With the local elections in 2006, new organisational schemes have been applied by the Flemish government to the 308 municipalities. These schemes included institutional reform aimed to change the outset of local government. In this article we studied the influence of these reforms in practice. Are they actually carriers of change? We thereby focused on the direct aspects linked to these reforms. While many instruments of reform had indirect intended effects, acceptance and action among key decision-makers (both voters and local governors to the extent of their discretion) to use them directly becomes crucial. Our analysis has shown that, with some exceptions, direct autonomous space for action was only limitedly used, hence reducing the chances for indirect change. While it is still too early to fully assess the foreseen indirect effects, in our opinion it is crucial to understand the nature of these local reforms within the central (Flemish) bargaining arena. The latter seems to have transformed the new schemes of local governmental organisation to the path-dependent art of the political feasible.

The last local elections in Wallonia were marked by the introduction of a new legislation concerning the way mayors are appointed. Before 2006, mayors were appointed by the regional government. Since the last elections, a new decree institutes that is automatically appointed mayor the candidate having most preferential votes from the list having most seats. In this article, we explore how the new legislation has affected the way voters behave in 2006 and also how it has affected parties when it comes to coalition agreements. What appears is that the new legislation has a fairly limited impact. Voters did not cast more preferential votes in 2006 than in 2000. The logics of coalitions have not been changed significantly. Finally, the only notable – even if not spectacular – impact of the new Walloon decree is that the elections have been much more focused on the candidates that were leading their list, the ones that were presented by their party as their candidate to become mayor. These candidates have more often been appointed mayors in 2006 than in 2000 and the proportion of preferential votes that they have received is higher in 2006. In that sense, even if the degree of change must not be exaggerated, Walloon local elections are slightly turning into a horse race between candidates leading their list.

The article analyses the gender balance in local Belgian politics in the aftermath of the 2006 local elections. It offers a comparative perspective on the position of candidates, representatives elected and those holding office including data from the last three decades. The article also discusses the attitude of local party sections towards measures to foster a gender balance. The data show that local party sections more easily agree on parity at the level of candidates than that they support a real sharing of power at the level of the executive power. Local party sections do not support more than a guaranteed minimum presence of wo/men in the councils of eldermen. Furthermore, both at the level of candidates and at that of the executive the data reveal a final glass ceiling. Women have difficulties reaching the top positions on electoral lists as well as the top positions in the local executive power.

In national elections the results tend to become more ‘nationalized’: a homogeneous party offer all over the territory, less variation in the results per constituency and more homogeneous electoral swings. This article investigates whether this nationalization can also be witnessed at local elections. It focuses on two indicators: the party offer and the voting behaviour. The party offer is the presence of the national parties on the local ballot paper, while the voting behaviour looks at patterns of homogeneity across the municipalities.
The answer to the question of nationalization is mixed. The Flemish and Walloon local elections display the same long-term trend as the national elections, but they keep their own local character. The heterogeneity of the local party offer clearly demonstrates the local specificity of the local elections, and consequently the voting behaviour also differs from the voting behaviour at national elections. We also find that the local elections in Wallonia are less nationalized than in Flanders.
Although the local character of the local elections remains important, the newer parties – Ecolo and Groen! – show until 2000 a clear trend towards nationalization. Especially the extreme right Vlaams Belang shows positive scores on all indicators of nationalization since its first local appearance in 1982.

It is often suggested that the nationalization of local elections has increased. As a result, a hypothesis could be that the mutual differences between policy programs of local divisions of the same national party decrease. In this contribution, we focus on the local election of October 8th 2006 in order to analyse these mutual differences. The aim of this contribution is to measure the homogeneity or heterogeneity of the policy positions of the local divisions of national political parties, on a range of substantive issues on which they have the freedom to differ. Therefore, we compare the opinions of local party agents within the different party families. We use the results of a survey among the representatives of the local departments of the different political parties in the run-up to the local elections. Our research shows that, regarding to the selected questions, in general the local divisions speak with one voice. Our analysis does not indicate that there is a large mixture of visions between divisions of the same national party. Besides, this analysis shows that in general, the size of the municipality can seldom be used in indicating the relative disagreement within political families. This level of agreement is the largest within the green party and the smallest within the liberal family. Those are also the two parties of which the local agents say that the influence of supra-local party levels is small, compared to representatives of other political families.

We analyse the presence of a cost of ruling and/or incumbency advantage in the municipal elections in Flanders of 1988, 1994 and 2000. The empirical work concentrates on vote share evolutions of the three main parties: CD&V, Sp.a and VLD. We find evidence that being in power leads to an incumbency advantage, especially for those parties who provide the mayor. The results indicate that this electoral advantage is not uniform among parties (the socialist party appears to benefit more) and, importantly, the incumbency advantage seems to dampen over successive elections. From the perspective of the local politicians, it is important to keep in mind that this incumbency advantage at the local level was insufficient to compensate for the general trend of the major parties to lose votes at all levels of government. Indeed, even in the 1988 elections – where a clear local incumbency (mayor) advantage could be identified for each of the three parties – it was the case that government parties on average lost votes.

The number of preference votes for the candidates running in the October 2006 local elections in the thirteen main cities of Flanders is largely determined by the position on the list and the previous political mandate. A multivariate analysis shows that an executive function on the local level yields a comparable electoral bonus as a national mandate. The campaign expenditures also have a significant effect. There is a spending limit, but the candidates on average spend only 22% of what they are allowed to. Christian-democratic candidates generally spend the most, with the liberals ranked second. The gender, age and professional status of the candidates have at most a very marginal effect on their electoral score, controlling for the other relevant variables. Candidates with a foreign name obtain a somewhat better result on average, but this is particularly the case with candidates running for the socialist party.

In discussing the results of the Belgian local elections of 8 October 2006 the media devoted most attention to Antwerp. Not only because it is Flanders biggest city, but especially because of the remarkable result of mayor Patrick Janssens and his socialist party (Sp.a-Spirit). They won over 35% of the votes and became the biggest party of the city, a position that was taken by the extreme right party for more than ten years. In this article we tried to map and explain the so called ‘Janssens-effect’. By analysing the election results and the data of an (inter-university) Internet panel we managed to prove that Patrick Janssens had a large share in the victory of his party. A lot of citizens of Antwerp that normally would not vote for the socialist party supported Janssens. To explain this personal success we suggested four plausible motives: the perceived qualities of Janssens as a mayor; his presidential campaign; the attention of and performance in the media; and finally the statement against extreme right. Our panel data suggest that a combination of these motives can explain the Janssens-effect. The appreciation for Janssens as a very capable mayor was present among the electorate long before the campaign started. This appreciation became the basis for his electoral success. The fact that many voters decided to switch to Janssens in the months and weeks before the election is related to his campaign and the media-attention it generated. The media strengthened the image of Antwerp as a polarised city. Especially the voters that changed their vote intention during the campaign were convinced that supporting Janssens was the most explicit signal against the extreme right image of their city.

Together with the city council elections, the citizens of Antwerp elected on 8 October 2006 for the second time their district councils. This new decentralised political level is primarily initiated to restore the confidence of the citizens in the city (and district) government(s). By analysing the results of the city and the district elections we try to find indications whether citizens feel closer to their new district governments or not. Firstly district elections resulted definitely not in less blank votes. Secondly, the number of list votes is higher on the district elections than on the city elections, while we would have expected a higher number of preferential votes. Thirdly, we see that the differences between the electoral results of the city elections and the district elections are becoming more pronounced. Although this last result seems to support the legitimacy of the decentralised district they merely reflect changes in the logic of the city elections. Mainly as a result of media coverage the city elections were direct elections of the mayor. Therefore voters used the district elections to vote for their preferred political party. This was not always possible at city level, because some parties did not have an eligible candidate for mayor. Generally spoken, we can conclude that the district elections do not give much proof of a closer connection between the citizens and the city government.

Official results of local elections are especially interesting when collected at the lowest level possible, to enable analyses of voting behaviour on the smallest social unit. In Antwerp, together with the city-level results, the voting results for the city districts are public too. Yet, still no possibility exists to review the official results at the neighbourhood level. This lack of data makes it difficult to relate specific neighbourhood characteristics to voting behaviour in those neighbourhoods and to compare them. Thus, to collect the data on neighbourhood voting behaviour, we organised an exit-poll in 18 neighbourhoods of seven districts in Antwerp, during the recent county, local and district elections. This article briefly sketches the exit-poll and its organisation and presents the main results. These results will be compared with the formal results on district level, and the neighbourhoods will be compared to one another. The results are – as expected – slightly biased, although in some districts more than in others. Voting behaviour in neighbourhoods turns out to vary strongly. We try to explain the voting behaviour with the help of several neighbourhood characteristics, like the proportion of immigrants and unemployed. The analysis confirms some important correlations, but refutes others.

Although the 2006 local elections can hardly be described as ‘historical’, there is sufficient evidence to distinguish remarkable characteristics associated with this elections. For the first time in decades, turnout has been growing. This evolution can be explained by several factors. This article emphasizes besides the impact of changes in the electoral rules, transformations in demographic structure of the population and the stake of the elections the importance of the media campaign surrounding the elections. However, in spite of this (national) campaigns, there are more than enough indications that local politics keeps its local ‘nature’. Secondly, the 2006 elections were the first ones organised after the transfer of the responsibility for municipality legislation from the federal state to the regions. This means that each region designed its own local government architecture and electoral rules. Yet, in practice, the consequences of this transformations seem to be very limited. Thirdly, and particular in the Flemish region, ‘strong mayors’ arose from the ballot stations (with the Antwerp mayor as the most spectacular case). The consequences of this trend will in the future be the issue of a new debate concerning the relations between council, board of alderman and mayor.