The Life Cycle of Party Manifestos: The Austrian Case

Abstract

Election manifestos are one of the most prominent sources of data for the study of party politics and government. Yet the processes of manifesto production, enactment, and public reception are not very well understood. This article attempts to narrow this knowledge gap by conducting a first investigation into the ‘life cycle’ of election manifestos from the drafting stage to their use in the campaign and post-election periods. Specifically, it investigates the Austrian case between 1945 and 2008 (with special emphasis on the 1990s and 2000s), employing a wealth of qualitative and quantitative data. While the research is thus mostly exploratory, it develops systematic expectations about variation between parties according to their ideology, organisation, government status, and characteristics of their electorates across the stages of the manifesto life cycle. Of those factors, organisational characteristics and status as government or opposition parties were found to be relevant.

Introduction

Among the different ways in which political parties communicate their interpretations of the current state of the world and their policy prescriptions to improve on it, electoral manifestos have become ubiquitous in political science analyses. This is largely due to the regularity with which parties dutifully produce these documents and the constant efforts of the Comparative Manifesto Project (CMP) to turn these texts into data (Budge et al.2001; Budge et al.1987; Klingemann et al.2006).1 Yet the dominant role of manifestos is not reflected in clarity about what manifestos actually are, how they emerge, what their precise role is in the campaign for which they are written, and what impact they have on post-election politics. The present article proceeds from the assumption that a process analysis can take us some way towards answering these questions and provides a framework for such an analysis.

Given the centrality of manifesto-related data for party researchers, it is surprising that only scant attention has hitherto been devoted to the intra-party practice of writing manifestos, as well as to their reception. This article aims at addressing this white spot on the map of party research by highlighting multiple routes that researchers might follow when examining what we term the ‘life cycle’ of manifestos, that is, the process constituted by the production, the formal enactment, and the public reception of manifestos. The life cycle comes to a close when the next elections loom and each of the parties begins to elaborate on a new manifesto, perhaps partly recycling the ideas contained in the previous one.

Our contribution is largely exploratory in nature and for obvious reasons it relates to the case we know best. We thus offer a first examination of different stages of the life cycle of manifestos in the Austrian case. In charting mostly unknown territory, we do not follow a theoretical model but trace the life cycle of manifestos along the schematic time line depicted in Figure 1.

The first section is thus concerned with the process of producing the manifesto. Here, we examine the actors who provide the policy content and are charged with shaping the manifesto. Next, we focus on the intra-party players that get to decide on the final document. We also investigate the way in which manifestos are enacted and then presented to the public. Next we shed light on some characteristics of manifestos that we expect to influence their reception. Then we turn to the publicity parties try to give their manifestos and examine how they are perceived by the media. Finally, we discuss the role of manifestos in the post-electoral processes of coalition bargaining and governing. The concluding section returns to the expectations formulated here and summarises our main findings.

Even though our article does not follow a theoretical model of the life cycle, we can formulate some general expectations about how party characteristics should play out over the different stages. More specifically, we expect that party ideology, party organisation, the parties' (expected) roles in government or opposition, and the nature of party target electorates exercise influence on party behaviour in the various stages and the resulting manifesto itself (Table 1).

With regard to ideology, we expect parties of the left to pay tribute to their greater programmatic tradition by taking manifestos more seriously than parties of the right (that traditionally have been more pragmatic and leader-oriented, see Epstein 1967). Paying tribute to the relevance of manifestos means taking time to elaborate them carefully, producing more substantial documents, legitimising them by a party congress vote, giving them a prominent role in the campaign, and taking them as the blueprint for government policy. At the same time, leftist and Green parties have a stronger tradition of intra-party democracy and hence should opt for more inclusionary ways of making and legitimising manifestos (Duverger 1954: 172; Harmel and Janda 1994).

The second feature that helps to distinguish political parties is their organisation. We take into account organisational size and complexity. With regard to size, we simply think that large organisations have more resources, which increases the number and quality of inputs to a manifesto a party can commission. Complexity means differentiation within the organisation (Dooley 2002; Hall 1991: 50–62). Greater complexity then increases the range of inputs a party can draw on in writing the manifesto and may have a similar effect to size. Yet we are more concerned with demands on the party to consult different power centres and coordinate the manifesto with them. Greater complexity means that more such processes need to be carried out, which, in turn, should lead to the drafting processes taking more time. Paying tribute to a greater range of demands and balancing them should tend to increase the length of manifestos. In terms of organisational resources, we can clearly distinguish the Social Democratic Party (SPÖ) and the Austrian People's Party (ÖVP) from the other parties. While the organisations of the two major parties are more complex than those of the smaller ones, the ÖVP – combining a fully-fledged functional with a fully-fledged territorial representation system (Müller and Steininger 1994) – clearly comes on top in terms of complexity.

Our third dimension is government status. Here, we expect both current and anticipated government participation to be relevant. Government parties simply have more resources available. It should allow them to write more substantial manifestos. In turn, a manifesto with more policy detail should have a greater impact on post-electoral behaviour, for instance in coalition negotiations. Government status and expected government status vary between elections. Yet the records of Austrian parties are very different, with the SPÖ and ÖVP spending long stints in government, the Freedom Party (FPÖ) and the Alliance for the Future of Austria (BZÖ) being confined to short periods in office (all ending unhappily), and none of the other parties ever coming close to government participation (Müller 2000).2

Finally, we might expect that a party's target electorate should impact on the manifesto. Such anticipation of voter interest, of course, will mainly affect the manifesto's contents. Yet in this article we refrain from subjecting the manifestos to detailed content analysis. However, we address one more process-related dimension: manifesto readability that we take as an indicator of manifesto sophistication. Clearly, some target electorates require more translation from policy specialist terminology to everyday language than others. Among the Austrian parties, the Greens stand out with regard to the formal education of their adherents, followed by the ÖVP. This would imply that these parties could afford to enrich their manifestos with more policy detail at the cost of readability.

Clearly, as we have just a few independent cases for most of the parties, the results will be over-determined by these variables. We therefore do not consider our exercise a proper test of the theorised factors but rather as a first check of their plausibility.

After having formulated some expectations regarding differences between the parties, it is apt to make several ‘technical’ clarifications. The first concerns the nature of the documents we examine. Throughout the article the term ‘manifesto’ refers to documents issued as policy statements in the run-up to parliamentary elections. At least for the German-speaking areas, it is necessary to distinguish such election manifestos from party programmes (Grundsatzprogramme) and action programmes (Aktionsprogramme, Spezialprogramme).

The most obvious differentiating criteria – besides the purpose for which such documents are written – are the documents' degrees of specificity and comprehensiveness. Party programmes are broad and general statements stipulating the core ideology of a party. Compared with election manifestos they not only refrain from specific policy proposals but are also much more long-lived.

In contrast, action or special programmes usually contain very detailed policy measures and are devoted to formulating quite specific policies. As might be expected, core areas such as economic policy have received considerable attention from parties writing special programmes over the post-war period. Yet Austrian parties have not confined such programmes to the main battlefields of politics. Note, however, that taken together the action programmes do not come close to providing a complete picture of party policies. Nor are they clustered in terms of date of origin. While election manifestos take an intermediate position between party and action programmes in terms of specificity, they are much more comprehensive thematically than action programmes and potentially as extensive in scope as party programmes.

In order to examine the nature and relevance of manifestos in Austria, our analysis draws on several data sources. The most obvious is the manifestos themselves: we draw on Austrian parties' manifestos for parliamentary elections from 1945 to 2008. However, we limit our analysis to those cases where parties were serious contenders, as indicated by parliamentary representation before or after the election (or both).

In addition to the manifestos, we consulted other primary party documents such as party statutes, party congress minutes, party communiqués in the party press, other press releases, and archival data shedding light on intra-party processes. In addition, we rely on 16 interviews with party leaders and leading party employees who served in party posts between the 1970s and the 2000s.3 As for the manifestos' relevance for the campaign, parties' press releases and newspaper reports serve as data sources for the quantitative analyses carried out. The information presented in this article refers mainly to the 1990s and 2000s. More fragmentary evidence for earlier decades suggests that manifesto-making processes were very similar in most respects.

Drafting Manifestos

The otherwise rich literature on the manifestos of European parties has given only scant attention to the process of manifesto writing. One recent theoretical approach models the production of a manifesto as a stochastic process, where the final document is one out of an infinite number of possible manifestations of the author's – essentially unobservable – ‘true’ policy position (Benoit et al.2009). While such a model arguably has its merits in the context of handling error and uncertainty in manifesto-derived estimates of policy positions, it provides little guidance for the researcher seeking to understand the genesis of manifestos in the real world.

TABLE 1
FACTORS EXPECTED TO INFLUENCE MANIFESTO-MAKING PROCESSES AND MANIFESTOS

Factor	Factor range	Impact on	Expected direction	Possible indicators	Placement of Austrian parties
Ideology	Left–right	Relevance attributed to manifesto	Left: + Right: −	– Length of drafting process – Length of document – Formal adoption – Campaign salience – Coalition bargaining relevance	Left: SPÖ, Greens Right: FPÖ, BZÖ, LF, ÖVP
Organizational size	Small–big	Resources available	Small: − Big: +	Number and quality of input providers the party can draw on	Small: FPÖ, BZÖ, Greens, LF Big: SPÖ, ÖVP
Organizational complexity	Low–high	Resources available/ need for consultation and coordination	Resources: Low: − High: + Need for consultation and coordination: Low: − High: +	– Number of actors consulted and coordinated with – Length of drafting process – Length of document	Complexity: Low: BZÖ, LF, FPÖ, Greens Medium: SPÖ High: ÖVP
Current government/opposition status	In/out	Resources available	Government: + Opposition: −	– Length of document – Coalition bargaining relevance	Situational (1990s and 2000s) (rank ordering according to govt. participation, from + to −): (1) ÖVP, (2) SPÖ, (3) FPÖ, BZÖ, (4) other parties
Expected government/opposition status	In/out	Demands on manifesto realism	Government: + Opposition: −	Coalition bargaining relevance	Situational (1990s, 2000s) (rank ordering according to govt. prospects, from + to −): (1) ÖVP, (2) SPÖ, (3) FPÖ, (4) other parties
Party target electorate: education level	Low–high	Demands on sophistication of manifesto	Low: − High: +	Manifesto readability	(Rank order according to target electorate's formal education, from − to +): (1) SPÖ, FPÖ, BZÖ, (2) ÖVP, (3) Greens
Note: BZÖ = Alliance for the Future of Austria, FPÖ = Freedom Party of Austria, LF = Liberal Forum, ÖVP = Austrian People's Party, SPÖ = Social Democratic Party.

FIGURE 1

THE LIFE CYCLE OF PARTY MANIFESTOS

Clearly, there is a consensus that Austrian manifestos emerge from an intra-elite process (Horner 1987: 374; Jenny 2006; see also Kadan and Pelinka 1979: 8–9). Yet ‘party elite’ remains an amorphous description for a set of decision-makers – a kind of ‘black box’. In the remainder of this section, we look into that ‘black box’ from two perspectives: a process perspective and an actor perspective. We turn to a power perspective in the next major section.

The Drafting Process 4

Manifesto drafting is not detached from the parties' political and programmatic work in general but can be seen as a continuous process that intensifies before an election and then focuses on writing a document that needs to meet specific requirements. In most instances, the manifesto's content is derived from other programmatic documents, especially the party programmes, and parliamentary and – in the case of government parties – ministerial documents, or the manifestos produced for earlier elections. Parties differ in terms of the timing of that process: while the ÖVP and the Greens start the drafting process about one year before the (presumed) election, the SPÖ does so about half a year prior to that date. In the case of a snap election, the drafting process is shortened accordingly.

In all parties the process begins with the strategic decision about the main message for the campaign and which issues will be particularly stressed. The manifesto is then arranged around these issues. In each party, a small task force or even a single person gathers input from various sources, coordinates the substantive contributions, and takes on the task of writing the first draft.

Important party officials are then asked to comment on the manifesto draft which is revised along these lines. This process may be repeated several times before a final draft is completed. Clearly, the major parties – the SPÖ and ÖVP – spend more time asking for feedback, whereas the smaller parties tend to adopt a tighter procedure.

At some point, the manifesto draft is considered final by the taskforce working on it and submitted to the party leaders, usually the party executive committee. The members of the party leadership may require further changes in pre-meeting communication or in the meeting designated to decide on the manifesto. Typically, the process has been nearly completed by then and such amendments relate to details and a few choices of phraseology. The process as practised today thus aims at ‘precooking’ the manifesto so that it is ready when the ‘chefs’ that make up the leadership body turn to it. Occasionally, they may still add a pinch of a particular ingredient or a few drops of water before the dish is served, meaning they may agree on paraphrasing a few sentences of the final draft before they vote on the manifesto. And typically the vote is unanimous. This decision concludes the real decision-making process. Yet it may be followed by another, much more formal decision by a more representative party body.

Among the Austrian parties, the ÖVP requires most time for the drafting process that is nevertheless typically completed earlier than those of the other parties. With a regular election approaching, the ÖVP aims at having the manifesto ready at least half a year before the election, irrespective of its eventual enactment date. The reason for such an ambitious timeframe is that, ideally, campaign planning should unfold against the background of the manifesto content. In contrast, in the other parties the campaign is organised parallel to the drafting of the manifesto.

The SPÖ only spends about three or four months on the whole drafting procedure, while the FPÖ needs even less time. Despite the different time requirements, interviewees in all parties agreed on the objective of having the manifesto ready and formally approved two months before the election at the latest.

In all parties covered here, the organisation of the drafting process is held constant even when an early election has been called. Yet such a scenario leads to feedback loops being cancelled, meaning that each group or individual is consulted only once rather than being invited to comment on subsequent drafts. Note that the amount of tightening differs across parties: while the major parties need to deviate substantially from their elaborate routine, the smaller ones simply speed up the process with respect to writing rather than cutting short the intra-party consultation process.

The Actors Involved

We now provide a more systematic perspective on the actors and their degree of involvement, distinguishing five generic types of actors and referring to the party specifics under these broad categories.

The party in central office

Typically, the party in central office drives the manifesto drafting process. This means that the party headquarters staff and one top political appointee (e.g. the central party secretary or party manager) act as the coordinators of the drafting process. The second central institution of Austrian parties is their respective (publicly funded) political academy that is in charge of the education and training of party cadres. Their directors are ex officio members of the parties' executive committees. The academies are routinely involved in programmatic work and their directors can also play a very active role in the drafting of manifestos. Thus, the current academy director drafts the manifestos of the Social Democrats. Though in less prominent roles, the directors of the other party academies have also been involved in the drafting of manifestos. The parties have to be careful, however, as the publicly funded academies are not allowed to spend their resources on election campaigns. The influence of the party in central office on the manifesto is high: the relevant individuals largely select the issues covered in the manifesto, coordinate the input from other bodies, and assume responsibility for putting together the draft.

Land party organisations hardly participate in that process and the party rank-and-file are not involved at all. Yet there are differences between the parties. In our observation period, the SPÖ and the Liberal Forum Land parties seem not to have contributed to the manifestos at all, while ÖVP manifesto positions were agreed with the Land party organisations. The Green and FPÖ Land party organisations cover the middle ground. In practice, Land parties confine themselves to checking issues that are ‘touchy’ from a Land perspective.

The party in public office

Parties in government rely on the personal staff of government ministers and the parliamentary party to provide policy specialists' input for the manifesto. Typically, the inputs of the former are the more important. The parliamentary party is the most important resource for providing policy expertise within opposition parties. The parliamentary party is even more important in the smaller parties that cannot rely on policy expertise in the central party organisation or on inputs from resourceful interest organisations. Therefore the FPÖ's experience after the party split in 2005 was traumatic, as only two MPs and a tiny number of specialists remained (Luther 2006: 374).

Organised interests

The SPÖ and ÖVP can rely on one resource their competitors lack: these two parties have ties to a large number of interest groups, the most important of which have powerful organisations and play major roles in the corporatist arena (Luther 1999; Müller 1994). Some of these organisations are under the leadership of party factions such as the ÖVP leagues or the FSG (the Group of Social Democratic Trade Unionists), whereas other interest groups lack such organisational ties but have traditionally been close to a party.

In the ÖVP, these organisations are directly involved in the drafting process. The Federation of Austrian Industries (Vereinigung der Österreichischen Industrie), for instance, provides economic policy input, as do the Austrian Economic Chambers (Wirtschaftskammer). However, our interviews have revealed only a very limited direct input coming from the Economic Chambers in the last two decades. Business interests are more directly articulated and fed into the drafting process by the Business League, the relevant constituency organisation of the ÖVP. The same applies to agricultural concerns and the Farmers' League while the Workers and Employees' League tries to ensure that this wing's traditional policy concerns are heeded when the manifesto is drafted.

As even a casual observer of Austrian politics knows, the SPÖ has close ties with the labour movement and its two organisations – the Trade Union Federation (ÖGB) and the Chamber of Labour (Arbeiterkammer). Yet these organisations and other interest groups close to the party play only a marginal role in providing inputs for the drafting of manifestos. When in government, the SPÖ's ministerial and parliamentary policy experts provide all of the policy content. When in opposition, the SPÖ draws more on the Chamber of Labour to fill the gap in expertise left by the loss of ministerial capacities. The lack of input from the Trade Union Federation is due to its traditional division of labour with the Chamber of Labour that leaves policy expertise mostly to the latter (Karlhofer 2006).

External policy experts

On a few occasions, external experts have been prominent in providing inputs for the parties' manifestos (Stockinger 1982). In the ‘Action 20’, the ÖVP organised input from 20 academics in 1966. In the next election, the SPÖ struck back by employing an alleged ‘1400 experts’ in working out its six policy programmes that together made up 477 printed pages and collectively formed the full version of the SPÖ manifesto (‘Für ein modernes Österreich’). In recent years, the SPÖ ‘Network Innovation’ was a deliberate attempt to play the same trick again, but the 2002 snap election intervened before discussions could be completed. Incorporating experts in an extended process of drafting policy proposals aimed as much at image transfer as at providing input, and overall was of marginal importance in the 1990s and 2000s.

Hired campaign consultants

The manifesto's function as a fundamental part of the overall campaign strategy notwithstanding, external experts, like hired consultants and advertising agencies, do not take part in the manifesto drafting process as such. In the two big parties, the SPÖ and ÖVP, consultants and advertisers take part in the strategic talks that precede the gathering of input for the manifesto, but they do not contribute to the document itself. Only when the final draft is issued do both parties ask their external consultants to help in the editing process, to ensure that the wording of campaign messages remains as uniformly as possible across party communications.

The varying relevance of different individuals and groups in the parties' manifesto drafting process indicates notable differences between the parties (Table 2). With substantial input provided not only by the party in central government and in public office but also by interest organisations within or close to the party, the ÖVP's drafting process marks the ‘low centralisation’ extreme of our universe. This characterisation is also reflected in the process of repeatedly sending back and forth manifesto fragments and drafts between the groups involved, each of which has de facto veto power. In the SPÖ, the drafting procedure is more centralised: with interest organisations and external consultants playing minor roles and the Land party organisations being involved to an even lesser degree, the number of influential players is smaller than in the ÖVP. When in office, it is mainly the ministries that provide content for the SPÖ manifestos. The parliamentary party steps in when the coalition partner is in charge of the relevant portfolio and no SPÖ junior minister shadows the minister. However, when in opposition, the parliamentary party provides most of the input. Overall, most inputs seem to come from the government side in our observation period. The drafting procedure in the SPÖ is divided into several steps and each of the groups involved is invited to provide feedback on at least a first draft. Although the SPÖ manifesto writing process is more centralised than that of the ÖVP, there are still quite a few players involved.

TABLE 2
RELEVANCE OF INDIVIDUAL AND COLLECTIVE ACTORS IN THE MANIFESTO DRAFTING PROCESS

	SPÖ	ÖVP	FPÖ	Greens	LIF
General secretary	No	High	High	No	No
Party executive	High	High	Medium	High	High
Ministerial cabinets	High	High	–*	–	–
Parliamentary party	Medium	High	High	High	High
Party academy	High	High	No	Low	Medium
Land parties	No	Medium	Low	Low	No
Interest organisations	Low	High	No	No	No
Hired consultants/agencies	Low	Low	No	No	No
Degree of centralisation	Medium	Low	High	High	High
Notes: Data generated from interviews, refers to the situation in the 1990s and 2000s, in the case of the FPÖ mainly to the period after the party split in 2005. No data available for the BZÖ. *Due to intra-party conflict that had left the FPÖ's government team largely divorced from the party organisation, the personal cabinets of this party's government members were not involved in the manifesto drafting process in 1986 and 2002. Relevance: High = central role in the drafting process, medium = being consulted with limited influence, low = being involved in the drafting process with marginal influence, no = not involved in the drafting process. Degree of Centralisation: High = only a small group of people being involved in the drafting process, medium = a larger group of people drafting the manifesto with some individuals and groups being consulted without giving substantial input, low = several weighty groups drafting or influencing the manifesto substantially.

This is not the case with regard to the smaller parties, where we find a much higher degree of centralisation: these parties have no ancillary organisations that need to be actively involved. Also, Land party organisations and party academies are less influential and hired consultants practically irrelevant. Hence, the drafting process is a matter of the party in central and public office that can draw on only a fraction of the human resources that the ÖVP and SPÖ command. In addition, our interviewees reported a less elaborate feedback process than in the SPÖ and ÖVP. In the case of the FPÖ, for instance, the current general secretary drafts the manifesto and is at the same time in charge of the party's campaign. An informal board of about 10 people decides on the substance of the manifesto as well as on its public presentation. We therefore classify the manifesto drafting process in the Green party, the Liberal Forum, and the FPÖ as highly centralised.

The remarkably high degree of centralisation in the FPÖ certainly represents an extreme, but in each party there is one person ultimately responsible for drafting and coordinating the manifesto. Typically, this role devolves to an individual leader who has internalised the collective goals of the party (see Cox and McCubbins 1993: 91) and possesses the skills to foresee problems that specific proposals or formulations may create internally or externally and who is good at writing coherent texts. Usually an experienced generalist of the inner leadership takes on this role. We know about deputy party chairmen and parliamentary party group leaders who have taken on this task in the SPÖ in the past. In the ÖVP, this job has been handed between successive leaders of the parliamentary party group and the party's general secretaries, depending on who was the most suitable candidate given the job description. These people often enjoy considerable standing in the party, which facilitates the process.

Enacting Manifestos: The Silence of Statutes and Political Practice

We now turn to the politics of manifestos. For this purpose, we rely on party statutes, archival material, newspaper reports, and interviews with members of the party elite. The various historical sources we have consulted indicate that the party manifestos received considerable attention from the party leadership even when they were short and sketchy from today's perspective. This begins with consideration of the strategic positioning of the party and the selection of issues that are considered important for the upcoming legislative term and which party leaders think to attract voters. Such discussions are typically held in the party's smallest leadership body and among informal leadership circles. The discussion continues in the party's formal leadership bodies once a draft manifesto has been completed. In some instances, the formal party bodies play an active role in shaping the manifesto.

The elaborate drafting processes described in the previous section aim at producing a manifesto that is both attractive to the party's target electorate and acceptable within the party. The parties have routinely succeeded with the latter task as manifestos have typically been accepted by unanimous vote in the executive bodies even when the preceding discussion was conflictual. In any material sense, the process of decision-making on manifestos is concluded by the decision of the relevant party executive body. Yet manifestos share this characteristic with many other decisions and it does not by itself settle the question of their formal status as a party document.

Many of the writings of the Comparative Manifest Project have generated expectations that the making of manifestos is precisely regulated by party constitutions (Budge 2001: 211) and that the resulting document is ‘unique in being the only authoritative party policy statement approved by an official convention or congress’ (Klingemann et al.2006: xvi). To the best of our knowledge, the assumptions of precise party rules of manifesto-making and approval by party congresses or other representative gatherings of party delegates has never been subject to empirical scrutiny for European countries. This certainly applies to the Austrian case. We therefore check these assumptions in this section.

Amazingly, the otherwise extremely detailed and extensive Austrian party statutes do not even mention the ‘electoral programme’ or ‘electoral manifesto’ and therefore do not contain explicit regulations about which party body decides on electoral manifestos. This holds true for the entire post-war period. Assuming that the content of manifestos does not contradict the party programmes, we can identify party executive bodies as the relevant decision-making institutions by default. Making the final decision on the manifesto in the party executive makes sense in parliamentary systems of government, where elections can occur at any point in time, as organising party congresses in the case of snap elections might waste time and energy needed for fighting the campaign.

Having established the party executives as the bodies entitled to decide on manifestos, of course, does not rule out the possibility that parties make the manifesto an issue to be decided by a party congress or similar representative gathering. Indeed, remaining flexible is the imperative governing this aspect in the design of intra-party rules. And parties do indeed use their representative bodies if such a move fits the timeframe and is politically welcome.

Yet we find patterns of behaviour independent of the timing of elections. As a general rule, the major parties (the SPÖ and ÖVP) opt for enactment by the party congress more often either when in opposition or when governing alone. Of 10 such instances, only two occurred under grand coalition governments.

Before 1971, the Social Democrats enacted all of their manifestos in elitist circles: either in the so-called Parteivertretung (abolished in 1967), the party's executive committee topped up with an equal-sized group of less senior leaders also elected by the party congress, or the ‘small party conference’ (Parteikonferenz, after 1954 named Parteirat) where, in addition to members of the Parteivertretung, the party's women's organisation and the Land party executives were represented – altogether amounting to some 250 people (Müller 1996: 246–60). While more representative than the party executive committee, these bodies were still largely confined to professional politicians. In the Kreisky era of SPÖ single-party government, the party broke with this convention: the 1975 and 1979 manifestos were the first to be enacted at a party conference (Parteitag). Even before the party lost its majority it returned to its previous practice of deciding on the manifesto in a more restrained group in 1983. Only during the party's recent stint in opposition (2000–2007) did the pattern of enactment change again: all manifestos since 2000 have been adopted at party conferences (though perhaps only because, after the return to government in 2007, the need to elect a new party chairman required the holding of a party congress anyway in 2008).

The ÖVP typically chooses smaller executive bodies to formally enact its manifesto. Interestingly, exceptions from this rule occur in a non-random fashion: enactments at party congresses never happen under grand coalition government but rather in times of increased political competition. In 1970, the ÖVP had to defend its achievements during the first period of single-party government in Austria. Likewise, the 2002 manifesto, written for the first election after the ÖVP had governed as part of a controversial coalition with the FPÖ, was enacted by a party congress. Two more such instances occurred in the early 1980s, when the party tried to re-enter government after having been in opposition for more than a decade and was in desperate need of generating attention-catching events.

For the opposition parties, we did not find evidence that any other body than the party executive committee enacted the manifesto. This perhaps is most amazing with respect to the Greens, who have a tradition of grass-roots democracy (Lauber 1995). Yet the decision on manifestos is formally in the hands of the party executive committee (Bundesvorstand) and the practice does not diverge from what the statutes say. Given that Green party congresses are much less predictable than those of the other parties, with delegates not willing to confine themselves to applauding their leaders' decisions, this would probably not be the kind of attention a party wants to generate during a campaign.

Manifesto Characteristics

Before turning to the role manifestos play in the campaign, we take a brief look at some of their characteristics that may influence that role. Clearly, document length and sophistication can enhance or hinder mass perceptions. Before we address these inherent qualities of the manifestos, we briefly review the form of their publication.

Until the 1960s all parties – the SPÖ, ÖVP, the Communist Party (KPÖ), and the League of Independents (VdU), succeeded by the Freedom Party (FPÖ) – published their manifestos as electoral proclamations in the parties' newspapers (the vanishing of the central party press put an end to this practice, see Müller 1992: 54–56). Then, the documents were short and also distributed as thin brochures or even flyers. As they grew in length, manifestos became more distinguishable from other campaign materials and had more limited circulations. Since the late 1990s, the internet has become the main medium of distribution, while print editions are confined to a few hundred copies. In 1971, the FPÖ was probably the first party to hold a press conference to present its manifesto. Later, the ÖVP and all the new parties adopted this practice. In contrast, the SPÖ resorted to holding special presentation events, typically coinciding with the manifesto's formal enactment. With respect to the timing of the manifesto's presentation, we find that almost all manifestos were published at some point between eight and four weeks prior to the election, no matter whether the election was at its regular date or early.

Figure 2 shows that there has been a marked increase in the average length of Austrian election manifestos over the course of the past 65 years. While early elections do not impact on the timing of manifesto publication, they tend to decrease their size. This is quite intuitive given that such a scenario leaves little time for an elaborate process of gathering inputs, consultation, and deliberation within the parties. The notable empirical exception here is the 2002 election; this may be explained by the novel political constellation at the time, with a centre-right government and a centre-left opposition. Moreover, the early and dramatic end of that government seems to have motivated the coalition parties in particular to write at length about their achievements in office.

FIGURE 2

AVERAGE LENGTHS OF ELECTION MANIFESTOS IN AUSTRIA (NUMBER OF WORDS)

In order to provide a more comprehensive account of the systematic variation in the length of election manifestos, we perform a negative binomial regression of the manifesto word count on a set of independent variables (Table 3). First, we include a time variable (measured in years, with 1945 = 0), thus capturing the trend depicted in Figure 2. Also, a variable named ‘truncation’ captures the number of months by which an early election shortened the legislative period, thus accounting for the fact that such elections put time constraints on parties. We have already seen that older election manifestos contain fewer words and expect this bivariate relationship to hold in a multivariate test. Government parties typically have more bureaucratic resources at their disposal than their opposition competitors. This should result in manifestos becoming more elaborate and extensive. A similar argument can be made with respect to party size (operationalised as the legislative seat share at the previous election). Larger parties tend to be better equipped in terms of finance and staff, what makes it easier to gather policy expertise.5 Finally, we include a dummy variable for parties of the left (SPÖ, KPÖ, Greens), thus accounting for the programmatic tradition that these parties have typically exhibited.

TABLE 3
NEGATIVE BINOMIAL REGRESSION ON MANIFESTO LENGTH, 1945–2008

	Model 1		Model 2		Model 3
Variable	Coefficient	S.E.	Coefficient	S.E.	Coefficient	S.E.
Time-related variables
Time (1945 = 0)	0.04***	0.00	0.04***	0.00	0.04***	0.00
Truncation (months)	−0.03***	0.01	−0.02***	0.01	−0.02***	0.01
Party-related variables
Party in government					0.49***	0.12
Party size			1.35***	0.35
Left-wing party			−0.22	0.13	−0.18	0.12
Constant	7.23***	0.14	6.92***	0.19	7.04***	0.15
N	75		70		75
Chi^2	79.15***		81.31***		94.70***
Pseudo R ^2	0.055		0.060		0.065
Note: Figures are negative binomial regression coefficients. Model 2 reports a lower N since the party size variable displays missing values for the first election to the Austrian parliament in 1945. All parties are included that were represented in parliament either prior or after the election or both. ***p-value < 0.001.

As the coefficients show, we find the hypothesised effects for the time and truncation variables, as well as the government and party size variables. We can therefore conclude that a large government party at a regular election in the 1990s is likely to produce a much longer manifesto than a small opposition party at an early election in the 1950s. Interestingly, ideological differences do not affect the length of manifestos.

The Manifesto in the Campaign

After having published and presented their manifestos to the public, these documents might become an integral part of parties' campaign efforts. With respect to this period in their life cycle, we will focus on three research questions: (1) Who reads these texts? (2) How important are they in the parties' communication efforts? (3) Does the media report on the manifestos?

The Manifesto Readership

From our interviews, it became clear that parties do not expect the average voter to read their manifesto. Several party representatives mentioned ‘elite voters’ – people who are more educated and much more interested in politics than the general electorate – as potential readers. Even optimistic estimates of this group's size by our interviewees never exceeded three-digit figures. Yet journalists and the media in general seem to be a very important ‘target group’ for manifesto authors. Almost all respondents referred to them.

A further group of manifesto readers mentioned by several interviewees are party officers who prepare themselves for questions from potential voters. Several interviewees also mentioned their party's political opponents as manifesto readers. Clearly, manifestos are one of the sources to screen when looking for weaknesses in competing parties. Such weaknesses can be blatant breaches of pledges from the last campaign or ill-conceived new pledges. Many of the party representatives we interviewed also referred to academic and educational institutions such as schools, where candidates appear in political discussions and which charge their students with reading the manifestos in preparation for the event.

One reason for the – presumably – rather small readership might be the style or the level at which manifestos are written. There has been some criticism of the language used by politicians as they often fail to appeal to a wider audience (e.g. Kercher 2010). If voters in contemporary elections really have ‘begun to choose’, as Rose and McAllister (1986) put it, the manifestos might provide the basis for a well-informed party choice. The way these political texts are written might therefore become a more important matter than in previous decades (Brettschneider et al.2009: 668). But so far, the researchers have not focused on the level at which manifestos are written and thus the ‘readability’ of these texts (for a notable exception see Brettschneider et al.2009).

The most widespread approach to assess the readability of a written text (or an oral statement) is the classic formula proposed by Flesch (1948), who based his calculation on the average length of words (measured in syllables) and the average length of sentences measured by the number of words. Flesch's readability scores normally range between 0 (difficult) and 100 (easy). Applications of this formula can be found in various fields of research as well as in more practical contexts, such as the phrasing of medical instructions (e.g. Hartley et al.2004; Hayden 2008; Heilke et al.2003; Lowrey 2006).

The readability scores reported in Table 4 indicate that party manifestos in Austria are very difficult to read. Many (potential) readers would have problems comprehending their messages if they were to use these texts as the basis for their vote choice. With an average Flesch score of only 14.5 points the manifestos' overall readability is by far lower than corresponding values we calculated for well-known texts in German: Franz Kafka's ‘Ein Bericht für eine Akademie’ (A Report to an Academy), for example, scores 61, and the Bible's story about the Creation (Book of Genesis, Chapter 1) reaches a Flesch score of 79. But the parties' scores are also below corresponding values calculated for quality newspapers that show Flesch scores between 20 and 40. In fact the manifestos resemble scientific texts normally ranging between –20 and 20. The results presented in Table 4 also do not show substantive differences between the parties or over time. A regression analysis (not shown) demonstrated that the scores do not follow any observable trend and do not systematically differ between the parties. Only the Greens, whose manifestos tend to be easier to read (p <0.1), stand out. Given this party's well-educated electorate (Dolezal 2010), the admittedly naïve expectation formulated above (which stipulated a relationship between the character of a party's electorate and the readability of its manifesto) can thus be dismissed.

TABLE 4
READABILITY SCORES OF PARTY MANIFESTOS, 1986–2008 (FLESCH'S FORMULA)

	SPÖ	ÖVP	FPÖ	BZÖ	Greens	LIF	Mean
1986	10	32	13	–	20	–	18.8
1990	11	9	0	–	9	–	7.3
1994	7	10	8^a	–	20	6	10.2
1995	26	3	16	–	16	11	14.4
1999	5	12	16	–	15	30	15.6
2002	11	4	6	–	9	–	7.5
2006	6	13	5	11	21	–	11.2
2008	10	26	16	27	32	–	22.2
Mean	10.8	13.6	10.0	19.0	17.8	16.0	14.5
Minimum	5	3	0	11	9	6	5.7
Maximum	26	32	16	27	32	30	27.2
Notes: Flesch readability scores range from about 0 (difficult) to about 100 (easy). ^aThis score was calculated after dropping a 49-page introduction authored by the then party leader Jörg Haider.

Manifestos as Campaign Tools?

Even though the parties' manifestos do not reach a wide audience directly, they still could serve various functions in their campaign. Manifestos, the practitioners told us, are important sources of information for candidates and party officers – especially when they meet specific voter groups or their representatives.

One way to estimate the relevance of manifestos during the campaign more systematically is to analyse the content of the parties' press releases. These are an important element of the parties' communication activities during the campaign. They are not directed towards the general public but are written for journalistic purposes as parties try to influence the campaign coverage in the mass media – sometimes with quite substantial success (Hopmann et al.2010). In order to assess the relevance of the manifestos in this regard, we compare the shares of press releases comprising at least indirect references to the programmes with the shares of releases containing references to the parties' top candidates. For each election, we identified all press releases published within two months of Election Day and calculated the relative shares of ‘programmatic releases'6 and of ‘leader releases’.

The percentages reported in Table 5 clearly demonstrate that candidates are far more important than the manifestos: about a third of all press releases include a reference to these candidates, whereas manifestos – even indirectly7 – are rarely mentioned. Clearly, the SPÖ's value for 2008 stands out. This is due to its five-point programme of fighting inflation (just a small fraction of the manifesto) that heated up party competition and led to legislative action even before the election.

TABLE 5
PARTIES' PRESS RELEASES REFERRING TO THEIR LEADERS AND MANIFESTOS, 1990–2008 (PERCENTAGES)

		SPÖ	ÖVP	FPÖ	BZÖ	Greens	LIF	Mean
1990	Leader	28.5	41.4	33.1	–	48.5	–	37.9
Manifesto	2.4	3.8	0.4	–	1.2	–	2.0
(n)	(776)	(914)	(278)	–	(167)	–
1994	Leader	24.3	28.9	27.5	–	38.2	26.6	29.1
Manifesto	3.1	1.8	0.2	–	0.0	0.5	1.1
(n)	(679)	(657)	(408)	–	(55)	(188)
1995	Leader	22.3	36.3	27.9	–	11.5	26.9	25.0
Manifesto	2.3	2.3	1.9	–	4.6	2.3	2.7
(n)	(871)	(695)	(462)	–	(174)	(175)
1999	Leader	26.1	34.5	28.7	–	36.6	11.3	27.4
Manifesto	3.4	1.4	0.8	–	1.7	2.7	2.0
(n)	(916)	(621)	(752)	–	(238)	(293)
2002	Leader	29.1	41.3	39.4	–	22.4	–	33.1
Manifesto	2.2	2.2	1.0	–	1.4	–	1.7
(n)	(1370)	(891)	(497)	–	(281)	–
2006	Leader	21.8	29.5	67.8	46.2	23.9	–	37.8
Manifesto	3.0	1.2	0.5	1.1	4.3	–	2.0
(n)	(1072)	(481)	(205)	(275)	(234)	–
2008	Leader	34.4	39.7	17.7	42.2	18.1	–	30.4
Manifesto	21.5	12	7.5	7	11.1	–	11.8
(n)	(864)	(758)	(372)	(554)	(271)	–
Note: Percentage of all press releases within two months before Election Day. Source: APA (https://www.aomweb.apa.at/).

The Media Reception of Manifestos

In order to assess the impact of parties' manifestos during the weeks before Election Day, we finally discuss their profile in the news coverage of the election campaigns. Journalists, almost all interview partners told us, belong to the rather small manifesto readership. But do they also report on them? As in the analysis of the parties' press releases, we compare the share of articles that refer to the manifestos with that of articles that refer to the leading candidates. Again, we deliberately overestimate the ‘programmatic shares’ as our full-text search also covers articles that do not precisely refer to the manifestos but to all statements that contain the selected keywords. As the universe of articles covering the election campaign we took all those that include a reference to at least one of the competing parties.

Figure 3 compares the relative shares of ‘programmatic’ vs. ‘candidates’ articles from 1994 to 2008. Note that our two measures are independent of each other as each individual article can relate to both objects of interest. As in the analysis of press releases, we focus on the last two months before Election Day. Because the databases available to us for such a comparison over time severely limit the number of newspapers, we can only focus on Die Presse and Der Standard, two Vienna-based quality papers.

FIGURE 3

MEDIA REPORTS ON MANIFESTOS COMPARED WITH REPORTS ON THE PARTIES' TOP CANDIDATES

When comparing the two lines it is obvious that the parties' manifestos – even though we definitely overestimate their impact – are far less newsworthy than the top candidates. Between 40 and 50 per cent of the articles include a reference to at least one of the top candidates. The manifestos or ‘programmatic statements', by contrast, are mentioned in less than 10 per cent of the coverage.

Even though the share of ‘programmatic’ articles is low it is interesting to compare these shares between the several newspapers that compete for Austrian readers. Given the differences between the readerships of quality papers and tabloids – especially with respect to education and also political interest – we expect quality papers to report comparatively more about the parties' programmes. We can check this for the 2008 elections, where we can draw on a much broader sample of eight newspapers.8 Our results (not shown) indicate that our expectation is perfectly met. The shares do indeed vary according to the character of the selected newspapers. Well-educated readers, the ‘qualified elite’ – to cite one of our interview partners – are indeed a potentially more important audience for the parties' manifestos, even when we take into account the mediated nature of their contact with the party messages.

Manifestos in Post-election Politics

Potentially, the post-election period provides us with the deepest insights into the nature of manifestos because it is only now that we can see how seriously the parties take their manifestos. Research into the post-election relevance of manifestos has followed two major strategies. One is to compare policy outcomes to the manifestos of the government (and sometimes also opposition) parties pledge by pledge (e.g. Thomson 2001), the other is to analyse – in a much more indirect way – their likely budgetary consequences (Klingemann et al.1994). In the following we do not look at observable implications of manifestos but again provide some direct information, sketchy as it may be, that speaks to the issue.

Leonard Ray's (2007) attempt at understanding the nature of manifestos is helpful here. Manifestos can be conceived as contracts between parties and the voters, party advertising, or statements of the parties' respective identity and philosophy. We leave aside the lattermost category as it is primarily the party programmes that serve this purpose in German-speaking countries (see above). As a contract, Ray suggests, ‘manifesto pledges should reflect a party's realistic assessment of the policies which they would expect to implement if elected. Policy preferences which are unrealistic would have no place in such a contract’. As party advertising, party manifestos would not carry such strong commitments and ‘voters may discount these documents heavily’. Manifestos would contain ‘unverifiable or meaningless claims that are nonetheless fair for advertising purposes’. They would exaggerate policy differences and contain ‘a proliferation of vague or unrealistic promises’ (Ray 2007: 17).

If we think that these categories cover the universe of actual manifesto purposes, can we generate expectations as to when parties will produce documents that resemble one of these categories or – applying these categories to the individual statements within a manifesto – a particular mix of them? In the following, we proceed from the assumption that political parties act strategically when drafting their manifesto and that behaviour is governed by anticipation. From there we derive the expectation that parties anticipating government participation are likely to produce contract-type manifestos while parties anticipating opposition are more likely to write advertisement-type manifestos.

Applying these considerations to the Austrian case, we have two parties – the SPÖ and ÖVP – that could always hope for government office and on many occasions could almost be certain that they would find themselves in the cabinet after the election. In contrast, the prospect of ending up in the government for the smaller parties was much bleaker throughout the post-war period, and indeed hopeless on many occasions.

Our interviews provide anecdotal evidence that politicians involved in both the drafting of and decision-making on the manifesto and subsequent coalition bargaining see the former as a baseline for the latter. A senior ÖVP politician argued that ‘ideally, the manifesto should fully enter the coalition agreement’ (ÖVP 3). Another representative of the ÖVP recalled specifically the coalition formations of 1999/2000 and 2002: ‘The bargaining positions of the ÖVP were more or less identical with the manifesto’ (ÖVP 2).

As the following quote from a senior FPÖ politician shows, even the vague chance of government participation can foreshadow the drafting of the manifesto. Explaining the party's moderation compared to earlier manifestos with regard to some of its core concerns, the politician pointed out: ‘We already had the goal of government participation in mind when drafting the manifesto’ (FPÖ 2). The FPÖ's hopes materialised: the party was indeed invited to the negotiation table and ended up in a coalition with the ÖVP in 2000.

Parties that do not consider government participation as a short-term option do not think about coalition bargaining when drafting the manifesto:

I don't think we would write a different manifesto when considering coalition bargaining a real option. You would not change the party programme either. We would not smooth down the manifesto. Such considerations are not relevant for the manifesto but rather when you decide who will represent the party in coalition bargaining and what should be the main points for a government programme. However, you would not put down that in the manifesto. The manifesto's task is to clearly position the party on the political map. That's the basis for all further steps. (FPÖ 1)

The above quotes are rare qualitative evidence on the thinking of politicians on issues of manifesto politics and coalition bargaining and their connection. While they cannot prove anything in any strict sense, they certainly lend credibility to our theoretical considerations and suggest that more work along these lines may be worth the effort.

A more technical dimension of the direct impact of manifestos on coalition bargaining relates to the advantage of having workable policies formulated. One interviewee has put it as a law-like proposition: ‘The one who can table a text always has a bargaining advantage’ (ÖVP 1).

The second way in which manifestos can be prominent in post-election politics is by taking a place in the ‘long campaign’ – the attempts by political parties to build up credit and undermine the standing of their competitors in the daily warfare of politics between elections. Manifestos and other electoral pledges then serve as a normative yardstick against which the actual behaviour of government parties is judged. In the hands of political competitors and critical media, the yardstick typically becomes a stick to beat the party that has made specific pledges. Of course, the pledges chosen by political competitors to receive so much attention must share some qualities. One is general popularity. It simply makes no sense to accuse a party of abandoning marginal and unpopular pledges. Another precondition is acceptability to the actor bringing up the issue. Otherwise, the actor would have a hard time explaining why not honouring a pledge is a bad thing. A third quality is that the relevant issue can indeed be closely associated with the target party. Just being contained somewhere in the manifesto may not be enough. Yet the party leader having been on air with a strong and unambiguous statement (that can be broadcast again and again) would fulfil this condition. Take the case of the short-lived Chancellor Alfred Gusenbauer. Probably nothing undermined his standing more than his inability to honour his strong campaign pledge to cancel the purchase of interceptor aircrafts and abolish tuition fees for students. The dominant interpretation was that he had traded these policies against one particular benefit – the office of Chancellor.

Our discussion of the post-election relevance of manifestos has closed the cycle: we are back to the campaign period. Although it would be fascinating to go on and see when campaign pledges disappear from the agenda and what makes this happen, we stop here. Our exploration of the post-election relevance of manifestos can be summarised as follows: first, the foreshadowing of the post-election period probably impacts more on the writing and content of manifestos than manifestos impact on anything that happens after the election. This foreshadowing of the parties' future role as either government or opposition, providing single-party government or being a coalition partner, in turn creates party-specific incentives concerning what to say in manifestos. Second, properly prepared manifestos impact on the government programme of single-party governments and the positions adopted by parties during coalition negotiations. Third, party pledges (that will almost inevitably be contained in the manifesto but do not get their prominence from simple inclusion) have the potential to survive as issues of the ‘long campaign’ that lasts until the next election. However, this will generally not be the choice of the party from which the specific pledge originates. Rather, providing long-lasting pledges (rather than policies) indicates that a party has been myopic in writing the manifesto and deserves to be reminded.

Conclusion

While the content of party manifestos in European democracies has been the subject of much research from different methodological angles, the process of how manifestos ‘emerge’ and the role they play in the different stages of the political process has remained largely uncharted territory. The prime goal of this article was to provide a first exploration into the ‘life cycle of manifestos’ in the Austrian case, by drawing on a wealth of qualitative and quantitative data derived from interviews, archival, and media research.

We found the intra-party process of drafting manifestos elite-centred and quite elaborate, involving stages of strategic discussion, aggregating policy input from various sources, writing drafts, and gathering feedback. The main actors in this process are the party in central and in public offices. While the former typically assumes a coordinating role, the latter's task is to provide policy content – a function that in some parties is also performed by organised interest groups within or close to the party. External policy experts and hired consultants play a minor role. While we found the drafting process to be highly centralised in the smaller parties, it is quite decentralised in the ÖVP, with the SPÖ taking an intermediate position.

Whereas party statutes do not specify the procedure of enactment, manifestos are typically adopted by a formal decision in the party executive (and rarely at party conferences). While adoption is usually unanimous, the document may have been subjected to substantial revision in the course of the drafting process. In general, the public presentation of manifestos coincides with the start of the election campaign, roughly four to eight weeks before Election Day – a pattern that is not altered even in the case of a snap election. However, early elections tend to decrease the length of manifestos, which otherwise has grown considerably over the 65-year period under study.

During the campaign, parties do not expect the average voter to actually read their manifesto. Rather, the document is aimed at a qualified elite and journalists. This is not only reflected in the low readability scores of manifestos when compared with other types of documents, it also becomes visible in the low number of references to manifestos in parties' press releases and in newspaper articles. Predictably, manifesto coverage is more extensive in the quality press than in the tabloids.

After the election, manifestos may serve as a blueprint for coalition bargaining and the drafting of government programmes. Depending on the likelihood of entering government, parties may therefore refrain from making unrealistic pledges in the manifesto.

At the outset of the article we also derived some general expectations on party differences with regard to the various stages of the manifesto cycle related to party ideology, party organisation, party placement along the government–opposition divide, and characteristics of party target electorates (Table 1). Clearly, our empirical findings cannot be more than a first check on the plausibility of these expectations as we have only a few parties but many potentially relevant variables and our sources do not always allow the parties' precise and time-specific placement relative to each other. We are, nevertheless, confident on the placement of those cases that distinguish themselves clearly from the others. The admittedly naïve expectation on manifesto readability as a function of the parties' target electorates is not borne out by the data. While party ideology may work in the expected directions, other influences on parties tend to be stronger. None of our more specific expectations related to ideology is clearly met in the Austrian case. In contrast, our expectations derived from party organisational characteristics and the government–opposition divide seem to work. Accordingly, parties with large and complex organisations receive more and better-informed inputs for their manifestos. Organisational complexity leads to greater coordination needs and efforts, a more time-consuming drafting process, and eventually longer manifestos. Likewise, government parties tend to produce longer documents that have more relevance as bargaining tool than opposition parties.

Finally, the shadow of the future indeed seems to impact on the drafting of manifestos. It encourages prospective government parties to draft manifestos that resemble a ‘contract with the voters’ while prospective opposition parties use manifestos more unashamedly for advertising purposes. We have formulated our expectations in general terms and expect that our findings generalise beyond the country studied. While nested in the Austrian case, the parties included in this study cover much of the range in the relevant dimensions (Table 1). We would expect other European parties to have more in common with their Austrian equivalents than with the parties in the United States, the only country where the making of manifestos has received considerable academic attention over the years (e.g. Maisel 1993–1994; Pomper 1990: 143–73; Shafer 1988).

Finally, what is the implication of our process analysis for the predominant use of manifestos in the discipline as the prime sources for party policy salience and positions? In short, parties take manifestos seriously. We have seen that they invest a great deal of care and resources in their writing. And manifestos have important ‘backstage functions’ in the campaign, providing information for the candidates and activists and serving as a compendium of policy statements that politicians and staff members draw on to answer questions from voters and interest groups. The fact that manifestos as such play a very limited role in the parties’ attempts to reach out to the electorate via the mass media does not imply that their contents share this fate. For instance, a preliminary analysis of quality press reporting on the 2008 Austrian election campaign (not included here) shows considerable overlap in the issue saliencies between election manifestos and party leader statements in the newspapers. Perhaps media logic even requires parties that want to push their issues to avoid redundant references to the manifesto document, as journalists may no longer consider such information newsworthy a few weeks after its publication. While our findings suggest that manifestos may not mean exactly the same thing to all the parties in the game and over time, these documents remain unmatched in providing point-specific raw information on party policy.

Analyses of the manifesto life cycle can take us some way towards understanding such inter-party and over-time differences. This article may serve as a first step towards a framework for the comparative analysis of how manifestos come into being and what role they take in the campaign and post-election politics in parliamentary systems of government. Future work should not only broaden the empirical basis by turning to cross-national analysis but also link the analysis of the manifesto life cycle to the contents of the manifesto and government policy programmes.

Acknowledgements

This research was carried out under the auspices of the Austrian National Election Study (AUTNES), a National Research Network (NFN) sponsored by the Austrian Science Fund (FWF) (S10903-G11). For helpful comments we are grateful to the participants in the AUTNES Supply Side staff seminar and the workshop ‘Why and How of Party Manifestos in New and Established Democracies’ at the ECPR Joint Sessions, University of St. Gallen, Switzerland, April 2011, and the reviewers of West European Politics.

Notes

1. While the present article is not concerned with analysing manifesto content, the authors of this article have developed a new coding scheme for the estimation of party policy positions from election manifestos. This new instrument will be the subject of future publications.

2. The Communist Party also participated briefly in the immediate post-war period that is outside our observational period for most of this article.

3. The interviewees held one or several of the following positions during the past 40 years: party leader, party deputy leader, leader of the parliamentary party group, general secretary, head of the party academy, head of party leader's personal staff, and head of the party's policy department. Our interviewees include five top party members each of the SPÖ and ÖVP, three of the FPÖ, two of the Liberal Forum and one from the Greens. The interviews were conducted on the understanding that they would not be personally attributable.

4. This section draws exclusively on our interviews with party elites (see note 3).

5. Due to the frequent occurrence of grand coalition governments in Austria the party size and the party in government variables correlate highly (r = 0.68). We therefore run two models including either variable at a time.

6. Here we used the following string for a full-text research: (#programm# or #manifest#).

7. Our electronic search deliberately not only includes the formal title of the manifestos but all explicit references to ‘programmatic’ stances.

8. This calculation is based on the quality papers Wiener Zeitung, Wirtschaftsblatt, Die Presse, Der Standard, and Salzburger Nachrichten. As tabloids we identified the Neue Kronen Zeitung, Österreich, and Heute.