5. PARTY LEVEL EXPLANATIONS FOR PARTY UNITY IN EUROPE
5.5 Accounting for change over the last decade
The analysis in this chapter has shown the static relation between the variables considered in the party unity model. The aim was to see if the relationships are there or not, and how strong they are. The model shown in chapter 2 theoretically implies a dynamic relationship, aimed especially at those parties which, in the light of conditions x, see themselves in the position of adopting measures y. The prediction was that, given the relative young age of the party systems in Central Eastern Europe,
CEUeTDCollection
the parties will not have a very high degree of attitudinal unity (for which has already been given evidence in chapter 4) and, in order to arrive at a very united vote in the parliament, they will increase their level of centralization and apply more disciplinary rules.
The 2007 expert survey provides information about the change in the voting unity over the last decade in programmatic cohesion and party centralization. The results of the regression analysis presented in table 5.9 favor the argument portrayed in the above mentioned model. The higher centralization has become over the last decade, both in Western Europe and in Central Eastern Europe, the more party unity has increased, both in ideology and MPs’ voting behavior. In an obvious manner, the more homogenous a party has become in terms of ideology and policy positions, the higher its agreement when voting in the parliament. As a logical consequence, the higher party unity of behavior has become by 2007, the fewer disciplinary measures have been applied over the last decade.
Table 5.9 Explaining the change in unity of behavior over the last decade
Variables Indicators
Change in unity of attitudes .302***
Change in party centralization .232**
Disciplinary measures applied over the last decade
-.121*
*p<.1, **p<.05, ***p<.01; dependent variable: change in unity of behavior over the last decade;
unit of analysis political parties; N=131; R2=.20; Adjusted R2 =.18;
Note: entries are standardized beta regression coefficients;
Data source: 2007 Borz, Enyedi, Janda expert survey on party unity;
As shown, both centralization and disciplinary mechanisms are tools for achieving unity of behavior. There is however a fundamental difference between the two. While
CEUeTDCollection
they are positively associated with each other, when it comes to change over time as shown in table 5.9, centralization positively affects unity while disciplinary measures affect it negatively. The negative coefficient allows for contradictory interpretations.
One is that using disciplinary measures tends to decrease party unity of behavior, while the other, and the more plausible one, is that disciplinary measures are not needed anymore when unity increases. As parties get more and more united they need disciplinary actions less and less. This however works only above a certain level and until that level of unity is reached disciplinary actions help. This interpretation requires us to differentiate between two different mechanisms which work behind centralization and disciplinary measures. Centralization appears to produce a “culture of reward” for unity, whereas disciplinary measures reflect a “culture of punishment”.
As MPs behave more and more according to the organizational reward structure, there is less need for punishment.
The changes in party centralization over the last decade go hand in hand with changes in programmatic cohesion. The correlations showed by the change in both variables are much higher (.37, table 6 appendix D) than the correlations between their 2007 snapshot values (.13, already presented in table 5.7). This correlation suggests that changes in centralization and unity of attitudes vary together (at least in the memory of experts) without implying any causal relation or temporal sequence. As suggested earlier in chapter 2, especially for Central and Eastern Europe, it is expected that in the conditions of low party ideological unity, in order to achieve the desired level of behavioral unity, the party will increase its centralization. The change could normally be generated by the party leadership or the parliamentary party and arise mostly from the desire to be re-elected.
CEUeTDCollection
Figure 5.4 Party centralization and programmatic change over the last decade
5.00 4.00
3.00 2.00
1.00
change in party centralization
5.00
4.00
3.00
2.00
1.00
programmatic cohesion change
UK UK
UK UK
UK Switzerland
Switzerland
Switzerland Switzerland
Slovakia Slovakia
Slovakia
Slovakia
Slovakia
Slovakia Slovakia Slovakia
Sweden Russia
Russia Russia Russia
Russia Russia Romania
Romania
Romania
Portugal Poland
Poland
Poland
Poland
Poland
Netherlands
Netherlands Netherlands Luxembourg
Luxembourg
Luxembourg
Italy Italy
Italy Italy Italy
Iceland Iceland
Iceland
Hungary
France
France France
France Denmark
Denmark
Czech Republic
Czech Republic
Belgium
R Sq Linear = 0.136
Figure 5.4 illustrates the change in party centralization as explained by the change in party ideological unity. The scatter plot presents the parties and their country affiliation. As observed, 13 percent of the variation in ideological unity is explained by the variation in party centralization over the last decade. Especially in the right upper corner of the figure we can observe cases (parties) from Central Eastern Europe and this verifies my initial proposition that, especially in these countries, the increase in party centralization is explained by the levels of unity of attitudes within parties.
Most of the remaining parties are situated in the centre (value 3 on both axes means no change), meaning that not much change occurred during the last decade and that parties do indeed change slowly. Some parties in Luxembourg, Iceland and Denmark have increased their programmatic cohesion while keeping the same level of
CEUeTDCollection
centralization, while others, mostly from Eastern Europe (Slovak, Russian, Polish parties and some exceptions in UK, Belgium or Italy), have maintained or decreased their programmatic cohesion but have increased their centralization level.
The disciplinary measures applied by parties always go hand in hand with the process of strengthening party centralization. The occurrence of disciplinary measures such as expulsion, loss of privileges or simple warnings is moderately correlated (.30) with leadership factionalism. This means that, with the exception of political parties which have factionalism officially acknowledged in their party statutes, a large number of the remaining parties are trying to avoid factionalism based on the personal attraction of individual party members. Besides centralization, one solution for that problem are the above mentioned disciplinary measures. Radu Comănici, a popular local leader and member of Greater Romania Party (PRM), was expelled from the party after he strongly criticized the central leadership and asked for major internal reforms in order to make the party more competitive in elections. The motive for the expulsion was the he tried to falsify the party list18 for the local elections in 2008 and did not take into account the position of central party executive on the list of candidates. Such measures and this example highlight the party’s determination to avoid leadership factionalism and emphasize how centralization and disciplinary measures go hand in hand in upholding unity of behavior.
18“500 de peremişti cer demisia de onoare a preşedintelui PRM”(“500 members of The Greater Romania Party are asking for the resignation of the Greater Romania Party Leader”) in Adevarul 16 May 2008.
CEUeTDCollection
If disciplinary measures19 are not imposed by the party soon enough, the dissatisfied members will leave the party on their own initiative. A very recent example is Sorin Oprescu, former member of the Romanian Social Democratic Party (PSD) who, on the grounds that he was not nominated to run as a party candidate for the local elections (after two previous unsuccessful candidacies), resigned from the party and ran as an independent candidate for the Bucharest position of mayor in June 2008 local elections. Another famous example is of the Slovak MP, who not only left his original party, but then joined its opponents. The MP did this on various occasions, and even declared in the Slovak Daily SME,20 in 2003, that changing parties is like changing any other job and going for the one which offers more. The MP defected first from the New Civic Alliance (ANO) and, after he had spent some time in the parliament as a non-affiliated deputy, in November 2003 joined the Slobodne Forum (Free Forum) which is a breakaway parliamentary faction of SDKU (Slovak Democratic and Christian Union). Similar cases occurred frequently in Russia where, as Mikhail Gorbachev mentioned about the 2003 elections, “A politician may be today in the list of one electoral bloc, tomorrow – of another, and the day after tomorrow - on the third list. Utter, cynical pragmatism. Only to get pork, nothing more” (Kulik 2007: 193).
19 Other measures against defection are however stipulated in the country constitutions, such as in Portugal, and discussions about similar constitutional amendments have taken place in the Romanian parliament as well.
20 "Opaterny: Zmena strany? Ako zmena zamestnania", SME (Slovak Daily), November 23 2003, accessible at
http://zaujimavosti.sme.sk/c/1178255/Opaterny-Zmena-strany-Ako-zmena-zamestnania.html
CEUeTDCollection