5. PARTY LEVEL EXPLANATIONS FOR PARTY UNITY IN EUROPE
5.1 Party unity of behavior versus unity of attitudes
This section will show that what politicians say and what they do are two different issues. Edmund Burke defined a political party as “a body of men united for promoting by their joint endeavors the national interest, upon some particular principle upon which they are all agreed” (Ayling 1988: 73). In keeping with Burke’s definition, parties have been long perceived as unitary actors in government and coalition formation theories (Baron 1993, Budge and Keman 1990), but lately this assumption has been challenged (Laver and Garry 2000, Laver and Shepsle 1990) by a greater emphasis on intra-party politics and individual preferences. Working along similar lines, I explore unity of attitudes and behavior and advocate a clear distinction between them. From one opinion to a specific action based on that opinion there is a
CEUeTDCollection
considerable distance with many intervening factors - the voters, the constituency party, the national party, the parliamentary party - which can all alter the final decision of an individual MP. Besides these, trade unions relations, or even the influence of the church in some countries, can be possible related factors.
While there is agreement that modern democracy is representative democracy, the representation process has its paradoxes. Most parliamentary democracies emphasize individual freedom for their MPs in their national constitutions. Still, in practice, if every MP is free to vote as he/she wishes the very process of representation and translation of programs into public policies will be jeopardized and will not take place efficiently.
Even if nearly all Western and Central Eastern democracies assume that members of parliament are free to vote as they want, in reality, the MPs vote in a united manner either because they have the same opinion as their party or because the parties have increasingly strengthened the apparatus whereby they control their parliamentary representatives. The figures presented on the next page, illustrate that unity in attitudes is not the same as unity in behavior. My research shows, therefore, that simplifications can distort reality more than commonly realized. Political parties are not always united and made up of like-minded individuals who share the same opinions.
As of 1996, attitudinal unity based on the elite surveys, is compared across party families in Europe in figure 5.1. Similarly, the same comparison is presented with the unity of behavior measured by the Rice index alone in figure 5.2. Ideological unity is
CEUeTDCollection
based on MPs’ declarations from 1996 elite surveys, on the question of their position on specific policy areas. Unity in attitudes is therefore measured by the standard deviation of the MPs’ specific policy positions. The higher the standard deviations, the less similar are the party MPs in their views about policy areas. Therefore the lower the scores are in figure 5.1, the higher the unity of attitudes within party families.
Figure 5.1 Unity of attitudes by party family in 1996 (standard deviation of MP's policy attitudes)
Figure 5.1 above illustrates different levels of party ideological unity across party families in Europe. Especially the ethnic regional, ecologist and communist parties were very much united around their policies at that time. A comparison of both figures 5.1 and 5.2 portrays the difference between unity of behavior and unity of attitudes as of 1996 when party family is taken into consideration. Although unity of
Bars show means
ecologist communist
social democratic liberal
Christian democrat conservative
nationalist agrarian
ethnic regional special issue
party family
1.00 2.00 3.00 4.00
1.19 1.16 2.15
1.72 1.77 2.18
2.45 2.85
1.62 3.87
CEUeTDCollection
behavior and unity of attitudes are correlated at 0.40, they are not the same thing and they do not always vary together. The agrarian and special issue parties appear as the most united parties in their behavior but they have the lowest attitudinal homogeneity in their MPs when compared with the other party families across Europe. In the two contradictory cases presented, there could be other factors that make the MPs very united in their voting: party restrictions or disciplinary measures imposed, the power status of the respective parties, institutional restrictions (like the electoral system) or because other outside agents impose unity, such as the church or the trade unions.
Figure 5.2 Unity of behavior by party family 1996 (Rice's index)
While there is a lot of variation in the attitudes towards party ideology, the same thing cannot be said about voting unity. There is not much variation in voting scores between party families. Voting along the party lines appears to be indeed very important for MPs and for parties. It is in parliament where public policy decisions are
Bars show means
ecologist communist
social democratic liberal
Christian democrat conservative
nationalist agrarian
ethnic regional special issue
party family
0.00 25.00 50.00 75.00 100.00
94.96 94.33 94.9592.52 92.45
88.19 87.25 96.16
93.7797.37
CEUeTDCollection
ratified and a parties’ policy platform put into practice, and it is also in the parliament where parties will shape policy and secure votes for future elections based on their ability to deliver the promised platform.
The situation after ten years has not changed much. The Borz, Enyedi, Janda 2007 party unity expert survey13 illustrates unity of behavior (voting) and unity of attitudes (ideological) correlated at .38, which clearly means that they cannot be confused with one another. Table 5.1 portrays the difference between the two, when factionalism is taken into consideration. All types of factionalism appear negatively correlated with unity of behavior but are not similarly correlated with unity of attitudes. This is an attestation that factionalism (ideological, leadership or substantive) is an indicator for the lack of unity, clearly pertaining more to behavior than to ideology. Out of the three types of factionalism considered, issue factionalism proves to have the highest association with a lack of voting unity. The result is not surprising, as recent issues such as redistribution, minorities’ rights, EU enlargement and integration, taxation, defense policy and also social rights of homosexuals or abortion law have caused much dissent over votes in European national parliaments. In the conditions of a financial crisis, Gordon Brown, who followed Tony Blair as the new British Prime Minister in June 2007, changed the taxation process by withdrawing the ten pence tax rate which was introduced by the same government ten years ago. The measure caused rebellion among the Labour MPs who, even after discussions and negotiations with the party whips, threatened to vote against future government bills. The internal
13 The expert scores on voting unity are correlated at .60 with the Rice Index values from 1997, which validates the usage of experts for the future in large comparative studies. Of course there is a difference between raw calculations of Rice index and the value judgments, but I expect the expert values to give a clearer image of voting unity, overlooking many of the problems the Rice index has as a generally accepted index of party unity (i.e. instances of abstentions, types of votes considered, time period considered, instances of disagreement when a vote is not actually reached in the House etc.).
CEUeTDCollection
party crisis led to a ‘U’ turn strategy from the government which then introduced measures meant to ease the situation of those affected by the new policy.
Table 5.1 Factionalism, unity of behavior and unity of attitudes correlations Unity of
behavior 2006/07 Leadership factionalism
Issue
factionalism Ideological
factionalism Unity of attitudes Unity of
behavior 06/07 1
N=175 -.437**
N=174 -.569**
N=174 -.439**
N=174 .381**
N=174 Leadership
factionalism -.437**
N=174 1
N=182 .702**
N=182 .637**
N=182 -.391**
N=178 Issue
factionalism -.569**
N=174 .702**
N=182 1
N=182 .877**
N=182 -.402**
N=178 Ideological
factionalism -.438**
N=178 .637**
N=182 .877**
N=182 1
N=182 -.345**
N=178 Unity of
attitudes .381**
N=174 -.391**
N=178 -.402**
N=178 -.345**
N=178 1
N=178
** correlation significant at 0.01 level; data source: 2007 Borz, Enyedi, Janda party unity expert survey
The social democrats in Finland constitute another such illustrative example of issue factionalism. While the party is united on all other policy areas, on foreign policy alone they were still divided in 2007. The conflicts in this area occur between the traditionalists and Atlantic reformist groupings formed either around a former foreign minister or a former party leader.
Issue factionalism is more negatively associated with unity of behavior (correlation coefficient -.56 significant at 0.01, table 5.1) than with unity of attitudes. The existence of issue factionalism does not exclude unity of attitudes with respect to the overall party ideology, and the same is the case for leadership factionalism. The persistence of the latter, for example in Austrian FPŐ or BZŐ parties, does not necessarily imply that their members are not united in their political beliefs. Apart from conflicts between the populists and the nationalists, which made the former leader Haider leave the party together with other highly positioned party personalities and form BZŐ, most of the FPŐ members and representatives agree on stronger
anti-CEUeTDCollection
immigration laws, stricter law enforcement, and more support for families in poverty.
Another clear example is the Belgian Front National Party which, according to the experts from our survey, exhibits the highest degree of leadership factionalism but still manages in practice to appear as an ideologically united party, with low levels of issue and ideological factionalism and a fairly high agreement among its MPs insofar as the party policies are concerned. Overall, this strengthens the argument that unity in attitudes and unity of behavior are distinct concepts, and their interaction with the surrounding institutional environment may be different. Whether unity of behavior results from party or institutional systemic restrictions will be clarified in the next sections and the next chapter where the major systemic and party level hypotheses will be tested.