The Significance of Leadership in the Evolution of Policy Styles: Reconciling Policy-Making in the Short and Long Term

Can relatively short-term political leadership affect long-term national policy styles, and if so, how? We examine these questions in order to explain the temporal dynamics in national policy styles observed in several recent studies whose findings have challenged the traditional view that policy styles lend continuity and stability to policy-making. In its original formulation, the concept of policy styles refers to the extent to which governments are committed to anticipatory or reactive policy-making, as well as to whether they seek to reach consensus with societal groups over policies or simply impose decisions on them. Following these definitions, we contend that political leaders can alter a long-held policy style and often do so due to strategic considerations, either in an attempt to improve their policy performance or in response to an exogenous shock, such as an economic crisis. We illustrate this argument by using longitudinal records of changes in country-level governance indicators which include data on dimensions of policy styles. Specifically we examine how two of the Sustainable Governance Indicators compiled by the Bertelsmann Foundation changed for Ireland, Italy, Malta, and Poland between 2014 and 2020. As the cases show, government leaders sought to alter long-standing processes and the arrangements of governance and policy-making. Our empirical findings support the overarching argument that the study of national policy styles should be combined with that of political leadership.

shown regularly to adopt, or prefer, anticipatory or reactive decisions and often to do so either by seeking to reach consensus with interest groups or by imposing decisions on them, notwithstanding opposition from such groups.
Evidently, this perspective highlights long-term constraints and orientations towards policy targets and publics, which lead to persistent forms of interaction and behavioural patterns, a phenomenon that more recent studies have also observed (e.g., Adam et al. 2017). While the origins of such preferences and practices remain understudied, some scholars have suggested that these distinct ways of formulating and implementing public policy involve historically entrenched institutional and politico-administrative routines that secure their own reproduction (Howlett and Tosun 2019a, b).
This approach is different from those commonly found in many studies on policymaking, including those that emphasise institutional constraints (e.g., Béland 2009;Hall and Taylor 1996;Peters 2019;Wiseman and Béland 2010), as it highlights the opportunities that arise from these same arrangements and the manner in which they shape processes as well as outcomes. It is also distinct from both group theories and institutional accounts that emphasise the significance of policy-maker behaviour in policy processes. Unlike the former approaches, which highlight the significance of agents and individual-level agency in creating variable outcomes (van Stokkom 2005), and thus implicitly deny longer-term patterns in such activities, the analysis of policy styles suggests that agents and individual-level agency play a much more limited role than do other factors in constraining policy-making activities and processes.
However, stressing an absolute conceptual dichotomy between agents and institutions is not helpful and limits the analytical potential of both the concept of policy styles and more recent work on behavioural insights and neo-institutionalism (Radaelli et al. 2012). While it is true that the concept of policy styles emphasises the importance of institutional lock-ins and routinisation, which are also parts of neo-institutional theory (see Hall and Taylor 1996), it remains open to the need for individuals to construct and follow dictates in their decisions and interactions. Similarly, even the most agent-based approach has to recognise that the range of adopted policies is not as wide as one would expect and that some mesolevel or macrolevel factors affect and constrain the range of possible policy actions and processes (Falleti and Lynch 2009).
In this study, we seek to bridge this gap by clarifying and expanding on the roles played by individuals in the creation and sustenance of policy styles. We do this by concentrating on the various roles that can be fulfilled by political leadership, which, according to Blondel (1987, p. 1), constitutes the single most visible and recognised element of political life, as it creates, perpetuates, and alters styles of politics.
The remainder of the article unfolds as follows. First, we discuss in depth the theoretical underpinning of policy styles and propose an operationalisation of the concept. Subsequently, we give an overview of the concept of political leadership and leadership "styles." While policy research has shown that the ideology and party affiliation of policy-makers affect policy outputs (e.g., Knill et al. 2010;Zohlnhöfer 2009), it has paid less attention to political leadership per se, which we define as the characteristics and behaviour of executive leaders and how they exercise power (Galanti 2021;Nye 2008). We hereby also strive to strengthen the analytical role of political leadership in policy studies, as recent studies have advocated for (Capano and Galanti 2018;Torfing and Ansell 2017), as well as help to develop and operationalise the concept of policy styles.
In order to illustrate these insights, we use a set of the Bertelsmann Sustainable Governance Indicators (SGI) (Schraad-Tischler and Seelkopf 2016) that have remained relatively consistent for most European countries over the past decade but did change for Ireland, Italy, Malta, and Poland between 2014 and 2020. Prima facie, this pattern suggests that national policy styles exist and can be measured. More detailed analysis of these four anomalous cases, moreover, reveals how different political leaders have altered aspects of existing styles.

Policy Styles in Theory
Policy-making is one of the key functions of any political system, irrespective of the nature of the political regime in place. Therefore, it comes as little surprise that for decades, scholarship in comparative public policy has been interested in identifying whether long-lasting patterns of policy-making can be found in different countries and sectors of activity, whether these are changing, and if so, how these can be classified and these changes explained Bennett 1991).
Policy studies have developed concepts such as "policy regimes" and "policy styles" to denote long-term policy patterns whereby a common set of policy ideas, a long-lasting arrangement of policies that have accumulated over time, a common or typical policy-making process, and a more or less fixed set of policy actors are found in many policy areas (Howlett and Tosun 2021a).
Some of the work on policy regimes has focused on policy outputs-that is, the nature of the efforts made by different governments-and on how (in)consistent or (in)effective they have been. Castles (1993), for example, identified clusters of countries, denoted as "families of nations," that opted for either more state-based or market-based public policies as welfare states developed in the post-World War II era. Investigating economic and social policy in advanced democracies, Obinger and Wagschal (2001) also showed that distinct families of nations exist in terms of the kind of effort they make in these policy sectors and that these patterns and relationships are stable over long periods of time.
However, another classic theme in the study of policy patterns has been whether policy-making processes in the different jurisdictions also follow a certain tradition, trajectory, history, or "style." Whether, and to what extent, all governments develop a distinctive style (Simmons et al. 1974) when making decisions and choosing which policy alternatives to adopt, and how those styles correspond to or affect outputs, have been outstanding questions in the field since the first major study of the subject, presented in the volume edited by Richardson (1982).
That volume provides the point of departure for any study on policy styles, especially those focused on national variations (Howlett and Tosun 2019a). As noted above, the book examined the contemporary similarities and differences among the countries of Western Europe in terms of their propensity to take anticipatory or re-K active decisions and, in either case, whether they tended to do so by seeking to reach consensus with organised groups or by imposing decisions on them, notwithstanding opposition from such groups (Richardson et al. 1982).
Identifying such styles and where they exist is important, for they describe not only how policies are typically made but also potentially the shape and character of policy outputs and outcomes. The extent to which (scientific) evidence informs policy choices, for example, has an important impact on the character and effectiveness of policy outputs and is directly related to the manner or style of their development and adoption. Some policy choices and solutions emerge from carefully crafted formulation processes, for instance, while others are more heavily influenced by other processes such as political or electoral bargaining (Howlett and Mukherjee 2014). It is essential for researchers to ascertain whether countries tend to process policy issues in one way or another, as this affects our understanding of the differences not only in those processes but also in policy outputs and outcomes. As Adam et al. (2019), for example, have shown, the accumulation of public policies over time can reduce a state's capacity to deliver on them, and one of the reasons why some countries are affected by such a "democratic responsiveness trap" can be found in their particular policy style(s).
Many open questions surround the concept of policy styles, however, as it was put forth by Richardson et al. (1982). While their work was pitched at the national level, Freeman (1985) argued that the differences found between policies in different countries were less apparent at the general, national level than at the sectoral one, thereby introducing the competing vision of sectoral rather than national-level policy styles. This argument has been supported by Cairney (2021), for example, who also advocates the notion of sectoral policy communities and their influence on the emergence and maintenance of a given policy. Another important reformulation focused on the implementation of public policies rather than their development and stressed the importance of administrative styles as a component of a national or sectoral style (Knill 1998;Painter and Peters 2010;Peters 1990). Howlett et al. (2020) have also attempted to advance the conceptualisation of policy styles, suggesting that researchers should conceptualise styles by drawing on the insights into the workings of each stage of the policy cycle rather than only decision-making or policy formulation, as was done by Richardson et al. (1982). The volume edited by Howlett and Tosun in 2021 showed that one can indeed identify styles regarding how agenda-setting takes place (Bevan et al. 2021), how policies are formulated (Howlett and Mukherjee 2021) and implemented , and how the performance of public policies is evaluated (De Francesco and Pattyn 2021).
Recent empirical work suggests, in contrast to Richardson et al. (1982), that longer-term patterns of policy-making can arise at the national, sectoral, or issue levels; that they may differ across the stages of policy-making; and that they can even change over time (Candel et al. 2020;Ertugal 2021;Howlett and Tosun 2019a;2021b). Why one style might emerge in one place and not another has been partially answered; it remains open as to why a style might change, for how long, and in what direction.

The Role of Political Leadership and Leadership Styles
One factor that may explain these variations and answer the "why" questions posed above is leadership, which can be defined as a process of social influence or, more specifically, as steering the policy process towards a specific policy goal of stability or change (Capano and Galanti 2018, p. 31). Leadership is embodied by structurally empowered individuals (e.g., ministers or secretaries) or collectives (e.g., cabinets or ministries), which can act either in accordance with an established style or not.
The study of political leadership, in particular, represents a classic lens for political studies on how power is exercised, including in policy-making (Blondel 1987). While the first sets of studies carried out on this subject in the 1940s concentrated on the personal traits of political leaders, the perspective shifted to the study of their behaviour in the 1960s, and then in the 1970s to the contingencies or circumstances that affected that behaviour (Nye 2008). The late 1970s and 1980s produced studies on political leadership that concentrated on the goals of leaders and the strategies or "styles" they adopted to pursue their interests. Particularly influential, in this latter period, for example, was the differentiation between "transactional" and "transformative" leadership put forth by Burns (1978), which hypothesises that many leadership traits can be combined into a "style" in the same way aspects of policy-making can.
This differentiation highlights a melding of structure and agency that was lacking in earlier eras. Transactional leadership is characterised by taking the initiative in making contact with others, with a view to exchanging "valued things" (Burns 1978, p. 83) within an existing political or administrative apparatus. In other words, it is a continuation of existing arrangements or styles. Transformative leadership, on the other hand, is characterised by engagement that sees "leaders and followers raise one another to higher levels of motivation and morality" (Burns 1978, p. 83) and highlights discontinuity and changes in style. Significantly, both transactional and transformative leadership can serve to change or maintain the policy status quo (Galanti 2021), although the maintenance of the status quo, in general, is more closely associated with a transactional style (Nye 2008).
Regardless of what purpose certain leadership styles fulfil, scholars have highlighted their (internal/external) interactions between agency and structure. Elgie (1995) stressed that the cross-country variations in how leadership is exercised depend not only on the personal characteristics of the leaders but also on their institutional environment. The latter involves the structure of the executive, the features of the party system, or the electoral system-institutional variables that affect the leaders' strategic options (Elgie 1995, p. 206). But while both leadership styles exist within specific kinds of institutional arrangements and societal relations, the idea is that a single political leader may shape or alter his or her own government's policy style by making their leadership style more personalised or collegial (Helms 2004), or more collaborative and integrative (Torfing and Ansell 2017).
These discussions on leadership styles add another dimension to the work on policy styles. But, more significantly, they also suggest a factor that might help to resolve the puzzle of policy style changes set out above, since governments exercise some control over aspects of a policy style when they change existing institutional K or processual arrangements (e.g., decisions regarding the practices of the prime minister office). Building on this, we postulate that decisions by political leaders can affect policy styles either directly or indirectly. Drawing loosely on Helms (2012), we argue that political leadership can impact political styles through (1) the leaders' beliefs and convictions, (2) the need to resolve a given policy problem, and (3) the need to respond to events and changing circumstances. However, and equally significantly, we expect any deviations from predominant policy styles to be short term, for administrative routines and institutional variables outlast individuals over the long term. These possibilities serve as hypotheses. We assess the plausibility of the hypotheses in the illustrative empirical analysis below.

Clarifications on Operationalisation, Data, and Case Selection
As sketched out above, there has been much discussion in the literature on whether policy styles refer in general to policy-making within a country or to policy-making in individual sectors within a country. In this study, we adhere to the formulation of policy styles put forth by Richardson et al. (1982) and investigate the stability or dynamic characteristics of national policy styles. We still find the idea of sectoral policy styles compelling and relevant, but for this analysis we limit ourselves to the original conceptualisation.
The overwhelming majority of empirical assessments of national policy styles is based on in-depth studies of single or a few cases that are subsequently compared to each other, as done by Richardson (1982) and more recently by Howlett and Tosun (2019a) or Zahariadis et al. (2022). Empirical studies by Liefferink and Jordan (2004) and by Candel et al. (2020), which build on an explicitly comparative logic, are exceptions. To date, there have been no attempts to operationalise national policy styles for a large number of countries over a longer time period. Consequently, there are also no datasets on national policy styles or how they have changed over time. From this perspective, the study of policy styles deviates from other research lines in comparative public policy, where the use of quantitative indicators has increased significantly over time.
Nonetheless, following Zohlnhöfer and Tosun (2021) and Tosun and Howlett (2022), we contend that the publicly available SGI dataset by the Bertelsmann Foundation offers a viable operationalisation of national policy styles. The SGI data is constructed on the basis of reports prepared by country experts (Schraad-Tischler and Seelkopf 2016). Although the dataset was not constructed to measure national policy styles (for an overview, see , it includes at least two indicators, strategic planning and societal consultation, that can be used as proxies for the two dimensions of the original concept of policy styles. Although the SGI dataset has been available since 2009, there have been changes to the methodology used to produce it, which limits the comparability of the data over time (Croissant and Pelke 2022). Therefore, we must start our analysis using data for the year 2014. We acknowledge that this is a relatively short time period, but it is satisfactory for the illustrative purpose of this study.
We approximate the first dimension of the original conceptualisation of policy styles-the anticipatory vs. reactive approach to problem-solving-using the SGI indicator strategic planning. The scores for strategic planning are gleaned from answers country experts provide to the following question: "How much influence do strategic planning units and bodies have on government decision-making?" The answers are grouped into four response categories, to which we will subsequently refer to as "high," "moderately high," "moderately low," and "low": "Strategic planning units and bodies take a long-term view of policy challenges and viable solutions, and they exercise strong influence on government decisionmaking" (scores 9-10); "Strategic planning units and bodies take a long-term view of policy challenges and viable solutions. Their influence on government decision-making is systematic but limited in issue scope or depth of impact" (scores 8-6); "Strategic planning units and bodies take a long-term view of policy challenges and viable solutions. Occasionally, they exert some influence on government decision-making" (scores 3-5); "In practice, there are no units and bodies taking a long-term view of policy challenges and viable solutions" (scores 1-2). Figure 1 shows the difference in governments' strategic planning capacities between 2014 and 2020. With most countries (almost 60%), the relative emphasis on strategic planning capacity did not change over the time period studied. Turning to K the countries where experts indicated a change in emphasis on strategic planning, the degree of change was minimal in most countries (an increase or decrease of 1 point). However, seven countries experienced greater degrees of change, an observation we will return to when explaining our case selection.
We approximate the second dimension of national policy styles-impositional vs. consensual relationship between government and interest groups-using the SGI indicator societal consultation. This is assessed using the country experts' responses to the following question: "Does the government consult with societal actors in a fair and pluralistic manner?" Answers are grouped into four categories, which we refer to as "high," "moderately high," "moderately low," and "low": "The government always consults with societal actors in a fair and pluralistic manner" (scores 9-10); "The government in most cases consults with societal actors in a fair and pluralistic manner" (scores 6-8); "The government does consult with societal actors, but mostly in an unfair and clientelist manner" (scores 3-5); "The government rarely consults with any societal actor" (scores 1-2). Figure 2 presents the country scores for this second indicator and how it changed between 2014 and 2020. Compared to strategic planning, societal consultation was more dynamic during the observation period, and only 40% of cases had a value of Fig. 2 Changes in societal consultation, 2014-2020 (Own elaboration, with data taken from https://www.sginetwork.org/2020/Publications) zero for temporal change. The interesting cases are those distributed outside the two vertical lines, as the changes in their scores are greater than 1.
We operationalise changes in national policy styles by examining trends in the SGI scores for the chosen indicators. In contrast, the operationalisation of a political leadership effect is more elusive and requires a two-step approach. First, we must determine whether there was any political leadership effect at all. If so, we must then assess the mechanisms by which political leadership resulted in a change in strategic planning and/or societal consultation. Potential mechanisms include political leaders' beliefs and convictions, the need or wish to resolve a given policy problem, and the need or wish to respond to events and changing circumstances (see Helms 2012).
Usually, obtaining this kind of data is challenging. However, the SGI dataset provides country scores as well as annual country reports in which the experts document their coding decisions and provide background information on the individual countries. Thus, together with secondary sources such as scientific publications, the country reports provide a rich empirical basis for determining whether and how political leadership mattered with respect to changing governments' approaches to strategic planning and societal consultation.
Interestingly, political leadership has been acknowledged and studied predominantly in presidential systems and has received less attention in the study of parliamentary systems (Helms 2004). However, it should be noted that not only the differences between parliamentary and presidential governmental systems matter for political leadership but also the degree of institutional complexity, as highlighted by Lijphart (2012) in his influential distinction between consensus and majoritarian democracies.
Because studies in political leadership have stressed the differences between presidential and parliamentary systems (Blondel 1987;Elgie 1995;Helms 2004), we limit our explorative and strictly illustrative analysis to one political regime type. The impact of political leadership can be perceived as less significant in parliamentary systems that lack presidents with executive powers. Consequently, an analysis of parliamentary systems offers a more demanding test of leadership effects, and therefore we concentrate on this political regime type in our study. If we find leadership effects on national policy styles in parliamentary systems, we can expect to find even more pronounced leadership effects in presidential systems.
Our case selection is based on the data presented in Figs. 1 and 2. When inspecting Fig. 1, the United States demonstrates the greatest change in its strategic planning between 2014 and 2020. However, the United States is a presidential system, and therefore we had to exclude it from our analysis. The same applies to South Korea in Fig. 2 on the changes in societal consultation between 2014 and 2020.
In Fig. 1, Italy and Poland stand out since they experienced a marked reduction in their strategic planning scores, while Malta changed considerably in the opposite direction. Further, we can see that societal consultation particularly deteriorated in Poland, whereas it became more pluralistic in Ireland and the United Kingdom. Of these cases, we chose Malta, Italy, Ireland, and Poland, as this gave us one country per dimension where the scores increased and one where they decreased between 2014 and 2020. Malta's score for strategic planning increased, whereas Italy's score decreased. Poland was assigned relatively low scores for societal consultation during the observation period. Lastly, Ireland is the country we investigated in depth to assess how political leadership has led to increased societal consultation. We decided against including the United Kingdom in our analysis due to Brexit-related institutional changes that have taken place since the June 2016 referendum and thus would likely have biased the findings (see Richardson 2018a;2018b).
Ireland, Italy, and Malta are pure parliamentary systems, but the classification of Poland's semipresidential system is more ambiguous (see Duverger 1980). The Polish constitution defines it as a parliamentary republic, but the president of Poland appoints the prime minister and can veto legislation. Nevertheless, the prime minister and the cabinet propose legislation and have the power to reorganise government and governance arrangements. Therefore, we argue that for the purpose of this study, it is acceptable to treat Poland as a parliamentary system (see also Schwindt-Bayer and Tavits 2016, p. 39).

Insights from the Country Vignettes
The SGI data consulted reveal instances of changes in both strategic planning and societal consultation, as well as pinpoint their temporal variations (see Figs. 1 and 2). That the variation is more marked for the dimension on the consultation of societal actors might come as a surprise, since the relationship between the governments and interest groups should, in principle, be more difficult to change than the degree to which a government is anticipatory or reactive. What confirms the overall expectation of national policy styles is that in the majority of cases, the changes observed in the SGI data on strategic planning and societal consultation are minor to moderate in degree.
Does political leadership explain these temporal variations in policy styles? And if so, what are the mechanisms underlying this relationship? In what follows, four country vignettes examine these questions. All vignettes contain the same set of elements, though not always in the same order since this permits us to emphasise different aspects. They report the SGI indicators of interest and how they changed over the observation period as well as identify the political leaders under whom the changes occurred and their corresponding mechanisms. The vignettes further capture the opportunity structure for political leadership and, if applicable, allude to additional variables that need to be considered in order to explain the changes in the SGI indicators.

Malta: Increased Strategic Planning
In 2014, Malta's strategic planning was considered low. However, it increased gradually though continuously, first from 3-4 points in 2015, then from 4-5 in 2016, from 5-6 in 2017, and then from 6-7 in 2018 before stabilising at this level. The increase in the country's strategic planning took place during the government of Joseph Muscat of the Labour Party (2013-2020) and represents part of a comprehensive and continuous public sector reform (Bezzina et al. 2021).
Each government ministry has a unit responsible for strategic planning, and the Management Efficiency Unit coordinates the separate ministry plans. However, the prime minister's office is the central strategy office. After 2014, strategic planning strengthened in the fields of domestic, education, and fiscal policy. Further, the government invested in vertical and horizontal coordination across the government entities (Bertelsmann Foundation 2017, p. 17).
Formerly ruled by the British, the political system of Malta demonstrates numerous similarities with the Westminster system. The most important feature is the pure and polarised two-party political system comprising the Labour Party and the Christian Democrats, which alternate in power (Harwood 2020, pp. 4-5). The prime minister has considerable room to manoeuvre, as he or she has profound executive dominance and is not obliged to form a coalition government with another party (Veenendaal 2019). Given the political opportunity structures, Malta presents a case in which the political leadership of the prime minister should be particularly effective. And indeed, the changes in strategic planning can be attributed to a deliberate decision made by the prime minister.
The government's investment in strategic planning was driven by the Maltese Presidency of the Council of the European Union (EU) between January and June 2017 (Pirotta et al. 2019(Pirotta et al. , 2020. The Presidency of the Council of the EU prepares and chairs the meetings of the various sectoral Councils of the EU. Governments need considerable capacity to be able to manage these duties. Small EU member states like Malta have fewer capacities and therefore must invest significantly to achieve the necessary level. At the same time, the Presidency of the Council of the EU offers a window of opportunity for governments of small member states to shape the EU's political agenda (Panke and Gurol 2018). Therefore, over one and a half years earlier, the Maltese government took steps to improve strategic planning and also undertook other reforms to strengthen the public management sector (Bezzina et al. 2021, p. 172).
During the 6-month presidency, the government was able to place several policy issues on the EU's political agenda, including issues related to asylum and migration (Panke and Gurol 2018). The government considered its Presidency of the Council of the EU a success and maintained the new structure for strategic planning after terminating its turn (Pirotta et al. 2020).
In addition to this external driver, there were some domestic drivers for investing in strategic planning. Most importantly, the government regarded improved strategic planning as a means of reducing public debt as well as of modernising and improving the country's education system (Bertelsmann Foundation 2017, p. 17).
Taken together, we can state that the prime minister reformed Malta's strategic planning to increase the government's problem-solving capacity and to prepare for the Presidency of the Council of the EU, which did not only entail burdens for the government bodies but also provided a window of opportunity to shape the EU's political agenda.

Italy: Decreased Strategic Planning
During the observation period covered in this analysis, Italy had three different electoral laws in place (Chiaramonte and D'Alimonte 2018). The country experimented with electoral reforms in response to the notorious instability of Italian multiparty cabinets (see Ceron 2016). Indeed, four different prime ministers were in power between 2014 and 2020: Enrico Letta (Democratic party; 2013-2014), Matteo Renzi (Democratic party;2016-2018), andGiuseppe Conte (independent;2018-2020. Within the Presidency of the Council of Ministers, the government body, there exists a special department that oversees governmental planning. In addition, the Treasury contributes to strategic planning since it defines the budgetary goals and priorities (Bertelsmann Foundation 2015, p. 15). Prime Minister Letta relied on traditional advisors, such as trade unions and entrepreneurial associations. When Matteo Renzi became prime minister in early 2014, he centralised strategic planning and appointed a small team of economic and legal experts to advise him on policy strategies (Cotta et al. 2015). This group became a de facto policy unit inside the Presidency of the Council of Ministers and was in charge of coordinating policy proposals in labour, school, and pension policies (Nannicini et al. 2019). Another noteworthy change was that the minister of finance had to negotiate with the prime minister before implementing strategic plans.
When Paolo Gentiloni came into power in 2016, the minister of finance began to play a more significant role in strategic planning (Cotta et al. 2018), but the overall approach to strategic planning remained the same and changed only after the electoral victory of the Five Star Movement and the (Northern) League. The Conte government paid little attention to strategic planning but focused instead on a rigid government contract negotiated between the parties forming the coalition government, namely the Five Star Movement and the (Northern) League (Cotta et al. 2020).
The developments detailed above are also reflected in the scores given to Italy by the pertinent experts, who rated it moderately high in strategic planning from 2014 to 2018 (score: 6). From 2018 to 2020, the scores for strategic planning deteriorated to moderately low (scores: 5 and 4).
It is worth noting that between 2014 and 2018, the governments' commitment to strategic planning did not change even though the actors in charge of it did. The SGI country experts argue that these changes in Italy's strategic planning can be attested to the leadership styles of the individual prime ministers. The most marked change took place when the prime ministership changed from Enrico Letta to Matteo Renzi, as the latter introduced a new, highly personalised leadership style characterised by rapid political choices and a willingness to undertake reforms and to engage directly with the Italian citizens (Capano and Pavan 2019;Salvati 2016). From this perspective, it is plausible that he centralised strategic planning and asked a small group of trusted experts to advise him. The Gentiloni government then reinstituted the classic approach to strategic planning.
In the Conte government, the political leaders were the deputy prime ministers-Matteo Salvini of the (Northern) League and Luigi Di Maio of the Five Star Movement-not the prime minister himself. The government mainly had to focus on balancing the interests of these two parties, which had very different ideologies and policy agendas (Cotta et al. 2020). The Five Star Movement is characterised by an ideologically eclectic mix of policy positions and does not clearly locate itself on either the left or the right of the party system (Caiani and Graziano 2019), whereas the (Northern) League is a far-right populist party with a heavy emphasis on antimigration policies (Caiani and Graziano 2016). However, it was not only the ideological differences that made this government constellation difficult, but the "position-taking" activities that Di Maio and Salvini undertook to signal to their own party constituency that they were defending their interests (Pedrazzani 2018, p. 8). Each of the two deputy prime ministers strived to obtain as much power as possible within government, which resulted, among other things, in a weakening of the role of the minister of finance, who had played an important role in strategic planning in the Gentiloni government.
Consequently, the Italian case shows how the absence of political leadership by the head of government in the presence of intragovernment tensions can result in a decrease in strategic planning.

Ireland: Increased Societal Consultation
In the Republic of Ireland, the prime minister is the head of a multiparty government. She or he can shape policies to the degree to which the political parties in the coalition government agreed (Debus 2007  The highest rise in the country's societal consultation scores took place between 2014 and 2015, when it increased from moderately low (score: 4) to moderately high (score: 6). In 2016, the country experts assigned Ireland the slightly lower score of 5, which increased again to 6 in 2017 and remained at that level until the end of the observation period. Therefore, Enda Kenny was the prime minister under whom the changes in societal consultation took place.
To understand the changes in societal consultation in Ireland, it is important to take into account the fact that Ireland was affected by the economic and financial crisis that began unfolding in 2007/2008. Consequently, the government had to adopt austerity measures and cut public spending. In this context, the government downsized the public sector by departing from the tripartite social pact involving private sector employers, which had been in place since the 1990s. Instead, the government reached a deal with the public executive committee of the Irish Congress of Trade Unions (Hardiman and Regan 2013). This represented a paradigmatic shift because until then, trade unions had played an important role in Irish politics, and their involvement was considered to stabilise the relationship between government and industry. How important the involvement of trade unions was can be seen by the fact that economic policy was made in collaboration with them and the Department of the Prime Minister. In 2008, however, the competences for making economic policy shifted back to the Department of Finance as a result of the economic and financial crisis (Regan 2017, p. 123).
When the economic situation began to improve, Prime Minister Kenny resumed consulting policies with societal actors more broadly, especially in relation to tax issues. A public sector strike and a threatened strike by the police force in 2016 induced the government to restrengthen the involvement of societal groups (Murphy et al. 2020). Fuelled by the improving economic situation, these groups demanded the successive governments led by Leo Varadkar and Micheál Martin to grant them an even more influential role in (economic) policy-making. However, both governments refused them (Murphy et al. 2020).
The Irish case illustrates the importance of exogenous events more than the importance of political leadership. The government had to take drastic steps to prevent the country from going bankrupt, which resulted in a change of the style of societal consultation. The prime minister restrengthened societal consultations because of (the threats of) strikes in the public sector. Political leadership, on the other hand, only played a role in entrenching the postcrisis consultation model and preventing the government from going back to an arrangement in which societal groups, especially employers and trade unions, would exercise substantial influence over (economic) policy.

Poland: Decreased Societal Consultation
Poland is a parliamentary system with some aspects of a presidential one. Specifically, it constitutes what Duverger (1980) calls a (weak) semipresidential system, which exists when a popularly elected president retains some powers over the nomination of a prime minister and/or the dissolution of parliament. During the observation period, Poland had four prime ministers: Donald Tusk (Civic Platform;Ewa Kopacz (Civic Platform;-2015; Beata Szydło (Law and Justice;2015-2017), andMateusz Morawiecki (Law andJustice;2017-present).
The arrangements as they were in place in 2014 form the departure point of our analysis. Matthes et al. (2014) characterise these as ones in which, during the post-Soviet period after 1989, the government had to consult with interest groups before proposing reform projects. To this end, the government even created a system of online consultations, which gave interest groups that participate in the process the chance to follow a policy proposal through the legislative process. In practice, however, the consultation of societal actors was often formalistic, and interest group relationships were more corporatist than pluralist in nature. This can be seen in the Council of Social Dialogue (which replaced the Tripartite Commission in 2015), which is composed of trade unions and employers whose members are appointed by the president.
With the government of Beata Szydło, the relationship between societal actors and the government changed significantly, in a more centralised direction. While the de facto participation of societal actors was already lower than stipulated by law, the Szydło administration began to deliberately bypass consultation procedures by introducing bills directly through members of parliament; in such cases, the "regular" consultation mechanisms become obsolete. Another strategy adopted was to push for the swift passage of important laws and to have the time constraint reduce the scope for involving societal actors. Consequently, even previously privileged societal actors, such as the NSZZ Solidarność trade union, have been excluded from the legislative process (Matthes et al. 2020). A factor that facilitated these institutional changes refers to Andrzej Duda of the Justice and Law party becoming elected as president. Thus, since 2015, both the prime minister and the president have belonged to the same party.
The Morawiecki administration continued with the trajectory that began with the Szydło administration. The experts reduced the score for societal planning in Poland from 7 to 5 (in 2017), to 4 (in 2018-2019), and even further to 3 (in 2020). The decline in the country's societal consultation is remarkable and is also paralleled by a sharp reduction in strategic planning, as shown in Fig. 1.
These changes align with the Law and Justice party's overarching aim of overcoming the "liberal consensus" in place since 1989 by instituting an executive-centred political regime (Bill and Stanley 2020). Thus, the mechanism through which the political leadership impacted societal consultation is the ideological convictions and beliefs of Beata Szydło and Mateusz Morawiecki.  Table 1 summarises the main findings of the four country vignettes. Political leadership explains the changes in strategic planning and societal consultation in Malta, Italy, and Poland. In Ireland, the more important factor was the austerity regime established in 2008. For Ireland and Malta, we were able to observe that the new institutional arrangements became entrenched. As concerns Italy, we lack data for the period after the Conte administration, which ended in 2021. Nonetheless, we expect strategic planning to have improved under Mario Draghi, who became prime minister in 2021. The future trajectory of societal consultation in Poland is more difficult to predict because it will depend on whether the next prime minister will come from a different party than Law and Justice.
To assess our expectation that deviations from predominant policy styles will be short term, it is useful to go back to Figs. 1 and 2. These have shown exactly this pattern, thus supporting our overarching argument. The cases of Malta and Ireland have further shown how new institutional arrangements become entrenched, which aligns with the "critical junctures" logic in historical institutionalism (e.g., Béland 2007). In our view, the occurrence of "lock-ins" is compatible with the notion of national policy styles, as they create those institutional equilibria that are necessary for sustaining certain styles.

Conclusion
For many years, the concept of national policy styles put forth by Richardson et al. (1982) lay dormant. However, in recent years a growing body of research seeking to understand the nature of policy trajectories and continuities has been taking a fresh look at the concept. This has involved efforts to extend the concept beyond the few Western European countries examined in early work (Howlett and Tosun 2019a) and efforts to clarify and reform many aspects of the original formulation (Howlett and Tosun 2021b).
While these efforts have greatly improved our understanding of the origins and impact of policy styles in maintaining similar policy-making processes and outcomes over time, one of the questions on which the original notion of policy styles was silent was whether, and how, policy styles changed. That is, from a conceptual viewpoint, the notion of styles stresses stability and routinisation, although a style also is a dynamic phenomenon, with an identifiable origin and the ability to evolve, albeit normally slowly, over time. Thus, for example, one of the phenomena observed in the styles literature is the role played by macrofactors such as Europeanisation (Candel et al. 2020;Liefferink and Jordan 2004;Richardson 1982) in affecting national patterns and practices of administration and policy-making and, in the specific case of the post-Brexit United Kingdom, de-Europeanisation (Richardson 2018a, b). Changes in the political regime were thus thought to influence national policy styles (Ertugal 2021), but exactly how this occurred remained largely unexplored.
In this study, we have treated these findings in the literature about the nature of short-term political and long-term institutional effects on policy styles as the starting point and assessed to what extent observable patterns of stability and dynamics of national policy styles can be explained by these factors. In particular, our study, following recent insights by policy research that draws from concepts in psychology (see, e.g., Leong and Howlett 2022;Sulitzeanu-Kenan and Zohlnhöfer 2019) and sociology (see, e.g., Hornung et al. 2019), examines the importance of individuallevel and, especially, behavioural factors in affecting both long-term and short-term changes or deviations away from an existing style.
Political leadership in particular provides a promising analytical lens for providing a microfoundation for understanding the interrelationships between structures and behaviours that produce, maintain, and alter national policy styles. Leadership is a dynamic concept, and it is well known that the ways political power is exercised varies from government to government as well as during the incumbency of the same government. Leadership, therefore, may well often work within an existing policy style but may also from time to time be able to alter or change it. That is, the decisions by political leaders can affect policy styles either directly or indirectly through leaders' beliefs and actions, their need to resolve a given policy problem immediately, and their need to respond to events and changing circumstances in a timely and effective way.
To probe the plausibility of our argument, we utilised the SGI database and its measurement of the countries' strategic planning and societal consultation to examine the two dimensions of national policy styles set out by Richardson et al. (1982). The data for the observation period, 2014-2020, as expected, revealed many instances of stability concerning these two indicators, reinforcing the idea that national policy styles exist and transcend sectoral and other sources of volatility and variation, but the data also highlighted some instances of substantial change in these two indicators. More detailed analyses of countries in which these variations occurred demonstrated that political leadership is a major factor helping to explain the changes observed in their approaches to strategic planning and societal consultation.
In our view, the findings reported here are the starting point of a new line of research on the patterns of change and stability in national policy styles and, indeed, other styles of administration and governance more generally (see Howlett and Tosun 2021b). Ideally, future research would extend the operationalisation and measurement of the two dimensions of national policy styles and produce additional country-comparative data on the effects of political leadership. Thus, for example, while the present study assessed only whether political leadership brought about changes in strategic planning and societal consultation, future research could not only operationalise national policy styles but also leadership styles more precisely to afford a more systematic assessment of the research questions addressed here. Similarly, starting with an influential volume by Streeck and Thelen (2005), institutionalist analysis has developed a range of tools for characterising and explaining institutional change (see, e.g., Béland 2007;Mahoney and Thelen 2010;Wiseman and Béland 2010). These conceptual tools could also be usefully targeted towards the study of national policy style dynamics and change.