimizing anti-partyism") was based on deprived economic resources and education, being most prevalent among older people and women -in terms of patterns of political Left-right self-placement "In politics, people sometimes talk about 'left' and 'right'.Card nº 11 defines this situation: 0 -stands for someone whose views are entirely to the 'left'; 10 -for someone whose views are entirely to the 'right'.Of course, there are intermediate positions to the degree that defines one's views as being more or less to the 'left' or to the 'right'.When you think about your own ideas on this, where would you place yourself on this scale?"Salience of left-right selfplacement Self-placement in left-right scale 1, no self-placement in left-right scale 0 Post-materialist values To what extent do you agree or disagree of the following statements: a) Stronger measures should be enforced to protect the environment; b) Immigrants are beneficial to Portugal's economy; c) Women should be free to decide about abortion; d) Marriage between individuals from the same sex should be allowed by law.The variable is the result of the sum of affirmative answers to each one of these items (completely agree + agree) Attitudes towards parties Diffuse support for parties "

Table 5
we see "pro-party attitudes" assuming a clear preponderance over "anti-party" ones: on average 84% of the Portuguese population recognizes that parties are necessary to democracy, supporting Schattschneider's (1942) view that "democracy without parties is unthinkable to most citizens".party alternatives due to their poor functioning.At the same time, little over 20% seemed satisfied with the way major parties had been carrying out their roles.bpsr 114 (2011) 5 (1) 105 -128 Public Attitudes towards Parties in Portugal

Table 7
Parameter estimates of the multinomial logistic regression of popular support for parties inPortugal (2008)

Table 8
Parameter estimates of the multinomial logistic regression of popular support for parties inPortugal (2008)