Citations

... In a survey from 2004 97.3 percent of the local politicians who participated in the investigation reported that they rarely or never vote against the party line without notifying the party about it in advance, while 83.1 percent report that they rarely or never vote differently than the rest of their follow party members in cases where they notify the party about the divergence beforehand (Blom-Hansen, Serritzlew & Skjaeveland 2004: 14). Due to a low barrier of representation (Elklit 2005), it is relatively easy even for small parties to win seats on the city councils, and the vast majority of parties with seats on city councils are also represented on the national level. In many municipalities, however, local branches of national parties are supplemented by local party lists that only run in single municipalities. ...
Article
Full-text available
The institution of committees in democratic legislatures has for years been said to bias policy making because the preferences of committee members differ from, and are more alike than, those of other legislators due to self-selection to the committees. Based on an analysis of United States Senate committees, Hall and Grofman suggested in 1990 that the preferences of committee members primarily diverge from, and are more alike than, those of non-committee members on policy issues that are salient to constituencies or at least to an easily definable segment of constituencies. This article argues that the logic of Hall and Grofman should in fact be reversed in legislatures characterised by highly cohesive parties. Accordingly, the main hypothesis is that in such legislatures the preferences of committee members are more likely to be alike than those of non-committee members in the committees that work with policy issues of less salience to constituencies. Using a large-scale comparative design comprising data collected in 2008 on the spending preferences of 1,348 Danish local politicians, evidence is found supporting this hypothesis. This finding points to the importance of considering the role of parties when assessing committee bias.
... c Local election statistics for Sweden and England do not distinguish between minor national political parties and local (non-partisan) lists; therefore, the total count is used as an indicator of the maximum size of this vote. d Numbers given are averages of the effective number of parties in the electorate and in parliament(Taagepera & Shugart 1989;Elklit 2003b; Särlvik 2002, 237; Norris 2004, 87). e It is not possible to calculate this figure, since in the British case local election districts are not always subsets of a general election constituency. ...
Article
The first ever simultaneous general and local elections in Denmark (November 2001) allow for a comparison of Danish voters’ inclination towards inter-level ticket splitting with similar phenomena in Sweden and England. Inter-level split-ticket voting occurs when voters cast their vote on two different parties in the two different (but simultaneous) elections; this happened far more often in Denmark in 2001 than in the two other countries. One hypothesis suggests that this owes to party system differences between the three countries, since both the number of parties running in the different elections and the discrepancy between the national and the local party systems are expected to influence the level of inter-level vote splitting. However, elec-tion statistics and survey data based analyses (Denmark in 2001, Sweden in 2002, and England in 2001) give only limited support to the hypothesis. It appears that Danish voters did in fact split their 2001 national and local votes more than Swedish and English voters did and more than party system differences can account for.
... 19 Five-to-seven po-17 The UUP did not call a time-out to consider its eighth-round choice, but it did not need to do so since it would have had time to consider its options when the DUP called a time-out after round six. 18 The entities seeking office under the d'Hondt list PR system used for Danish local elections may be either national political parties presenting local lists or local groups of concerned citizens (Elklit 1997(Elklit , 2002b. 19 In addition to the defined executive body positions and the mayor, the city councils in Aalborg, Aarhus, Copenhagen, and Odense elect two deputy mayors. ...
Article
Full-text available
Some proportional representation (PR) rules can also be used to specify the sequence in which each party in a parliament or each member in a multiparty governing coalition is given its choice about (unique) desired resources, e.g., "indivisible goods" such as cabinet ministries or executive positions, thus providing an algorithmic method for determining "fair" allocations. Divisor rule sequencing using the d'Hondt method was recently used to determine the ten cabinet positions in the Northern Ireland Executive Committee created under the 1998 Belfast ("Good Friday") Agreement; and such sequential allocation procedures have been used in some Danish municipal governments, and for determination of committee chairs in the European parliament. Here we examine in some detail the procedures used in Northern Ireland and Denmark, with a focus on special features such as the option in Denmark to form post-election alliances.
... 19 Five-to-seven po- 17 The UUP did not call a time-out to consider its eighth-round choice, but it did not need to do so since it would have had time to consider its options when the DUP called a time-out after round six. 18 The entities seeking office under the d'Hondt list PR system used for Danish local elections may be either national political parties presenting local lists or local groups of concerned citizens (Elklit 1997Elklit , 2002b). 19 In addition to the defined executive body positions and the mayor, the city councils in Aalborg, Aarhus, Copenhagen, and Odense elect two deputy mayors. ...
Article
Some proportional representation (PR) rules can also be used to specify the sequence in which each party in a parliament or each member in a multiparty governing coalition is given its choice about (unique) desired resources, e.g., “indivisible goods” such as cabinet ministries or executive positions, thus providing an algorithmic method for determining “fair” allocations. Divisor rule sequencing using the d'Hondt method was recently used to determine the ten cabinet positions in the Northern Ireland Executive Committee created under the 1998 Belfast (“Good Friday”) Agreement; and such sequential allocation procedures have been used in some Danish municipal governments, and for determination of committee chairs in the European parliament. Here we examine in some detail the procedures used in Northern Ireland and Denmark, with a focus on special features such as the option in Denmark to form post-election alliances.
... The immediate consequence is that it is not particularly important for the overall proportionality of the outcome of the election if we use one or another of the various PR formulas available (on the same vote distribution, of course), and especially if the number of seats to allocate by PR is not too small. Examples of such calculations are plentiful in the scholarly literature (see, e.g., Nohlen, 1990: 85, 193-194; Elklit, 2002: 10). ...
... If this argument is taken to its logical conclusion, then we will end with the MMP model – used e.g. in Lesotho and Germany – where some seats are allocated in single-member constituencies (sometimes 50 per cent of the total number of seats), while the rest are used as compensatory seats, to allocate to those parties who have won less than their proportional share of the seats in the constituencies. This system has often inspired reflection and thinking in this country (see, e.g., Faure 1999), but a main problem is that it requires that a fairly considerable share of the seats are compensatory seats, if full proportionality is still to be achieved (cf., e.g., the outcome in the May elections in Lesotho (Elklit, 2002, forthcoming)) – and then there is only a correspondingly smaller number of seats to be used as single-member constituency seats. The unavoidable consequence is that the constituencies become too big to provide the close linkage between voters and representatives, which most of us probably consider the main raison d'être for having single-member constituencies, and which would most certainly also be the problem here in South Africa. ...
... The next step is then to include another country, and Denmark is an obvious choice – not because I'm familiar with that system (Elklit, 1999; 2002), but because we find some illustrative differences from the South African system, which demonstrates how legislators can be attracted by different solutions when they design electoral systems. ...
... Den enkle varianten er å telle antallet partier som konkurrere om stemmene (valgpartier). Problemet med dette er at vi får ett mål som ikke forteller oss noe om partienes relative størrelse, og da heller ikke om partikonkurransen i lokalpolitikken (Elklit 1998). Teller vi alle partier som «en», ender vi naturlig nok opp med at flerpartisystemet dominerer i samtlige kommuner. ...
Article
This article examines the effect of Mogens Lykketoft’s and Anders Fogh Rasmussen’s tours of the electoral campaign in 2005. The purpose of the tours was to use the local media in a more easy and unedited way to carry through the politics of the parties and at the same time to expose the two prime minister candidates. The conclusion of the article is that the behaviour of the local media does not correspond to the expectations, which the parties had prior to the tours. The media’s coverage of the tours is less than expected, more critical and with a less distinctive focus on politics than on process and persons. This means that the central party organisations ought to look for other ways to influence the local media – than by sending top politicians on tour.
Article
Full-text available
In this article we look closer at the overall question of how more than a quarter-century's EU membership has affected national decision-making processes. The specific questions to be addressed concern membership's effects on delegation and accountability relationships between voters, MPs, the cabinet and the civil service. We use the general approach selected for the country contributions to the present volume. For each principal-agent link (voters-MPs, MPs-cabinet, cabinet-ministers, minister-civil servants) in an ideal type chain of delegation and accountability, we are interested in the ex ante and ex post instruments of control available to Danish principals, and in how these instruments and their use have changed with EU membership. Thus, while the main focus is on accountability as a process of representation, we also refer to accountability as a policy outcome of representation.
Article
When collective choices are made in more than one round and with different groups of decision-makers, so-called election inversions may take place, where each group have different majority outcomes. We identify two versions of such compound majority paradoxes specifically, but not exclusively, relevant for systems of proportional representation with governing coalitions: The 'Threshold Paradox' and the 'Federal Paradox'. The empirical relevance of the two paradoxes is illustrated with examples from three Danish elections (1971, 1990, 2011), where a majority of the voters voted for one bloc of parties but where a majority of the seats fell to another.
Article
When public budgets have been drafted and approved, a new and equally important game begins. Budgets are routinely broken when they are implemented, causing higher public spending than politicians originally intended. Institutional, political, and economic causes of budget overruns are examined in an empirical analysis of more than 17,000 cases. While political factors have only marginal effect, it turns out that tractability explains budget overruns well; if users and producers are strong, budget overruns tend to be large. In a reform perspective, the effect of institutional rules is especially interesting. No quick fixes to problems with budget overruns exist, but rules prohibiting supplementary appropriations for non-mandatory spending can reduce overruns, provided that politicians are willing and able to credibly commit themselves not to waive them later.