Michael Gallagher and Peter Marsh: Candidate Selection in Comparative Perspective. The Secret Garden of Politics, Beverly Hills and London: SAGE Publications, 1988 (294 pp.).

Henry Valen, the only Scandinavian contributor to this book, makes - to me at least - an important reflection on how political scientists tend to perceive the process of candidate selection in political parties. According to Valen, political scientists are strongly influenced by Robert Michels' 'Iron Law of Oligarchy' and therefore tend - without any hesitation - to apply the law of oligarchy to all party activities including candidate selection. although inspiring and fascinating to read, the generalizations forward in Michels' seminal work are based mainly on research by the German Social Democratic Party at the turn of this century. Therefore, all systematic party studies that explicitly or implicitly challenge the old theory construction, including the subject of my current review, are welcome.

Gallagher and Marsh very pertinently subtitle their book 'The Secret Garden of Polities', revealing to the reader the wide gap in scientific knowledge of the candidate-selection process as well as the secrecy surrounding selection process before the final candidate lists are presented to the voters in the electoral campaigns. The editors have, at least to some extent, opened the lock to secrecy, by first convening a workshop (ECPR) on the theme and then giving permanence to their aims by publishing the contributions from the workshop. However, the secrecy of candidate selection should not be overestimated; parties are much more open for scientific studies today than they were yesterday. Moreover, scientific research has been undertaken, at least to some extent, during the period when parties have contested general elections. In the first section of the book, Michael Gallagher introduces the problem by examining theoretically the most relevant studies written about candidate selection. From his overview we learn: that the research of candidate selection is not supported by an all-embracing theory on how candidates are recruited; what prerogatives valued by the nominees; sources of intra-party conflicts in the process and how they are resolved; and finally what consequences the candidate selection has on campaigning, party cohesion, and the political standpoints of the office holders. Generally, however, theories are developed what impact candidates have on organizational strength and activity in parties.

For practical purposes, the common theme of the book is confined to a few core issues, since all the nine countries included are evaluated separately and presented by different authors. This design is perhaps the most common way of publishing books in political science of today. True, this method may have many advantages, but without a distinct comparative approach

the individual sections will merely comprise a series of system specific studies. However, in the volume of Gallagher and Marsh the authors have succeeded in raising the system-specific discussion dominating the book to a comparative and more general level.

From the concluding section some notable findings should be emphasized. First of all, candidate selection in Western Europe is exclusive to the party organizations and their members. By doing so non-members are excluded, and thereby the vast majority of eligible people are excluded from the process of nomination. By contrast, within the party organizations party leaders or the national executive have little if any power in the candidate selection process. Instead, the power is mainly decentralized to members at the constituency level in the party organization. Though the methods of selection differ, rank-and-file members are given the initiative to nominate and select candidates without supervision from party leaders.

The significance of this volume is related to the finding that membership power clearly outbalances that of the party leadership in the selection of candidates. However, it is important to note that there are also other sources of influence in the selection process. In some democracies, electoral laws set the broad limits to parties on how the process should be designed. More important, though, is the question to what extent the structure of electoral systems have an impact on the selection process. The selection process is said to be organizationally a simpler affair in single member constituencies than in multi-member constituencies under proportional representation. More importantly, we should distinguish between candidates in systems with preferential lists and those who are selected in systems with non-preferential lists. In the former case, party members are given the power to rank the candidates according to the selectors' preferential system which the voters are obliged to accept when casting their ballots. However, in the latter case, party members select the candidates the right to rank them. Clearly, by casting the ballot voters rank their candidate according to a preference system that might considerably differ from the preference system prevailing in party organizations. indication of different preference systems between voters and party members is the tendency for M.P.s to be re-elected more often in systems with preferential candidate lists. Should this tendency be significant, incumbency is a less continuous commitment for those elected in systems with non-preferential lists. The impact of these two types of lists on governing cannot be evaluated here. However, it seems plausible that the effects are present when a considerable part of the M.P.s in one parliament are constantly unskilled and when the remainder never can be sure of winning a prolonged mandate, compared to another parliament where the M.P.s are skilled party people and a majority of the office holders are guaranteed a future in politics.

Moreover, in democracies where the process of nomination is not regulated law the difference between party practices may be considerable. It is a well-known fact from the writings of Maurice Duverger that party cultures are systematically different. Although parties no longer - if ever - have strictly fitted into the conceptualization of mass, cadre and devotee parties, cultural differences may still be observable. Evidently, party culture is a very broad concept that must be problematized. Nevertheless, the organizational heritage must necessarily be different in social democratic parties compared to liberal or conservative parties. Clearly, in selecting candidates social democratic parties generally must be very reluctant to include candidates launched by the closely linked labour union. Often it is the union that is more powerful than the party in inter-organizational conflicts. Also, candidates launched from other auxiliary organizations are usually included, thereby candidates representing different sectors of the labour movement are included on the party lists. In liberal and conservative parties a similar organizational link to a specific trade union is unknown. Therefore, unions cannot claim to have their members nominated on party lists. Parties, on the other hand, have the option to include union people as well as people from other organizations on their lists suggesting that they fulfil the formal criteria of party membership. Not surprisingly, these parties often seek to nominate members who are mutually organized in associations with a large membership in order to maximize votes, but the process of selection is exclusively an intra-party decision.

What consequences the process of candidate selection may have on parliament policy is not evaluated in the book. Therefore, we can only make some tentative suggestions. Firstly, it seems more or less apparent that it is the selectors and not the voters who primarily determine the composition of parliament. The American system with open primaries gives the voters extreme influence in choosing candidates, whereas preferential systems at the other extreme (such as Norway and Sweden), give party members almost exclusively the role to choose. Electors are only given the option to accept the list of candidates by casting a ballot or alternatively to abstain from voting. Indirectly, by voting for a competing party the voter may influence the selection of candidates to the next election. However, this might not materialize since it is always the party members and their leaders who decide whether it was wrong candidates or something else that caused the electoral failure. To be sure, it seems evident that in systems where party members make the selection, candidates are more in concert with the party doctrine, compared to systems where the selection is open to voters. Comparing these two extremes might produce empirical evidence for the suggestion that the process of selection does matter. Further research on the process of nomination could also add more knowledge about party discipline, cohesion, and decision making in parliament.

Finally, a good book should raise more questions than it answers. This could be said about this volume. Firstly, the process of nomination and its consequences (to parties and parliaments) must be more systematically studied before we can make any more definite statements about which variables in the process of selection cause effects, to what extent, and with what consequences. Secondly, if our main interest is in studying what consequences the process of selection have on parties and their members, our information remains incomplete unless we include elections at the local level. In Scandinavia not only municipal but also community council elections should be included. By widening the theoretical population from parliamentary elections to elections at the local level, the process of selection its consequences becomes considerably different. However, my intention is not to rewrite an already skilfully written volume. Rather, I want to make a plea for more systematic Scandinavian studies on the topic that has been my interest for a couple of years.