Assuming that the political information provided to the voters directly determines the communication efficiency of the political system in a democratic framework, this paper therefore proposes and tests some hypotheses explaining the politician's behaviour in that respect. Those hypotheses are tested on statements made by major Belgian parties' leaders at the eve of the 1978 and 1981 national elections. Having first related the degree of reserve on the politician's statement with his party's share in the constituency, a positive relationship is found. The second hypothesis states that the degree of reserve in the politician's statement is explained by his reaction either offensive or defensive triggered by the variation in his party's share given the outcome of the precedent election. It is observed that a defensive attitude is adopted by parties having known a reduction of their market share. The third question dealt with examines the infiuence of the size of the party on the abovementioned relationship between the variation of the market share and the degree of reservedness. The results show that apparently the elimination of the size effect does not affect the findings concerning the second hypothesis.

This paper describes the common techniques used in the constitution of candidates-lists for parliamentary elections. A common feature of these techniques is the consultation of party members. But the way in which this is done differs among the parties. AGALEV, the ecologist party, offers every member the possibility to have his say about every candidate. The socialist party (SP) uses this system in two constituencies; in the other constituencies a special congress decides. The christiandemocrats (CVP), the liberals (PVV) and the Flemish nationalists(VOLKSUNIE) constitute their lists by indirect vote: the leadership performs a first selection and the members then either approve or disapprove of this choice. Our research indicates that there are no essential differences between these two techniques from the point of view of membership participation. In both the ecologist and the socialist party the majority of members does not participate in the candidates-elections. Therefore, as with the indirect vote, candidates-lists are constituted by an active, militant and elitist group of members. Whatever the technique used the infl,uence of the rank-and-file-member is small.

The main characteristics of the elections of 13th October 1985 seem to be the following. The turnout, as appears from the number of laid down ballot papers in relation to the number of registered voters, is slightly declining compared with 1981. It indeed amounted to 93.59 % against 94.56 % four years before. This rate of participation averages those of previous elections. The number of blank and spoilt ballot papers is rising very slightly. It totalled 7.45 % for the House of Representatives, against 7.37 % in 1981. This is again an average percentage compared with the previous general elections. It is considerably lower than the number of blank and spoilt ballot papers at the European elections of both 1979 and 1984. The extent of the shifts in votes is much lower than in 1981 and also averages those of the last year's elections. The votes are moving in a very similar way in the three regions of the country, although with one notable exception: the loss suffered by the PVV in Flanders. Apart from that, the Christian Democrats are registering a gain in the northern and southern parts of the country as well as in Brussels as far as the CVP is concerned; the extent of the Socialist advance is the same in the three regions; the loss suffered by the nationalist parties is general and the «green» lists are reinforcing their breakthrough everywhere. The parties of the outgoing coalition are improving their result, which very seldom happens at Belgian elections: Christian Democrats and Liberals are indeed progressing by 2.26 %. Another noteworthy element, however, is the fact that the ahead movement of the main oppositionparties is even more important than the majority's as the Socialists obtained 3.25 % more of the valid votes. Of course this double success took place at the expense of the nationalist parties, the Communists and some smaller lists. The shift in seats does not exactly reflect the shift in votes, because of a different distribution of the seats between the provinces, on the one hand, due to the variations in the number of inhabitants, and because of the concentration on certain lists of the allocation of the last seat, on the other.

At the general elections in Belgium, the voter bas the possibility to bring out a vote for a party or a vote for a candidate of the party (a preferential vote). At the general elections of October 13, 1985, for the House of Representatives, the voters have voted for 48 % by preference, whereas, for the Senate 38 % of them have done so. The use of preferential votes is varying from one electoral district to another: from 66 % in the district of Oudenaarde to 34 % in Mechelen. There is also a difference from one political party to another. The highest percentage is obtained by the christian-democratic parties: 58 %; the lowest by the ecologists: 22 %. The preferential votes individual candidates obtain differ widely, pointing out not only their popularity but also to a certain extend their political power.

This article is based on a quantitative evaluation of the shifts in voting behaviour of the Flemish population, from the national elections in 1981 to those in 1985. Using the variables sex, age and occupational status a further analysis is made of the voting behaviour of the different subgroups. Firstly the article describes the methodology by which the mobility tables were calculated using statistical information from poll surveys and the results of the elections. In the second part the results of the analysis, using log-linear modelling, correspondence analysis and cluster analysis,are presented. The results indicate significant shifts between the different parties. Furthermore they indicate that the sex of the voter has no significant influence on his voting behaviour. Finally, the odds ratios are calculated for the different subgroups with regard to the likehood of a voter moving (changing his voting behaviour) or staying.

During the Belgian legislative elections on 13 october 1985, the Centre Liégeois d'Etudes de l'Opinion (CLEO), University of Liège, polled voters leaving the voting booths. The objective was to determine the characteristics of electors of the various parties according to the usual criteria of identification (sex, age, socio-economic status, education, religion, revenue, previous votes). The motives and wishes of the voters were also recorded (problems influencing the final vote decision, timing of the decision, attitude toward the incumbent government, type of coalition desired for the new government). The poll consisted of a representative sample of one thousand French speaking voters in the province of Liège. Most of the article presents the results obtained by cross-tabulating various identifying and motive variables with the political parties. In this way, we measure and demonstrate the relative importance of all the independent variables on the distribution of votes, except for sex which seems have little influence.

At the provincial council elections of the 13th of October 1985 in Belgium two of the three traditional politie al families made a general progress, namely the socialists and the christian democrats. The liberals had a serious decline in Flanders, but for the greater part this was counterbalanced by the liberal progress in Wallonia. The ecologists also made progress, mostly in the Flemish part of the country. Beside the Flemish liberals, the big losers of these elections were the regional parties, namely the Flemish nationalists and for sure the Walloon Rally who disappeared from the electoral scene.

Our research concerns political knowledge among eighteen-year-old pupils (last year secondary school) in Belgium (sample size: 600) and the variables that may cause a differential level of political knowledge. The results point out that the average level of political knowledge is really low (54 %), though some elementary items (like the prime minister, Mitterrand and Greenpeace) score up to 98 %. The studyreveals as the most significant independent variables: the sex (boys score better than girls), political participation of the parents, mass media, and (in a negative way) school. Mass media are most frequently mentioned as the most important sources of political information. Moreover, respondents with a high level of newsmedia exposure do indeed score best. Political education at school shows to have low impact, on the contrary, respondents that mention school as their most important source of political information tend to score worst.