Jurisdiction,time,and money

Abstract

The present article examines the fulfillment of election pledges. More specifically, it focuses on the effect of resources (jurisdiction, time, and money) on the parties’ abilities to fulfill their promises. Based on the case of Austria (1990–2013), the analyses cover almost 1700 pledges made in the run-up to six national elections. The multivariate models show that resources have an impact on a government’s policy outcome. The chances of pledge fulfillment decrease if the Austrian Länder is involved in the policy-making process. Furthermore, while higher levels of pledge fulfillment are associated with regularly ended legislative periods, snap elections reduce the amount of fulfilled pledges. Finally, the likelihood of pledge fulfillment increases in times of stronger economic growth. Especially pledges to expand government spending are more likely to be acted upon when the economy performs well.

Keywords

coalition governance mandate theory representative democracy pledge fulfillment Austria

Introduction

Over the past decades, more and more scholarly attention has been focused on parties’ election promises and their fulfillment during the subsequent legislative period (e.g. Naurin, 2011; Royed, 1996; Thomson, 2001; Thomson et al., 2014). These studies are relevant from a mandate view of political representation and their analyses of the factors that influence pledge fulfillment enhance the understanding of the mechanisms of representation.

From a normative perspective, mandate theory presumes a visible connection between pre- and postelection behavior (APSA, 1950; Friedrich, 1963; Ranney, 1954). The theory holds that at least two parties present distinct policy programs in the run-up to an election, voters cast their votes based on the alternatives presented, and afterward the winning party forms a government and starts implementing its program (Roberts, 2010: 38). These theoretical assumptions originate from a classical two-party system with majoritarian voting rules that enforces single-party governments with stable government majorities. Yet, in contemporary democracies, especially with multiparty systems and proportional representation, this hardly describes the situation of governing parties. Instead, they often need to negotiate their policy proposals not only with a coalition partner but also with other actors while simultaneously being constrained by both time and money.

These considerations, however, have not been a fixed component of studies on pledge fulfillment. The present article argues that resources may affect a government’s policy performance and therefore need to be included in a test of mandate theory. Resources are understood as jurisdiction (i.e. the national government’s responsibility over pledge fulfillment compared to the involvement of local governments, the European Union (EU), and other international actors in policy making), time (i.e. governments serving a full term in office compared to short-lived cabinets), and money (i.e. the country’s economic condition during the legislative period).

The present article moves beyond existing research in two ways. First, it looks at jurisdiction as a new explanatory factor of pledge fulfillment. In so doing, it develops a coding scheme for classifying pledges according to the extra-governmental actor involved in their fulfillment. Second, it reevaluates some of the most prominent hypotheses on pledge fulfillment with new empirical evidence. The test is based on coalition governments in Austria and extends the previous study on that country by including three additional cabinets (Schermann and Ennser-Jedenastik, 2014). The data covers almost 1700 promises across all policy areas addressed within the later government parties’ manifestos in the run-up to six national elections during a period of more than 20 years. To my knowledge, the following represents the most extensive single-case study on pledge fulfillment of coalition governments.

Moreover, Austria seems especially interesting when it comes to analyzing resources: With regard to jurisdiction, policy making requires collaboration at the local, the European, and other international levels. With regard to time, there are not only incidents of short-lived cabinets but there is also a change in the regular legislative period from 4 years to 5 years. Finally, with regard to money, the comparatively good economic performance makes it a least likely case in that respect and a positive finding would suggest an even stronger effect under more volatile economic conditions.

In line with previous studies (Thomson et al., 2014), the present article discusses the explanatory factors of pledge fulfillment alongside three categories, which are resources, characteristics of pledges, and institutional context.

Expectations on pledge fulfillment

Resources

In the run-up to an election, parties decide on a campaign strategy that most commonly includes making promises for the future legislative period. If elected, the fulfillment of large parts of these pledges will lie within their jurisdiction. Yet, sometimes enactment will require joint efforts by the national government and extra-governmental actors such as governments at the local level, institutions of the EU, or other international actors. If parties promise an action or outcome over which they have only limited control, then it seems most reasonable that the likelihood of pledge fulfillment will be affected negatively. This rationale rests on Tsebelis’s (1999, 2002) veto player theory that defines a veto player as an actor whose agreement is required for policy change. Tsebelis argues that as long as the actors’ policy preferences diverge, adding another veto player will increase policy stability. Hence, the involvement of extra-governmental actors will render pledge fulfillment more difficult.

Even if parties are aware that voters may use their vote not only to select future representatives but also to sanction poor policy performance (Fearon, 1999; Fiorina, 1981), the inclusion of pledges over which they have only limited control seems comprehensible. Understanding parties as rational actors, it can be assumed that they will address policies which they believe are important to the electorate. The actual fulfillment of these proposals may involve further actors either because the electorate is ignorant about different areas of responsibilities or because they expect government activity irrespective of that.

However, to date, we know little on the impact of extra-governmental players on the policy making of a national government. Empirical evidence is largely missing, not least because pledges for which fulfillment would require support by another nonnational actor were often excluded from the analyses (e.g. Royed, 1996 excluded foreign affairs policy; Thomson, 2001 looked only at socioeconomic pledges; but see Kostadinova, 2013 who looked at pledges related to Bulgaria’s EU-membership negotiations and Naurin, 2014 who controlled for EU pledges). In times of supranational cooperation between states, especially between EU member states, these types of pledges may become both more relevant and frequent and therefore, excluding them could lead toward a bias in the results. Moreover, most of the literature on pledge fulfillment is explicitly based upon mandate theory which states that parties are elected because of their self-stated policy program (e.g. Roberts, 2010: 38). Consequently, their performance should be judged against this agenda. If a party promises to introduce a European Currency Transaction Tax, the fulfillment of this measure should be included in the evaluation of mandate responsiveness.

H1: A pledge is less likely to be fulfilled if its fulfillment requires joint efforts by extra-governmental actors. Extra-governmental actors are understood as local governments, the EU, and other international actors.

Even though only a small number of pledges are directly related toward each other (Costello and Thomson, 2008: 253; Mansergh and Thomson, 2007: 315; Royed, 1996: 66), they all take away from the same scarce resource time. Neither cabinet, nor bureaucracies, nor parliament may deal simultaneously with all policy proposals. Consequently, coalition partners have to agree upon an agenda according to their preferences, which then may be completed one after the other. In case of a snap election, however, the envisaged time to do so is shortened. As a result, policy programs further down the list would be left undealt with. H2 therefore reads as follows.

H2: A pledge is more likely to be fulfilled if a government serves a full term in office.

The saliency approach of testing the mandate theory compares issue emphasis in manifestos with financial emphasis in corresponding state budgets and thus builds on the simplifying assumption that policy making requires money (e.g. Klingemann et al., 1994). This being the case, a greater financial leeway should enable parties to act upon their policy proposals more freely. Since financial leeway is mainly determined by a country’s economic development (Huber and Stephens, 2001; Huber et al., 2008), H3 postulates a positive relationship between economic growth and pledge fulfillment.

H3: A pledge is more likely to be fulfilled the better the economy performs.

Characteristics of pledges

Parties’ abilities to follow through on their mandates might not only be influenced by resources but also by the nature of pledges themselves. The following discussion focuses on the most probable prominent characteristic of a pledge: the type of change promised. Based on Royed (1996; also Royed and Borrelli, 1997, 1999), researchers of the Comparative Party Pledges Group (Thomson et al., 2014) have, on the one hand, differentiated between pledges to maintain the status quo and, on the other, pledges to alter government spending or tax policies as well as proposals to initiate any other, nonbudget-related change.

There is a broad consensus among scholars that status quo pledges will most likely be fulfilled (Costello and Thomson, 2008: 250; Kostadinova, 2013: 12; Mansergh and Thomson, 2007: 319; Royed, 1996: 61–62). Again, Tsebelis’s (1995, 2002) veto player theory supports this position. Similar to the veto power of extra-governmental actors, coalition partners themselves hold the position of a veto player. If a party has promised to keep the current policy, then it already possesses the power to enact its pledge by vetoing on any unwanted reform proposal.

H4: A pledge is less likely to be fulfilled if it promises a reform. A reform is understood as a proposal to alter government spending or tax policies as well as a proposal to initiate any other, nonbudget-related change.

H3 expects higher rates of pledge fulfillment in times of stronger economic performance due to a greater financial leeway. However, based on the distinction between different types of pledges proposed by Royed (1996; also Royed and Borrelli, 1997, 1999) and the comparative literature on pledge fulfillment (Thomson et al., 2014), the following two state measures seem to have the greatest impact on a government’s budget: an increase on the expenditure side and a decrease on the revenue side. While the former can be associated with an expansion of the welfare state, that is, the state takes on more tasks such as rising child care benefits or lowering the retirement age, the latter manifests itself through tax cuts, that is, people contribute less to the state’s budget due to, for example, lower income or sales taxes.

Therefore, the following hypotheses refine H3 and state that these costly promises are fulfilled more easily in times of economic growth.

H3a: A pledge to expand spending is more likely to be fulfilled, the better the economy performs.

H3b: A pledge to cut taxes is more likely to be fulfilled, the better the economy performs.

Institutional context

Under the assumption that politicians act as if they were perfect agents to their parties, the next two hypotheses transfer descriptive ideal models about the internal working of governments to expectations on pledge fulfillment. In the prime ministerial model, the prime minister plays a distinctive role and influences policy decision making according to his or her own preferences (Dunleavy and Rhodes, 1990: 5–8). If the prime minister holds a privileged position in terms of a government’s policy-making, then this should be seen in the party’s ability to follow through on its pledges compared to its coalition partners. This reasoning is supported by models of government formation (Austen-Smith and Banks, 1988; Baron, 1991; Diermeier and Feddersen, 1998). The party of the chief executive has the power to initiate coalition talks and, therefore, benefits from a first-mover advantage when deciding on its negotiating partners. Furthermore, the largest party usually claims to hold prime ministership and higher fulfillment rates by that actor would accord with Gamson’s law (1961) on proportional distribution of (office) government payoffs.

However, in the ministerial government or fragmented government model (Andeweg, 1997), executive power is dispersed among the cabinet ministers, which enables them to exercise controlling influence over their own jurisdiction. Laver and Shepsle’s (1990, 1994, 1996) model on government formation builds on this assumption and adopts such a division-of-labor system. If the portfolio allocation is an essential mechanism to influence policy decisions, then ministerial control should favor pledge fulfillment within that area (Mansergh and Thomson, 2007; Schermann and Ennser-Jedenastik, 2014; Thomson, 2001; but see Costello and Thomson, 2008). Having an awareness of potential agency loss over the coalition partner’s portfolios, however, parties agree on several control mechanisms. In order to keep track of policy making within the most important ministries, they often appoint junior ministers, who belong to the other government party than the respective minister (Falcó-Gimeno, 2014; Greene and Jensen, 2014; Lipsmeyer and Pierce, 2011; Thies, 2001). The appointment of a watchdog junior minister should equalize an effect of portfolio allocation.

H5: A pledge is more likely to be fulfilled if the pledge-making party holds the prime ministership.

H6: A pledge is more likely to be fulfilled if the pledge-making party exclusively controls the corresponding portfolio.

The nomination of junior ministers is one of several control mechanisms that parties have at their disposal in order to minimize the risk of potential agency loss over the coalition partner’s domains (e.g. Strøm et al., 2006). Another instrument, which is also very common in Western European systems, is a written coalition agreement (Müller and Strøm, 2000, 2008; Strøm and Müller, 1999). Within these “most authoritative documents” (Müller and Strøm, 2000: 18) parties specify broad policy guidelines, working procedures, or certain policy measures that represent a coalition’s common ground of future governance. Thus, it is most reasonable to assume that pledges that are included in these agreements are much easier to implement than other policy proposals (also see Schermann and Ennser-Jedenastik, 2014; Thomson, 2001).

H7: A pledge is more likely to be fulfilled if it is included in the coalition agreement.

The case of Austria and data on pledge fulfillment

The article is based upon pledge fulfillment in six Austrian coalition governments that have been in office between 1990 and 2013.¹ Since the political dominance of the Social Democratic Party (SPÖ) and the People’s Party (ÖVP) was weakened considerably by the end of the 1980s, the country can be perceived as a typical case among Western European democracies with multiparty systems and coalition governments.

Austria provides variation on both government composition and the balance of power between the actors involved. After the elections in 1990 and 1995 as well as after the 2006 and 2008 elections, Austria was governed by the so-called “grand coalitions” between the SPÖ and the ÖVP. The landslide election of 1999 paved the way for two center-right coalitions between the ÖVP and the Freedom Party (FPÖ). The case of the FPÖ (2000–2003) seems especially interesting, since the populist right-wing party entered government after a long period of opposition and with only little government experience. Of the six cabinets analyzed, three bound together two parties of almost equal electoral strength (two SPÖ–ÖVP cabinets and one ÖVP–FPÖ cabinet).

Similar to Royed (1996: 79) and Thomson (2001: 180), I understand a pledge as a party’s commitment to a future policy plan that is written down in the election manifesto and its fulfillment in the course of the upcoming legislative period is objectively testable with a criteria provided by the party itself. This definition includes promises to act (“action pledges”) as well as promises to achieve an outcome (“outcome pledges”). For example, if a party pledges to set a higher legal retirement age, then it announces a specific action planned for the next legislative period (i.e. the implementation of a pension reform act). If a party promises to reduce youth unemployment, then it signals a certain outcome of its political performance. In order to fulfill its promise, the party may then choose freely between different policy measures and must hope that they lead to the desired result. Both action and outcome pledges are also relevant in the comparative research endeavors (e.g. Thomson et al., 2014).

The data collection was carried out under the auspices of the Austrian National Election Study (AUTNES, www.autnes.at). More precisely, the identification of pledges was part of a quantitative content analysis of election manifestos. In order to code these texts, the AUTNES team split all sentences into smaller statements based on Noam Chomsky’s (1957) phrase structure model (for a detailed discussion see Dolezal et al., 2015). Building up on these units, researchers coded relevant information such as the policy issue addressed, the party’s position toward this issue, or references of another (political) actor (Dolezal et al., 2015). Although the identification of pledges requires the reading of the original text, the actual coding of the pledges was done at the level of these statements as well.²

Having identified all pledges within the election manifestos, a list of different pledges per year was created³ and the fulfillment of each item on that list was evaluated. The method to evaluate a pledge depends on its type. Action promises are rated by the action itself, outcome pledges are tested upon the development of the relevant indicator. On the basis of the examples described above, if a party promises a higher legal retirement age, a coder checks whether a retirement reform act that rises the minimum retirement age passed parliament. If a party promises to reduce youth unemployment, the evaluation is based on a comparison between the actual rate of youth unemployment at the beginning of the legislative period and the average rate during a government’s term in office.

Table 1 gives some examples of pledges written down in the 2008 manifestos together with the coders’ evaluation. Enactment was assessed on a three-stage scale that distinguishes between fully, partly, and not fulfilled promises. A pledge is fully fulfilled if the party enacted the relevant measure (like the ÖVP did by implementing the Tax Reform Act) or if the desired outcome was reached to the extent promised (like the reduction of the pay gap in line with the SPÖ’s pledge). Again, evaluation is based on the wording of the pledge in the manifesto. Since the SPÖ only hold out the prospect of a reduction of the pay gap without specifying an exact amount, the pledge was fully fulfilled due to a decrease in the relevant statistic. Pledges are only partly fulfilled if the result falls behind the promised output or outcome. Note that the ÖVP’s pledge on the research and development budget was only partly fulfilled because the budgetary growth remained lower than the 3% promised. The same logic holds true for the SPÖ’s pledge to introduce a month-long paid paternity leave for all employees. This approach not only simplifies the coding process but also satisfies the assumption of mandate theory that parties must be judged upon the program they have set for themselves. Finally, pledges are not fulfilled if no action was taken or if the relevant statistic did not change accordingly. For example, neither the compulsory education age was lowered (SPÖ’s pledge) nor did the government announce a family allowance (ÖVP’s pledge). The main data sources were the legislative database of the Federal Chancellery, newspaper articles in media archives, and official databases such as Statistics Austria or Eurostat.

Table 1.

Examples of pledges and their fulfillment, 2008–2013.

Party	Pledge	Coding decision: explanation (data source)
SPÖ	Reduction of the pay gap between men and women	FF: Compared to 2008, the differences in income between full-time employed men and women decreased during the legislative period (Statistics Austria).
ÖVP	Tax deductibility of childcare costs	FF: Tax deductibility of childcare costs was introduced by the Tax Reform Act 2009 (Steuerreformgesetz 2009; legislative database of the Federal Chancellery).
SPÖ	Month-long paid paternity leave	PF: Only in public service, a month-long and paid paternity leave was introduced by the Budget Accompanying Act 2011 (Budgetbegleitgesetz 2011; legislative database of the Federal Chancellery).
ÖVP	Increase in the budget for research and development (3% of GDP)	PF: The increase in the budget for research and development falls below the reference value of 3% (Statistics Austria).
SPÖ	Compulsory education from the age of 5	NF: The compulsory education starting age remained at 6 years. No relevant amendment of the Compulsory Education Law (Schulpflichtgesetz).
ÖVP	Introduction of a family allowance	NF: Family allowance was not introduced. No relevant amendment of the Income Tax Act (Einkommenssteuergesetz).

SPÖ: Social Democratic Party; ÖVP: People’s Party; FF: fully fulfilled; PF: partly fulfilled; NF: not fulfilled; GDP: gross domestic product.

The coding process of the dependent variable resulted in a list of, on average, 153 different and objectively testable pledges by party and year. On average, half of these were fulfilled at least partially (Table 2). This performance is largely in line with the literature on pledge fulfillment in coalitions (Kostadinova, 2013; Thomson, 2001) and scores well below the results of single-party governments (Artés, 2013; Artés and Bustos, 2008; Naurin, 2011; Royed, 1996).

Table 2.

Pledge fulfillment by period and party in percent (N = 1696).^a

	Faymann I (2008–2013)		Gusenbauer (2007–2008)		Schüssel II (2003–2007)
	SPÖ (N = 146)	ÖVP (N = 70)	SPÖ (N = 139)	ÖVP (N = 177)	ÖVP (N = 288)	FPÖ (N = 223)
Fully fulfilled	48.0	52.9	36.0	47.5	54.2	45.7
Partly fulfilled	7.5	7.1	3.6	6.2	5.9	6.7
Not fulfilled	44.5	40.0	60.4	46.3	39.9	47.5
Total	100	100	100	100	100	100
	Schüssel I (2000–2003)		Vranitzky V \| Klima (1996–2000)		Vranitzky III (1990–1994)
	ÖVP (N = 249)	FPÖ (N = 178)	SPÖ (N = 19)	ÖVP (N = 70)	SPÖ (N = 152)	ÖVP (N = 122)
Fully fulfilled	45.0	19.7	57.9	60.0	48.0	41.0
Partly fulfilled	3.6	5.1	15.8	11.4	8.6	9.8
Not fulfilled	51.4	75.3	26.3	28.6	43.4	49.2
Total	100	100	100	100	100	100

SPÖ: Social Democratic Party; ÖVP: People’s Party; FPÖ: Freedom Party.

^aDeviations from 100% possible due to rounding.

The lowest level of pledge fulfillment is shown by the FPÖ (2000–2003). Most obviously, the cabinet was terminated early; however, the score is considerably lower than the result of the FPÖ’s coalition partner. This comparatively weak performance in terms of pledge fulfillment may be explained by the party’s populist vote-maximization strategy during the electoral campaign, a high personnel turnover (four ministers resigned during the government’s first 2 years), and a limited number of FPÖ partisans among civil servants (Luther 2010: 88–89).

In order to test the hypotheses, the following independent variables were generated:

The categorical variable jurisdiction distinguishes between the actors necessary in order to fulfill a pledge. The baseline category are pledges that fall under the sole jurisdiction of the national government (and the parliament). Then, there are pledges that need some action of the federal states. Policy proposals within that category may either lie in the exclusive jurisdiction of the Austrian Länder (e.g. housing subsidy, regional policy, waste management, kindergarten, and nature and landscape protection) or require joint work by both national and local actors (electricity, hospitals, maternity protection, and child protection). The next category subsumes pledges that touch the context of EU. These are pledges that fall under the exclusive area of responsibility of the EU (e.g. customs union, common commercial policy, and monetary policy for euro area states) as well as proposals in which parties explicitly demand European efforts or hold out the prospect of results at the EU level. The last category depicts all pledges for which fulfillment requires efforts of international actors such as other national governments or international organizations.

The variable full term in office identifies governments that served until the regular end of their term. The economic performance is measured as the average growth in the gross domestic product (GDP growth) during the legislative period (OECD, 2014).

Another categorical variable describes the type of change promised. A pledge may contain a promise to maintain the status quo (including pledges to review the status quo), to cut spending (including pledges to increase taxes), to expand spending, to cut taxes, or to introduce other change, that is, a reform in any other way as just discussed (Royed, 1996; Royed and Borrelli, 1997, 1999).

The variable chancellery identifies all pledges made by the party that holds the chancellery. Another dichotomous variable indicates promises for which a party solely controls the relevant portfolio, that is, portfolios without a junior minister who belongs to the other governing party, then the corresponding minister (portfolio, no watchdog). Then, the variable coalition agreement distinguishes between pledges that are at least partially written down in the coalition agreement and other pledges. A pledge was coded as “at least partially written down in the agreement” if either the pledge was mentioned in the agreement or if the pledge was mentioned but the wording fell short of the promised action or outcome (e.g. tax cut by 2% instead of 5%). Already the incomplete mention of a pledge is understood as a negotiation success and should therefore favor pledge fulfillment.

Finally, the models control for the types of government that are tested in the analyses. The variable grand coalition distinguishes SPÖ–ÖVP coalitions from ÖVP–FPÖ coalitions.

Factors explaining pledge fulfillment

The following analyses are based upon four binary logistic regressions on pledge fulfillment. While the first three models in Table 3 each focus on one explanatory category each (resources, characteristics of pledges, and institutional context), the last model simultaneously tests all influencing factors. In each design, the dependent variable differentiates only between promises that were at least partially fulfilled and those that were unfulfilled.⁴

Table 3.

Binary logit models: Explanatory factors of pledge fulfillment in Austria, 1990–2013.^a

	Model I Resources		Model II Characteristics of pledges		Model III Institutional context		Model IV Full model
Jurisdiction: National government	Reference category						Reference category
Federal states	0.616**	(−2.35)					0.604**	(−2.27)
European Union	0.658**	(−2.09)					0.711	(−1.48)
International actors	0.475*	(−1.94)					0.733	(−0.77)
Full term in office	1.922***	(6.31)					2.115***	(6.51)
GDP growth	1.176**	(2.07)					1.099	(0.86)
Type of change: Status quo/review			Reference category				Reference category
Cut spending/increase taxes			0.169***	(−7.27)			0.145***	(−7.47)
Expand spending			0.192***	(−9.32)			0.164***	(−9.80)
Cut taxes			0.080***	(−9.26)			0.070***	(−9.03)
Other change			0.111***	(−12.46)			0.106***	(−12.18)
Expand spending × GDP growth							1.408*	(1.92)
Cut taxes × GDP growth							0.699	(−1.07)
Chancellery					1.264**	(2.27)	1.202	(1.60)
Portfolio, no watchdog					1.313***	(2.68)	1.282**	(2.23)
Coalition agreement					2.141***	(7.61)	2.198***	(7.21)
Grand coalition	1.199*	(1.77)	1.424***	(3.38)	1.166	(1.53)	1.291**	(2.24)
N	1696		1696		1696		1696
Pseudo R ²	0.024		0.094		0.035		0.147
Cases correctly predicted	58.4%		63.8%		59.8%		69.9%
Difference to modal category (50.5%)	7.9 pp		13.3 pp		9.3 pp		19.4 pp

pp: percentage points; GDP: gross domestic product.

^aFigures represent odds ratios (z-statistics in parentheses). Due to multicollinearity between the interaction effects and GDP, I used a centered GDP growth variable in model IV. Please note that the interpretation changes accordingly.

*p < 0.1; **p < 0.05; ***p < 0.01.

The results show that resources have explanatory power of pledge fulfillment.

H1 expects lower pledge fulfillment in cases that involve extra-governmental actors. In order to test this hypothesis, pledges that fall under the jurisdiction of the National government are tested against pledges that require efforts by the federal states, the EU, or international actors. As hypothesized, all coefficients take on values smaller than 1 and thus, pledge fulfillment becomes less likely. However, in the full model, only pledges that involve the Austrian Länder are significantly less likely to be acted upon. Compared to the other factors analyzed, the corresponding effect size is quite substantial. The coefficient indicates a decreased probability of at least partial pledge fulfillment by 0.13 (from 0.52 to 0.39; Figure 1). The insignificant results of both European and international actors seem surprising. Of the 116 pledges that require joint efforts by the EU, 42% are fulfilled, and of the 30 international pledges, only 33% are acted upon (at least partially). Both rates are considerably lower than the overall pledge fulfillment rate of more than half of the pledges. Especially with regard to international pledges, the insignificant result might be additionally driven by the unevenly distributed variable. Overall, there is some evidence in support of H1.

Figure 1.

Predictions on (partial) pledge fulfillment. Predictions are based on model IV while holding everything else constant (mean values). Vertical lines represent 95% confidence intervals that were calculated by the delta method. All differences within a variable are significantly different from zero (see Long and Freese, 2006: 95 for detailed information on calculating predictions and testing their difference).

H2 postulates that parties fulfill more promises if the government serves a full term in office. The results in both the simple and the combined model support this claim. All else being equal, the odds of pledge fulfillment increase by a factor of 2.1 (model IV) if the legislative period ends regularly. Simply put, the predicted probability of at least partial pledge fulfillment in times of early termination is estimated to be 0.41 compared to 0.59 in times of full terms (Figure 1).

The test of H2 rests on the comparison of pledge fulfillment between the two short-lived cabinets, Schüssel I (2000–3) and Gusenbauer (2007–8), with all other coalitions in the sample. Both Schüssel I and Gusenbauer were terminated after only 37 and 23 months, respectively, and the governing parties were left with very little time to fulfill their programs. Thus, this finding seems very plausible and is in line with the existing studies on short-lived cabinets in Italy (Moury, 2011) and Ireland (Thomson et al., 2014). From an institutional perspective, however, the result raises the question on the legislative cycle per se. That is whether serving for 4 years, or 5 years, or even 6 years effects pledge fulfillment. Austria provides an interesting test case to tackle this issue. The regular term in office was increased from 4 years to 5 years in 2007, and therefore, the cabinet Faymann I (2008–13) had one additional year to fulfill its promises. A simple chi-square test between pledge fulfillment by all 4-year cabinets and Faymann I shows that there is no significant change in parties’ abilities to fulfill their pledges (χ ² (1, N = 997) = 0.32, p = 0.57). An extra year at the end of a legislative term had no effect on the policy outcome in terms of pledge fulfillment.

The variable GDP growth in model I highlights a positive and statistically significant impact on pledge fulfillment. This indicates that pledges are more likely to be fulfilled in times of better economic conditions. The support of the corresponding H3 remains valid in a full model without the interaction effects (results not shown). However, the theoretical section concluded that not all policy proposals are equally costly and the significant interaction effect between pledges to expand spending and GDP growth supports this claim (Model IV). Figure 2 plots the probability of at least partial pledge fulfillment depending on the economic state. For example, at 2.4 percentage points less growth than during the investigation period, the probability of (partial) fulfillment of pledges to expand spending lies at 0.10 (95% confidence interval 0.035, 0.166). At 2.4 percentage points more growth than the overall average level, this figure increases up to 0.47 (95% confidence interval 0.289, 0.653). H3a is supported by the model. The same does not hold true for H3b. Pledges promising to cut taxes are fulfilled to a lower extent and a positive economic performance does not boost pledge fulfillment within that area (Cut taxes × GDP growth).

Figure 2.

Predictions on (partial) pledge fulfillment related to GDP growth. Predictions are based on model IV while holding everything else constant (mean values). Grey lines represent 95% confidence intervals. An x value of zero denotes an average economic growth during the legislative period that corresponds to the average growth rate over all periods analyzed. The x values vary from two standard deviations below to two standard deviations above the average growth rate over all periods analyzed. Plotted differences are significantly different from zero (see Long and Freese, 2006: 95 for detailed information on calculating predictions and testing their difference).

In line with previous studies (e.g. Costello and Thomson, 2008: 250; Kostadinova, 2013: 12; Mansergh and Thomson, 2007: 319; Royed, 1996: 61–62; Thomson et al., 2014: 24), the results show that not only promises to expand spending or cut taxes but also promises to cut spending/increase taxes or to introduce any other change are significantly less likely to be fulfilled than promises to maintain or review the status quo. This supports H4.

Up to now, scholars’ discussions would stop with the conclusion that status quo pledges have a good chance of being acted upon. Clearly, the identification of such a strong and statistically significant predictor is valid in itself. However, without further evaluation, this might contribute less to a deeper understanding of political representation than it might cast doubt on the causal claims in research designs based on electoral promises. It is almost impossible to know whether a government party managed to fulfill its status quo pledge because it acted on behalf of the mandate or because the preferred result has not been challenged by someone else. Moreover, a high percentage of fulfilled status quo pledges made by opposition parties even points to the second reasoning (Praprotnik, 2015).

In order to better understand this finding in the light of mandate theory, I propose to look at conflictual pledges or, more specifically, at situations in which one party promises to keep the current policy and the other party promises a reform. These pledges are in direct disagreement between the coalition partners and, consequently, only one party can push through on its proposal. Since actors in multiparty systems with proportional electoral rules are generally very reluctant in making conflictual promises (Costello and Thomson, 2008: 253; Håkansson and Naurin, 2014: 8; Mansergh and Thomson, 2007: 315; Royed, 1996: 66), there are only a few examples of this kind. Yet, these cases provide a quite clear picture.⁵ For example, one prominently debated topic within the Austrian context is the question of tuition fees. While Austrian right-wing parties usually support tuition fees in order to finance tertiary education, left-wing parties oppose this idea. In the run-up to the 1990 election, the ÖVP pledged to implement tuition fees that have been abolished under a social democratic single-party government in the 1970s. Even though the party entered a SPÖ–ÖVP government after the election and fulfilled large parts of its program during the subsequent period (see Table 2), it failed to act on this specific promise. The SPÖ used its veto power and hindered a reform of the preferred status quo. After the election of 1999, the ÖVP formed a coalition with the right-wing FPÖ and implemented tuition fees. Being in opposition, the SPÖ could not block the government’s proposal. In 2006, SPÖ and ÖVP renewed their collaboration in government. In the following term, the SPÖ failed to abolish tuition fees that were now part of the status quo. Finally, tuition fees were abolished only when the grand coalition collapsed in 2008 and parliament took a vote without coalition pressure.

This empirical example indicates that the predictive power of the status quo indeed tells us something in the light of mandate theory. Whenever challenged, a ruling party uses its veto power in order to protect a desired state. Opposition parties, on the other hand, fail to secure the current situation.

H5 expects a positive effect of holding the chancellery on pledge fulfillment. Contrary to previous studies (e.g. Thomson et al., 2014), the hypothesis is not borne out by the data. While the coefficient is statistically significant in model III, the effect disappears in the full model (see also Table 2 which shows that in half of the coalitions the junior partner actually fulfilled a higher share of its mandate).⁶

If the characteristics of the Austrian party system are taken into account, the nonfinding appears less surprising. The SPÖ who traditionally ruled out a coalition with the FPÖ had no real first-mover advantage against the pivotal party ÖVP (Müller and Fallend, 2004: 810). Entering a coalition with the ÖVP was its only chance to form a majority two-party government. Since Austria has almost no experience with either single-party minority or multiparty cabinets, these alternatives were never seriously considered. Moreover, three of the six elections analyzed resulted in coalitions between partners of almost equal electoral strength (a difference of less than 5 percentage points). The fact that none of the often equally strong coalition partners were able to push through significantly more pledges is actually in line with Gamson’s (1961) law of proportional payoffs.

While there is no significant effect of the chancellery, having control over the portfolio seems to influence pledge fulfillment (portfolio, no watchdog). If a party solely controls a ministry, the odds of pledge fulfillment increase by a factor of 1.3, keeping all other variables constant (model IV). Or, more intuitively, the probability of at least partial pledge fulfillment increases from 0.48 to 0.55 if the pledge-making party holds the relevant portfolio (Figure 1). H6 is supported.

Furthermore, the analyses corroborate H7. All else being equal, being part of the coalition agreement increases the odds of pledge fulfillment by 2.2 (model IV). Again, Figure 1 shows this result in terms of predicted probabilities. Partial pledge fulfillment increases from 0.41 to 0.61 if the proposal is part of the agreement.

The significant effects of both the portfolio, no watchdog and the coalition agreement variables highlight the importance of the coalition formation process. The distribution of office as well as policy payoffs largely influences the policy outcome of a coalition (similar conclusions are reached by Mansergh and Thomson, 2007; Schermann and Ennser-Jedenastik, 2014; Thomson, 2001).

Conclusion

The present study on mandate responsiveness set out with two objectives. On the one hand, it aimed to stimulate research in this field by focusing on resources and proposing jurisdiction as a new explanatory factor. On the other hand, it aimed to reevaluate established hypotheses by presenting new empirical evidence on coalition governance.

The analyses show that resources have an impact on a government’s policy outcome. The results indicate that the involvement of local governments in the policy-making process hinders pledge fulfillment. Moreover, while full terms are associated with higher levels of pledge fulfillment, a comparison between pledge fulfillment in 4-year- and 5-year-long periods revealed no significant difference. In times of stronger economic growth, parties fulfill more pledges in general and more pledges to expand government spending specifically. Due to the country’s relatively good performance throughout the investigation period (and therefore little variation on the independent variable), Austria must be perceived as a least likely case in that respect. The fact that the hypothesis still holds suggests that this result might be even more pronounced when variation in economic performance increases.

Most of the previous conclusions regarding institutional setting and characteristics of a pledge remained valid when including resources into the models. Yet, contrary to the results from other countries, there was no significant difference between the party holding the chancellery and its coalition partner. The negative finding might be due to the fact that viable coalition options were limited and the difference in electoral strength between the government parties was generally small. Furthermore, looking at conflictual situations revealed that governing parties do use their veto power in order to prevent an unwanted reform. Since conflictual pledges are a rare phenomenon, further research is needed to support this conclusion.

Studies on pledge fulfillment expand our knowledge of the congruence between policy programs and policy outcomes as well as of the mechanisms on policy making in various political systems. Nevertheless, they focus only on one aspect of political representation and are based on the simplifying assumption that higher levels of pledge fulfillment indicate better representation. Obviously, there are other ways in which politicians may act responsively and defection from a pledge does not necessarily mean defection from the mandate (Stokes, 2001). Addressing these issues, future studies could understand mandate responsiveness not only in terms of pledge fulfillment but also in terms of actions in line with a mandate. Especially when evaluating pledge fulfillment by opposition parties (Artés and Bustos, 2008; Kostadinova, 2013; Praprotnik, 2015; Royed, 1996; Thomson, 2001), it seems plausible to take politicians’ behavior in general into account (e.g. legislative initiatives or voting behavior in parliament).

Footnotes

Acknowledgements

The author would like to thank Andreas Frössel, Anita Bodlos, Laurenz Ennser-Jedenastik, Matthias Kaltenegger, Vera Pober, Bernd Schumenjak and Daniel Strobl for their collaboration during the data-generating process. The author would also like to thank the anonymous reviewers for their helpful comments.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: 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); and the project “Political Economy of Reforms” (SFB 884: C1) financed by the German Research Foundation (DFG).

Notes

References

Andeweg

(1997) Collegiality and collectivity: cabinets, cabinet committees and cabinet ministers. In: Weller

Bakvis

Rhodes

RAW

(eds) The Hollow Crown. Countervailing Trends in Core Executives. London: Macmillan, pp. 58–83.

Artés

(2013) Do Spanish politicians keep their promises? Party Politics 19(1): 143–158.

Artés

Bustos

(2008) Electoral promises and minority governments: an empirical study. European Journal of Political Research 47(3): 307–333.

Austen-Smith

Banks

(1988) Elections, coalitions, and legislative outcomes. American Political Science Review 82(2): 405–422.

Baron

(1991) A spatial bargaining theory of government formation in parliamentary systems. American Political Science Review 85(1): 137–164.

Brant

(1990) Assessing proportionality in the proportional odds model for ordinal logistic regression. Biometrics 46(4): 1171–1178.

Chomsky

(1957) Syntactic Structures. The Hague: Mouton & Co.

Committee on Political Parties American Political Science Association (APSA) (1950) Toward a More Responsible Two-Party System. New York and Toronto: Rinehart & Company, Inc.

Costello

Thomson

(2008) Election pledges and their enactment in coalition governments: a comparative analysis of Ireland. Journal of Elections, Public Opinion and Parties 18(3): 239–256.

10.

Diermeier

Feddersen

(1998) Cohesion in legislatures and the vote of confidence procedure. The American Political Science Review 92(3): 611–621.

11.

Dolezal

Ennser-Jedenastik

Müller

. (2015) Analyzing manifestos in their electoral context: a new approach applied to Austria, 2002-2008. Political Science Research and Methods 4(3): 641–650, DOI: 10.1017/psrm.2015.38

12.

Dunleavy

Rhodes

RAW

(1990) Core executive studies in Britain. Public Administration 68(1): 3–28.

13.

Falcó-Gimeno

(2014) The use of control mechanisms in coalition governments: the role of preference tangentiality and repeated interactions. Party Politics 20(3): 341–356.

14.

Fearon

(1999) Electoral accountability and the control of politicians: selecting good types versus sanctioning poor performance. In: Przeworski

Stokes

Manin

(eds) Democracy, Accountability, and Representation. Cambridge: Cambridge University Press, pp. 55–97.

15.

Fiorina

(1981) Retrospective Voting in American National Elections. New Haven and London: Yale University Press.

16.

Friedrich

(1963) Man and His Government: An Empirical Theory of Politics. New York: McGraw-Hill.

17.

Gamson

(1961) A theory of coalition formation. American Sociological Review 26(3): 373–382.

18.

Greene

Jensen

(2014) Manifestos, salience and junior ministerial appointments. Party Politics 22(3): 382–392, DOI: 10.1177/1354068814549336.

19.

Håkansson

Naurin

(2014) Promising ever more: an empirical account of Swedish parties’ pledge making during 20 years. Party Politics 22(3): 393–404 DOI: 10.1177/1354068814549338.

20.

Huber

Mustillo

Stephens

(2008) Politics and social spending in Latin America. The Journal of Politics 70(2): 420–436.

21.

Huber

Stephens

(2001) Development and Crisis of the Welfare State. Chicago: The University of Chicago Press.

22.

Klingemann

H-D

Hofferbert

Budge

(1994) Parties, Policies, and Democracy. Boulder: Westview Press.

23.

Kostadinova

(2013) Democratic performance in post-communist Bulgaria: election pledges and levels of fulfillment, 1997–2005. East European Politics 29(2): 190–207.

24.

Laver

Shepsle

(1990) Coalitions and cabinet government. American Political Science Review 84(3): 873–890.

25.

Laver

Shepsle

(1994) Cabinet government in theoretical perspective. In: Laver

Shepsle

(eds) Cabinet Ministers and Parliamentary Government. Cambridge: Cambridge University Press, pp. 285–309.

26.

Laver

Shepsle

(1996) Making and Breaking Governments. Cabinets and Legislatures in Parliamentary Democracies. Cambridge: Cambridge University Press.

27.

Lipsmeyer

Pierce

(2011) The eyes that bind: junior ministers as oversight mechanisms in coalition governments. The Journal of Politics 73(4): 1152–1164.

28.

Long

Freese

(2006) Regression Models for Categorical Dependent Variables Using Stata. College Station, TX: Stata Press.

29.

Luther

(2010) Governing with right-wing populists and managing the consequences: Schüssel and the FPÖ. In: Bischof

Plasser

(eds) The Schüssel Era in Austria. New Orleans: University of New Orleans Press, pp. 79–103.

30.

Mansergh

Thomson

(2007) Election pledges, party competition, and policymaking. Comparative Politics 39(3): 311–329.

31.

Moury

(2011) Italian coalitions and electoral promises: assessing the democratic performance of the Prodi I and Berlusconi II governments. Modern Italy 16(1): 35–50.

32.

Müller

Fallend

(2004) Changing patterns of party competition in Austria: from multipolar to bipolar system. West European Politics 27(5): 801–835.

33.

Müller

Strøm

(2000) Coalition governance in Western Europe: an introduction. In: Müller

Strøm

(eds) Coalition Governments in Western Europe. New York: Oxford University Press, pp. 1–31.

34.

Müller

Strøm

(2008) Coalition agreements and cabinet governance. In: Strøm

Müller

Bergman

(eds) Cabinets and Coalition Bargaining. The Democratic Life Cycle in Western Europe. Oxford: Oxford University Press, pp. 159–199.

35.

Naurin

(2011) Election promises, Party Behaviour and Voter Perceptions. Basingstoke: Palgrave Macmillan.

36.

Naurin

(2014) Is a promise a promise? Election pledge fulfilment in comparative perspective using Sweden as an example. West European Politics 37(5): 1046–1064.

37.

OECD (2014) Aggregate National Accounts: Gross domestic product. OECD National Accounts Statistics (database). Available at: https://dx-doi-org.web.bisu.edu.cn/10.1787/data-00001-en (accessed 14 April 2015).

38.

Praprotnik

(2015) Die Policy-Bilanz der Regierung Faymann I. Eine Studie zur Anwendbarkeit der Mandatstheorie. Österreichische Zeitschrift für Politikwissenschaft 44(2): 1–14.

39.

Ranney

(1954) The Doctrine of Responsible Party Government: Its Origins and Present State. Urbana: The University of Illinois Press.

40.

Roberts

(2010) The Quality of Democracy in Eastern Europe. Public Preferences and Policy Reforms. New York: Cambridge University Press.

41.

Royed

(1996) Testing the mandate model in Britain and the united states: evidence from the reagan and thatcher eras. British Journal of Political Science 26(1): 45–80.

42.

Royed

Borrelli

(1997) Political parties and public policy: social welfare policy from Carter to Bush. Polity 29(4): 539–563.

43.

Royed

Borrelli

(1999) Parties and the economic policy in the USA. Party Politics 5(1): 115–127.

44.

Schermann

Ennser-Jedenastik

(2014) Coalition policy-making under constraints: examining the role of preferences and institutions. West European Politics 37(3): 564–583.

45.

Stokes

(2001) Mandates and Democracy. Neoliberalism by Surprise in Latin America. New York: Cambridge University Press.

46.

Strøm

Müller

(1999) The keys to togetherness: coalition agreements in parliamentary democracies. Journal of Legislative Studies 5(3): 255–282.

47.

Strøm

Müller

Bergman

(2006) Delegation and Accountability in Parliamentary Democracies. Oxford: Oxford University Press.

48.

Thies

(2001) Keeping tabs on partners: the logic of delegation in coalition governments. American Journal of Political Science 45(3): 580–598.

49.

Thomson

(2001) The programme to policy linkage: the fulfilment of election pledges on socio-economic policy in the Netherlands, 1986–1998. European Journal of Political Research 40(2): 171–197.

50.

Thomson

Royed

Naurin

. (2014) Explaining the fulfilment of election pledges: a comparative study on the impact of government institutions. Annual Meeting of the American Political Science Association. Washington, D.C.

51.

Tsebelis

(1995) Decision making in political systems: veto players in presidentialism, parliamentarism, multicameralism and multipartyism. British Journal of Political Science 25(3): 289–325.

52.

Tsebelis

(2002) Veto Players. How Political Institutions Work. New York and Princeton: Russell Sage Foundation, Princeton University Press.