Promissory representation lies at the heart of modern representative democracy. It is also a valuable lens through which to examine the complex interplay between economic globalization and democratic processes. In this chapter, we provide an overview of promissory representation. Promissory representation refers to the idea that during election campaigns, political parties and candidates make promises to voters, which they are then expected to fulfill if elected to office. We explore how this idea is crucial for understanding the relationship between voters and their representatives, as well as the mechanisms of accountability in democratic systems. By focusing on promissory representation, we offer a detailed, nuanced, and policy-relevant understanding of the impact of economic globalization on the core functions of democratic governance. This approach provides a valuable complement to broader studies of democracy, offering insights that might be missed by more general measures or conceptualizations of democratic quality.
We begin by introducing the principles of promissory representation. We draw on the work of influential scholars, including Ian Budge, Jane Mansbridge, Hannah Pitkin, Bingham Powell, and Susan Stokes, who have made significant contributions to understanding promissory representation and its limitations. This form of representation aligns with traditional theories of accountability, creating a forward- and backward-looking principal-agent relationship between voters and their representatives. We will describe the mandate and trustee versions of promissory representation and argue that the former provides stronger accountability. Mandate theory emphasizes the role of political parties in organizing coherent policy proposals and fulfilling campaign promises.
We delve into several critical perspectives on promissory representation, addressing both its assumptions and the challenges it faces. These critiques range from questioning the substantive importance of campaign promises to evaluating voters’ capacity to hold governments accountable. To address these concerns, we present evidence from comparative research, particularly the Comparative Pledges Project (CPP), which highlights the significance of campaign promises in shaping party competition and democratic representation. We also explore how ideology and partisanship influence the relationship between voters and parties, offering insights into how alignment between voters’ preferences and party promises can occur, even when citizens lack detailed policy knowledge. Additionally, we examine how voters assess the fulfillment of promises and how this shapes their voting behavior. Finally, we consider fundamental critiques of mandate theory, especially William Riker’s challenge to the concept of “the will of the people” from social choice theory. In response, we explore adaptations of normative mandate theory, including the median voter theorem and the consensus vision of democracy.
This chapter underscores that while promissory representation is elusively complex and contested, it is a key touchstone for democratic theory and practice (Thomson Reference Thomson, Robert Rohrschneider and Thomassen2020). We set the stage for further exploration of the constraints and variations in promise keeping across different political contexts, providing a solid foundation for our book’s main theoretical arguments about the effect of economic integration on democratic representation and accountability.
The Principle of Promissory Representation
Parties make promises to voters during election campaigns, including promises to enact specific policies or achieve particular outcomes, as well as more general promises to address certain problems or to act in the interests of their constituent groups. Parties that gain control over the levers of government after elections, primarily parties that hold executive power that enables them to initiate legislative and spending proposals, are able to fulfill those promises or can at least attempt to do so. The extent to which parties make, break, and keep their campaign promises are key characteristics of modern democratic representation in theory and practice. Jane Mansbridge, one of the most prominent theorists of contemporary democracy, uses the term promissory representation to denote the idea that “during campaigns representatives made promises to constituents, which they then kept or failed to keep” (Mansbridge Reference Mansbridge2003: 515).
Promissory representation is the form of representation best suited to the traditional theory of accountability, which implies that representatives are morally bound by what they promised to voters and are answerable to voters (Pitkin Reference Pitkin1967). According to Pitkin, the idea that elections impose accountability on representatives is diametrically opposed to the idea that elections confer authority on them. The imposition of accountability places constraints on politicians, while the granting of authority frees them from constraints. Promissory representation also implies a forward-looking principal-agent relationship between voters and representatives. At the time of an election, voters (the principals) send an instruction to representatives (the agents) by casting their vote for the candidate whose promises align closest with their (the voters’) preferences. During the subsequent governing period, the selected representatives are morally obliged to follow this instruction by keeping their campaign promises. In addition to this moral imperative, representatives are compelled to keep their promises by the prospect of future elections. Representatives anticipate rewards and punishment for keeping or breaking their past promises.
Depending on the specificity of the promises made by candidates, promissory representation comes in both mandate and trustee versions (Mansbridge Reference Mansbridge2003). It is the mandate version – also called the mandate theory of democracy and the delegate version of promissory representation – with which we are most concerned here, as this provides the potential for the strongest accountability. In the mandate version, during election campaigns candidates make promises that align with constituents’ preferences on substantive policies. This means that candidates make specific and testable promises to enact policies or achieve outcomes during the subsequent governing period. Because of the specificity of the promises made, candidates who receive sufficient votes to govern are in effect delegates who follow voters’ instructions. “Testable” means that by the end of the governing period following the election, citizens – as well as opposing politicians, relevant civil society organizations, journalists, and even political scientists – can gather relevant evidence and would generally agree on whether those promises had been kept. Specific and testable promises are also called “election pledges,” and as we will discuss, political scientists have created a veritable cottage industry of research on campaign promises in a broad range of countries, some of which is now loosely coordinated in the CPP.
In contrast to the mandate version of promissory representation on which we focus, in the trustee version, candidates confine their promises to general commitments to address certain problems or to act in the interests of particular constituency groups or even in the interests of the country as a whole (Mansbridge Reference Mansbridge2003). While Mansbridge is correct to point out that the trustee version is theoretically compatible with promissory representation, the trustee version is a quite different category of representation, one that gives preeminence to representatives’ freedom, rather than to their constraints, which is the basis of accountability. The trustee version is an antiquated version of representation commonly associated with Edmund Burke, an influential eighteenth-century Irish Member of Parliament and philosopher. When standing for office in Bristol, England, in 1774, Burke famously argued that representatives should exercise their independent judgment on behalf of voters, rather than functioning as mere delegates who are obligated to follow explicit instructions. A semblance of traditional accountability may be possible in the trustee version of promissory representation, but it is a pale reflection of that found in the mandate version. In the trustee version, representatives may still have a moral obligation to act in accordance with their general or vague promises, and voters may still reward and punish them for perceived successes and failures in keeping those promises. But vague promises are at best weak constraints on representatives and cannot effectively bind representatives to voters’ substantive policy preferences. As we discuss in more detail in Chapter 9, vague statements are compatible with a broad range of subsequent policies, which gives representatives considerable room for maneuver. Moreover, vague statements are not only less specific but are also less testable. Consequently, citizens and other observers are unlikely to be able to agree upon relevant evidence with which to assess whether these promises were fulfilled. In short, if representatives are pure trustees rather than delegates, elections grant at least as much authority as accountability on them, which is difficult to reconcile with the original understanding of promissory representation.
The mandate version of promissory representation is also associated with the responsible party model, which gained widespread attention in 1950 following a high-profile and controversial report by the American Political Science Association’s (APSA’s) Committee on Political Parties (American Political Science Association Committee on Political Parties 1950). The most relevant point from this report and subsequent discussion of the responsible party model is that it puts the focus on political parties, as distinct from individual candidates. Modern representative democracies are party democracies, in which most candidates are associated with parties that have comprehensive election platforms, manifestos, or equivalent documents. APSA’s 1950 report was titled Toward a More Responsible Two-Party System and contained a set of sweeping recommendations for reforming and improving American democracy. The report called for political parties with more internal cohesion and clearer and more distinct policy platforms that offered citizens sharp policy alternatives. Moreover, parties should commit themselves to enacting those policy platforms if the election results put them in a position to do so. In other words, parties should follow through on their campaign promises.
APSA’s 1950 report was subsequently criticized, even by some of its authors, for failing to distinguish adequately between empirical and normative statements, as well as failing to engage with some of the existing and emerging research that cast doubt on some of its underlying assumptions (Kirkpatrick Reference Kirkpatrick1971). For instance, Kirkpatrick lamented the fact that the report did not engage directly with research, which indicates that voters’ decisions are often largely shaped by factors other than parties’ policy stances. A more recent report from APSA on the state of American democracy is more circumspect and less proselytizing in tone (APSA Presidential Task Force on Political Parties 2023). Nonetheless, the 2023 report shares the 1950 report’s emphasis on the critical role played by parties in organizing political ideas and policy proposals coherently, including in the form of campaign promises. Given the current state of American democracy, the 2023 report is more sensitive to the dangers of polarization and divisiveness that can accompany the existence of parties with sharply distinct visions for the future of the country. The more recent report is also more attuned to the reality of parties being loose coalitions of diverse interests and to empirical research on voting behavior.
Downs’ influential book, An Economic Theory of Democracy, contains a concise formulation of mandate theory and the responsible party model (Downs Reference Downs1957). The book was an important milestone in empirical democratic theory in several respects: First, in line with modern political science practice, Downs explicitly stated his assumptions concerning the motivations of political parties and voters. Second, he stated these assumptions in empirical rather than normative terms, which was one of the criticisms leveled at APSA’s 1950 report. Third, he outlined a theory that, with some adaptations, has had major impacts on a diverse range of phenomena, including party competition, coalition formation, and policymaking.
In the simple theory formulated by Downs, the political landscape is populated by two sets of actors: voters and parties. Voters hold clear preferences concerning the future trajectories of public policies. In stark contrast, parties are primarily motivated by the pursuit of the material benefits and status associated with holding government office. Parties are therefore characterized as office seekers rather than genuine policy seekers. Parties adopt policy positions strategically to win votes, not out of an intrinsic commitment to certain ideals. Voters, in turn, select the party whose policies align most closely with their preferences, while at the same time holding the governing parties accountable for successes and failures in delivering on their previous campaign promises. In the context of a two-party system, the party securing a majority of votes takes office. Once in power, the governing party has strong incentives to fulfill its campaign promises, not because of an inherent belief in the virtue of those promises, but out of a fear of electoral reprisal in the form of retrospective sanctions at the next election should it fail to do so. The Downsian model demonstrates that one need not have an overly optimistic, some might say naïve, view of politicians’ motivations to expect there to be a robust linkage between campaign promises and subsequent public policies.
To the extent that campaign promises are clear, there are at least two ways in which voters can hold politicians accountable. As illustrated in Figure 1.1, political parties make promises during election campaigns, after which voters assess these promises against their own preferences and typically support parties whose promises align with their views. This creates a strong incentive for parties to craft promises that resonate with their constituents, as citizens reward appealing promises with their vote and punish unappealing ones by withholding their vote. We call this prospective accountability. At the same time, retrospective accountability allows voters to hold governing parties accountable for their past performance. In doing so, voters can evaluate the extent to which the parties’ promises were kept. Voters can reward parties that fulfilled their commitments and punish those that did not. While some promises may be broken for legitimate reasons, consistent failure to deliver on commitments can damage a party’s credibility and the overall legitimacy of the democratic process. This retrospective accountability creates a powerful incentive for parties in power to work diligently on keeping their promises, as their future electoral success depends on it. Both prospective and retrospective accountability are crucial to ensuring that political parties remain responsive to public opinion.
Critiques of and Adjustments to Promissory Representation
We need to consider several critiques of promissory representation in its various forms. These criticisms include direct attacks on mandate theory as well as theories and findings that lead us to question some of the assumptions and propositions outlined herein. None of these criticisms have led to the demise of mandate theory either in political science research or in the practice of democratic representation. Some criticisms, such as the assertion that parties do not make important campaign promises, can be dispatched easily. Other criticisms, such as the demonstration that election results rarely provide unequivocal signals of the will of the people, if such a will exists at all, have led to more fundamental adaptations of the theory.
Is Democratic Representation under Promissory Representation Too Inflexible?
Under promissory representation, the link between voters and government policies is generated through the assumption that political parties in government feel obliged to implement the promises they made before they were elected and that they expect to be voted out of office if they do not fulfill their promises. This view of democratic representation may appear inflexible. To the extent that politicians must adhere to the promises they made before elections, promissory representation can be rigid, limiting a representative’s ability to adapt to new circumstances or changing public opinion. The world is unpredictable, and the needs or priorities of the electorate may shift between elections. Sticking too rigidly to campaign promises might lead to suboptimal outcomes if the context changes.
This could be particularly problematic in globalized environments where opportunities and constraints constantly influence and shift policy priorities. For example, during the 2008 global financial crisis, several European countries, like Greece, were forced to implement harsh austerity measures due to conditions set by international lenders such as the International Monetary Fund and the European Central Bank. In these cases, political parties may have campaigned on promises to increase social spending or reduce taxes, but the economic crisis and bailout terms made it impossible to fulfill these promises. For another example, the Comprehensive and Progressive Agreement for Trans-Pacific Partnership and similar trade agreements require signatory countries to adhere to specific trade and regulatory standards. A political party may promise to protect domestic industries through subsidies or tariffs during an election. However, after coming to power, the government might find that such policies violate existing trade agreements, forcing them to abandon their promises. While often a matter of perspective, cases like these suggest that voters may ultimately be better served when parties break campaign promises that prove infeasible or harmful under changed circumstances.
An alternative concept that is tied to the functioning of representative democracy, but that allows for more flexibility, is the idea of political responsiveness (Arnold, Franklin, and Wlezien Reference Arnold, Franklin and Wlezien2010; Franklin and Wlezien Reference Franklin and Wlezien1997; Hagemann, Hobolt, and Wratil Reference Hagemann, Hobolt and Wratil2016; Hobolt and Klemmensen Reference Hobolt and Klemmensen2005; Hobolt and Wratil Reference Hobolt and Wratil2020; Lax and Phillips Reference Lax and Phillips2009; Meijers, Schneider, and Zhelyazkova Reference Zhelyazkova, Bolstad and Meijers2019; Powell Reference Powell2004; Rauh Reference Rauh2016; Schneider Reference Schneider2019, Reference Schneider2020; Tausanovitch and Warshaw Reference Tausanovitch and Warshaw2014; Wlezien Reference Wlezien1995; Wlezien and Soroka Reference Wlezien and Soroka2011; Wratil Reference Wratil2017; Zhelyazkova, Bolstad, and Meijers Reference Meijers, Schneider and Zhelyazkova2019). In this framework, the quality of representation depends on the ability of elected officials to respond to the preferences and needs of the electorate during their time in office. Rather than being strictly bound to the promises made during campaigns, democratic responsiveness is adaptive and emphasizes the need for representatives to adjust their actions based on evolving public opinion and changing circumstances. It involves both retrospective and prospective elements, where elected officials may adapt policies to reflect changing voter preferences, even if these diverge from the original promises.
Democratic responsiveness emphasizes flexibility and the ability of representatives to respond to changes in public opinion or new circumstances. This can be particularly important in fast-changing global or domestic environments where the priorities of the electorate may evolve after elections. A focus on responsive representation provides justifications for political parties in government to change their policies compared with their previous promises. As long as they align with changing public opinion on this issue, the link between voter preferences and policy outcomes is sustained. In addition, rather than being judged solely on campaign promises, democratic responsiveness requires representatives to continuously engage with and respond to voters throughout their time in office. This fosters a more dynamic and ongoing relationship between voters and representatives. A governing party that adapts to crises and evolving public preferences may be seen as more legitimate than one that sticks steadfastly to what it promised.
We do not argue that promissory representation is superior to democratic responsiveness. Neither promissory representation nor democratic responsiveness is inherently more legitimate in all circumstances, and a balance between promise keeping and responsiveness may offer the most robust model for ensuring both accountability and adaptability in democratic governance. It is important to consider these different approaches to understand both the value and the limitations of analyzing democracy through the lens of promissory representation.
Although a perspective of democratic responsiveness might be seen as more legitimate in environments that value flexibility, ongoing voter engagement, and the ability to adapt to changing circumstances, it can dilute accountability if representatives are constantly shifting their positions based on fluctuating public preferences. Voters may find it difficult to judge their representatives’ performance if there are no clear benchmarks (such as promises) against which to evaluate them. At the same time, representatives might shift their positions opportunistically to gain or maintain power, catering to short-term trends or specific interest groups rather than following a coherent policy agenda aligned with their long-term vision.
These drawbacks also highlight some of the main benefits of analyzing democratic representation through promise making and keeping:
1. Clarity and accountability: Promissory representation creates an implicit contract between voters and representatives. Parties or candidates make explicit promises, and voters can hold them accountable for fulfilling those promises. This strengthens the accountability mechanism, as voters have a concrete basis (the promises) on which to evaluate their representatives.
2. Voter control: Promissory representation offers a straightforward way for voters to influence policy. By voting for the party or candidate whose promises align most closely with their preferences, citizens exercise control over the policy agenda. If representatives don’t deliver, voters can “punish” them in future elections.
3. Predictability: The model of promissory representation ensures a degree of predictability in politics. Voters know what to expect from the parties they vote for, which can contribute to political stability. It also strengthens trust in the system when promises are kept. Promissory representation not only benefits voters but also increases predictability for political parties. When parties commit to a set of promises, they provide a clear roadmap for their policy agenda, which can help strengthen internal party cohesion and strategy. It allows parties to maintain a consistent message and agenda throughout their time in office, reducing internal conflicts over the right policy direction and making it easier to manage party members’ expectations and discipline. Moreover, predictable promises allow parties to build stronger, long-term relationships with their core voter base by delivering on clearly defined commitments.
4. Moderate flexibility: While promissory representation is less flexible in adapting to changing preferences and circumstances, there are at least two reasons to assume that promissory representation offers a balance between consistency, predictability, and flexibility. Promissory representation does not imply that politicians are slaves to all the promises they make. It is generally accepted that political parties may not keep all their campaign promises, especially as public opinion or circumstances change. The accountability constraint only bites when parties consistently fail to keep large shares of their promises. At the same time, the promises provide a benchmark that makes changing policies less arbitrary.
Are Campaign Promises Substantively Important?
A few researchers have claimed incorrectly that campaign promises are unimportant. For instance, Hofferbert and Budge (Reference Hofferbert and Budge1992: 153) argued that “specific pledges are made in peripheral or unimportant areas of policy which are easy to alter. On central areas of policy such as the economy, for example, no specific commitments are made” (see also Klingemann, Hofferbert, and Budge Reference Klingemann, Hofferbert and Budge1994). The claim that promises do not matter is refutable on the basis of evidence that was available in 1992 when it was made and even more so on the basis of comparative evidence that was published in later years. Two important early studies on campaign promises are particularly noteworthy, one on the United States and one on the United Kingdom and Canada. Pomper and Lederman (Reference Pomper and Lederman1980) conducted an authoritative study of campaign promises made by US parties in their platforms for presidential elections from 1944 to 1975. They concluded that the distribution of promises across subject areas corresponded to voters’ concerns, whereby promises were concentrated in more electorally salient policy areas. They found a concentration of promises in economic policy, social policy, and foreign policy in parties’ election platforms (1980). Furthermore, they identified many substantively important promises, such as the Democrats’ 1976 promises to introduce strip-mining legislation in environmental policy, to maintain the value of welfare benefits, and to provide construction workers with new picketing rights (1980). Such policies had significant implications for the groups affected, including businesses, environmentalists, employees, and welfare recipients.
Similarly, Rallings (Reference Rallings, Budge, Robertson and Hearl1987) studied the fulfillment of campaign promises made by parties in the United Kingdom and Canada between 1945 and 1979. He concluded: “Parties clearly feel under some obligation to make promises on action on those matters to which they have given general prominence in their manifestos” (Rallings Reference Rallings, Budge, Robertson and Hearl1987: 5). Manifestos were also replete with substantively important promises, such as the UK Conservative Party’s 1970 promises to reduce numbers of civil servants, to cut income taxes and other taxes, and to implement detailed measures to reduce trade unions’ rights. Specific measures such as these are, of course, means to larger ends, which generally include macroeconomic outcomes. Parties are generally more circumspect about making firm commitments on outcomes over which they have less direct control, such as achieving low inflation, low unemployment, and strong economic growth, and voters undoubtedly evaluate governments based on the extent to which they achieve these greater goals. Rallings (Reference Rallings, Budge, Robertson and Hearl1987) points out that the making and fulfillment of campaign promises may have little to do with the extent to which voters’ expectations in relation to these outcomes are met. This does not, however, mean that promises are of little consequence.
More recently published comparative research on campaign promises confirms the importance of specific campaign promises. Early studies of campaign promises by pioneers such as Pomper and Lederman and Rallings laid the foundation for an international comparative research program on campaign promises known as the CPP. The CPP is now an open and expanding network of researchers who are conducting research on campaign promises in a broad range of countries, including research on the promises that parties make and public opinion on those promises. The book Mandates and Democracy summarizes some of the comparative research into over 20,000 campaign promises made by parties during fifty-seven election campaigns in twelve countries (Naurin, Royed, and Thomson Reference Naurin, Royed and Thomson2019; see also Thomson et al. Reference Thomson, Royed, Naurin, Artés, Costello and Ennser‐Jedenastik2017).
Apart from offering a broader coverage of countries and parties than was available in previous research, CPP researchers have paid closer attention to formulating common standards for defining and coding promises and to calculating measures of inter-coder reliability. Among the key findings from the CPP research are that parties of all kinds make substantial numbers of promises (on average over 100 promises in each manifesto), regardless of their incumbency status, ideological position, and institutional context (Thomson et al. Reference Thomson, Royed, Naurin, Artés, Costello and Ennser‐Jedenastik2017). The contents of parties’ promises are important in that they help give substance to parties’ ideological principles and distinguish parties from each other. Furthermore, detailed descriptions of the promises made by parties reveal many examples of substantively important promises (Naurin, Royed, and Thomson Reference Naurin, Royed and Thomson2019). It also appears that once in government, parties that hold executive power make serious efforts to fulfill the promises they made during the previous election campaigns.
A separate criticism of the Downsian model of party competition – not of mandate theory or the responsible party model more generally – is that parties are not pure office seekers as Downs’ model assumes. Rather, they are policy seekers too. This critique not only questions the accuracy of Downs’ portrayal of party behavior but also suggests an additional rationale for anticipating that parties make important campaign promises and attempt to follow through on those promises if they have the chance to do so. Parties, as historical entities, are deeply embedded in specific societal cleavages, such as the historic division between church and state and the processes of industrialization (Lipset and Rokkan Reference Lipset and Rokkan1967). Consequently, parties are not the malleable entities envisioned by Downs; rather, they are historically rooted organizations that are embedded in historical contexts. Their identities are tied to distinct ideological principles and policy positions that have evolved slowly over time. This policy-seeking perspective on political parties implies that parties receive intrinsic rewards from pursuing and implementing policies associated with their identities. Unlike the transactional approach posited by Downs, which implies a focus solely on gaining power, this perspective suggests that parties derive direct satisfaction from adhering to their policy commitments. This inherent policy-seeking motivation further strengthens the argument for a robust manifesto–policy linkage. It implies that parties, driven by a historical legacy and commitment to specific policies, are likely to make promises in their programs and attempt to fulfill those promises.
The cartel party thesis makes the opposite claim to the policy-seeking critique, by claiming that parties have become estranged from their historical roots (Katz and Mair Reference Katz and Mair1995). The cartel party thesis does not challenge the proposition that parties’ campaign promises are important, but it does suggest a different perspective on why they are important and what we would expect to find in terms of the nature of those promises. The kernel of the cartel party thesis is that mainstream parties in Western democracies have grown simultaneously closer to the state and further from the social groups they were established to represent. The thesis contends that parties have become dominated by elites, whose main concern is the pursuit of government office, in other words with Downs’ office-seeking motivation. The proponents of the thesis argue that modern parties offer little choice in terms of stark ideological alternatives. Electoral competition has been reduced to the selection of which team of government managers appears to be most competent. The cartel party thesis implies that parties need to be programmatic and to offer voters blueprints for government that would include detailed policies on the most important issues facing the country. After all, parties that failed to do so could not credibly claim to offer competent teams of government managers. However, the cartel party thesis implies that parties’ campaign promises, at least those of mainstream parties, should not differ markedly from each other in terms of their respective underlying ideological principles. There are certainly compelling parts to the cartel party thesis, as it is based on a large amount of comparative qualitative evidence on party organization in different countries. However, the claim that parties have become ideologically indistinguishable does not correspond to the increased polarization that characterizes party politics in at least some Western democracies in recent years. We return to this point in Chapter 7, where we examine the possible impact of globalization on political parties’ ideological convergence.
Do Citizens Hold Governments Accountable?
Critics of promissory representation, at least naïve versions of it, contend that voters generally do not have clearly defined policy preferences on many issues, and so their votes for parties cannot be interpreted as clear instructions to those parties to carry out the specific policy proposals they made prior to those votes being cast (Achen and Bartels Reference Achen and Bartels2016; Bartels Reference Bartels1996, Reference Bartels and Leighley2010; Berelson, Lazardsfeld, and McPhee Reference Berelson, Lazardsfeld and McPhee1954; Gabel Reference Gabel1998; Magee, Brock, and Young Reference Magee, Brock and Young1989). In other words, this criticism holds that Downs’ policy-seeking voters do not in fact know which policies they prefer. This line of criticism derives from early work on public opinion, which arguably had unrealistically high expectations of the general public’s knowledge of and interest in politics and policies that most citizens did not live up to (Campbell et al. Reference Campbell, Converse, Miller and Stokes1960; Converse Reference Converse2006). However, Downs’ model is an analytical exercise, an elegant model, and is reduced to a combustible straw man if taken literally. A realistic version of promissory representation recognizes the limits to voters’ ability to formulate nuanced policy preferences, due in part to the constrained time that most people have available to immerse themselves in the intricacies of public policies, as well as their limited interest in grappling with such details.
Leadership, ideology, and partisanship are the main mechanisms through which congruence between voters’ policy preferences and parties’ campaign promises can be achieved even when citizens are not well informed about the details of public policies (Fiorina Reference Fiorina1981; Gabel Reference Gabel1998; Gabel and Scheve Reference Gabel and Scheve2007; McKelvey and Ordeshook Reference McKelvey and Ordeshook1986; Page and Shapiro Reference Page, Gilens and Shapiro1992; Ray Reference Ray2003; Robertson Reference Robertson1976; Wittman Reference Wittman1989). Leadership is a cornerstone of the responsible party model in that parties must take the lead in identifying issues and formulating proposals that voters view as relevant and in line with their interests (Pomper and Lederman Reference Pomper and Lederman1980). Many voters do not know what they want until party leaders tell them what they want. For democratic theory, the alignment between voters’ views and the election platforms of the parties for which they vote is more important than the direction of the causal relationship between the two variables. This does not imply that party leaders could propose any policies and their supporters would blindly follow. Parties are constrained by the bonds of ideology and partisanship.
Ideology plays a vital role in addressing the informational deficits of voters. Ideologies are sets of general principles that help people to interpret the world as it is and how it should be (Denzau and North Reference Denzau, North, Lupia, McCubbins and Popkin2000). While citizens may not have clearly defined policy preferences on specific issues, their ability to cast votes is grounded in perceptions of the ideological proximity between themselves and various political parties (Van Der Eijk, Schmitt, and Binder Reference Van Der Eijk, Schmitt and Binder2005). Ideological congruence between citizens and parties implies agreement between citizens and parties at the more abstract level of principles and a consistency between those general principles and the specific campaign promises made by parties. Similarly, the term macro-polity was coined to denote how shifts in aggregate public sentiment contain signals on the desired direction of policy (Erikson, Mackuen, and Stimson Reference Erikson, Mackuen and Stimson2001). Such shifts in broad sentiment do not require voters to be familiar with the details of policies.
Partisanship or party identification is an affective bond between citizens and the parties with which they have most affinity, and these bonds form voters’ political identities. This offers a complementary linkage between citizens and parties that does not require citizens to be aware of the details of parties’ specific promises. Partisan attachments are linked to expectations that those parties will pursue the interests of particular social groups and related ideological principles. Insofar as parties’ campaign promises align with their general ideological principles and partisan traditions, they are consistent with voters’ expectations about what those parties would do in office. This alignment between voters’ expectations and their parties’ performance is found in the range of different ways in which partisanship has been understood by political scientists. According to the classical conception, party identification is a personal characteristic that changes little over time. People develop partisan attachments in early life, and these remain relatively stable throughout their lifetimes. According to this view, people who identify with a party evaluate the performance of that party in representing their interests more favorably. In accordance with this classical view, Campbell et al. (Reference Campbell, Converse, Miller and Stokes1960: 133) wrote: “Identification with a party raises a perceptual screen through which the individual tends to see what is favorable to his partisan orientation.” The revisionist conception of party identification is that it is a consequence rather than a cause of how citizens evaluate parties. In this revisionist view, citizens’ partisanship is a “running tally” of their present and past evaluations of parties (Fiorina Reference Fiorina1977, Reference Fiorina1981; Johnston Reference Johnston2006). Studies on party identification continue to be framed in terms of the contrast between traditional and revisionist views on party identification (Bartels et al. Reference Bartels, Box-Steffensmeier, Smidt and Smith2011; Thomson Reference Thomson, Marsh, Farrell and McElroy2017). A realistic version of promissory representation recognizes the importance of partisanship in shaping citizens’ support for parties, which is intertwined with their support for those parties’ policies and related campaign promises.
While acknowledging the limits to most citizens’ knowledge of parties’ detailed policy proposals, research suggests that voters often do know which important promises governing parties manage to fulfill or break. In a series of national public opinion surveys, researchers asked citizens to assess the fulfillment of a range of campaign promises made by parties in Ireland (Thomson Reference Thomson2011), Sweden (Naurin and Oscarsson Reference Naurin and Oscarsson2017), Portugal (Belchior Reference Belchior2019), Canada (Duval and Pétry Reference Duval and Pétry2018), and the United Kingdom (Thomson and Brandenburg Reference Thomson and Brandenburg2019). The questions put to voters concerned quite specific promises, such as to increase numbers of police officers, reduce public school class sizes, and implement certain tax cuts. Even though answering such questions requires detailed knowledge, many respondents were able to distinguish accurately between fulfilled and unfulfilled promises. Researchers found that governing parties’ actual records of promise keeping are among the most important factors in shaping citizens’ evaluations of promise keeping.Footnote 1
Research also indicates that the extent to which governing politicians keep their promises significantly affects voters’ evaluations of those politicians and their support for those politicians. The evidence for this insight comes in two broad types: first, macro-level associations between promise keeping and electoral support and, second, detailed micro-level work, including survey experiments, that isolates the causal impact of broken promises on citizens’ evaluations of politicians and their electoral behavior. Matthieß (Reference Matthieß2020, Reference Matthieß2024) provides one of the most comprehensive studies of the macro-level kind. She integrates data on promise fulfillment by governing parties with data on changes in electoral support for those parties between the election year in which they made those promises and the election year at the end of their term in office. The key finding is that voters punish parties that fail to deliver on their campaign promises through lower vote shares in subsequent elections. This relationship holds even after controlling for the range of factors commonly found to influence vote share such as economic performance.
Similarly, other important studies have reported strong relationships between promise keeping and breaking and electoral support. Stokes’ (Reference Stokes2001) work on Latin America uses a broad conception of promise keeping. Her work provides overall qualitative assessments of whether governments’ economic policies were sharp departures from the promises they made during previous election campaigns. Her research concludes that when governing parties, in particular presidents, fail to deliver on their promises, voters perceive this as a breach of democratic trust, leading to electoral punishment. In a related study, Johnson and Ryu (Reference Johnson and Ryu2010) use Stokes’ assessments of promise breaking in seventy-eight Latin American governments to examine the interaction between promise breaking and economic performance. They find that when presidents break their campaign promises, the effect of economic performance has a particularly strong effect on those presidents’ electoral outcomes at the end of their term in office. In other words, when presidents abandon their promised economic policies, often to pursue more liberal market agendas, voters punish them particularly harshly if these unannounced policies do not work out and reward them if they do. This implies that promise breaking is a very risky course of action for governing parties.
Micro-level research that focuses on individual citizens confirms that people generally respond positively to promise keeping and negatively to promise breaking. This micro-level research comes in a variety of forms. One study was based on a large-scale survey experiment in Sweden (Naurin, Soroka, and Markwat Reference Naurin, Soroka and Markwat2019). The researchers developed and tested a sophisticated theory in which voters reward and punish parties for promise keeping and breaking, as well as for carrying out policies they either agree or disagree with. One of the main insights of this study is that governing parties are generally punished for breaking promises, especially if voters agree with the content of the promises that were made. An earlier study uses a more qualitative approach to interview citizens on their perceptions of campaign promises (Naurin Reference Naurin2011). The study finds that voters’ perception of politicians as promise breakers is strongly linked to a lack of trust in politicians and voters’ negative assessments of them.
Survey-based research in other countries also confirms the relevance of governing parties’ records of promise keeping on citizens’ assessments of those parties as well as citizens’ voting behavior. Matthieß (Reference Matthieß2022) followed up on her macro-study of promise breaking and electoral outcomes with a more narrowly focused survey experiment in Germany. When citizens are presented with information that a governing party broke a promise, they lose confidence in that party and are less likely to vote for that party in federal elections. These findings reflect similar findings in other studies about the effects of governing parties’ performance in terms of pledge fulfillment on citizens’ voting behavior both in experimental settings and in fieldwork (Bonilla Reference Bonilla2022; Born, Van Eck, and Johannesson Reference Born, Van Eck and Johannesson2018; Elinder, Jordahl, and Poutvaara Reference Elinder, Jordahl and Poutvaara2015; Simas, Milita, and Ryan Reference Simas, Milita and Ryan2021).
Citizens differ in how strongly they respond to their representatives’ records of promise keeping and breaking. Some of these differences are based on citizens’ attitudes and partisanship. For instance, Naurin, Soroka, and Markwat’s (Reference Naurin, Royed and Thomson2019) survey finds that voters who disagreed with a party’s campaign promise may punish a party even if the party kept its promise. Similarly, another experimental study by Simas et al. (Reference Simas, Milita and Ryan2021) in the United States shows that voters disapprove of politicians who take stances with which they disagree, regardless of whether those stances were foreshadowed by clear or ambiguous language during the election campaign. This dynamic is particularly relevant in polarized political environments, where voters’ preferences are strongly aligned with their ideological positions. Bonilla (Reference Bonilla2022) presents similar results in her experimental study of US citizens. Moreover, she finds that voters object to promise breaking in and of itself, even when breaking a promise results in an outcome they prefer. In other words, some voters simply don’t like promise breakers, regardless of what they think of the content of the promise being broken.
The impact of keeping or breaking promises on citizens’ evaluations of parties is also shaped by partisanship. Bonilla’s (Reference Bonilla2022) experimental work shows that voters evaluate promises through the lens of their partisanship, meaning that co-partisans (those who support the governing party) are more forgiving of broken promises, while out-partisans are less forgiving. This finding aligns with earlier survey research finding that although people are generally able to distinguish between promises that are kept and those that are broken, partisanship leads to systematic biases. People in Ireland (Thomson Reference Thomson2011), Sweden (Naurin and Oscarsson Reference Naurin and Oscarsson2017), Portugal (Belchior Reference Belchior2019), Canada (Duval and Pétry Reference Duval and Pétry2018), and the United Kingdom (Thomson and Brandenburg Reference Thomson and Brandenburg2019) are more likely to say that their favorite parties kept their promises, regardless of those parties’ actual record of promise keeping.
Overall, these studies provide strong evidence in support of a key assumption in our argument, namely that voters can assess whether governing parties kept their campaign promises and that they use these assessments to inform their voting decisions. This does not mean that voters’ assessments are perfect or that their voting decisions are not also informed by factors other than parties’ records of promise keeping and the contents of the promises made in any given election campaign. All that our argument requires is that voters’ capabilities to assess promise keeping and act upon these assessments are sufficiently strong that parties worry about this. Therefore, as globalization reduces the ability of many governing parties to fulfill their pledges, those parties feel pressured to adapt to mitigate the negative electoral consequences. Because this assumption is such a crucial part of our argument, Chapter 6 presents new analyses, based on both observational and experimental data, to show that citizens are indeed capable of holding their governments accountable for promise keeping and breaking. The analyses we present move beyond the existing literature by demonstrating that these accountability mechanisms are even stronger in the context of globalization.
The Arbitrariness of Aggregating Citizen Preferences
Finally, the normative elements of mandate theory have been subject to a sophisticated critique, which questions the existence of “the will of the people,” based on social choice theory (Riker Reference Riker1982). Riker uses Arrow’s (Reference Arrow1963) theorem to demonstrate that election outcomes cannot reliably be interpreted as clear signals of “the will of the people,” as elected representatives often claim they do. Arrow’s theorem shows that, under certain reasonable assumptions, the aggregation of individual citizens’ policy preferences across multiple issues into a single collective outcome is essentially arbitrary and subject to change through endless voting cycles. Furthermore, given that parties make packages of pledges, there is no guarantee that each individual pledge enjoys broad support, even among the parties’ supporters. The problem is compounded in many electoral systems where distinguishing between winning and losing parties is difficult. In first-past-the-post majoritarian systems, the “winning” party often has a minority of votes cast even if those were (usually) more votes than were cast for the next largest party. In more proportional electoral systems, parties’ entry into governing coalitions is as much about their ability to haggle their way into a coalition as it is about the size of their electoral support. All this makes the prescription of normative mandate theory that governing parties “ought to” fulfill as many of their campaign promises as possible seem unconvincing.
In response to the Rikerian critique of normative mandate theory, some scholars have adapted normative mandate theory. One such modification emphasizes the policy preferences of the median voter and the alignment of those preferences with a party’s election platform (Huber and Powell Reference Huber and Bingham Powell1994; Powell Reference Powell2000). McDonald and Budge (Reference McDonald and Budge2005) stress that the median voter has a distinctive claim to hold policy preferences that are the most democratically justifiable policy outcomes. This claim is founded in Black’s (Reference Black1958) median voter theorem, which demonstrates that the position of the median voter stands as the policy triumphing over all others in a series of pairwise competitions, satisfying various theoretically crucial conditions. For this theorem to hold, it must be possible to effectively summarize voters’ policy preferences as points somewhere on a singular ideological dimension. The position of the median voter is the position of the voter who occupies the fiftieth percentile when counting voters from either end of this continuum. This ideological alignment serves as a stabilizing force, mitigating instability and voting cycles. Moreover, as discussed previously, ideology plays a vital role in addressing the cognitive and informational deficits of voters.
An alternative defense of normative mandate theory in response to the Rikerian broadside is contained in the consensus or proportional vision of democracy. This vision broadens the set of parties deemed deserving of a substantial congruence between their election platforms and subsequent policies. Proponents of this vision contend that public policies should, to the greatest extent possible, reflect the proposals and priorities outlined in the election appeals of all parties that secured significant portions of the popular vote in preceding elections. This emphasizes the need for an equitable representation of diverse viewpoints in democratically representative policymaking (Lijphart Reference Lijphart2012; Powell Reference Powell2000). Recognizing that election outcomes seldom yield clear-cut winners and losers, this acknowledges that parties without executive authority can still command substantial electoral support. If the campaign promises of multiple parties are to be translated into tangible public policies, a prerequisite is their compatibility, or at least lack of incompatibility, ensuring that the realization of one party’s manifesto does not inherently hinder the fulfillment of others. This theoretical prerequisite for the consensus vision is at least in part supported by the fact that parties’ campaign promises are mostly either not directly related to those of other parties or essentially agree with other parties’ pledges. Only a small minority of pledges are in direct conflict (Naurin, Royed, and Thomson Reference Naurin, Royed and Thomson2019).
Discussion
The framework of promissory representation provides a lens through which we can focus on the impact of globalization on democracy. Like all lenses, it sharpens our view on certain parts of the process and blurs others. Promissory representation is particularly relevant to the impact of globalization in established democracies with developed party systems consisting of parties that issue relatively sophisticated election programs (manifestos or platforms) that can be considered blueprints for government. These are systems in which the mainstream parties are real contenders for government and feel compelled to outline policies for a comprehensive government agenda. These are also systems with relatively affluent and educated citizens, who can hold governing parties to account not only for broad economic performance but also for the policies those governments implemented to achieve a range of desirable – or undesirable – outcomes. While this boundary condition for our argument still includes a very broad range of democratic political systems and parties, this limits our inquiry to the contexts in which globalization’s effects on promise making and keeping can be observed directly.
While the scope of our argument and empirical inquiry is broad, globalization’s impact on democracy is even broader than the domain of promissory representation in established democracies. Expanding the scope of our study even more would, however, make it excessively shallow. As noted in Chapter 1, the debate on globalization covers a wide spectrum of concerns, which although important, are not part of our focus. For example, some scholars argue that globalization concentrates power in the executive branch, as government leaders and ministers play key roles in negotiating international treaties (Meyerrose Reference Meyerrose2020, Reference Meyerrose2024). This can potentially tip the balance of power away from legislatures, reducing the ability of parliaments to influence or control policy. On the other hand, more optimistic perspectives highlight that international integration can strengthen democracy by promoting the spread of international laws and norms that protect human rights and enhance accountability in countries where these safeguards may be fragile (Keohane, Macedo, and Moravcsik Reference Keohane, Macedo and Moravcsik2009). While these broader debates are noteworthy, they are tangential to what we argue is the core mechanism of modern democratic representation: that parties’ ability to make and keep promises is fundamental to maintaining a strong link between public opinion and policy outcomes.
By focusing the scope of our study on promissory representation in established democracies, we can conduct a detailed and nuanced examination of the full chain of democratic representation. In doing so, our study moves beyond individual dimensions in isolation: such as voters’ evaluations of parties, electoral behavior, governing parties’ performance in terms of promise keeping, and mainstream parties’ adjustments to the content and form of their electoral appeals. Instead, we consider the entire chain of democratic representation with a coherent analytical framework. This holistic approach allows us to examine how each component of the process interacts with others, creating a more comprehensive understanding of the system as a whole and how it is affected by globalization. These interactions involve feedback mechanisms that influence how representation evolves over time in response to globalization. Parties from different ideological traditions are likely to respond in distinct ways to globalization’s constraints, and our research design can detect such distinct responses. Our approach is also flexible enough to capture both short-term effects, such as a party’s ability to fulfill promises within a single electoral cycle, and long-term shifts in party ideologies over multiple elections. In doing so, this integrated perspective gives us a more robust understanding of how globalization is reshaping the process of accountability in established democracies.
Open access