Bibliographic Essay on Party Politics in Sweden, 1950-1962
The search for material on Swedish political parties proved to be more frustrating than initially expected. Although numerous citations to Swedish politics were located without difficulty, few treated political parties per se, describing and analyzing their structures and activities. Where political parties were specifically involved in the discussion, concern was directed to the interactions among parties over issues, elections, and formation of the government. Although our file contains thirty-five documents, its size is inflated by the inclusion of several brief items from American Scandinavian Review and The Christian Science Monitor. More than half the total number of pages came from only six documents as noted in the bibliography listing.
The table of indexing codes applied to our 575 pages on Swedish party politics reveals the coverage of the literature according to our coding categories. The unusually heavy reliance on code 300 (party supporters) over all the other codes indicates the literature's preoccupation with electoral studies. Unfortunately, much of this information about sources of party support went unused in coding the relevant "attraction", "concentration" and "reeflection" variables in the ICPP conceptual framework, for it proved less suitable for our purposes than obtaining cross-tabulations of raw sample survey data according to our own specifications.
The second most frequent code (530), pertained to "issue orientations" of the parties in Sweden, and the literature on this topic was both ample and exceedingly helpful. We were able to code the parties on most of the thirteen issues in the framework, although we had better luck with the Social Democrats and Center Party than with the Liberals and Conservatives. Following 530, the most frequent codes are 720 (social conditions), 830 (interparty competition), and 850 (origin of party system). All of these categories, and indeed all the other codes used 35 times or more, pertain to "systemic" features of Swedish politics rather than political parties as units of analysis. Consequently, they were only of marginal utility in coding the parties on the variables within our interest.
The limitations of our file for coding Swedish parties can be seen more dramatically in the frequencies of usage of the major coding categories. The two sets of codes used most frequently to index the literature are 6-- (political environment) and 8-- (party system). In the ideal case, these codes only furnish contextual information to aid coding, but they dominate the Swedish file. The lowest usage codes, on the other hand, are 4-- (party organization), 1-- (party origins), and 2-- (party activities). These categories pertain specifically to political parties, and they were seldom evoked by the literature. The "party organization" codes, for example, were used only 66 times. Because two or more codes could be assigned to the same page, fewer than 66 pages in our file discuss aspects of party organization. As a result, our coding of variables in the "degree of organization" and "centralization of power" clusters was done with low degrees of confidence. Clearly, the literature in our file is woefully weak on the organizational characteristics of parties.
Our experience in the ICPP Project has found a relationship between the scholarly quality of the literature on party politics and the amount of attention given to party organization. A scholarly literature usually implies thorough treatment of party organization. The file on Sweden, however, stands in stark exception to the rule. Although the literature is relatively low in attention toward party organization, it stands high in scholarly treatment. Considering the heavy usage of codes indexing the "study of parties" (0--), the Swedish literature must be characterized as highly analytical. Some 45 pages, for instance, discuss the "methodology of studying parties" (070), and another 28 pages are indexed for "general theory about parties" (040). These codes and the others in the category reflect a literature distinguished by rigorous research. The rigorous research, however, is only obliquely related to political parties. An examination of the 11 pages indexed for "explicit propositions about parties" (030) reveals that six deal with the consequences of proportional representation (insuring the existence of more than two parties, producing a weak executive, and leading to compromise politics), three deal with other aspects of the electoral system as practiced in Sweden (those which discourage the growth of minor parties), and two pertain to explanations of voting behavior.
Another guide to the quality of the literature is provided by the table of data quality codes, which have been assigned to only 31 of the 35 documents--four items from The Christian Science Monitor being added after this analysis was completed. Even allowing for the exclusion of the Monitor articles, we find that most of the documents are journal articles, sections in books, or books themselves, and most of the authors are academics and of Swedish nationality. About half of the documents involve some quantitative data and almost a third feature some theoretical treatment.
These characteristics of the quality of the literature, and others reported in the table but not mentioned here, support our overall evaluation of the literature as scholarly and analytical. It is unfortunate from the interests and needs of the parties project that the literature is not more directly oriented to political parties as units of analysis. Perhaps the Swedish literature, which we were unable to manage, does contain key studies of the structure and activities of the individual parties. Perhaps the English literature published since 1969, when this file was completed, now contains such studies. But at the time we conducted our bibliographic search, we were unable to locate any.