Windows of opportunities

In the original framework, windows of opportunity' narrowly refer to events in the problem stream such as crises or disasters or changes in the political stream such as a shift in national mood, change of administration, a change in personnel, or a shift in the partisan or ideological distribution of seats in Congress (Kingdon 1984: 176-8). However, the explanatory power of the multiple-streams framework can only be strengthened if the concept of windows of opportunity can also include other types of environmental pressures. This study enhances the concept of windows of opportunity' by bringing in the concept of environmental triggers developed by Cortell and Peterson (1999). Environmental triggers refer to domestic or international events, whether a crisis or non-crisis, alone or in combination, that open windows of opportunity for policy makers to make institutional change possible (Cortell and Peterson 1999: 183-5). Such events discredit existing institutions and policies and hence produce pressure for change or reform (Cortell and Peterson 1999: 184). International environmental triggers include (1) war, (2) geopolitical conflict, (3) a changing balance of power, (4) macro-economic change, dislocation, and shock, (5) international norms, organizations and treaties, and (6) technological change (Cortell and Peterson 1999: 185). Domestic environmental triggers include (1) revolution, (2) civil war, (3) coup d’etat, (4) election/change of government, (5) economic growth rate, (6) demographic change, and (7) social movement or conflict (Cortell and Peterson 1999: 185). According to Cortell and Peterson (1999: 186), environmental triggers may generate costs measured in terms of political, social, or economic dislocations that lead to demands for changes in policy practices. They may also shift

Population ageing, financing, and ETC in Asia 15 national mood and increase policy makers’ autonomy from short-term political constraints to innovative (Cortell and Peterson 1999: 187). Nevertheless, whether policy makers would act on a window of opportunity depends on their perceptions of the desirability, feasibility, and scope of change (Cortell and Peterson 1999: 188) and whether the prevailing institutional arrangements ‘create opportunities for, or place limits on, an official’s ability to translate her desire for structural change into policy’ (Cortell and Peterson 1999: 190).

Historical institutionalism

In the original framework, there is an insufficient attention to the role of institution in the policy and political streams. While the role of institutions may be less decisive at the agenda-setting stage (Zohlnhôfer et al. 2016: 248), the role of institutions matters most at the decision-making stage because ‘formal institutions are those that are assigned decision-making or veto powers’ (Zohlnhôfer et al. 2016: 244). Institutions ‘define which actors participate in which constellation in the adoption of policies and which policy options are institutionally permissible’ (Zohlnhôfer et al. 2016: 247). The role of institutions is particularly important in the political stream because it determines whether policy cntrepreneur(s) can gain the majority support needed to get their pet project adopted (Zohlnhôfer et al. 2016: 249).

In response to the criticisms that the multiple streams framework is ahistorical and neglects the impact of institutions on policy-making, this study combines the framework with the theoretical lens of historical institutionalism to examine the distinct role of institutions in affecting policy entrepreneurs’ goal-setting, policy preferences, and strategies; defining the channels in which ideas can be absorbed and translated into policy; and responding to the effects of environmental pressures or shifting contextual conditions. Historical institutionalism emerged in the early 1990s as a distinct tradition of institutional analysis to examine the roles of institutions and history in structuring and shaping political behaviors and outcomes over time (Fiorctos et al. 2016; Steinmo 2008: 118). In general, historical institutionalists define institutions as “macrostructures and institutional assemblages like the welfare state or national economic systems” (Fiorctos et al. 2016: 6), formal organizations, the rules or conventions promulgated by formal organizations, standard operating procedures, routines (Hall and Taylor 1996: 938), informal rules and norms (Steinmo 2008: 124). Institutions define ‘appropriate behavior for specific groups of actors in specific situation[s]’ (March and Olsen 1998: 948).

Institutions play three important roles. First, institutions play a determinant role in distributing power among actors in a given polity (Luk 2014a: 8). Historical institutionalism emphasizes asymmetrical power relations (Luk 2014a: 8). It ‘sees different institutional structures as setting different political rules of the game’ (Shih el al. 2012: 307) that privilege some actors and groups in society' while putting others at a disadvantage (Thelen and Steinmo 1992: 2). Institutional position, responsibilities, and relationships with others affect how actorsinterpret and pursue their self-interests, define their goals, choose their policy preferences, and determine their strategies (Luk 2014a: 8). Due to uneven distribution of power, some actors ‘are central to the development of policy whereas others are ineffectual, forced to accept gains and losses determined by others’ (Pierson 1994: 40).

Second, historical institutionalism sees political institutions as structuring the flow of ideas and shaping the absorption and diffusion of ideas (Hall 1992; King 1992; Thelen and Steinmo 1992; Weir 1992). The structure of political institutions, the formal rules and procedures governing policy-making define the channels and mechanisms by which ideas penetrate the policy-making process and are promoted, adopted, and implemented as policy (King 1992; Thelen and Steinmo 1992: 24; Campbell 2002: 30). For example, decentralization and institutional fragmentation of the American political system empowers business actors to oppose progressive economic and social ideas (Beland 2009: 708).

Third, institutions respond to changes in the external environment by changing their meaning and functioning (Thelen and Steinmo 1992). Open-systems theory, which assumes considerable interdependence between institutions and their environments, sees changes in the latter triggering adaptive responses within the institutions (Milakovich and Gordon 2013: 164). Institutions constantly need to balance pressures, responses and demands, and resources to ensure their stability in the long run, which permits continued functioning in the manner expected by leaders, workers, and people outside the institutions (Milakovich and Gordon 2013: 165). Crises or shifts in external conditions do not necessarily lead to dramatic institutional change, such as institutional breakdown (Thelen and Steinmo 1992). Instead, gradual or incremental institutional change can take place. The concept of‘institutional dynamism’ indicates that changes in the broader political or socioeconomic context enable previously latent institutions to suddenly become salient or existing institutions being put in the service of different ends when old or new actors are empowered to pursue new goals or adopt new strategies (Thelen and Steinmo 1992: 16-18).

Path dependence is another important concept in historical institutionalism. Path dependence refers to ‘specifically those historical sequences in which contingent events set into motion institutional patterns or event chains that have deterministic properties’ (Mahoney 2000: 507). ‘With path dependence, each step in a particular direction makes it more likely that a unit will continue to follow that same direction’ (Mahoney et al. 2016: 82). It means that there is ‘an inertial tendency towards continuing along the same path rather than changing direction regardless of the advantages of the alternatives’ (Luk 2014a: 9). The inertial tendency can be explained by one or a combination of several types of self-reinforcing mechanisms, including investment effects, learning effects, coordination effects, and adaptive expectation effects (Pierson 1994, 2000). Investment effects mean that high set-up or fixed cost provides individuals with a strong incentive to stick with a single option in order to exploit the option to the full extent, and further investment will be made to support this option. Learning effects is learning by using leads to accumulate knowledge, skills, and experience. Efficiency gained

Population ageing, financing, and ETC in Asia 17 from iterations and cost reduction because of a more skillful performance (faster, more reliable, and with fewer errors) creates a strong incentive for individuals to stick with the chosen option (Sydow et al. 2009: 700). Coordination effects ‘build on the benefits of rule-guided behavior’ (Sydow et al. 2009: 699). More actors adopting a specific institution such as the right-hand traffic rule can significantly reduce coordination costs because their behaviors can be anticipated and reactions can be considered in advance (Sydow et al. 2009: 699). Adaptive expectation effects ‘relate to the interactive building of preferences’ (Sydow et al. 2009: 700). Individuals are willing to adopt a particular practice because they expect others to do the same and wish to end up on the winning side (Sydow et al. 2009: 700). The subsequent self-fulfilling prophecies make a particular practice more attractive (Sydow et al. 2009: 700). These self-reinforcing mechanisms create positive feedback loops that ‘essentially remove certain alternatives from the menu of political possibilities’ (Pierson 2004: 12) and hence ‘forsaken alternatives become increasingly unreachable with the passage of time’ (Pierson 2004: 13). This indicates that timing and sequencing of particular events matter because subsequent self-reinforcing processes generate irreversibility, affecting the consequences of later developments (Pierson and Skocpol 2002: 701).

When the concept of path dependence is applied to understand a policy path, one needs to know that ‘a policy passed at time Tj may significantly change the range of possible options at time T2’ (Hacker 2002: 53). While it is important to know which decisions are taken, it is also important to know when those particular decisions are taken because the same decisions will not necessarily have the same effect at different stages of the path (Oliver and Mossialos 2005: 11). There are five conditions pushing policy development down a particular path. These include circumstances when ( 1 ) a policy generates the creation of large-scale institutions with substantial set-up costs; (2) a policy benefits sizable organized interest groups; (3) a policy embodies long-lived commitments; (4) institutions are of necessity densely interwoven with the broader features of an economy or society; or (5) features of the environment within which a policy is implemented make it harder to respond to unanticipated policy outcomes (Hacker 2002: 55). These conditions make policy reversal difficult to obtain.