Crisis Response via Dynamic Capabilities: A Necessity in NPOs’ Capability Building

Dynamic capability research increasingly seeks to identify mechanisms founding dynamic capabilities because this microfoundation provides options for influencing the application of dynamic capabilities. Dynamic capabilities in turn support organizations to achieve continuous organizational change. We pursue deeper insight into the microfounding mechanisms of dynamic capabilities with regard to management-related variables. Survey data from hot spots of refugee crisis 2016 all across Austria allow us to examine how managerial mechanisms on microlevel and dynamic capabilities on macro-level are linked and particularly how managers can influence the application of dynamic capabilities. Data evaluation is based on mediation analysis. Study reveals that participative leadership fosters dynamic capabilities of NPOs active in refugee aid. Further, we provide evidence that the manager´s perception, how self-determined he/she is and how impactful his/her actions are, enhances the application of dynamic capabilities.


Introduction
Times of crisis and political upheavals reinforce NPOs' roles as advocates, service providers as well as place for social engagement and thus highlights NPOs' position as central hub between government, affected individuals and other civil society actors (Simsa and Zimmer 2014;Meyer et al. 2010;Anheier 2005). The significance of NPOs' as civil society actor became especially evident in the 'refugee crisis' of 2015/16. Due to Austria´s geographic position and particularly its closeness to the 'welcome-nation' Germany, the country with a population of 8.7 million was confronted with more than one million refugees from Syria, Iraq and Afghanistan mostly transiting, but with over 100.000 also seeking asylum in Austria. 1 This huge amount of refugees by far exceeded Austria's governmental structures and its capacity in crisis and emergency management (CEM) 2 (Roth 2017, 6;Maduz and Roth 2017;2 ff.;Gratz 2016, 167 ff.). Particularly, Austria lacked an appropriate national CEM strategy, which regulates CEM across federal states. Hence, coordination of refugee aid across federal states was mainly informal (Gratz 2016, 91;Roth 2017, 7).
NPOs, particularly established rescue and aid organizations, strived to close this gap together with other civil society actors, such as grass-roots initiatives, self-organized volunteers (Simsa 2017, 79) and for-profit organizations, mainly infrastructure organizations and security firms (Gratz 2016, 142 ff.). In this context, NPOs were active in strategic political as well as in operative CEM. They substituted governmental deficiencies concerning strategic leadership and coordination of refugee aid as well as operative CEM issues, such as public security at national boarders (registration, transport, etc.) and primary care for refugees and asylum seekers (food supply, medication, etc.). NPOs also compensated privatized public duties, such as providing accommodation for asylum seekers (Maduz and Roth 2017;Frühwirth and Lachmayer 2015).
In sum, the 'Austrian refugee crisis' centered on coping with transit refugees and it was handled by a multiplicity of actors (Simsa 2017;Roth 2017;Maier and Ortner 2017). NPOs and other civil society actors predominantly fulfilled tasks, which exceeded governmental capacity in CEM or tasks, which have not been legally determined and were left to informality (Roth 2017, 7;Gratz 2016, 91).
NPOs success in coping with the refugee crisis, particularly its capacity to handle the dynamics during these times, was substantially based on their experiences in international disaster relief and on their huge pool of human resources, especially volunteers (Simsa 2017;Gratz 2016). In addition to resources, NPOs displayed organizational capabilities, which were substantial to handle refuge crisis. NPOs were able to provide time-critical status reports including the identification of opportunities and threats, even though there was a lack of official communication structures. They managed to set up coordination and cooperation structures between diverse actors, and they were able to combine different logics such as spontaneous activism of volunteers and planning approach of rescue units in decision-making processes (Gratz 2016, 29 u. 63 f.). NPOs also succeeded in (re-)combining resource pools innovatively to set up critical infrastructure (Maduz and Roth 2017).
Being able to cope successfully with the situation described strongly suggests the presence of specific organizational capabilities-dynamic capabilities (DCs). This is because DCs are, firstly, capabilities which enable an organization to sense changes-opportunities and threatsand define possible resource actions (Teece 2007(Teece , 2014. Secondly, DCs help an organization seizing opportunities by selecting the appropriate resource actions and then implementing these decisions. Thirdly, DCs transform or rather align resources to the environmental dynamics (e.g., clients', government's, other partners' needs) (Teece 2014;Jantunen et al. 2012). In short, DCs support organizations to adapt to changing conditions by contributing to an environment-organization fit and hence support organizations to compete in dynamic environments (Di Stefano et al. 2014, 315). Correspondingly, DCs are prevalently related to positive performance effects (Teece et al. 1997, Teece 2007Wang and Ahmed 2007).
While there is (some) research regarding performance effects of DCs in NPOs (Piening 2013;Kaltenbrunner 2018), it remains largely unclear, which intra-organizational mechanisms foster the application of DCs (Pavlou and El Sawy 2011;Sprafke et al. 2013). Particularly, the link between organizations as agents on a macro-level and managers as agents on a microlevel needs to be explored in more detail (Di Stefano et al. 2014). Gaining insights into this link-the microfoundation of DCs-is crucial because it provides options for influencing the application of dynamic capabilities (Felin et al. 2015).
Hence, we aim to explore the link between mechanisms on microlevel and DCs by answering the research question of how management affects the application of dynamic capabilities in NPOs active in refugee aid. A quantitative study among managers deployed in refugee camps all over Austria was conducted to answer this question. Based on the dynamic capability approach, we expect that NPOs coping successfully with the dynamics of refugee crisis applied dynamic capabilities and that managers' behavior affected their application.
Our paper makes the following contributions: Firstly, it provides evidence on how management and dynamic capabilities are linked, and thus adds to the highly claimed microfoundation of DCs in general (Sprafke et al. 2013;Teece 2007Teece , 2014 and in the context of NPOs in particular. We examine the microfoundation of DCs in NPOs, where research regarding the microfoundation of DCs is even scarcer. Secondly, we provide an integrative illustration of management-related microfounding mechanisms and introduce participative leadership as a hardly researched leadership style potentially enhancing DCs. Thirdly, the illustration of microfounding managerial mechanisms provides opportunities for influencing or rather managing the application of DCs. This is particularly important because DCs enable continuous organizational change, which in turn supports NPOs to meet civil society´s changing needs. Fourthly, referring to NPO practice, we point out how managers can govern DCs. The remainder of this paper is structured as follows; we first discuss the nature of dynamic capabilities and elaborate what these DCs could look like in refugee aid. We then illustrate the microfoundation of DCs as our theoretical background and explain how management can orchestrate the application of DCs in refugee aid. In the third section, we present the empirical results of our survey using a mediation analysis. The final section contains a brief discussion. Voluntas (2018Voluntas ( ) 29:994-1007 Framing Dynamic Capabilities in Theory and Practice

Nature and Relevance of Dynamic Capabilities
Despite difference in understandings of dynamic capabilities (Hsu and Wang 2010;Wang et al. 2015), there is consensus in the literature that the purpose of DCs is balancing the organization's, the employees' and the customers' needs as well as other 'business opportunities' (Baretto 2010;Eisenhardt and Martin 2000) in order to 'maintain leadership in continually shifting business environments' (Teece 2014, 329 f.). Scholars also agree that dynamic capabilities represent 'higher-level activities' (i.a. Teece 2014; Güttel and Konlechner 2009) or 'meta-capabilities' (Baretto 2010;Eisenhardt and Martin 2000) because they can govern other capabilities. DCs 'enable the firm to integrate, build, and reconfigure internal and external resources' (Teece 2014, 335). 'Higher-level activities' or 'meta-capabilities' also refer to the fact that DCs are not directly observable or visible. DCs become visible as soon as 'ordinary' capabilities 3 and activities, such as knowledge or information processing, learning, coordination and reconfiguring activities 'operationalize' them. 'Dynamic' refers to how the resource base is changed (Ambrosini and Bowman 2009). A capability is dynamic, provided the velocity of capability (re-)configuration corresponds to the velocity of environmental dynamics (Teece 2007;Eisenhardt and Martin 2000). DCs are relevant because they provide two 'added values' in relation to ordinary capabilities (Kaltenbrunner 2018). The first 'value added' of DCs is that DCs facilitate organizations' adaptation to changing environmental conditions repeatedly (Teece 2007(Teece , 2014 or systematically and over time (Hsu & Wang 2010). This is what Teece (2014) defines as 'dynamic fit.' In the case of NPOs active during the refugee crisis, this fit consisted in 'translating' the needs of refugees into services such as providing food, setting up camps or lobbying as fast as needs were changing. The second 'added value' of DCs emanates from their potential to generate a 'multi-perspective fit', not only a 'strategic fit' in terms of Chandler (1962). This implies that DCs can cope with dilemmas 4 in terms of multi-dimensional, partly conflicting interests, such as discrepant dominant logics of diverse internal and external stakeholders. During the refuge crisis, NPOs were, e.g., challenged to combine standard procedures of CEM disaster management (logic one) and activism, experimental search logics, such as intuition (logic two). By enhancing dynamic as well as multi-perspective fit, DCs support the organizational capacity to adapt and hence to compete in dynamic environments over time (Di Stefano et al. 2014, 315) and is reflected in strong evidence for a DC performance link (Teece et al. 1997, Teece 2007Wang and Ahmed 2007).

Conceptualization of Dynamic Capabilities and Refugee Aid
Following Teece's et al. (1997Teece's et al. ( , 2007Teece's et al. ( , 2014 understanding, DCs consist of three clusters of higher-level activities. These are sensing, seizing and transforming activities. As previously mentioned, sensing refers to identifying, scanning and monitoring environmental changes-opportunities and threats-and defining possible resource actions. This includes, e.g., having a good observation and judgment ability, a profound understanding of the impact of internal and external environments, a sense for major opportunities and threats and a sound information management (Li and Liu 2014). Transferred to refugee crisis, sensing could manifest in the ability to estimate the direction and quantity of arriving refugees and to provide status reports, to detect the refugees' needs and to anticipate their impact on infrastructure and resources. Seizing opportunities as the second cluster of activities mean choosing among possible resource actions and implementing the selected decision(s). Thus, seizing refers to the ability of absorbing and assimilating information in order to make (timely) client-oriented decisions, to design systems and structures and to mobilize the necessary resources for addressing the needs and opportunities for implementing decisions (Teece 2014;Li and Liu 2014;Jantunen et al. 2012). Seizing in the crisis situation could be found in defining appropriate decision-making and coordination structures; in activating and mobilizing (existent) (slack) resources, particularly volunteers and partners on a local level; and in delineating a kind of 'business model' in order to guarantee food supply and medical care. Thirdly, the transforming cluster encompasses the ability of aligning tangible and intangible resources to the environmental dynamics. This is mainly done by reconfiguring or recombining different resources (Teece 2014;Jantunen et al. 2012). There are various forms of reconfiguration, i. e., bundling existent resources and capabilities in new ways (Den Hertog et al. 2010). Resource bundling could become evident in the NPOs' ability to bundle their capabilities for setting up camps in a new way in order to build temporary homes. Reconfiguring also includes enriching existing resources and capabilities with new resources and capabilities. Transferred to the refugee crisis, this would be the NPOs' ability to enrich emergency supply of clients with new resources and capabilities for providing continuous care. Furthermore, it refers to replication of resources and capabilities in new organizational and market domains, e. g., NPOs active in (international) tracing services could stretch these capabilities in order to provide migration services.

Theoretical Framework of Dynamic Capabilities
As illustrated above, DCs-particularly in the notion of Teece (2007Teece ( , 2014-are based on resources or bundles of resources. These bundles of resources determine essentially, if an organization can compete over time or not. This view goes in line with the resource-based approach of Wernerfelt (1984), which correspondingly represents one dominant theoretical foundation of the dynamic capability approach 5 (Di Stefano et al. 2014). Complementarily a 'dynamic dimension' is added to enrich the dynamic capabilities approach because scholars regard bundles of resources as important, but not sufficient for gaining competitive advantage in dynamic environments. The 'dynamic dimension' refers to the organization's capability to adapt to environmental changes (Teece et al. 1997). This perspective roots in evolutionary economics (Nelson and Winter 1982). Evolutionary economics becomes particularly apparent in the dynamic capabilities approach via routines as fundament or building blocks of DCs. Thus, Teece (2016) regards the dynamic capabilities approach as 'capabilities enriched economic theory.' While the understanding of routines and capabilities is relatively elaborated, there is a lack of research regarding the microfoundation, the microlevel origins of routines and capabilities (Felin et al. 2012(Felin et al. , p. 1351. Microfoundation movement in the field of strategy and organization theory deals with the definition of underlying elements and mechanisms, which shape dynamic capabilities encompassing processes, activities, structures, rules as well as conditions of individual behavior, such as cognitions and affects and individual-level or group-level actions (cf. Teece 2007;Eisenhardt et al. 2010). Felin et al. (2012) cluster microfoundation of routines and capabilities into three categories, which are individuals, processes/ interactions and structure. The first category individuals as micromechanisms can be deconstructed in a cognitive and motivational dimension (Nelson and Winter 1982;Becker et al. 2005;Mäkelä et al. 2012). Eisenhardt et al. (2010) highlight the significance of managers, particularly their cognitions, for balancing efficiency and flexibility. Similarly, Zahra et al. (2006, p. 9) stress the key role of managers as decision-makers in applying dynamic capabilities. The second category, interactions and processes, relates to formal (e.g., procedures, rules) and informal forms of coordination (e.g., norms, experience) and illustrates how they shape routines and capabilities (Felin et al. 2012). Organizational routines as building blocks of capabilities are established with an increased level of interactions (Nelson and Winter 1982). Interaction-related micromechanisms serve two different roles. Firstly, they can illustrate how DCs are applied ('operationalized'), e.g., knowledge and information processes constituting sensing. 6 Secondly, processes as micromechanisms can describe how to govern or orchestrate DCs (cf. Coleman, 1990). In contrast to his earlier work, also Teece (2014) attributes managers an important role as micro-origin of dynamic capabilities. He emphasizes the role of managers in asset orchestration and leadership.

Microfoundation of Dynamic Capabilities in NPOs: Research Model and Hypotheses
Even if DCs are scarcely discussed in nonprofit research, scholars provide evidence that DCs are appropriate for NPOs, too. Pablo et al. (2007, 691) state that DCs are 'providing synergistic benefits through internal processes irrespective of market structures, a condition that could apply to either private or public organizations'. Bryson et al. (2007) note that DCs particularly enable NPOs to meet their stakeholders' needs. Garrido and Camarero (2014) claim that NPOs' organizational existence-similar to FPO-depends on sustainable competitive advantages, which in turn are influenced by DCs.
However, there is hardly research, which provides evidence regarding the microfoundation of DCs in NPOs (Piening 2013;Kaltenbrunner 2018). As previously mentioned, micromechanisms can refer to individuals, processes and interactions as well as structures (Felin et al. 2012). Process-related micromechanisms can either illustrate how DCs are applied ('operationalized'), e.g., knowledge and information processes constituting sensing, 7 or they 5 Based on bibliometric methods, Di Stefano et al. (2014) identified seven theoretical foundations of dynamic capabilities approach. These are the resource-based view (Wernerfelt, 1984), the knowledge-based view (Kogut and Zander 1992), behavioral theory (Cyert and March 1963), evolutionary economics (Nelson and Winter 1982), network theory (Granovetter 1992), transaction cost economics (Williamson 1975) and the positioning view (Porter 1980). In line with the two seminal dynamic capabilities, conceptualizations of Eisenhardt and Martin (2000) and Teece et al. (1997) two perspectives dominate. These are behavioral theory promoted by Eisenhardt and Martin (2000) and resourced-based view in combination with evolutionary economics endorsed by Teece and colleagues. 6 See illustration of DCs constituting processes and activities in chapter 2. 7 Compare for the illustration of DCs constituting processes, activities in chapter 2. Voluntas (2018Voluntas ( ) 29:994-1007 describe how they orchestrate or govern DCs (cf., Coleman, 1990). In this context, Teece (2014)  In light of these observations, this paper seeks to explore how management affects the application of dynamic capabilities in NPOs active in refugee aid. As illustrated in Fig. 1, we assume that managerial behavior (participative leadership) and managerial characteristics (psychological empowerment) affect DCs. The illustration of microfounding managerial mechanisms is highly relevant because it provides opportunities for influencing and managing the application of DCs

Participative Leadership and Dynamic Capabilities
We consider participative leadership as joint decisionmaking process, or at a minimum as shared influence of a manager and his or her employees in decision-making (Somech 2006;Koopman and Wierdsma 1998). Participative leadership describes managerial behaviors that 'use attentive listening and careful observation of nonverbal cues of member needs, feelings, etc.; serve as a consultant, advisor, teacher, and facilitator, model, and encourage appropriate leader behaviors; establish a climate that is conducive to expression of both feelings and ideas; encourage the group to address its maintenance needs and process problems in its regular group meetings; and relinquish control, allowing the group to make final decisions on appropriate issues' (Yukl 1981, 246 f.).
We regard participative leadership as a pivotal informal coordination mechanism between the manager and his/her team substituting missing formal structures and coordination mechanisms in the context of refugee crisis.
In contrast to transactional and transformational leadership, there is hardly research dealing with the effects of participative leadership on dynamic capabilities; exceptions are i.a. Fillipini et al. (2012) or Saunila et al. (2014). Fillipini et al. (2012) provide evidence that i.a participative leadership promotes ambidextrous initiatives. Saunila et al. (2014) found that participative leadership, especially in small enterprises, has positive effects on dynamic capabilities in terms of innovation capability. Similarly, Leskovar-Spacapan and Bastic (2007) show that participative leadership promotes creativity, which in turn facilitates innovation.
Scholars state that participative leadership opens communication channels, enhances interactions and thus increases the flow and use of information and knowledge (Schreiber and Carley 2006, 69). Individuals can gather information relevant for fulfilling their tasks and turn their ideas and individual knowledge into innovative services, products or procedures (Somech 2006, 152). In other terms, participative leadership provides conditions like enhanced interactions, which can facilitate the evolvement of human capital, such as cognitions (e.g., expertise) as well as the evolvement of social capital (Schreiber and Carley 2006, 71). The enhanced human (and social) capital in turn can result in enhanced learning and adaptive outcomes (Schreiber and Carley 2006, 69, 72), which are strongly related to DCs. Increased levels of interactions enhance organizational routines which are essential building blocks of (dynamic) capabilities (Nelson and Winter 1982). We propose that participative leadership affects DCs via enhanced outcomes and increased levels.
Participative leadership relates positively to dynamic capabilities.

Psychological Empowerment and Dynamic Capabilities
It is well established that individual cognitions affect the application of dynamic capabilities (Helfat and Peteraf 2015;Adner and Helfat 2003). Psychological empowerment represents a set of cognitions. Following Thomas and Velthouse (1990), Spreitzer (1995Spreitzer ( , 1443 defines psychological empowerment as 'intrinsic task motivation Fig. 1 Research model manifested in a set of four cognitions reflecting an individual to his or her work role'. Empowerment fosters individuals to launch initiatives and to handle uncertainty and risks (Spreitzer 1995). These effects are fundamental for coping with dynamic environments. Moreover, psychological empowerment relates positively to learning and innovation (Sears and Baba 2011) as well as to creativity (Deci et al. 1989). Learning, innovation and creativity are capabilities, which in turn are positively associated with dynamic capabilities. With regard to dynamic capabilities, Sprafke et al. (2013) provide evidence that psychological empowerment positively affects dynamic capabilities. This is due to the fact that psychological empowerment improves individual-level competences and perceived empowering working conditions enhance organizational capabilities. Psychological empowerment also increases self-efficacy of individual actors because highly competent individuals enjoy increased autonomy and self-determination and less competent individuals are provided with confidence in acting (Sprafke 2016). Phoocharoon (2011) provides evidence that psychological empowerment enhances the relationship between team learning behavior and the dynamic capability 'absorptive capacity'. Further, Hopkins et al. (2013) discuss the relationship between psychological empowerment, commitment and strategic renewal and prove that empowerment positively effects strategic renewal. Hence, we propose that managers active in refugee aid, who feel self-empowered foster the application of dynamic capabilities because they feel more competent to handle risks and uncertainties in refugee aid and they perceive the working conditions as appropriate to do their jobs. We therefore hypothesize that The manager's psychological empowerment is linked positively to dynamic capabilities.

Participative Leadership, Psychological Empowerment and Dynamic Capabilities
Following the discussion above, we propose that the direct relationship between participative leadership and dynamic capabilities is mediated by psychological empowerment, which is illustrated by the manager´s perception of his competence, impact and self-determination. We assume that the participation and integration of team members in decision-making processes allow the manager to benefit from the team´s knowledge, creativity, ideas in terms of expanding knowledge and acquiring new skills (Srivastava et al. 2006). This in turn enhances the manager´s perceived empowerment, especially his/her self-efficacy and impact of his/her actions.
Thus, we assume a positive relationship between participative leadership, psychological empowerment and dynamic capabilities which results in the following hypotheses: H 3a(+) The positive effect of participative leadership on DC is mediated by self-determination.
H 3b(+) The positive effect of participative leadership on DC is mediated by the leader's self-perception of his/her impact.
H 3c(+) The positive effect of participative leadership on DC is mediated by the leader's self-perception of his/her competence.

Research Design Sample
For gathering information on participative leadership, empowerment and dynamic capabilities during the refugee crisis in 2016 we used a survey instrument developed in close cooperation with a variety of experts. In a first step, interviews of two NPO managers plus two scientists in the fields of strategic management and human resource management, respectively, were conducted about the instrument itself and the most appropriate group of respondents for the information needed. In a second step, four NPO practitioners pretested the instrument resulting in small adaptions of the questionnaire. The experts consulted shared a strong opinion on people holding management functions as the only eligible respondents being able to assess dynamic capabilities. Also, this very dynamic situation differed significantly from a standard way of organizing that would allow for a clear and stable allocation of employees to managers. Instead, during the peak of the refugee crises the majority of people not holding management positions were volunteers with partly very short lived commitment leading to enormous turnover rates that made the subordinate level very hard to grasp. We thus approached people holding management positions via an online platform connecting NPOs active in refugee aid.
In total, 340 respondents started the survey of which 90 held management functions and could thus continue with the survey. Out of these 90, 16 showed more than 30% 'missing values' (cf. Cohen et al. 2003) and were thus excluded. Our final sample consists of 74 managers active in refugee aid. Table 1 describes the sample. In general, managers active in refugee aid were mostly male (79%). On average, the managers were about 40 years old and more than twothird were married or live in a relationship. 45% of the Voluntas (2018)

Measures Dependent Variables
For measuring dynamic capabilities as dependent variable, we draw on the scale of Li and Liu (2014). We use this DC scale because in contrast to most other DC scales, it has already been used in a setting other than competitive market structures (e.g., no complete market conditions and much political influence) which best approximate the conditions of refugee crisis. Li and Liu (2014) define DCs as constructs of three dimensions. These are 'sense-making capacity' (six items, e.g., 'we can fully understand the impact of internal and external environments'; 'we have good observation and judgment abilities'), 'timely decision-making capacity' (four items, e.g., 'under many circumstances we can make timely decisions to deal with strategic problems'; 'we can quickly deal with conflicts in the strategic decision-making process') and 'change implementation capacity' (five items, e.g., 'our strategic changes can be carried out efficiently'; 'we help each other in strategic change implementation').

Independent Variables
For capturing the predictor variable participative leadership, we draw on the scale from Hoch et al. (2013). The scholars applied the scale in an innovation context, which is similar to a DC context. This scale encompasses four items (e.g., 'I decide, together with my team, which tasks have to be done'; 'my team and I sit down together and reach agreements on the tasks to do') measuring participative goal setting leadership. Following previous DC research (cf. Sprafke et al., 2013), we chose psychological empowerment according to Spreitzer (1995) as mediator variable. This scale measures empowerment by drawing on four dimensions with three items each. These are 'competence' (e.g., 'I am confident about my ability to do my job'), 'self-determination' (e.g., 'I can decide on my own how to go about doing my job'), 'impact' (e.g., 'I have significant influence over what happens in my department') and 'meaning' (e.g., 'The work I do is meaningful'). Due to the fact that refugee crisis represents an emergency situation, where reflections about the meaning in terms of the 'fit between the needs of one´s work role and one´s beliefs, values and behavior' (Spreitzer 1995, 603) are hardly possible, we decided to exclude 'meaning' as empowerment dimension. The study includes 'environmental dynamics' and 'task complexity' as control variables, which might affect the relation between participative leadership and DCs, particularly because these variables are context dependent. We measured 'environmental dynamics' using the item 'our clients regularly ask for new products and services' suggested by Jansen et al. (2006). Following Gaitanides and Stock (2004), 'task complexity' was assessed through the item 'the tasks of our team mainly involve solving complex problems'.

Factor Analysis, Reliability and Validity
Due to the explorative character of DCs, we did a factor analysis of the DC scale as well as on empowerment. For DCs, the three factor structure proposed by Li and Liu (2014) could not be confirmed. The 'capacity for timely decision-making' did not show. Two of the items neither showed high enough loadings on the two factors identified nor formed their own factor. The remaining two items 'under many circumstances we can make timely decisions to deal with strategic problems' and 'We can remedy quickly to unsatisfactory customers' had high factor loadings on the other two dimensions. The factor 'change implementation' could be reproduced. KMO values for these two dimensions exceed the recommended value of 0.6 (Weiber and Mühlhaus 2014, 133;Backhaus et al. 2016, 398 f.). Reliability analysis produced Cronbach's α values ranging from .659 to .808. Thus, scales are above the minimum level of 0.6 (Bagozzi and Yi 1988), which indicates internal consistency. The factor analysis of the items used for psychological empowerment neither showed the suggested three factor structure (Spreitzer, 1995). While the dimension 'competence' could be replicated 'impact' and 'self-determination' merged into one factor that we name 'impact_self-determination.' Again, KMO values of the scales exceed the recommended value of 0.6. Reliability analysis produced Cronbach's α values ranging from .838 to .877. Finally, reliability analysis of the fouritem participative leadership scale confirms one factor with Cronbach's α = .829. Also, AVE, examining validity, exceeds the cutoff value of 0.5 (Weiber and Mühlhaus 2014, 64).

Findings
In order to test for linear direct and mediated effects, we employed ordinary least square regression combined with bootstrapping. We tested for model assumptions using scatterplots, VIF and a Kolmogorov-Smirnov test for normal distribution and did not find hints pointing to multicollinearity, heteroscedasticity or nonnormal distribution. Table 2 contains means, standard deviations and correlations between the variables used. DCs significantly correlate with participative leadership and with one dimension of psychological empowerment, i.e., 'impact_autonomy'. The relationship with the competence dimension is strong, but not significant (p = .06). The control variable environmental dynamics are significantly associated with DCs, while task complexity does not show any effect. Participative leadership is significantly correlated with 'im-pact_autonomy', but not with 'competence' dimension of empowerment. Table 3 shows the results of the ordinary least square regression for the hypothesized direct effects. This includes the effects of participative leadership and psychological empowerment on dynamic capabilities (model 3) as well as their effects on the specific DC 'sensemaking' (model 1) and the DC 'change implementation capacity' (model 2). The illustration of these effects corresponds to H 1 and H 2 .
In order to substantiate the effects of participative leadership on dynamic capabilities, we added an analysis of the direct effects separating DCs into the two dimensions 'change implementation capacity' and 'sense-making capacity' identified in the factor analysis.
Model 2 illustrates the effects on the DC 'change implementation capacity'. The effect of participative leadership on this DC is positive and significant (ß = .369*; SE = .161; p ≤ .05), thus supporting H 1 . Further, the direct effect of the empowerment dimension 'impact_autonomy' on this DC is significant (ß = .268*; SE = .115; p ≤ .05), whereas the direct effect of the empowerment dimension 'competence' is nonsignificant In sum, models 1-3 consistently support H 1 . Concerning the effects of psychological empowerment, the findings vary depending on which dimension of psychological empowerment and which DC is analyzed. Thus, the reported direct effects of empowerment are less appropriate for interpretation.
In order to test mediated effects as proposed by H 3, we started with the procedure described by Baron and Kenny (1986). This stepwise path procedure starts with a regression of the dependent variable (Y), dynamic capabilities, on the independent variable (X), participative leadership [c-path]. To continue mediation, the independent variable has to affect the dependent variable significantly. In a next step, the mediators (M), dimensions of psychological empowerment, are regressed on the independent variable, participative leadership [a-path]. Again, the relationships should be significant. Third, we deployed a regression of the dependent variable, DCs, on both: on participative leadership as independent variable [c-path] and on the dimensions of psychological empowerment as mediators [b-bath]. Here the mediators' effects on the dependent variable should be significant, while the independent variable should not be affected (nonsignificance for full mediation, less significance than in c-path for partial mediation). Table 4 shows the mediated effects of the independent variable participative leadership via the proposed mediators 'impact_autonomy' (H 3ab ) 8 and 'competence' (H 3c ) on dynamic capabilities as dependent variable.
Accordingly, based on a positive and significant a-path and b-path, findings imply partial mediation for participative leadership through 'impact_autonomy'; this supports H 3ab . Due to the lack of significance in a-path and b-path, H 3c is rejected.
As Table 5 shows, the bootstrapped results support these findings because the confidence intervals for 'impact_autonomy' is above zero (.0170 to .2746), whereas the

Discussion and Conclusions
The Refugee crisis of 2016 with its ambiguous, multi-dimensional and dynamically changing environmental conditions made evident that dynamic capabilities represent a crucial asset for NPOs because they facilitated NPOs to achieve the required continuous organization-environment fit. In other terms, DCs enabled NPOs to meet continuously changing civil society´s needs. In the case of NPOs active in refugee aid, this fit consisted in translating the needs of refugees into services, such as providing food, setting up camps or lobbying as fast as needs were changing. Dynamic capabilities are said to accomplish such a fit due to their property as meta-capabilities and thus governing other capabilities systematically and over time (Hsu and Wang 2010;Teece 2014).
In order to better understand the nature of DCs and particularly to provide options for influencing the application of dynamic capabilities in refugee aid, we aim to explore mechanisms on microlevel, which found DCs. This is what research defines as microfoundation of DCs (Teece 2007;Felin et al. 2015). Following Di Stefano et al. (2014) and Teece (2014), research particularly lacks of an integrated consideration of management-related aspects, e.g., managerial behavior and cognitions. Hence, we analyzed the link between micromechanisms and DCs on macrolevel by examining how management affects the application of dynamic capabilities in NPOs. In this context, we include managerial behavior-by drawing on participative leadership-as well as managerial characteristics-by integrating psychological empowerment-into our research model.
Concerning the relationship between participative leadership and dynamic capabilities, we assume that participative leadership positively influences managers' application of dynamic capabilities. This is because team members' knowledge, ideas, and creativity-gained through participative leadership-are enablers for DCs. We provide evidence for direct effects between participative leadership and dynamic capabilities, illustrated via H 1 . This is in line with previous research in the field of dynamic capabilities (cf. Saunila et al. 2014;Filippini et al. 2012) as well as with strategic management research in general (cf. Leskovar-Spacapan and Bastic 2007;Gottschalg and Zollo 2007;Lindenberg and Foss 2011).
We suppose that NPOs active in refugee aid (or under comparable ambiguous, dynamically changing and multidimensional emergency situations) use participative leadership in terms of interactive decision patterns as substitute for nonexisting organizational structures, respectively,  coordinative mechanisms. For instance, participative leadership can serve as substitutive for routines, e.g., standard operation procedures, which do not or rather only partly exist in such emergency situations. In this context, participative leadership is comparable to task forces or staff work. Further, we deployed a differentiated analysis and explored the effects of participative leadership on the DC dimensions 'sense-making capacity' and 'change implementation capacity' separately. Whereas the findings indicate a significant and positive effect regarding 'change implementation capacity', it was nonsignificant concerning 'sense-making capacity'. This could be due to the fact that team members can only restrictively serve as 'sense-makers' because they are not primarily cognizant of the overall political, economic or humanitarian situation and its dependences; they are experts, though, in delivering services which makes team members predestinated to provide information regarding 'changes in implementation'.
In order to explore managerial influence on dynamic capabilities profoundly, we also explored the effect of psychological empowerment. Findings show that the empowerment dimension 'impact_self-determination' affects dynamic capabilities significantly. Thus, the managers' perceptions, how self-determined (s)he is and how impactful his/her actions are, foster dynamic capabilities. This is because managers perceiving such cognitions are more motivated to take initiatives, to innovate as well as to increase performance (Sprafke et al. 2013). Psychological empowerment shows that cognitions are strongly linked with motivation; Spreitzer (1995) actually defines psychological empowerment as intrinsic task motivation. It is particularly important for NPOs to consider psychological empowerment in their capacity building because mainly intrinsic motivated volunteers constitute their work force. The effect of the second dimension 'competence' was not significant, though.
Our research focus was to explore the linkage between participative leadership, psychological empowerment and dynamic capabilities. In this relationship, psychological empowerment is considered to mediate the effects of participative leadership on dynamic capabilities. Comparable to findings mentioned above, we could not find evidence for the effect of 'competence' (H 3c (+)) on DCs, but for the effect of 'impact_self-determination', supporting H 3ab . 'Impact_self-determination' mediates participative leadership partially. This implies that this empowerment dimension enhances the effect of participative leadership on DCs. Table 6 provides an overview regarding the tested hypotheses.
Finally, we also provide evidence that DCs have a highly significant effect on organizational performance (cf. Table 2). Applying DCs increases NPOs performance in refugee aid or rather supports them to positively impact civil society´s needs.
The profound empirical evidence provided by our paper contributes significantly to the understanding of interdependences of DCs, management behavior and managerial cognitions. We examine this microfoundation in NPOs, where DC research is even more underrepresented. Hence, our study enlarges the research context of DCs. We found that managers represent a meaningful agent of DCs on microlevel and thus contribute to the behavioral foundation of organizations, which is an essential research stream of microfoundation to further explore (Felin et al. 2015). By analyzing managerial behavior (participative leadership) as well as managerial cognitions (psychological empowerment), we even provide an integrative illustration of management-related microfounding mechanisms. In contrast to previous research, which particularly focused on transactional and transformational leadership, we introduced and empirically proved participative leadership as a 'new' DC enhancing leadership style. Our study shows that The positive effect of participative leadership on DCs is mediated by the manager's self-perception of his/her competence 9 ✓ Hypotheses supported-, 9 hypotheses not supported; partly supported participative leadership represents a crucial, foremost informal coordination mechanism that can serve as substitute for barely existing formal coordination mechanisms in highly dynamic and even crisis situations. We show strong associations between participative leadership and DC, mediated by psychological empowerment. This suggests that actively stimulating a certain leadership style can have an impact on DCs and thus on organizational survival.