When less is more: ‘Negative resources’ and the performance of presidents and prime ministers

Abstract

From the perspective of resource-oriented approaches to political leadership, leaders are powerful and likely to be seen as eminent if they command a decent set of resources and are able to use these resources adroitly. Constraints are usually conceptualized as the exact opposite to resources and considered to shackle leaders, limit their capacity, and make them less eminent. This article launches two related criticisms of this widely held view: First, many constraints, usually conceptualized as the opposite to resources, tend to be ambiguous in character; some of them may even provide opportunities to demonstrate an actor’s particular leadership qualities. Getting hold of this phenomenon is facilitated by reconceptualising constraints as potential ‘negative resources’. Second, while leaders commanding limited resources and facing major constraints may, after all, be still less predominant than resource-rich ‘strong leaders’, the ongoing transformation of contemporary notions of ‘good leadership’, which values collaboration and exchange over domination and control, looks set to benefit this type of leaders in terms of recognition and perceived performance. Empirically, the focus of this article is on political chief executives, that is, presidents and prime ministers.

Introduction

The idea of thinking about political leaders in terms of the resources they command is an old one and tends to provoke little principled opposition. To some extent, this would appear to be so, because ‘resources’ can refer to fundamentally different things. The basic ideas of a resource-centred understanding of leadership can be equally well adopted by, for example, supporters of the ‘great man’ approach of leadership and advocates of institutionalist approaches to leadership.

While the former would contend that it is personal resources, and more specifically personality traits, that turns an individual into a leader and potentially makes a leader ‘great’, the latter argue that the status and performance of political leaders, as to be captured from a broadly comparative perspective, depend mainly on the disposal of particular institutional resources. To be sure, both of these approaches have long developed more complex concepts that seek to avoid simplistic conceptualizations and reductionist explanations, such as focusing on personal ambition and physical strengths, or the institutional powers of office available to a given leader. While agency- and personality-focused approaches have come to emphasize the importance of particular qualities beyond personal stamina and physical or psychological strength, such as the ability to persuade others on the basis of advanced rhetorical strategies and skills (see, for example, Kane and Patapan, 2012: ch. 4), more recent institutional approaches seek to identify complex patterns of institutional resources, rather than a fixed set of institutional devices, that empower leaders (Elgie, 2014). ¹

There is obviously a wealth of other kinds of resources, alongside personal and institutional resources, whose availability may benefit a given leader or would-be leader. For example, in particular in parliamentary democracies, stable support of a leader by his or her party is usually conceptualized as a key ‘political resource’ that leaders may possess to varying degrees (Verge and Claveria, 2016). In other studies, time (alongside money and staff) is highlighted as a key resource that allows a leader to develop accounts of ‘resourceful leadership’ (City, 2008). Furthermore, resources of different nature can combine into a complex set of, for example, personal, political, and institutional resources. Resources are generally conceived as the opposite to constraints, which are widely believed to limit the opportunity of leaders to bring about successful leadership. However resources are being defined exactly, different resource-oriented approaches share the overarching belief that the ample availability of resources of one kind or another, or the combination of which, helps leaders to get their way and tends to shape their overall performance favourably.

This article seeks to revisit the established narrative of ‘resourceful leaders and leadership’. It poses the question if, and under what circumstances, ‘less can be more’, or, more properly speaking, if possibly even some constraining factors, may benefit a leader and his or her perceived performance, and how? The next section takes a closer look at the state of the art of resource-focused approaches in the study of political leadership, and more specifically, executive leadership. Section 3 then revisits the ambiguous effects of factors traditionally perceived as constraints on the overall performance of political chief executives, that is, presidents and prime ministers in the major Western democracies.

Resource-oriented approaches in the study of political leadership

Few of the more recent comprehensive studies on political leadership contain an entry on resources in their index (see, for example, Elgie, 2015; Foley, 2013; ‘t Hart, 2014), which would appear to suggest that resource-related approaches have not figured particularly prominently in recent leadership research. At the same time, there are few, if any, chapters of these books that do not refer prominently to ‘resources’. Indeed, thinking about leaders and leadership in terms of ‘resources’ seems to have become so natural to most contemporary scholars of political leadership that approaches focusing on resources are rarely explicitly referred to as ‘resource approaches’. ²

Resources in the study of politics and leadership

Historically, some of the most influential approaches focusing on the various kinds of resources of political leaders emerged from a critique of structural power approaches. According to them, power, and leadership capacity, of individual or collective actors was mainly determined by the possession of particular powers of office. Such an understanding of the determinants of power and leadership capacity not only allowed the identification of the key power holder(s) within a given regime but also provided a convenient basis for broad comparative assessments of more or less powerful leaders in different countries or regimes.

The criticism of these approaches focused, in particular, on the inherent determinism of these approaches and their empirical reductionism. Before looking at this critique in some more detail, it is important to note that such structure-centred approaches are not generally incompatible with resource-oriented notions of leadership. Indeed, if one follows Keith Grint’s suggestion to distinguish between four basic concepts of leadership – ‘leadership as position’, ‘leadership as person’, ‘leadership as a result’, and ‘leadership as process’ (Grint, 2010: 4) – the first one, which basically assumes that it is where leaders operate that makes them leaders, stands out for putting resources at the very centre. From this perspective, as Grint explains, ‘we can define leadership as the activity undertaken by someone whose position on a vertical, and usually formal, hierarchy provides them with the resources to lead’ (Grint, 2010).

Scepticism about the suggested explanatory power of institutional powers, even when conceptualized as position-related resources, has been voiced in different ways. A moderate critique can be found in the well-known work by Richard Heffernan on the British premiership. Heffernan, while believing in the importance and power of institutions, pointed out that in order to bring to bear institutional resources, leaders need other resources that may not flow from their office-holding or their positional status in the wider sense. In fact, prime ministerial power is based on the combination of institutional and personal resources, and successful prime ministerial leadership emerges only, ‘when personal power resources are married with institutional power resources, and when the prime minister is able to use both wisely and well’ (Heffernan, 2003: 350). Even an advanced status of institutional pre-eminence of an actor creates little more than a decent prospect for actual political predominance (Heffernan, 2013: 6–7). For Heffernan, prime ministerial power and leadership capacity are both ‘locational’ (or as we could also say ‘positional’) and ‘relational’. The latter aspect concerns not only the necessary combination of personal and institutional resources but also the relationship between different actors to be found in the executive territory. ‘Executive actors and institutions live in an environment in which resources are exchanged, and where the resolution of problems is dependent on co-operation among resource-dependent actors and institutions’ (Heffernan, 2003: 348).

Resources, both positional ones and others, also figure prominently in Thomas Poguntke and Paul Webb’s influential study on the ‘presidentialization of politics’, in which different types of democratic regimes are distinguished on the basis of different structural configurations of resources (Poguntke and Webb, 2005a). Specifically, they argue that, other things being equal, ‘presidential systems offer far more executive power resources to the leader of the executive’ (Poguntke and Webb, 2005b: 4) than parliamentary democracies do. Effectively, also their key hypothesis – that parliamentary democracies tend to become more similar to presidential democracies – is based on the observation that some resources (such as in particular party-related power resources) lose importance, while others (such as in particular media-related power resources) have gained in importance over time. This notwithstanding, Poguntke and Webb do not dispute that potential power resources must be activated and used in order to empower political leaders in whatever type of political regime.

The distinction between different kinds of resources, including those tied to the particular position held by a leader or would be-leader, is also at the centre of the extremely influential work by Joseph Nye on leadership and power (Nye, 2008). Nye acknowledges that power resources change over time and context and examines how, and to what effect, leaders use hard and soft power resources. Generally, as Nye argues, ‘those without formal authority tend to rely more on soft power, whereas those in formal positions are better placed to mix hard and soft power resources’ (Nye, 2008: 38), but there are countless options to mix different power resources in the pursuit of effective leadership. Leaders’ styles can be distinguished by how they use hard and soft power resources; however, styles can change according to changing circumstances.

While Heffernan and others acknowledge that power is at least to some extent locational, or structural, in character with some actors profiting from their locational status and the structural powers attached to it, other scholars have argued that all power is essentially, and virtually purely, relational. This is true of the exceptionally influential ‘core executive’ approach, as originally put forward by Patrick Dunleavy and R.A.W. Rhodes (1990), and successively refined by the latter (see, for example, Rhodes 1995). For Dunleavy and Rhodes, ‘power is contingent and relational; resources are not fixed. They do not inhere in any position. Instead, the resource relationships vary between core executive actors’ (Elgie, 2011a: 66). Actors and the resources they can employ depend on other actors; that is, all resources are relative resources.

Other scholars have contributed further important elements: Martin J. Smith, whose book-length study on the core executive in Britain marks a modern classic in this field, has specifically emphasized that ‘power depends on how resources are exchanged, and hence it is about dependence not control’ (Smith, 1999: 31). Consequently, understanding the operation of the core executive is seen to be essentially about identifying the structures of dependency (Smith, 2003: 62). In particular, the debate about ‘resource-dependency’ has been prominently resumed and continued in the more recent literature on the evolving core executive (see, for example, Craft, 2015; Diamond, 2013: 108–109).

Rhodes’ ‘court politics’ approach – a strictly interpretive approach launched about two decades after the ‘core executive’ approach

gives a distinctive twist to the resource-dependency model of the core executive. Resources cease to be given by (say) institutional position. Rather, they are inter-subjective, constructed by actors against a backcloth of traditions … Dependence is not determined by control over resources … but also by inherited beliefs and individuals’ webs of significance. (Rhodes, 2013: 322)

Thus, here, even the relative weight of ‘relational resources’, once placed at the centre of the ‘core executive’ approach, is effectively reduced to the benefit of other factors.

Drawing the threads together

Certain differences in the conceptualization of resources apart, most scholars conceive of resources as dynamic and variable sources or supplies that do not cause any particular effects or outcomes on their own unless being used by an actor who brings those resources to bear. Some authors also mingle different kinds of resources in an attempt to gauge a leader’s overall amount of leadership capacity or authority. For example, in the presentation of their Leadership Capital Index (LCI) approach – a sophisticated and valuable attempt to measure the performance of leaders on the basis of a carefully designed set of hard and soft indicators relating to a leader’s relations, reputation, and skills – Bennister, ‘t Hart and Worthy define leadership capital as ‘the aggregate of a leader’s political resources’ (Bennister et al., 2015: 418). Furthermore, whenever an enquiry is essentially about capturing the actual power and authority of a given leader, rather than just providing a static analysis of power resources, thinking about leaders and leadership in terms of resources implies a committed focus on agency. Individual leaders do not only command different types and/or different amounts of resources; they may display a fundamentally different ability to make effective use of these resources. Even the same leader may have different, or different amounts of, resources at his or her disposal, and may be more or less good, or lucky, at exploiting them at different stages of his or her respective tenure. Indeed, extended tenures of political leaders are usually made up of notably more or less successful terms (see, for example, Theakston, 2013: 235).

Finally, it is important to note that even a careful analysis of the distribution of resources and agents’ ability to employ them may not provide perfectly ‘realistic’ assessments of either actual power or a leader’s leverage and clout. For one thing, leaders do not have to make use of the resources they command, and this may not be simply out of mere inability to do so, but may flow from strategic considerations. Moreover, it is easy to forget that the status of agents as power-holders and leaders is ultimately, at least to some extent, a social construction. Therefore, disregarding the actual distribution of structural power resources among actors, ‘sometimes agents can be more powerful than their resources apparently justify because they have a reputation for being powerful’, which reminds us that ‘perceptions of power become actual power as the perceptions become part of the belief set of each person in a power game’ (Dowding, 2017: 9), and in the leadership process for that matter.

‘Negative resources’ and the performance of political chief executives

Everything else being the same, both political leaders themselves and scholars of political leadership would appear willing to agree that it is best to have as many resources as possible at hand when it comes to being, and being seen as, an efficient and successful leader. As Heffernan has put it bluntly, ‘the more resources, the more powerful and predominant the prime minister is; the fewer resources, the less powerful and predominant they are’ (Heffernan, 2003: 347).

The empirical limits of ‘the more the better’ assumptions

There is good reason to believe, though, that this assumption may be too simple, and not just because commanding resources is not the same as wielding them adroitly. Specifically, more resources may not necessarily translate into more authority and power even if being used with reasonable skill. Indeed, it seems that an abundance of resources carries particular risks.

For example, in the United Kingdom, Tony Blair and his government famously had to cut back on their personnel resources for managing the media in response to widespread public criticisms of their obsession with ‘spin doctoring’, which had come to seriously weaken the government’s status in the wider political process (see Rawnsley, 2002). In a similar vein, Richard Neustadt, in one of his very late papers, identified the ever-growing staff working for the US president as a key force behind the perceived ‘weakening of the White House’ (Neustadt, 2001). Even the ample use of established institutional or constitutional resources may come at a high price, possibly ‘eating up’ other resources. Recent empirical evidence comes from France: French governments have unique devices for putting pressure on parliament at their disposal, including the combination of a vote on a bill with a confidence vote, but their unrestrained use tends to be costly. As the findings by Becher et al. suggest:

prime ministers experience a considerable drop in approval after their use of the confidence vote that is not accounted for by standard economic and political covariates. The effect size is similar to a 1 per cent decline in economic growth. The findings … also suggest that political costs constrain the bargaining power conferred by the confidence vote. (Becher et al., 2017: 252)

Introducing ‘negative resources’

These and other experiences with particularly resource-rich leaders come as a powerful incentive to rethink conventional notions and interpretations of resources and constraints and their effects on leaders and their perceived performance. For this purpose, I draw on an important but little-noticed suggestion by Gordon Smith when writing on the resources of a German chancellor, namely that leadership environments may accommodate ‘negative resources’ (Smith, 1991: 49), understood as structural factors that normally constrain leaders and may only exceptionally be used by them to further their own ends. In Smith’s own work, this important element of thought is not further developed in any way, and thus, a more precise conceptualization of ‘negative resources’ is needed. The following specifications would seem useful: A ‘negative resource’ is a constraint successfully transformed into a positive source that may benefit the status and performance of a leader. This possible transformation is the result of a complex process which involves in particular a leader’s skills, yet also a wealth of highly contingent contextual factors as well as the perception of that leader by others. There are obvious limits to transforming constraints into a source of authority and power. For example, vast unpopularity is unlikely ever to be turned into a positive resource. However, many constraints may, indeed, have a (if hidden) positive potential, and while some features of a given leadership environment may not be turned into a ‘negative resource’ proper, they may still do considerably less harm to the status and performance of a leader than widely assumed.

Capturing the performance of presidents and prime ministers

Before looking at some of the structural and contextual factors that have conventionally been considered to constrain leaders and impact negatively on their performance, a few remarks on ‘performance’ would not seem to go amiss. Performance-focused perceptions of politics tend to permeate the political arena as a whole; indeed, recent empirical research on retrospective voting suggests that ‘both government and opposition parties can expect credit and blame for their conduct’ (Plescia and Kritzinger, 2017: 156). In the same vein, other recent research suggests that while citizens may have particular expectations of leaders, approval of and trust in leaders are more directly related to leaders’ actual performance than to prior expectations (Seyd, 2014). The challenges of assessing and measuring the performance of political leaders in more detail are immense. For the purposes of this article, it may suffice to note that longevity and the re-election record of a given political chief executive (i.e. president or prime minister) have been widely acknowledged as key indicators, even though they are effectively little more than important proxy-indicators for measuring the leadership performance of individual office holders. Popular opinion and expert job approvals that usually cover both politics- and policy-related aspects of leadership performance may draw more complex pictures, yet there is unavoidably a strong element of subjectivity involved (see Strangio et al., 2013). Generally, and importantly, leadership assessments, even at the level of sophisticated scholarly evaluations, are largely based on the perceived performance of leaders (see Bennister et al., 2015: 422).

Can the assets widely associated with a successful leadership performance be achieved even in the face of powerful constraints, potentially to be turned into negative resources? The remainder of this section looks into this in light of the more recent experience of presidents and prime ministers in the major Western democracies.

‘Negative resources’ and executive leadership in Western democracies

Writing about presidential leadership in the United States back in 2002, Michael E. Bailey advanced the thesis that ‘divided government’ generally may carry important strategic advances for the president, allowing him to act heroically without having to deliver (Bailey, 2002). This is a bold assumption, and the recent experience with the Obama presidency tells a different story. However, there have been presidents for whom the logic behind this argument did indeed materialize nicely, with some of the most impressive evidence relating to the leadership performance of Bill Clinton (see Peele et al., 1998: 5). While the prospects for ‘presidential heroism’ during periods of divided government are mixed and contingent, the early experience with the Trump presidency suggests that forceful leaders, even when operating under the conditions of ‘unified government’, may end up ‘unheroically’; after all, Trump’s job approval ratings after 6 months in office were the lowest of any president since surveys started being conducted in the 1940s. Moreover, there is substantive empirical evidence which suggests that constraining circumstances may, indeed, have some beneficiary effects for the president. In a recent major study on public opinion and the presidency in the United States, Jeffrey Cohen found that ‘presidents are considered stronger under divided as opposed to unified government’ (Cohen, 2015: 81), and ‘divided government presidents are more popular than unified government ones’ (Cohen, 2015: 81).

The core of this phenomenon marks by no means an American peculiarity. It is echoed, for example, by the experience of French presidents and specifically those who had to operate during ‘cohabitation’ (i.e. when the president is from a different political party than the majority of members of parliament). Under ‘cohabitation’, the president’s powers are circumscribed in many ways. Most importantly, he or she has to appoint a prime minister that will be acceptable to the majority party in parliament, so that the president and the prime minister are most likely to belong to two different and politically competing parties. Remarkably, this has not at all undermined the president’s status with the French public; the historical lows of presidential popularity did not relate to ‘cohabitation’ presidents but to presidents enjoying strong majority support in both chambers of the parliament, such as Nicolas Sarkozy and François Hollande. Moreover, ‘cohabitation’ has not destroyed the re-election fortunes of French presidents. As Jean-Louis Thiébault has pointed out, ‘it is one of the paradoxes of the Fifth Republic’ that ‘a president is able to rebuild popularity in the context of “cohabitation” and its ensuing weakening of presidential powers’ (Thiébault, 2016: 521). Indeed, both presidents experiencing a ‘cohabitation’ (François Mitterrand and Jacques Chirac) were re-elected, while several others failed to secure re-election (such as Valéry Giscard d’Estaing and Nicolas Sarkozy) or did not even dare to seek re-election (François Hollande).

Similar patterns can be identified in several of Western Europe’s parliamentary democracies: In Germany, Chancellor Angela Merkel fared, overall, significantly better in terms of public popularity and job approval when serving as the head of two strongly power-dispersing ‘grand coalitions’ (made up of the two major parties Christian Democratic Union of Germany (CDU)/Christian Social Union in Bavaria (CSU) and Social Democrats (SPD), 2005–2009, 2013–2017) than as the leader of an ideologically more cohesive Christian-Liberal governing coalition in which the CDU/CSU held more than two-thirds of the cabinet seats (2009–2013; see Helms and van Esch, 2017). The two very different terms of the Aznar government (1996–2004) in Spain can be considered a particularly glaring case in point: In 1996–2000, the lack of a stable parliamentary majority, little experience, and a modest reputation induced Aznar to adopt a strongly consensual approach to dealing with some of his major opponents and to promote a notably modest image of himself. In 2000–2004, after having gained a parliamentary majority for his government, he abandoned this moderating approach and adopted a much more adversarial style, which soon came to undermine his appeal with the voters (see Tusell, 2004).

The hidden resource structures of different leadership environments

Such spectacular examples apart, many contextual variables shaping the environments that political chief executives encounter, and which have conventionally been considered to shackle the leader, are of a highly ambiguous nature. To begin with, this is true of coalition agreements which have become a widespread phenomenon in West European parliamentary democracies and beyond. His or her constitutional prerogatives tend to be effectively circumscribed by formally acknowledged rights in particular of the coalition partner. However, recent research suggests that, overall, it is the departmental ministers, rather than the head of government, whose room for manoeuvre is severely limited by coalition agreements (see Moury, 2013). Prime ministers may, in fact, profit from the drawing-up of coalitions agreements in at least two ways: First, several major and particularly contested issues will be sorted out during the government-building process, and thus, unpopular decisions will therefore not be closely associated with, and possibly blamed on, the prime minister’s leadership after the government has assumed office. Furthermore, comparative research suggests that coalition agreements not only rationalize inter-party relations but also help party leaders (and prime ministers) to keep their own troops together (see Müller and Strøm, 2008).

This argument can be extended from coalition agreements to coalition parties, famously conceptualized by George Tsebelis as ‘partisan veto players’, alongside the more manifest ‘institutional veto players’ (Tsebelis, 2002: 2; passim). Unlike in the original version of the veto player theorem, in real-world politics, veto players may be ‘constructive’ players, potentially willing to make status-quo-breaking decisions happen (see Benz, 2003). Such constructive behaviour is facilitated by the existence of suitable channels for communication, cooperation, and exchange (Benz, 2003: 232); thus, prime ministers may advance the prospects for fruitful cooperation by creating consensus-inducing decision-making procedures and arenas. Yet, coalition parties in complex multi-party governments may even help the prime minister in advancing his or her own overall status. As Poguntke and Webb note, ‘leaders in consensual systems … may be able to justify their decisions by referring to the constraints imposed upon them by veto players. In this sense, the very nature of consensual politics provides them with additional power resources’ (Poguntke and Webb, 2005b: 12).

As to the parliamentary arena, in contrast to conventional wisdom, small majorities do not always circumscribe the government’s room for manoeuvre. It has long been observed that it is often, perhaps paradoxically, more difficult to govern with a particularly sizeable majority than with a tight majority that cannot afford any dissenters or dissent (see Helms, 2005: chs 6 and 7). Party loyalty on voting behaviour tends to be strongest when the outcome is anticipated to be close and consequential to the success or failure of a given bill (Raymond, 2017; see also Stecker, 2015). In parliamentary systems in which governmental ministers are to be drawn from the pool of parliamentarians, relatively small parliamentary majorities may also help to keep the number of members of parliament (MPs) who see their ambitions to be appointed to ministerial office disappointed relatively small. There is some evidence that the number or size of cabinet reshuffles and, thus, the prospects for MPs to gain ministerial offices do indeed correlate with the level of governmental party discipline and government support in parliamentary voting (see Kam, 2009). Moreover, there is some evidence that the higher transaction costs of coalition governance at the level of executive politics may be compensated for by an easier ride of the government through the parliamentary arena. As José Magone notes in his recent comparative study on five (post-)consensus democracies, ‘Coalition governments condition Parliament; that is, coalition parties impose discipline when voting on issues related to the coalition agreement’ (Magone, 2017: 275).

Looking beyond the parliamentary arena, the spectacular recent attacks of the Polish and Hungarian government on their country’s constitutional courts strongly suggest that constitutional review is a major thorn in the flesh of determined political leaders. While traditionally having been conceptualized as counter-majoritarian institutional features of a political system, more recent research has pointed to the possible quality of constitutional courts as veto players (Volcansek, 2001), or more specifically, ‘conditional veto players’ (see Brouard and Hönnige, 2017). The latter notion implies that the role of courts in the political process is contingent, and that the veto player potential of courts may be effectively ‘absorbed’. In that case, there is no need for governments to practice ‘self-restraint’ to avoid defeat at the court. However, more than that, it is well possible to identify several concrete benefits of constitutional courts for governments and governmental leaders: Most importantly, ‘an independent judiciary creates opportunity to shift blame for unpopular decisions from the executive and legislative branches to the courts’ (Vanberg, 2015: 172). Political decision-makers may even feel an incentive ‘to adopt vague statutes that avoid concrete policy choices and instead rely on judicial decision making to fill in the details down the road’ (Vanberg, 2015: 172). Moreover, courts may help governments to ‘weed out’ legislation that turns out to be undesirable ex post by striking it down by judicial means, thereby sparing governments the costs of ‘undoing’ complex legislative bargains which could damage their potential for forging future legislative coalitions (Vanberg, 2015: 171).

Federal systems are also infamous for checking the room for manoeuvre and limiting the leverage and scope of national political leaders. The very definition of genuine federal systems is about the existence of (at least some) constitutionally guaranteed independent decision-making powers of the individual member states belonging to a federation, and even in institutionally weak federations there may be powerful actors operating at the sub-national level at least indirectly shaping the nature of politics and leadership at the national level. The enormous variety of different federal systems, as well as the trend in more recent research to no longer distinguish strictly between genuine federalism, on one hand, and devolution in unitary states, on the other (see, for example, Erk and Swenden, 2010), makes it immensely difficult to arrive at any more specific contentions that could claim to do reasonable justice to the conditions of exercising political leadership in ‘multi-level systems’.

However, at the very least it should be acknowledged that the institutions of vertical power-sharing are no static features of political systems that simply circumscribe political leaders. Leaders can play a role in shaping the public perceptions of multi-level governance, and they can adapt their positions to changing circumstances or the given issues at stake. For example, most American presidents have sought to gain political capital from ‘DC-bashing’ (in particular, yet not only during their presidential campaign), but have not hesitated to centralize control in the national capital whenever expedient. On other occasions, they have generously relocated responsibilities to the states, thereby easing their own burdens at the federal level (see Helms, 2005: 146–147). In other types of federations, such as Germany, the ‘interlocking’ character of the federal institutions, including the second chamber and the so-called mediation committee, creates a highly specific nature of inter-governmental decision-making, which tends to be complex, cumbersome, and opaque. Depending on the patterns of party control, the complex structure may be used by chancellors and their governments for forging complex package deals and/or engage in ‘blame avoidance’, if usually within limits and at a certain price (see Zohlnhöfer, 2007: 1134). Similarly, in different types of federation, the chief executives at state level tend to have ample opportunity to divert criticism and escape accountability by pointing to decisions made at the federal level and the possibly unwelcome effects thereof.

As Vivian Schmidt has argued, ‘compound democracies’ are also marked by particular discursive patterns. The ‘communicative discourse’, which involves political actors who communicate the ideas developed through coordinative discourse to the public, tends to be thin; leaders are expected to communicate with the general public only in vague terms, since detailed discussion risks unravelling compromises reached in private (Schmidt, 2005: 774). This may well be considered to mark a specific opportunity structure: Political leaders, even those operating at the top level, do not have to possess particular rhetorical skills to survive and succeed in this environment. The chancellorship of Angela Merkel demonstrates impressively that political leaders in compound democracies may live well and secure even in the absence of rhetorical mastery – to an extent that would be inconceivable for any British prime minister –, and Merkel marks by no means a unique exception among German chancellors (see Helms, 2016: 294, 300).

Finally, just as a reputation for being powerful can make a leader more powerful than he or she actually is in terms of resource richness (see above), perceptions of powerlessness, or resource scarcity of a leader, may have serious and wide-ranging effects – in particular, if this turns out to be a misperception that leads other actors to underestimate leaders. Some of the most obvious cases at the level political chief executives relate to female heads of government. Both Margaret Thatcher and Angela Merkel effectively benefitted from being seriously underestimated by their fellows and possible leadership rivals before and during the early stages of their tenure (Evans, 1997: 42–43; Helms, 2014: 116). It may even be doubted with hindsight if either of them would have been able to accumulate leadership capital in the way they did, had it not been for the misperceptions and complacency-driven misjudgements among their respective peers. We still know too little about the effects of particular recruitment patterns of women leaders on peer-group perception and their related effects in terms of resources and constraints. Yet, the findings of a recent comparative study on prime ministerial careers in Europe (1979–2015), including that ‘all women prime ministers in Europe have gained ministerial experiences in medium- or low-prestige portfolios while most of their male counterparts held high prestigious cabinet posts prior to becoming prime minister’ (Müller-Rommel and Vercesi, 2017: 256), would appear to be in line with the above-made suggestions concerning the relationship between the constraints – and the opportunities – of a perceived status of an incoming leader. While by no means all women prime ministers have been able to make capital out of this, many of them do have been considered to be political ‘lightweights’ and ‘transitory phenomena’.

Conclusion

The observations presented above suggest that constraints are not simply constraints but may be transformed into resources that benefit a leader’s performance. The term ‘negative resource’ seeks to highlight the hidden potential of an apparent constraint. Turning a constraint into an effective resource is a complex and uncertain process which is shaped not just by a leader’s particular skills (whose usefulness are invariably dependent on the ever-changing nature of contingent contexts), and possibly a fair amount of luck, but also by the changeable perceptions of a leader’s actual performance.

Complex structures may provide leaders with particularly rich opportunities for different and distinct blame avoidance strategies (see Hinterleitner and Sager, 2017). Overall, though, dealing with constraints is invariably more difficult and demanding than simply pulling the levers of power in power-concentrating regimes. Thus, it would seem reasonable to suggest that leaders operating under severely constraining conditions should deserve more, rather than less, credit for what they accomplish – which has, however, rarely explicitly been acknowledged. That said, some exceptions can be found: For example, Genovese et al. distinguished between ‘high-opportunity’ and ‘low opportunity’ political leaders in an attempt to establish what different leaders made of their opportunities, acknowledging that the ‘net leadership performance’ of some leaders may have been better than that of others, even though they may have achieved less in absolute terms as they operated under less favourable circumstances (see Genovese et al., 2014).

More generally, effective leadership may be, after all, more dispensable in power-concentrating regimes than in environments marked by a maximum dispersion of power and resources. Alongside all the other reasons given by Michael Foley for his powerful contention that the British Westminster democracy ‘leaves very little room for either the idea, or the practice, of leadership’ (Foley, 2013: 11, passim), the very fact that the British political system is, indeed, ‘a set of arrangements facilitating governance’ (Jordan, 1994: 196) marks an important factor in its own right. In the strongly power-concentrating context of Westminster, holders of the premiership are in a position to give some direction to the government even in the absence of exceptional leadership skills. By contrast, the so-called consociational democracies, once famously conceptualized as the outright opposite to Westminster democracies, and empirically marked by a maximum dispersion of power and resources (Lijphart, 2012), are by no means leader- or leadership-free environments. Indeed, in Lijphart’s early work, the quality of leadership is identified as the key element on which the well-being of deeply divided polities hinges, in both theory and practice (Lijphart, 1968: 211), and leadership has been subsequently highlighted as one of the key explanations, alongside the existence of suitable institutions, for successful consociationalist governance (see, for example, Deschouwer, 2006).

This notwithstanding, ultimately, leaders facing many powerful constraints, even if they are good at turning them into resources, are unlikely to be seen as genuine ‘strong leaders’ (in the conventional sense of ‘leaders who get their way’, see Brown, 2015: 2). However, importantly, this does not necessarily undermine the prospects of a leader for a successful and favourably perceived performance. While mass publics in many countries of the world may continue to display a certain yearning for ‘heroic leaders’, there has been a silent revolution going on in international scholarship on political leaders and leadership over the past decades that may eventually leave its mark on the wider public: ‘good’ leaders and leadership are less and less equated with ‘strong’ leaders and leadership. Some scholars have gone as far as to consider ‘strong’ leaders a major threat to democratic governance (see, for example, Brown, 2015, 2016; King, 2016). Good leaders are increasingly considered to be people with a proven ability to bring people together in order to work together on viable solutions. ‘Collaborative leadership’ has emerged as a key paradigm of public and political leadership for the 21st century (Smolovic Jones and Grint, 2013). Governmental leaders, while enjoying some particular resources that other actors lack, are part of a complex system of ‘interactive governance’ (Sørensen and Torfing, 2016).

Changing normative standards for evaluating leaders are accompanied, and supported, by empirical evidence: While leaders operating under the conditions of complex or compound democracy may feel bound to make only modest promises, they are likely to eventually benefit from this institutionally induced modesty. As Neil Malhotra and Yotam Margalit conclude on the basis of a major comparative empirical enquiry, ‘respondents prefer politicians who set a low expectation and exceed it even when their expectation is explicitly described as “pessimistic”’ (Malhotra and Margalit, 2014: 1014). Other things being equal, it is presidents in presidential systems that are most likely to promise too much and then suffer from facing towering expectations all but impossible to be met. While presidential democracies are by definition separation-of-power systems, ‘presidentialism is characterized by a broadly shared public perception that places the president at the center of the nation’s politics and views him (or her) as the person primarily responsible for dealing the challenges before the country’ (Mezey, 2013: 8–9). This comes close to a paradox, if one so deeply entrenched in the constitutional practice of most presidential democracies that it is rarely noticed. Other regimes, especially from the family of semi-presidential democracies, have known their own paradoxes which point to the limits of purely institutional understandings of politics, yet without questioning the fundamental importance of different institutional architectures (see Elgie, 2011b). Therefore, future research into ‘the politics of negative resources’, in particular when it comes to further illuminating the ‘when and how’ dimensions of this phenomenon from a broader comparative perspective, would seem well advised to observe the defining institutional features of different forms of government.