It would be inaccurate to characterize election campaigns as a generally neglected theme in political science. In fact, a survey of the literature under this heading conveys quite the opposite impression: there is substantial work in the field, and there even exists a special journal exclusively devoted to the topic.¹ Even so, it would be equally exaggerated to suggest that the 'state of the art' is satisfactory from a scholarly point of view.

If we look at the research on campaigns at large, one major imbalance has to do with the emphasis on candidates rather than on parties. The typical campaign study, irrespective of whether it is cast in mainly descriptive or theoretical terms, sees the individual candidate as the main vehicle behind any election campaign (cf. Blydenburgh 1971; Wemaere 1973; Ferejohn & Noll 1978). This is largely a reflection of the exceptionally strong US dominance in the literature on campaigns. In America, election campaigns are about individual candidates rather than about parties; in fact, a major part of political campaigning in the US takes place inside the parties in connection with primary elections. Moreover, even after the party candidates Congress have been selected for each constituency, they differ widely among each other on many central issues depending on local and regional circumstances. This being the case, the party organizations can hardly engage in specific 'party campaigns' that would decisively limit the room to maneuvre of individual candidates. In a recent study, two US scholars have suggested that party campaigning in the US could be characterized an 'oxymoron' rather than a research topic (Katz & Kolodny 1990).

In West European multiparty systems the role of the party organization as the central unit in election campaigns is naturally much more selfevident.Many European parties are highly centralized, and party organizations are often influential both with respect to the selection of candidates and the style and content of their campaigns. Still, there is little in the way of systematic, theoretically founded cross-national research on party campaigns in Western Europe (for a notable exception, see Farrell

& Wortmann 1987). It is difficult to escape the impression that the particularitiesof election campaigns, which permeate the bulk of the literature, have led European scholars to regard election campaigns as a field with limited theoretical interest.

On the other hand, to the extent that there are systematic studies of election campaigns from the point of view of political parties, the campaign as such often receives but secondary attention on the part of the researcher. The campaign is a suitable context since it produces a wealth of empirical material of various kinds: manifestos, debates and spectacular events. Typically, however, the theoretical focus of these studies is on party strategy in general, ideological change or political communication rather than on campaigns per se (Budge et al. 1987; Katz 1971). In other words, the campaign constitutes a mirror or an arena in which theoretically interesting questions can be observed in a systematic manner. These studies certainly tell us a great deal about many central features of party politics. However, they do not necessarily offer similar insights into the intrinsic dynamics of the 'campaign qua campaign' (Farrell & Bowler 1990, 1).

Comprehensive research on party campaigns has most likely also been hindered by the fact that the dynamic character of election campaigns makes it a cumbersome task to study them systematically; needless to say, a comparative cross-national effort would be even more demanding. Party campaigns are carried out through a multitude of channels at all societal levels. Nationwide television, regional newspapers and local billboards are used side by side; national party leaders, professional media consultants and local party activists all play significant roles. Just to tell 'the whole story of a campaign' in purely descriptive terms requires so much time and effort that comparisons over time or across systems are normally not attempted. In fact, perhaps the most typical comprehensive account is the journalistic one (e.g. Tyler 1987) written shortly after an election to meet a current demand in the market. But even when written by scholars, detailed analyses of election campaigns are frequently presented in the form of descriptive accounts rather than case studies cast in comparative or theoretical terms (Knox 1990).

To sum up, while the results of elections are subject to theoretically and methodologically refined, frequently comparative analyses, what precedes them - the campaign - is a field of party studies still in its infancy. Where the studies are abundant, they are about candidates rather than about parties. Where they are cast in terms of party theory, they are not focused primarily on the campaign as a phenomenon. And where the campaign is the primary focus, the studies leave much to be desired in terms of theoretical relevance and comparability across systems and over time.

Nordic Research - A Historical Perspective

Any review of Nordic political research should start with a word of caution: the community is small, and what may look like a 'major trend' may in fact be the work of merely two or three individuals. Bearing this caveat in mind, we may nevertheless attempt a characterization of Nordic research on election campaigns. In summary, there is a smattering of relatively early directly campaign-orientated research followed by a larger body of literature campaigns are an empirical arena for research rather then the main theoretical focus. Finally, we can point to a recent major study indicating that election campaigns can well be made the central focus of more ambitious and systematic research efforts in Scandinavia.

Although rather special in its temporal focus, Gunnar Wallin's thesis (1961) on Lower House elections and election campaigns in Sweden, 1866-84, be noted as an early research effort of relevance to the present context. The author presents a detailed account of opinion formation in connection with election campaigns. Among other things, he discusses the role of newspapers and the use of election meetings, manifestos and brochures by candidates and parties. As for postwar Nordic politics, Henry Valen & Daniel Katz (1964) devote two chapters to a systematic and fairly detailed examination of the 1957 Storting election campaign in their book on political parties in Norway. After a brief discussion of the political situation and main issues in connection with the election, the authors embark on a systematic account of the role of local party leaders as campaign organizers. With the aid of survey data, they describe the leaders' perceptions of their tasks, their efforts to come into contact with the voters, their varying roles and functions as well as their evaluation of the various campaign methods used. Moreover, comparable local interviews were carried out with samples of the electorate in order to assess the effects of campaign efforts on the voters. The results turned out to be rather disheartening from the point of view of the parties, since the campaign either directly missed most of the voters or failed to alter their party identification as determined by factors such as occupation and interest group affiliation (ibid., 145). As for Finland, the History of the Finnish Parliament contains a useful descriptive account by Tuttu Tarkiainen (1971) about parliamentary election campaigns in Finland between 1907 and 1963.

From the late 1960s on, the growing scholarly interest in the study of party propaganda on the one hand, and mass communication theory on the other, increased the output of campaign-related research in Scandinavia.Much this research was directly or indirectly inspired by the theoretical work by Gunnar Sjöblom (1968). His model for the study of party propaganda continues to be a major theoretical contribution. In terms of empirical analysis, a landmark study was presented by a team of Swedish

researchers (Isberg et al. 1974). Party propaganda found in connection with parliamentary elections in Sweden in the period 1960-66 was analysed extensively from a party strategic point of view. Special attention was paid to among other things the orientation of parties to various categories of voters, their attempts to reach floating voters and the effect of party propaganda on the voters' choices. A sophisticated technique for content analysis of propaganda material was established in this connection (Isberg et al. 1972). This theoretically and methodologically advanced study produceda of empirical conclusions, many of them in the form of testable hypotheses for further research. It was accompanied by a number of other Swedish studies, notably Kronvall's dissertation (1975) on the 1970 parliamentary elections. Taken together, studies of strategic party propaganda in connection with electoral campaigns represents an impressivechapter Swedish political science. Gronmo's study (1975) on cleavages in Norwegian party propaganda in connection with the 1969 and 1973 Storting election campaigns constitutes an exception to the pattern of Swedish dominance during this period.

Nordic research on mass communication and mass media systems has produced several major studies of direct relevance to the study of election campaigns.₂ Karen Siune's dissertation (1982) is a borderline case between the study of party strategy and mass communication theory. It focuses on the use by political parties of radio and television - which were a government monopoly in Denmark - in connection with the Folketing elections of 1971, 1973 and 1975. At the same time, the interaction between politicians and journalists is given special attention, thus also directing the theoretical focus towards such topics as communicator roles and journalistic attitudes.

More directly linked to mass communication theory was an earlier dissertation by Seppo Sisättö entitled Television and Elections (1977). The author investigated the viewing of election debates and the TV coverage of election results in Finland in 1960-75. His main interest, however, was the role of television in the system of political communication rather than party strategy or political campaigns.

Similarly, Kent Asp's study (1986) on the political campaign in the face of the 1980 Swedish referendum on nuclear power hinged on theories of mass media effects on political opinion formation. He found that the power of mass media over both the audience and the content of the campaign was considerable during the first phase of the opinion formation process. During the 'referendum stage', however, he found that the campaign makers themselves - the parties - increased their influence on the content of the campaign.

In addition, the Norwegian Electoral Research Program should be mentionedin
context. Since 1957, this research has scanned, among other
things, the audience of electoral debates on radio and television as well as

the process of individual opinion formation in connection with parliamentaryelections Bjørklund's contribution to this issue). On the whole these studies, although only secondarily cast in terms of 'campaign research', represent an important part of the relevant Nordic iiterature.

A major event - if not the major event - in Nordic research on electoral campaigns is marked by Peter Esaiasson's dissertation (1990) on Swedish electoral campaigns 1866-1988. This massive volume is an impressive piece of scholarship. Since Esaiasson's contribution to the present issue draws heavily upon it, there is no need to dwell on its contents in any great detail. Suffice to say that it proves that systematic analyses of 'campaigns qua campaigns' are possible over time; Esaiasson's book covers every single election (n = 42!) during the period studied. It systematically records a number of relevant aspects of these campaigns, suggests a periodization of Swedish electoral history, and finally proposes a theoretical model aimed at explaining major shifts. In many ways, Esaiasson has broken new ground in both Nordic and international research on election campaigns.

About this Issue

The ECPR Joint Sessions of Workshops in Bochum in 1990 included a workshop on election campaigning. The impetus behind it reflected many of the critical points mentioned above; the ad hoc character of much of earlier research, the lack of systematic and comparative studies focused on campaigns as such, etc. (Farrell & Bowler 1990). With the exception of Bjørklund's article, versions of the contributions to this issue were presented the workshop.³

Lars Bille begins by noting perhaps the most fundamental feature of election campaigning in Denmark - i.e. their frequent occurrence. Denmark had 22 election campaigns (national and local) since 1968! Moreover, are frequently faced with a short preparation period; a minimum of merely three weeks has to elapse between the call for new elections by the Prime Minister and the election day. The Danish party system is a highly fragmented one, including up to a dozen parties in the Folketing plus a number of minuscule parties that seldom or never gain parliamentary seats. As Danmarks Radio - the Government Broadcasting Company - gives all contesting parties equal time, this means that it is very difficult to organize meaningful TV debates between party leaders. Still, the campaign period is an increasingly important time of choice for the voters, and television is the most important source of campaign information.

Based on interviews with party secretaries, Bille's empirical analysis
focuses on the 1988 parliamentary election campaign. Two major parties
(the Conservatives and the Social Democrats) and the small Christian

People's Party are included. The findings indicate that a fairly traditional form of campaigning still prevails in Denmark. Newspaper advertisements continue to account for the main part of campaign expenditures, although sizeable sums are spent on the production of the 'party presentations' broadcast by television. None of the parties conducted pre-election surveys. The Conservatives and the Christian Party still lacked adequate fax and computer facilities. Compared with the Social Democrats, they also had weaker manpower and fewer organizational resources at hand.

The frequent occurrence and unexpected timing of elections is the major reason behind the relatively low degree of campaign 'professionalization' in Denmark. It should also be borne in mind that the Danish parties have relatively limited financial resources at their disposal. In a small country, parties are never likely to engage in the lavish use of consultants and advanced technology typical of US campaigns.

In a second Danish study, Jörgen Elklit looks at an area which is rarely highlighted in campaign studies - local elections. In contrast to parliamentary elections, local elections in Denmark have fixed dates, thus making for a degree of 'predictability' absent at the national level. Moreover, local election campaigns cannot utilize national mass media in the same fashion as national campaigns. Still the author finds that local elections are strongly affected by national politics and the national image of the parties. In empirical terms, Elklit presents results from interviews with local party leaders and campaign officers in connection with the 1989 local election campaign. Five parties and seven municipalities in the Aarhus region are included. He finds, somewhat surprisingly, that direct communication voters through door-to-door canvassing is rare in Danish local campaigns. He notes, moreover, that the character of local politics sets local elections apart from national ones. Most parties hold at least a few local offices, so almost no one is entirely out of power. He concludes that local elections matter for local politics, and local campaigning matters for local elections. Hence, local party branches have a function independent of their national party organizations, and they are likely to survive despite the varying fates of parties at the national level.

The remaining three contributions to this issue have a longitudinal focus. The article by Karvonen & Rappe marks a deviation from the 'campaign qua campaign' orientation of the rest of the issue. Their article follows much of the Nordic campaign-related research in that campaigns studied are one aspect rather than the primary object of investigation. Surveying party newspapers in connection with four parliamentary campaigns in Finland between 1954 and 1987, the authors test the thesis concerning the ideological convergence of political parties. Finland is a particularly interesting case since the link between party and class has traditionally been strong. Yet the country has undergone one of the most dramatic

structural transformations in postwar European history. Findings from this study indicate that the Finnish parties have reacted to the leveling-off of class distinctions by shifting their focus from class to 'nobody in particular'. At the same time, offensive propaganda techniques have all but disappeared from campaign language. The Finnish case indeed supports the gist of the 'end of ideology' thesis. The reaction of the electorate seems to be a continuing decrease in voting activity.

Peter Esaiasson presents a systematic account of the use of direct (campaign door-to-door canvassing, workplace activity) as well as indirect (pamphlets, posters and mass media) campaign channels by Swedish from 1866 to 1988. He seeks to explain changes in the use of these campaign methods with reference to six contextual factors: party resources, the media situation, changes in the electorate, constitutional regulations, the content of politics and general features of societal development. finds six characteristic periods in Swedish electioneering, ranging from the nationwide diffusion of politics around the turn of the century to the decisive intensification of party competition from the late 1960s on. Other periods are marked, by among other things, the breakthrough radio and television as dominant media, followed by the increased importance of the corps of independent mass media journalists as an actor in the election game. Overall, Esaiasson's account conveys a strong impression of the development of the mass media as a driving force in the transformation of election campaigns over the years.

Much in the same vein, Bjørklund demonstrates that Norwegian electioneering evolved hand-in-hand with changes in the position of the Norwegian mass media. By and large, Norwegian development seems to corroborate the notion about 'the decline of parties'. As to electoral campaigns, mass media have replaced the party organizations as the main vehicle in the campaign process. As in most other countries, the advent of television entailed a major transformation. Of equal importance, however, was the fairly dramatic change in the policy of NRK (the Government Broadcasting Company) some 20 years ago. From that point on. the role of independent journalists as agenda-setters has been of crucial importance. An example of the earlier more restricted position of mass media is that until 1969 they were prohibited from publishing the results of opinion polls which they frequently conducted before elections. Thus, there were two kinds of 'secret' opinion surveys in Norway before 1969 - polls carried out by mass media and surveys regularly conducted by the political parties for internal use. After 1969, opinion polls rapidly gained their 'natural" position as a central focus in the pre-election news coverage in Norway.

Bjørklund also presents some results pertaining to the effects of electioneeringon
activity and attitudes of the voters. His data indicate that
television may have helped to erase the major differences in turnout

between center and periphery. Moreover, it is clear that an increasing proportion of the electorate wait until the campaign before they decide on their party preference. The percentage of these iate deciders' is particularly large among the youngest segments of the electorate, i.e. the 'TV generations'.

The contributions to this issue do not originate from a coordinated Nordic research effort, so systematic comparisons are not possible. Some general trends nevertheless seem quite clear. The decline of the party press and the simultaneous rise to prominence of television and independent journalism central features all over Scandinavia. This corresponds to greater electoral volatility throughout the region. Despite this, national pecularities such as constitutional provisions still make a difference. The chief example is surely the difference between Denmark and Norway concerning the date for parliamentary elections. In Denmark, only the Prime Minister knows this for sure; in Norway only regular elections occur since the Storting cannot be dissolved. Consequently, Danish parties apparently find it less rewarding to conduct surveys, whereas in Norway polls have been part and parcel of party activity ever since the end of the Second World War.

By and large, the studies presented in this issue indicate that there is a promising mix of general similarities and unique national characteristics which would seem to provide fertile soil for a rewarding cross-national study. It is to be hoped that this issue can function as a step on the way towards a more comprehensive study of election campaigns in Scandinavia.

---

NOTES

1. Campaigns and Elections.

2. For an early Norwegian study not described here, see Torsvik (1967). Bjørklund's contribution to this issue deals with Torsvik's research at some length.

3. Comments and criticism provided by the workshop participants are gratefully acknowledged.

REFERENCES

Asp, K. 1986. Mäktiga massmedier. Studier i politisk opinionsbildning. Stockholm-, Akademilitteratur.

Blydenburgh, J. 1971. 'A controlled experiment to measure the effects of personal contact
campaigning', Midwest Journal of Political Science 15, 365-381.

Budge, 1., Robertson, D. & Hearl, D., eds., 1987. Ideology, Strategy and Party Change: Spatial
Analyses of Post-War Election Programmes in 19 Democracies. Cambridge: Cambridge
University Press.

Esaiasson, P. 1990. Suenska valkampanjer 1866-1988. Stockholm: Allmanna forlaget.

Farrell, D. & Wortmann, M. 1987. 'Party Strategies in the Electoral Market: Political

Ferejohn, J. & Noll, R., 1978. 'Uncertainty and the Formal Theory of Political Campaigns',
American Political Science Review 72, 492-505.

Grönmo, S. 1975. 'Skillelinjer i partipolitikken 1969-1973', Tidsskrift for Samfunnsforskning
16, 119-153.

Isberg, M., Wettergren, A. & Wittrock, B. 1972. 'A Technique for Structural Content
Analysis of Party Propaganda', Scandinavian Political Studies 7, 82-106.

Isberg, M, Wettergren, A., Wibble, J. & Wittrock, B. 1974. Partierna inför väljarna. Svensk
valpropaganda 1960-1966. Stockholm: Allmänna förlaget.

Katz, E. 1971. 'Platforms and Windows: Broadcasting's Role in Election Campaigns', Journalism
48, 304-314.

Katz, R. & Kolodny, R. 1990. 'Party Campaign Strategy in the United States: Research Topic
or Oxymoron?', paper prepared for the ECPR Joint Sessions, Bochum. Germany. 2-8
April 1990.

Knox, C. 1990. 'Election Campaigning - A Case Study of the 1989 N. Ireland Local
Government Elections', paper prepared for the ECPR Joint Sessions, Bochum, Germany,
2-8 April 1990.

Kronvall, K. 1985. Politisk masskommunikation i ett flerpartisystem. Sverige- en fallstudie.
Lund: Studentlitteratur.

Sisättö, S. 1977. Televisio ja vaalit. Helsinki: Politiikan tutkimuksia 17.

Siune, K. 1982. Valgkampe i TV og radio. Politiske partiers anvendelse af tv og radio, herunder
samspillet med journalisterne. Aarhus: Forlaget Politica,

Sjöblom, G. 1968. 'The Study of Party Scandinavian Political Studies 3, 15-46.

Propaganda",
Tarkiainen, T. 1971. 'Eduskunnan valitseminen 1907-1963", in Suomen kansanedustuslaitoksen
IX. Helsinki: Eduskunnan historiakomitea, 9-465.

Torsvik, P. 1967. 'Valgkampen i fjernsyn og radio'. Tidsskrift for samfunnsforskning 8. 143—
162.

Tyler, R. 1987. Campaign! The Selling of The Prime Minister. London: Grafton Books.

Wemaere, J. 1973. 'Reflexion sur la "communication politique'". Etudes (Paris) 689-699.