Under which conditions can shocks stimulate transformative recovery: the Strategy Shock Implementation Reaction (SSIR) framework

The first phase of the SSIR framework maps out how policy strategies are developed. In the 1970 s, the strategy concept became popular in management studies, focusing on organizations (van Assche et al. 2021b). It is defined as a ‘consciously intended course of action,’ such as a plan or guideline to address a specific problem (Mintzberg 1987, p. 11). Strategies are found in all spheres of society, and the strategizing of individuals, organizations, communities, or states never stops (van Assche et al. 2020), such as the strategy ‘living with floods’ developed across the globe some decades ago or ‘the behavioral turn’ strategy in risk management (Sayers et al. 2015; Kuhlicke et al. 2020).

The strategy concept has gained more attention in recent years, particularly in the fields of sustainable development and environmental policy which require long-term policy solutions (Matheson 2009; Nilsson et al. 2016; Sachs et al. 2019; van Assche et al. 2020; Assche et al. 2021a, b). Strategies are described as “a vision for a desirable longer-term future” (van Assche et al. 2020, p. 696) of different public and private actors that is based on collectively binding decisions (e.g., consensus or a democratic majority) on how to get there. Examples of higher-level international strategies are the UN Sustainable Development Goals, the Sendai Framework on Disaster Risk Reduction and the COP21 Paris Agreement on climate change.

Strategies need not always result from a deliberate plan, however. While intended strategies are purposefully developed to address a policy problem (like the ‘living with floods’ strategy), strategies can also emerge in the absence of intentions as a “pattern in a stream of actions” (Mintzberg 1987, p. 12). Emergent strategies also result from unexpected opportunities or ad-hoc solutions to isolated cases of a policy problem that are consistently replicated and transferred to other similar contexts (Mintzberg and Waters 1985). Most strategies contain both intended and emergent elements, however (Neugebauer et al. 2016). Some strategies remain unrealized and are discarded; for instance, because they cannot compete with other strategies that are better aligned with the dominant policy narratives or better address the policy problem.

To operationalize the strategy concept, van Assche et al. (2020) introduce the institutional and the discursive dimensions of strategies. The institutional dimension proposes that strategies depend on and are shaped by institutions (e.g., a national or regional authority), which coordinate and integrate different discourses and actions of public and private actors in different policy domains. The discursive dimension describes a collective narrative, vision, or perspective about a desirable future and how to get there (van Assche et al. 2021b). At higher political levels and for cross-cutting and transboundary policy problems such as climate change, food security, or sustainable development, more coordination and integration towards a shared vision or goal is required (Candel and Biesbroek 2016; Peters 2018; Biesbroek and Candel 2019).

Policy narratives reflect worldviews of different actors, which structure and convey their understanding of the nature and causes of a policy problem and the potential policy solutions that could be included in a strategy, but also connect these specific views with their general ideology, identity and policy beliefs (Jones et al. 2014). Narratives come to the fore under unfamiliar conditions (e.g., during a shock event) or when facing highly uncertain future developments (e.g., because of impending environmental or socio-economic changes); they may provide orientation and align strategic action by diverse actors (Constantino and Weber 2021). Policy narratives can create enduring collaboration following a vision and can be used for monitoring progress and achievements (Mintrom and Rogers 2022). In the case of natural hazards, policy narratives could appear in the mental models of experts (risk as the combination of hazard, exposure, and vulnerability; IPCC 2014) and laypeople (perception of risk as dreadful, unknown, or uncontrollable; Slovic 1987).

Policy narratives are rooted, inter alia, in policy beliefs, that is, a set of values that follows a systematic pattern (Sabatier 1988; Jones et al. 2014). Policy beliefs reflect core principles and commitments actors have based on their own norms, values, and ideas that determine their preferences regarding a specific problem (Sabatier 1988; Biesbroek 2021; Kammermann and Angst 2021). Policy actors take action based on the shared set of beliefs (Zhou et al. 2021). Policy beliefs and narratives tend to be more stable than policy strategies, which may constitute a challenge for policy coordination and integration if new policies run counter to established beliefs (Hall 1993; Pierson 1993); yet, there are examples of narratives changing over time, such as the adoption of fishery concerns within EU development cooperation policy which realized mutual synergies towards improving livelihoods and food security (Candel et al. 2015) or the implementation of Nature-based Solutions, which can provide multiple co-benefits such as risk reduction, improving biodiversity or carbon storage (Debele et al. 2023).

In strategy development, different policy actors want their beliefs and worldviews reflected in policy outcomes at different levels (Zhou et al. 2021). Policy actors operate at multiple political levels in order to increase policy flexibility and efficiency (Marks and Hooghe 2000). Multi-level governance can be organized into two different types. The first type is characterized by shared power between governments representing clearly defined territories such as nation states and regional governments (Liesbet and Gary 2003). In this type, the development of policy strategies typically follows a nested approach such that lower-level strategies aim to be coherent with higher-level strategies. This type is common in the European Union, where policy reforms initiated and pushed at the supra-national level are to be transposed to national and regional levels (Cunha and Swinbank 2009), such as the Floods Directive or the Nature Restoration Law. The second type of multi-level governance starts from the policy problem, and power and decision-making are dispersed between multiple concerned levels. This type is characterized by low entry thresholds into the governance structure and highly flexible organization (Liesbet and Gary 2003). It results in problem-driven policy strategies that emerge from close collaboration between concerned actors, while external interactions are kept to a minimum. A prominent example of the second type is the governing of common property resources such as communal tenure in alpine meadows or water resources (Ostrom 2015).

However, policy actors rarely act alone. Actor coalitions are groups of policy actors from different institutions and backgrounds such as elected officials, individual citizens, civil servants, or representatives of interest groups, trade associations or corporations who share a set of policy beliefs and have decision power (Sabatier 2007). For instance, two actor coalitions formed in a recent reform of the Common Agricultural Policy conveying either core agricultural or environmental policy narratives (Loacker et al. 2025). Some individuals or corporate actors – so-called policy entrepreneurs (Kingdon 1984), change agents (Mintrom and Rogers 2022) or catalyzing agents (New et al. 2022) – lobby for their favored policy narrative, build actor coalitions and networks, and seize the opportunity to initiate action after a policy window has opened (Zahariadis 2007; Oborn et al. 2011), like the implementation of planned relocation to adapt to extreme weather events (Thaler et al. 2020). While policy entrepreneurs do not have decision power themselves, they possess a variety of skills to collectively shift dominant beliefs and norms (Westley et al. 2013). Policy entrepreneurs play a crucial role in directing the development of policy strategies, providing a common vision, building social networks and trust, leveraging resources, capturing the attention of policymakers, and bridging local, narrow problems to broad, cross-cutting issues (Zahariadis 2007; Oborn et al. 2011; Westley et al. 2013; Candel and Biesbroek 2016).

Most policy problems do not appear entirely unexpectedly but evolve and are addressed over time. System performance reflects the history before the shock happens – how a widening discrepancy between revealed conditions and stated objectives grows into a policy problem, and how actions to deal with the problem are taken as long as profound shocks are absent (Farley et al. 2007; O’Donovan 2017). On a floodplain for instance, the system performance may keep up by expanding dams and pumps as riverside wetlands are drained to erect additional settlements; however, once a flood event exceeds these protective measures and delivers a shock, flood risk management turns into a policy problem that necessitates more extensive measures, like early-warning systems or planned relocations. System performance comprises the repeated occurrence or accumulation of small hazard events, as well as the incremental policy steps taken to cope with these small events. In addition, system performance includes processes of policy learning, if earlier policy decisions are revisited in the light of new information (O’Donovan 2017), policy integration, if strategies are shaped by related or adjacent policy domains (Solecki and Michaels 1994), and mission-oriented policies (Mazzucato 2018), if strategies are expanded to span large societal challenges such as climate change.

The second phase of the SSIR framework describes how strategies are implemented after a shock. A shock is a sudden, rare, harmful, disruptive, and urgent event that (almost) overstretches current coping capacities (Grossman 2015; Dolan 2021). Shocks related to climate change are large-scale extreme weather events like hurricanes, floods, droughts or heat-waves. A shock originates outside the socio-political system under study (Cairney and Jones 2016; Hanger-Kopp et al. 2022); thus, routine events such as periodic elections or regular turnover in institutions do not qualify as shocks (Beland and Howlett 2016; Dolan 2021). A shock may be anticipated, e.g., with long-term monitoring systems, but its exact timing is random. A shock conveys urgency, as it speeds up the development and application of strategies (Grossman 2015). Shocks should not be overrated in their relevance; a shock alone rarely suffices to propel an issue to enduring prominence (Solecki and Michaels 1994).

Some policy problems do not culminate in singular shock events, however. Slow-onset issues like climate change persist as background stressors, incurring continuous degradation or system underperformance that amounts to a shock once the accumulated shortfalls become unbearable (Grossman 2015). The SSIR framework is tailored to rare and outstanding events, however; slow-onset issues often give rise to reciprocal framing contests between system performance and policy narratives, which makes it hard to disentangle directional influences (Dolan 2021).

Shocks typically open policy windows for transforming governance regimes, creating new visions, enabling learning and innovation, or changing behavioral patterns (Broto et al. 2014; Herrfahrdt-Pähle et al. 2020; Kanda and Kivimaa 2020). A policy window is a temporary period of rapid policy implementation; during this period, existing policy arrangements become instable and fluid if they are designed only for managing routine fluctuations in system performance (Penning-Rowsell et al. 2006; Jones et al. 2014). For instance, key personnel may extend their daily working hours or introduce shortcuts in decision-making when managing a shock. During a policy window, the likelihood of developing new strategies is higher than usual as niche actors may step up and propose alternatives if the regime turns dysfunctional under the changed conditions (Solecki and Michaels 1994; Herrfahrdt-Pähle et al. 2020). For instance, insurance companies may use a recent flood to lobby for public co-financing of insurance schemes. Policy windows are considered particularly relevant for diffusion, acceleration and upscaling of innovations for sustainability transitions (Köhler et al. 2019). However, policy windows might also have negative implications, for instance if the shock doctrine encourages extensive privatization and deregulation after an event as it occurred after Hurricane Katrina in 2004 in the United States (Klein 2008).

When the window opens, public and policy attention turns to the policy problem and may challenge or reframe the dominant policy narratives (Bubeck et al. 2017; Rose et al. 2020). Actors strive for fast recovery and symptomatic relief and may draw on large resource inflows that become available after the shock, for instance, international aid funds or private donations (Penning-Rowsell et al. 2006; Birkmann et al. 2010; Brundiers and Eakin 2018). Once the most pressing issues have been resolved, actor engagement winds down because continued activity no longer yields significant returns and media attention decreases (although the problem may still be unresolved), and the policy window eventually closes (Solecki and Michaels 1994; Farley et al. 2007; Grossman 2015).

Radical and catalytic change that questions core beliefs and agendas happens rarely during the timespan from the opening to the closing of a policy window after a shock (Solecki and Michaels 1994). Instead, policy windows tend to accelerate or modify ongoing change, by amplifying existing ideas and pre-signals (e.g., by transferring policy instruments from other contexts; Johnson et al. 2005; Thaler et al. 2020), by instigating renegotiation of leadership and institutional roles (e.g., by speeding up administrative decision-making for permits and budgets; Birkmann et al. 2010), or by shifting policy narratives to favor a particular approach (e.g., by promoting a specific promising technology; Kulmer et al. 2022).

Once a shock has occurred, policy actors must act (Mintzberg 1987) as we can see on different examples of extreme weather events across the globe. An implemented strategy includes a series of activities undertaken by the government to tackle the policy problem revealed by the shock, realizing the policy solutions prescribed by the strategy. Implemented strategies include land use plans, laws, or government subsidies and directly affect households or businesses. The implemented strategy is either unchanged with only marginal amendments or omission of selected policy solutions; revised with substantial reorientation or expansion; or newly emerging due to a shock. A strategy can also fail and not be implemented, for instance, if it is expected to underperform when applied to the current shock.

Policy strategies can be implemented at multiple governance levels. Coordination and integration between and across levels are important to avoid redundancies, fragmentation, incoherence, or inconsistency (Peters 1998; Candel and Biesbroek 2016). National governments need to coordinate to align efforts, prioritize risk management, and implement measures that enhance resilience and address systemic vulnerabilities (Lal et al. 2012). Coordination and integration are achieved by either restricting the number of involved actors (in the first type of organization in multi-level governance) or limiting the engagement with unconcerned actors (in the second type; Scharpf 1994; Liesbet and Gary 2003; see Section 3.1). In both types, competencies are distributed such that the costs of heterogeneity are balanced by economic benefits of scale. Transaction costs of gathering and disclosing information, reaching consensus and decisions, and maintaining infrastructure should be kept low (Marks and Hooghe 2000). To some extent, however, redundancies ensure the reliability and flexibility of an implemented strategy (Landau 1969; Peters 2018).

The third phase addresses how an implemented strategy translates into reactions of households and businesses to the problem. The SSIR framework posits an indirect relationship: implementing the strategy does not directly incur individual reactions, but changes the underlying motivations and perceptions, which subsequently lead to specific individual reactions. The SSIR framework draws on Protection Motivation Theory (PMT; Rogers 1983) to describe how individual reactions result from motivations and perceptions. The PMT has been applied to a range of natural hazards, including droughts (Truelove et al. 2015), landslides (Mertens et al. 2018), tornados (Weinstein et al. 2000), volcanic hazards (Covey et al. 2019), wildfires (Martin et al. 2008) and most prominently floods (Grothmann and Reusswig 2006; Kuhlicke et al. 2020). The PMT may explain reactions from diverse individual actors such as households (Babcicky and Seebauer 2019), farmers (Mitter et al. 2019; Kropf et al. 2024), or local authorities (Grothmann et al. 2013).

According to the PMT, individual reactions arise from the interaction between risk appraisal, coping appraisal, and non-protective responses. Risk appraisal indicates how threatened a household or business feels by a certain risk. It is composed of a cognitive (risk perception as the expectancy value of probability and severity of negative outcomes) and an affective (feelings of fear and worry) subcomponent. An implemented strategy may influence risk appraisal, for instance, by countering misperceptions how likely and how severely the shock may return, or by contextualizing individual shock experiences. Coping appraisal indicates how an actor evaluates possible responses to reduce the perceived risk. Coping appraisal includes the three subcomponents response efficacy, self-efficacy, and response costs. Response efficacy describes how effective a reaction is considered in reducing the expected negative outcome; for instance, whether investing in irrigation infrastructure reduces damages of recurring droughts. Self-efficacy refers to the perceived ability to carry out this reaction; for instance whether the farmer has the expertise and capacity to carry out the entire investment process. Response costs include the required financial resources, time, and effort for the planned reaction, including investment and operating costs. An implemented strategy may influence the efficacy and costs of individual reactions by means of financial schemes (e.g., subsidies for energy-efficient investments), voluntary or mandatory guidelines (e.g., energy-efficiency standards), counselling, or other instruments.

The risk and coping appraisal processes interact: Without a significant risk appraisal, actors would not see any need to take action. If high risk appraisal coincides with high coping appraisal, protective reactions are initiated. If high risk appraisal meets low coping appraisal, actors may feel overwhelmed and instead turn to non-protective responses. Non-protective responses are avoidant or suppressing reactions in order to downplay the risk (e.g., denial, wishful thinking) or to shift responsibility (e.g., fatalism, overreaching reliance on external support; Bubeck et al. 2013; Babcicky and Seebauer 2019). For instance, if farmers recognize the drought risk but cannot imagine an effective response they could take themselves, they may instead consider the government responsible for compensating drought-related yield losses. An implemented strategy that takes comprehensive governmental steps to mitigate the shock impacts may make people disregard their personal contribution, or policies of promise and appeasement may encourage denial and wishful thinking.

The PMT is well suited to explain reactions that include a subjective balancing of costs and benefits (Kuhlicke et al. 2020); however, the PMT should not be misconstrued as covering all motivational determinants of individual reactions. In the example of floods, further relevant factors are past experiences of hazard events (Osberghaus 2017; Thistlethwaite et al. 2018), role models and normative expectations conveyed by the social environment (Poussin et al. 2014; Bubeck et al. 2018), or social capital when informal community networks provide emergency assistance or facilitate access to external charity and support (Aldrich and Meyer 2014). The SSIR framework focuses on risk appraisal, coping appraisal, and non-protective responses, though, because these factors are well confirmed in the pertinent literature and can be directly targeted in an implemented strategy.

The PMT is open to being applied to any kind of individual reaction; however, the strength of influence of risk appraisal, coping appraisal, and non-protective responses may vary between individual reactions. For instance, in the case of flood preparedness, the impact of risk appraisal is small (Bubeck et al. 2012; Bamberg et al. 2017), whereas coping appraisal is essential (van Valkengoed and Steg 2019; Noll et al. 2021). While risk appraisal refers to the general, overarching threat, the impact of coping appraisal depends on the specific reaction considered (Babcicky and Seebauer 2019; Noll et al. 2021).

The SSIR framework does not specify which particular individual reactions stand at the endpoint of the framework’s three phases; however, for the purpose of applying the SSIR framework in climate resilience research, we propose transformation, maladaptation, backfire, and inaction as prototypical individual reactions. These four prototypical reactions may illustrate possible (mis)directions in integrating climate change adaptation and mitigation.

Transformation (‘build back better’) aims to improve an actor’s resilience to withstand shocks (Folke et al. 2016) or fundamentally change their capacities that are no longer adequate (Walker et al. 2004; Park et al. 2012). Transformation is also characterized by catering to previously underrepresented goals such as climate resilience or sustainability transition (Elmqvist et al. 2019). It expands the scope of action to indirect consequences of the shock and to possible overlaps with other domains by considering an actor’s double exposure (O’Brien and Leichenko 2000). Examples of transformation reactions are modifications of production systems or income diversification.

Maladaptation (‘build back short-sighted’) can be defined as intended actions that restrict their scope to the direct, short-term consequences of the shock and increase vulnerability or deteriorate the conditions for sustainability transition (Juhola et al. 2016). It includes quick end-of-pipe fixes to remedy the most pressing problems as well as implementing corrections that cannot counter ongoing degradation. An example of a maladaptation reaction is business operations focused on optimizing gross margin and gaining competitive advantage in narrow market segments at the expense of overall resilience.

Backfire (‘build back worse’) refers to reactions that are not just insufficient to accommodate future shocks, as maladaptation does, but that make the system worse off than it was before the shock. These reactions undermine instead of reconcile climate change adaptation and mitigation, or increase negative environmental or social externalities. Examples of backfire reactions are expansions in business operations to compensate for losses, resulting in additional greenhouse gas emissions or higher assets at risk. Backfire also manifests as rebound effect, when savings from more efficient provision of services by technology modernization are (over)compensated by subsequent increases in consumption (Sorrell 2007; Gomez and Perez-Blanco 2014).

Inaction (‘build back as before’) refers to the active or implicit decision to not perform actions (or to postpone actions indefinitely) that go beyond those already agreed upon before the shock happened. Uncertainties are often used as an excuse for remaining inactive (Howden et al. 2007; Gifford 2011). Inaction is characterized by restoring an actor’s pre-shock status, by depreciating damages or ignoring sunk costs, by remaining in inertia or with entrenched habits, and by continuing business-as-usual despite evidence that shocks will return in the future.

Planned relocation is usually organized between different political-administrative levels (Schindelegger 2019): National and regional authorities design and the local level implements the strategy. Policy entrepreneurs may play a core role if they show transformative leadership by encouraging and managing local agents such as affected households, neighborhoods, or grassroots organizations (Thaler et al. 2020).

During phase 1 of the SSIR framework, the policy problem manifested in an increasing number of buildings constructed in floodplains, indicating insufficient system performance. Moving buildings from high-risk flood hazard zones to no-risk zones was considered a policy strategy for adapting to this risk (Schindelegger 2019). Planned relocation is used across the globe, for example, in Austria, the United States of America, Australia, Vietnam, Japan, and Mozambique (Correa et al. 2011; de Sherbinin et al. 2011; Sipe and Vella 2014; UNHCR 2014; Kloos and Baumert 2015; Bukvic and Owen 2017; Barnett and McMichael 2018; Seebauer and Winkler 2020). Most communities resist this radical strategy, though, and instead turn towards other, less contested risk reduction measures (Thaler et al. 2020). The dominant policy narrative in flood risk management centered on cost-benefit assessments (Slavikova et al. 2021). Planned relocation adhered to this narrative because the one-time investment for removing old and constructing new buildings clearly pays off by precluding any repeated flood damages in the future. Planned relocation also matched with the narrative of shifting responsibility to homeowners as the state is no longer willing to provide comprehensive protection for everyone (Kuhlicke et al. 2020), and with the narrative of ‘room for the rivers’ and ‘making space for water’ (Warner and van Buuren 2011). The narrative in flood risk management was based on hierarchical policy beliefs, where the voluntary relocation of individual residents was combined with centralized technical engineering solutions (Thaler et al. 2020).

In line with phase 2, planned relocation was usually implemented after a shock, such as the floods in Austria in 2002 or 2013, New Orleans in 2005, Germany in 2016, or post-tsunami in Japan in 2011 (Iuchi 2014; Mayr et al. 2020; Thaler et al. 2020). In Austria, policymakers used the 2013 flood event as a policy window to implement the strategy as a financial compensation scheme (Schindelegger 2019; Thaler et al. 2020). The Austrian government offered all residents in the relocated Danube community up to 80% of the value of the old building and the demolition costs (Thaler et al. 2020). The strategy used in Austria in 2013 was mostly unchanged as it had already been implemented in adjacent downstream areas in 2002. By contrast, in Simbach in Germany in 2016, households were relocated as an emergent strategy as no previous plans had existed but the public administration used the momentum after the flood (Mayr et al. 2020). In Abruzzi in Italy, planned relocation was revised to also increase biodiversity in the abandoned areas, improve individual wellbeing, and decrease carbon emissions (Knobloch 2005; Micangeli et al. 2013).

In phase 3, relocation programs met the individual considerations of flood-prone households. In Austria, the risk and coping appraisal in the decision to leave or remain in the flood risk zone centered on emotional reasons on the one hand, such as traumatic flood memories and a personal bond to the place, and financial restrictions on the other hand, such as uncertain income and family prospects (Seebauer and Winkler 2020). Non-protective responses appeared as feeling helpless and impotent against an uncertain yet overwhelming flood threat and against powerful state actors (Thaler et al. 2020). The potential individual reactions span all four prototypical reactions mentioned in Section 3.3. As a transformation reaction, the new building could be constructed as a zero- or low-emission building that is also hardened to natural hazard events, for example in the sustainable energy supply system of the relocated L’Aquila municipality (Italy; Micangeli et al. 2013). Individuals may act in a maladaptive way when rebuilding their homes with only piecemeal adaptation measures or at the fringe of, or even inside the flood risk zone (Nalau and Handmer 2018). Backfire occurred in the Austrian case when new buildings were developed at a larger size, including swimming pools or other energy and carbon-intensive amenities (Thaler and Fuchs 2020). Both maladaptation and backfire reactions could be avoided by mandating flood-proof, energy-efficient rebuilding standards in the relocation strategy, or by linking financial compensations not just with risk reduction but also with energy efficiency. Most voluntary relocation programs result in partial inaction if some residents refuse to relocate, thus remaining at risk, raising public costs for maintaining infrastructure for the few who stay in the floodplain, and blocking the use of the floodplain for water retention. Restoring a building destroyed by flooding to its previous state without any improvements towards climate change adaptation or mitigation would also represent an inaction reaction.

Building renovation is typically promoted by national policy-makers, but implementing the strategy by paying out refurbishment subsidies or by enforcing building standards and the phasing-out of fossil fuel heating systems is allocated at the local governance level. Craftspeople for construction, plumbing and electrical systems are important intermediate actors who translate government guidelines into specific products and services they market to individual residents (Hecher et al. 2017).

According to phase 1 of the SSIR framework, the system performance clearly highlighted the policy problem that the thermal performance of the existing housing stock was insufficient and that widespread refurbishment was necessary, including insulating walls and roofs, installing double- or triple-glazed windows or retrofitting fossil-fuel heating systems. National climate monitoring continued to report that current efforts did not comply with carbon emission reduction targets (Umweltbundesamt 2021). In parallel, repeated heat waves increased heat stress and morbidity, in particular among vulnerable groups such as the elderly or people with cardiovascular and respiratory diseases (Kjellstrom et al. 2016). Well-insulated residential housing provides the double benefit of reducing heating demand in winter and keeping the building interior cool in summer. Consequently, policy strategies in Austria, the focus country of this example, called for extended efforts in building renovation, in climate and energy strategies (BMWFJ and BMLFUW 2010, Bundeskanzleramt 2013) as well as in heat protection plans (BMLFUW 2012; City of Vienna 2015). The dominant policy narrative was that the public sector should incentivize the renovation of privately-owned buildings. In contrast to the construction of new buildings, where strict energy standards were mandated (OIB 2019), renovating existing buildings was voluntary. This stemmed from policy beliefs in market forces, on the one hand, arguing that the role of policy just lies in providing subsidies to shorten the amortization period of efficiency investments, and in the sanctity of private property, on the other hand, arguing that policy should not impose restrictions on private possessions that had already been taxed and approved.

Phase 2 commenced with the shock of the 2008 global financial crisis. This shock was sudden and disruptive, as most market actors had not anticipated when and how hard the crisis would hit across all economic sectors. The Austrian government implemented a subsidy program for building retrofitting, mainly to support the construction industry (Amann et al. 2014); thus, as the economy recovered, subsidy budgets were cut substantially after 2014 (Seebauer et al. 2019). The original strategy remained basically unchanged: Because of voluntary participation and funding rates of ca. 30% of total renovation costs, the implemented strategy mainly reached mid- to high-income homeowners who could afford upfront investment costs (Schleich 2019). Despite continued subsidy provision, the retrofitting rate is still below policy targets (Umweltbundesamt 2021). In 2021, the program was revised by offering a 100% funding rate to low-income households to provide equal access to the program; however, it is still only homeowners who may apply, even though most low-income households are tenants (Seebauer et al. 2021). A similar renovation subsidy program in Italy even offered a 110% funding rate (Governo Italiano 2020). A possible future emergent strategy could be the inclusion of innovations in building technology in the funding criteria, for instance offering higher subsidies if recyclable and locally produced insulation materials made from straw are used.

In Phase 3, the renovation subsidy program intersected with the motivations and perceptions of households and thereby induced specific individual reactions. The risk appraisal was low: At this time, heating costs did not heavily burden household incomes, and energy poverty rates were lower than in other European countries (Statistik Austria & E-Control 2021); heat waves were not considered a severe threat (Babcicky and Seebauer 2016). The coping appraisal for renovating one’s home depended on available capital for covering upfront investment costs, ownership status, and replacement opportunities when the old fossil heating system broke down (Hecher et al. 2017). Thus, a frequent non-protective response was to shift the responsibility for renovation to landlords or housing associations. The implemented strategy may have led to different individual reactions which advance or hinder climate resilient outcomes: A transformation reaction would be retrofitting the entire building envelope, combined with a heat pump that also cools in summer and with façade greening or shading. Linking the renovation subsidy program with funding schemes for urban greening could incentivize this transformation reaction. A maladaptation reaction would be quick and partial insulation while retaining the old heating system, which is then overdimensioned for the reduced heating demand; since 2011, the Austrian subsidy program also funds partial retrofits, which may have unintentionally promoted this reaction. Other possible maladaptation reactions are installing a more fuel-efficient burner but still heating with oil or natural gas. Backfire reactions would be heating to warmer room temperatures because it is cheaper after the retrofit (Sorrell 2007), or taking the subsidy as a windfall profit when the home would have been remodeled anyway because of changed domestic needs (Wilson et al. 2015; Seebauer et al. 2019). Inaction reactions would be refraining from any structural improvement of the building and bearing the full brunt of climate change, heat waves, and energy prices. Another inaction reaction would be adopting only everyday behaviors for energy saving (e.g. dressing warmer rather than turning up the heating) or heat-adapted routines (e.g., shifting leisure activities to cooler daytimes); however, everyday behaviors are much less enduring and effective than structural modifications. Moreover, self-imposed austerity such as cutting heating to a bare minimum may hit vulnerable and marginalized households disproportionally (Eisfeld and Seebauer 2022).