La estrategia de saneamiento en las reestructuraciones empresariales

UNIVERSIDAD DE SEVILLA DOCTORADO EN ADMINISTRACIÓN Y DIRECCIÓN DE EMPRESAS TESIS DOCTORAL LA ESTRATEGIA DE SANEAMIENTO EN LAS REESTRUCTURACIONES EMPRESARIALES (RETRENCHMENT STRATEGY IN TURNAROUND SITUATIONS) Doctorando: Dª Alicia Ramos García Directores: D. José Carlos Casillas Bueno D. José Luis Barbero Navarro Mayo, 2022 2 AGRADECIMIENTOS Me gustaría mostrar mi agradecimiento a Dios por las personas que ha puesto en mi vida y, ahora concretamente, por aquéllas sin las cuales este trabajo no hubiera sido posible. GAD. En primer lugar, gracias por mis directores de tesis José Carlos y José Luis, porque me han mostrado el valor de la investigación y por la gran paciencia que han tenido conmigo en mis momentos de desánimo, porque me han hecho ver que el camino merecía la pena y por todo lo que he aprendido con ellos. Gracias por mis padres, que siempre me enseñaron el valor del esfuerzo y de las cosas bien hechas y que tanto, tanto, me han dado. Gracias por mis hijos Alfonso, José Luis, Miriam y Pablo, que en estos años han sufrido las privaciones del tiempo que yo he necesitado para acabar este trabajo. Gracias por el resto de mi familia y por los amigos que me acompañan y comparten conmigo este camino. Y finalmente, mi agradecimiento a este Departamento de Administración de Empresas y Marketing de la Universidad de Sevilla y a la Comisión Académica de Doctorado, por los recursos facilitados para hacer posible esta tesis. 3 ABSTRACT During the last decades and starting the “great recession”, businesses have suffered one of its most volatile periods in history. Technology radical changes, fiscal and monetary policies, social changes and the advance of the financial industry have brought for many businesses difficulties which have translated in a situation of bankruptcy or lower profitability for many businesses. Survival is more challenging than ever. During this decade turnaround research has lagged behind. The 80´s and 90´s turnaround researchers found a number of important insights that allowed for the advance of the research area. However, the number of important pieces of research since then has been scant. One of the most important ideas in the turnaround literature is the two stages model, proposed by Pearce and Robbins (1992). This model proposes a turnaround should start with a retrenchment stage followed by a recovery stage. Although the model has received many critics and the evidence is mixed, the model remains actual. This dissertation focuses on the retrenchment stage of the two-stages model. The retrenchment stage is the most researched topic and one of the most important topics in the turnaround literature. First, during retrenchment firm survival, rather than a mere increase in profitability, is at stake. Second, the mixed evidence on the stage has been the main obstacle preventing progress of turnaround studies. 4 We study three topics that have received increased attention in the management literature during the last two decades, which turnaround research has remained silent on. These three topics are extremely relevant to the development of the retrenchment strategy. These topics are dynamic capabilities, CEO change and the temporal dimension of retrenchment. Specifically, we study three topics in this dissertation: 1) The relationship of dynamic capabilities and the two retrenchment strategies (asset retrenchment and cost retrenchment). 2) The conditions under which an early CEO replacement influences turnaround success. 3) The relationship between the time and volume of retrenchment and turnaround success. In the first study, we posit that despite the assumed importance of environmental dynamism with respect to organizational decline, theory and evidence of its effect are lacking. Integrating work from firm survival and turnaround literature within a dynamic capabilities framework, we posit that dynamism positively moderates the relationship between firm retrenchment and performance. We collect data from SABI on a sample of 230 Spanish declining firms. We find that dynamism positively moderates the relationship between firm performance and both asset and cost retrenchment. However, as expected from the dynamic capabilities framework, the moderating effects of dynamism differ according to the chosen retrenchment strategy. 5 They are stronger for the asset retrenchment strategy than those for cost retrenchment. In our second study, we argue that prior turnaround studies have investigated “whether” CEOs should be replaced and the related boundary conditions. Given how critical time is for declining firms, we shift this CEO succession research question to “when” should CEOs be replaced. We study the effects of the timing of CEO replacement on firm performance. With the support from the downward spiral perspective, we hypothesize the conditions (i.e.: internal versus external causes of decline, CEO tenure and duality) under which early CEO replacement leads to positive turnaround outcomes. Using data from Compustat, we find support for our hypothesized relationships. In our third study, we posit that recent turnaround literature has shown evidence of the importance of time considerations on turnaround success. We take this topic forward and argue there is a relationship between the volume of retrenchment and time of retrenchment. Declining firms can act aggressively on the depth of the cuts (volume aggressiveness), or on the timing and speed of retrenchment (time aggressiveness). Based on the survivor syndrome and the downward spiral perspectives, we argue and show evidence that time aggressiveness influences volume aggressiveness which will determine the outcome of the retrenchment process. 6 INDEX 7 Acknowledgements ............................................................................................... 2 Abstract ..................................................................................................................... 3 FIRST CHAPTER INTRODUCTION 1.- Background ...................................................................................................... 12 2.- Decline and turnaround ................................................................................. 14 3.- Retrenchment ................................................................................................... 16 4.- Retrenchment and dynamic capabilities .................................................. 18 5.- Timing of CEO replacement and turnaround success ............................ 19 6.- Retrenchment and time ................................................................................. 21 7.- Methodology, samples and variables ........................................................ 25 8.- Research questions and contributions ........................................................ 29 9.- Structure ............................................................................................................. 31 8 SECOND CHAPTER (PAPER I) RESTRUCTURING IN DYNAMIC ENVIRONMENTS: A DYNAMIC CAPABILITIES PERSPECTIVE 1.- Introduction ....................................................................................................... 34 2.- Theorical background .................................................................................... 38 3.- Hypotheses ........................................................................................................ 54 4.- Method .............................................................................................................. 74 5.- Results ................................................................................................................. 83 6.- Discussion ........................................................................................................... 86 7.- Limitations and directions for future research ........................................... 91 References.......................................................................................................... 93 THIRD CHAPTER (PAPER II) TIMING OF CEO DISMISSAL UNDER CORPORATE CRISIS. A TEMPORAL APPROACH TO CEO SUCCESSION 1.- Introduction ..................................................................................................... 110 2.- Theorical background .................................................................................. 114 3.- Hypotheses ...................................................................................................... 116 4.-Method.............................................................................................................. 124 5.- Results ............................................................................................................... 134 9 6.- Discussion ......................................................................................................... 136 7.- Limitations and directions for future research ......................................... 140 References........................................................................................................ 142 FOURTH CHAPTER (PAPER III) THE MEDIATION OF VOLUME AGGRESSIVENESSS ON TIME AGGRESSIVENESS IN RETRENCHMENT PROCESSES 1.- Introduction ..................................................................................................... 161 2.- Theorical background .................................................................................. 165 3.- Hypotheses ...................................................................................................... 167 4.-Method.............................................................................................................. 172 5.- Results ............................................................................................................... 179 6.- Discussion ......................................................................................................... 182 7.- Limitations and directions for future research ......................................... 183 References....................................................................................................... 185 16 common CEOs creating a gap between perception and reality on the severity of the situation (Slatter, 1984). 3. Retrenchment A turnaround involves a series of actions as response to a situation of decline. The aim of these actions is to reverse the situation of decline (Robbins & Pearce, 1992). In the past, authors have proposed several models of turnarounds (Arpi, 1999; Bibeault, 1982; Schoenberg, Collier, & Bowman, 2013; Slatter, 1984). The model most used by researchers has been the Robbins & Pearce (1992). This model consists of two stages: retrenchment and recovery. During the initial retrenchment stage the firm introduces actions to achieve cost and asset reductions (Michael & Robbins, 1998). These actions have been argued to have an operating nature (Trahms et al., 2013). There are two types of retrenchment actions or strategies: asset retrenchment and cost retrenchment. Asset retrenchment involves the net reduction of assets (Lim, Celly, Morse, & Rowe, 2013). Some examples of asset retrenchment are closing plants, selling property or equipment, reducing inventory (Lim et al., 2013; Morrow, Johnson, & Busenitz, 2004). Cost retrenchment deals with the reduction of costs. It involves the reduction of Selling, General and Administration (SGA), interest, salaries, perks, advertising and marketing costs (Lim et al., 2013). Cost retrenchment is generally used when the situation of decline is more severe (Pearce II & Robbins, 1993). Recent research has argued that retrenchment actions also have a strategic nature, beyond the 17 operational nature, given some of these measures are implemented with the intention to set new directions for the firm (Barbero, J., Di Pietro, F., Chiang, 2017). During the recovery stage, the firm adjusts the domain of the firm to align it to the environment. The actions carried out during this stage possess a strategic nature (Trahms et al., 2013) because the aim is to reposition the company with the objective of growth and profitability (Barker III & Duhaime, 1997; Pearce II & Robbins, 1993; Schmitt & Raisch, 2013). The two stage model has been widely criticized. Literature has argued that the implementation of both type of measures, retrenchment and recovery, are not always necessary (Barker & Mone, 1994). Also, it has been argued that both stages can be simultaneously implemented (Schmitt & Raisch, 2013). The focus of our dissertation is the retrenchment stage and the actions carried out during this stage. There are three reasons for such. First, the retrenchment stage has been one of the most controversial ideas in the turnaround literature (Barker & Mone, 1994; Pearce & Robbins, 1994; Robbins & Pearce, 1992). The evidence on the value of retrenchment actions vary widely. Studies have only managed to get consistent results under a very contingent approach. Hence, contributions in this research niche will help the advance of turnaround research, which has remained stalled during the last decade. Second, the retrenchment stage is more unique to turnarounds. The recovery stage is typically when the organization is zero or positive cash flow, and the firm degree of urgency is lower. The study of recovery is not unique to 18 turnarounds but can be and has been studied in other firm situations beyond decline. Third, the aim of retrenchment actions is corporate survival (Pearce II & Robbins, 1993). From our perspective, survival is a more important activity than a mere improvement in corporate results. 4. Retrenchment and dynamic capabilities The dynamic capabilities framework has been one of the most researched areas during the last two decades. Only in the last decade more than one thousand papers have been published on the topic (Peteraf, Di Stefano, & Verona, 2013). This is due to the role of dynamic capabilities on the generation and sustainability of a firm´s competitive advantage (Teece, 2014). Firms possess broadly two types of capabilities: ordinary and dynamic capabilities. An ordinary capability allows firms to deploy resources to earn a living in the present (Collis, 1994; Schilke, 2014). A dynamic capability allows an organization to purposefully create, extend, or modify its resource base (Helfat et al., 2007). Dynamic capabilities fall under three clusters of activities: sensing, seizing, and transforming (Teece, 2007, 2012). Given that dynamic capabilities may be rooted in certain change routines and turnaround management is composed of transformational capabilities, the question arises as to whether turnaround strategies involve dynamic capabilities. Limited works have directly linked turnarounds with dynamic capabilities 19 (Danneels, 2011; Zúñiga-Vicente & Vicente-Lorente, 2006), which remains a strong candidate for research (Teece, 2012) Addressing turnarounds through a dynamic capabilities framework is important for two reasons. First, dynamic capabilities represent a framework used in change situations. Setting the boundaries for the framework is important to delineate its application. Given that turnarounds represent situations of extreme change, the evaluation of whether the framework is applicable to turnarounds represents the testing of the framework boundary conditions. Second, the literature is concerned that turnaround research has been mainly phenomenon-driven rather than theory-driven (Trahms et al., 2013). Finding links between turnaround and other theoretical frameworks is key for the advancement of research in the turnaround field. In line with other studies which test whether certain capabilities are dynamic (Drnevich & Kriauciunas, 2011; Schilke, 2014), we replicate their methodology to assess how dynamic are retrenchment actions. We test whether the two type of retrenchment strategies, asset retrenchment and cost retrenchment influence turnaround performance in an environment characterized by its high dynamism. 5. Timing of CEO replacement and turnaround success The reduction of personnel is part of the cost retrenchment strategy (Lim et al., 2013; Morrow et al., 2004). One of the most important decisions by a 20 declining firm is about leadership (Bibeault, 1982; Slatter, 1984): should the existing CEO be removed and under what conditions should the CEO be removed. Beyond the area of turnarounds, CEO succession has been studied with inconclusive results (Krause, Semadeni, & Cannella, 2014). Again it has been proposed a contingent approach to its study given the mixed results (Berns & Klarner, 2017). In the turnaround area few studies have covered this important topic (Barker & Patterson, 1996; Chen & Hambrick, 2012; Davidson, W; Worrell, D; Dutia, 1993). Those studies have also shown more conclusive results. However, the models proposed tend to be underspecified and little we know about the conditions under which CEO replacement should occur under organizational decline. The study of time has become a promising avenue for the advance of turnaround research. We use the temporal approach to study the timing of CEO replacement. Based on this line of research, we shift the question not to whether CEO should be replaced but when should a CEO be replaced. In other words, we study the timing of CEO replacement in a situation of decline and the conditions under which the CEO should be replaced early. Specifically, we study and argue that if the causes of decline are external, and hence less controllable by the CEO, replacing management later improves turnaround success. Contrastingly, we also argue that in the case of internal causes of decline and for these same reasons, a CEO should be replaced early. Second, we also study whether the length of the tenure during which the CEO has remained within the firm has any interactive effect 21 on the relationship between early replacement and turnaround success. Tenure has been one of the variables showing more consistent results in the study of CEO succession (Barker & Patterson, 1996; Hambrick & Fukutomi, 1991; Henderson, Miller, & Hambrick, 2006). Finally, we study how if the firm has a dual structure (CEO and President are same person), the later the CEO is replaced, the longer the dual structure is maintained, the greater the turnaround success. This is argued on the basis of the need for unity of command in a situation of decline, which will allow fast decision and fast action (Dowell, Shackell, & Stuart, 2011; Finkelstein & D´Aveni, 1994). The effects of time have not been studied in the CEO succession literature or the turnaround literature. Additionally, decline is a very valid context to study the effects of time (Barbero, J., Di Pietro, F., Chiang, 2017) and also the study of CEO replacement (Chen & Hambrick, 2012). 6. Retrenchment and time The study of the general patterns of change requires a focus on the temporal context (Pettigrew et al., 2001; George & Jones, 2000). A few temporal dimensions of change–timing, speed and rhythm (Ancona et al., 2001; Huy, 2001)–have been the focus of studies in a diverse range of management areas such as internationalization, mergers and acquisitions, product development, new ventures, and so on (Vermeulen & Barkema, 2002; Bauer & Matzler, 2014; Atuahene-Gima, 2003; Klarner & Raisch, 2013; Gersick, 1994; Amis et al., 2004; Pacheco-de-Almeida et al., 2014). The study of these temporal dimensions has shown promising results. In addition to 22 showing evidence of how these variables influence firm performance, prior studies such as the seminal paper by Vermeulen and Barkema (2002) and the extant research on firm internationalization have brought new perspectives to the research on change. Thus, the study of the effects of time is well developed in other change-intensive areas of management research with very positive results. In few areas are these temporal dimensions of change -timing, speed, and rhythmas critical as in the context of turnarounds. The turnaround literature overwhelmingly agrees that early timing and a fast speed of change are critical to firm survival (Pearce II & Robbins, 1993; Slatter, 1984; Hambrick, 1985; Arogyaswamy et al., 1995). Researchers consistently posit that timing and speed are most important during retrenchment, which is the initial stage of a turnaround, when the survival of a firm is at stake (Arogyaswamy et al., 1995; Pearce II & Robbins, 1993). Thus, the literature on turnaround is strongly grounded on the assumptions of the need for early timing and a fast pace of change due to the urgency of the situation. Recent research has shown promising results on the study of time in a turnaround context. Tangpong, Abebe, & Li (2015), Barbero, J., Di Pietro, F., Chiang (2017) and Barbero, J., Martinez, J.A. and Moreno A.M. (2019) have studied with positive results the effects of time on turnarounds. The first piece of research studies whether time (early/late retrenchment) influences turnaround success. They use a very sophisticated methodology and use a large number of dependent variables to confirm the positive influence of time on the effect of 23 retrenchment; this is, an early timing of retrenchment positively influences turnaround success. The second piece of research is a much more sophisticated study that breaks down the time construct into three subdimensions (timing, speed and rhythm of retrenchment) and analyzes the influence of each on turnaround success. This study also tests the effects of a munificent and of a dynamic environment on the three dimensions of time. Finallly, the last piece of research has introduced the idea of retrenchment aggressiveness dividing such into volume aggressiveness and time aggressiveness. This research studied and found a positive effect of time aggressiveness and a negative effect of volume aggressiveness on turnaround performance. Finally, the study found an interaction effect between volume aggressiveness and time aggressiveness on turnaround success. Those pieces of research study the topic using the “downward spiral” and “survivor syndrome” perspectives (Hambrick & D´Aveni, 1988; Brockner, 1992). These perspectives are well-fitted frameworks to the area of turnarounds and have been used very often in research. Management literature has used other frameworks to explain the effects of time on change. One of the most used frameworks has been the speed of decision and speed of action frameworks (Baum & Wally, 2003; Eisenhardt, 1989; Forbes, 2005; Rajagopalan, Rasheed, & Datta, 1993). Given the complexity of turnarounds and time, it is advisable to fine grain more the origins of a positive relationship between time of retrenchment and turnaround success. In lieu of a multi 24 theory approach necessary to explain the complexity of decline and turnarounds, we use those two frameworks to argue a mediation effect of volume aggressiveness on the positive effect of time aggressiveness on turnaround outcomes. As a third question of research we analyze the subtle but important relationship between volume aggressiveness, time aggressiveness and turnaround outcomes. 25 7. Methodology, samples and variables The methodology used in most cited turnaround studies tends to be very complex. First, the selection of the firms in decline includes a large number of firm characteristics (industries, degree of concentration, number of employees, etc) that makes the sample selection a gruesome process. Second, these studies either use a dichotomic variable -turnaround success/non success- (Barker III & Duhaime, 1997; Chen, 2014; Ndofor, Vanevenhoven, & Barker, 2013; Tangpong et al., 2015) or use panel data (Chen & Hambrick, 2012; Morrow et al., 2004). We have followed the methodology used by top researchers. Below we include a table describing for each of the three studies the methodology, sample and variables: 32 DYNAMIC CAPABILITIES PERSPECTIVE". The second study entitled "TIMING OF CEO DISMISSAL UNDER CORPORATE CRISIS. A TEMPORAL APPROACH TO CEO SUCCESSION” has been included Chapter 3 ". Finally, the study entitled "THE MEDIATION OF VOLUME AGGRESSIVENESS ON TIME AGGRESSIVENESS IN RETRENCHMENT PROCESSES” is part of Chapter 4. Each of these chapters includes different sections: introduction, hypotheses, methodology and conclusions (results, discussion and limitations). Chapter 5 completes the dissertation by summarizing a set of conclusions, contributions, limitations and future lines of investigation following this study. References have been included in each chapter. 33 SECOND CHAPTER RESTRUCTURING IN DYNAMIC ENVIRONMENTS: A DYNAMIC CAPABILITIES PERSPECTIVE 1. Introduction The dynamic capabilities research is still relatively new despite the number of published articles (Peteraf et al., 2013; Teece, 2012). The dynamic capabilities framework has been applied to many types of firms and it has been suggested that future research should identify new types of firms for which the concept could be useful (Barreto, 2010). Limited knowledge exists concerning the role of dynamic capabilities in organizational survival because few studies have focused on firms in decline (Danneels, 2011; Zúñiga-Vicente and Vicente-Lorente, 2006). Moreover, no existing studies address turnaround change capabilities and dynamic capabilities, which “remains an obvious candidate for future research” (Teece, 2012, p. 1397). Addressing turnarounds through a dynamic capabilities framework is important for two reasons. First, dynamic capabilities represent a framework used in change situations. Setting the boundaries for the framework is important to delineate its application. Given that turnarounds represent situations of extreme change, the evaluation of whether the framework is applicable to turnarounds represents the testing of the framework boundary conditions. Second, the literature is concerned that turnaround research has been mainly phenomenon-driven rather than theory-driven (Trahms et al., 2013). Finding links between turnaround and other theoretical frameworks is key for the advancement of research in the turnaround field. 35 During recent years, and spurred by the economic crisis, turnaround and survival as a research area has gained momentum (Boyne and Meier, 2009; Lim et al., 2013; McKinley et al., 2014; Ndofor et al., 2013; Schmitt and Raisch, 2013). However, research on the subject remains essential (McKinley et al., 2014), particularly research associated with retrenchment (Lim et al., 2013; Morrow et al., 2004). Retrenchment is the most common strategy for underperforming firms and the most crucial stage in the turnaround process (Lim et al., 2013; Morrow et al., 2004). Retrenchment research has evolved from initial studies that analyzed the effects of retrenchment (Barker III and Duhaime, 1997; Robbins and Pearce, 1992), to recent studies that emphasize the effects of contingency factors on turnaround, especially those exerted by the environment (Boyne and Meier, 2009; Lim et al., 2013; Ndofor et al., 2013). The environment plays a critical role in firm survival because many of the causes of a firm´s decline are based either on environmental maladaptation or on environmental hostility (Arogyaswamy et al., 1995; Barker III and Duhaime, 1997; Cameron et al., 1987; Trahms et al., 2013). Recently, scholars have studied the environment in a turnaround context (Boyne and Meier, 2009; Lim et al., 2013; Morrow et al., 2004; Ndofor et al., 2013); however, the effects of retrenchment actions as the most important strategy in a turnaround within a dynamic environment remain unaddressed. 36 The environment is also critical to the dynamic capabilities framework (Davis et al., 2009; Dixon et al., 2010; Teece et al., 1997). The dynamic capabilities view represents an appropriate framework for the development of a theory of organizational change in turbulent environments (Dixon, Meyer, and Day 2010, Eisenhardt and Martin 2000, Teece, Pisano, and Shuen 1997). Scholars argue that dynamic capabilities are applicable not only to more stable environments (Zollo and Winter, 2002) but to rapidly and moderately changing environments (Eisenhardt and Martin, 2000). Our research uses the dynamic capabilities framework to analyze the impact of dynamism on organizational survival actions such as firm retrenchment. We use one of the most commonly applied dynamic capabilities model (Barrales-Molina et al., 2013; Peteraf et al., 2013), which is proposed by Teece (1997) and Teece and Shuen (2007) to address the following two questions: 1) In what way do asset and cost retrenchment affect a firm’s performance under conditions of high dynamism? 2) Do firms using asset retrenchment, the retrenchment strategy most linked to dynamic capabilities, perform better under dynamic environments than firms using cost retrenchment? In summary, we study the relationship between retrenchment, representing extreme firm change, and dynamism through a lens – the dynamic capabilities framework – suitable for the study of both. Our research builds on the dynamic capabilities framework and develops arguments for the effects of retrenchment in an environmentally dynamic 37 setting. Although the concept of retrenchment has been widely debated, there is considerable ambiguity concerning its strategic nature. Retrenchment is often characterized as operational actions. We argue and theorize, consistent with other scholars (Arogyaswamy et al., 1995; Barker III and Duhaime, 1997), that retrenchment is a stage in which firms not only implement operational actions but strategic actions. Strategic action is recognized as a dynamic capability (Bowman and Ambrosini, 2003; Eisenhardt and Martin, 2000; Katkalo et al., 2010; Moliterno and Wiersema, 2007). Next, we theorize the nature of retrenchment as a cluster of dynamic capabilities recognized by literature and identify them beyond strategic actions. In arguing the dynamic capability nature of retrenchment, we use the sensing-seizing-transforming framework by Teece (2007). We hypothesize first that dynamism has a positive effect on the relationship between retrenchment and performance. Second, we hypothesize that the effects of dynamism depend on the chosen retrenchment strategy. We argue that the effects will be lessened in the case of cost retrenchment compared to asset retrenchment because the nature of cost retrenchment is less related to dynamic capabilities (Morrow et al., 2004). We test the arguments on a group of Spanish firms involved in asset or cost retrenchment between the years 2000 and 2005 and the effects on performance two years later. Our dynamic panel data analysis allows us to conclude that both types of retrenchment have a positive effect on firm 38 performance in dynamic conditions. However, asset retrenchment has a greater impact than cost retrenchment. Our study enhances knowledge concerning the effects of dynamism and retrenchment actions on firm performance. However, most importantly, our primary contribution is to articulate the nature of retrenchment actions and turnarounds by building on the dynamic capabilities framework (Danneels, 2011; Teece, 2012). Additionally, this study identifies new types of firms for which dynamic capabilities can be of use (Barreto, 2010). 2. Theoretical background 2.1 Clarifying the nature of retrenchment: operating actions and strategic actions First, we argue that firms affect operating actions as well as strategic actions during retrenchment. The discussion is significantly important in order to categorize retrenchment correctly as a set of dynamic capabilities. A turnaround is achieved when a firm facing life-threatening performance declines over a period, succeeds in reversing the performance decline, and achieves sustained profitability (Barker III and Duhaime, 1997). The literature has established the distinction between two types of actions to achieve a successful turnaround. These are “operating” and “strategic” actions 39 (Bibeault, 1982; Ndofor et al., 2013; Schendel et al., 1976). Operating actions address efficiency achievements and seek short-term cost reductions (Ndofor et al., 2013; Trahms et al., 2013). Strategic actions address adjustments in a firm´s domain or the way in which the firm competes within those domains (Ndofor et al., 2013) and are oriented towards sustained long-term profitability (Barker III and Duhaime, 1997). The turnaround model proposed by Robbins and Pearce (1992) has been accepted mostly in the literature, although it has some critics (Arogyaswamy et al., 1995; Barker and Mone, 1994; Pearce and Robbins, 1994). The model distinguishes between retrenchment and recovery as the two distinct stages in a turnaround. During retrenchment, firms execute actions to ensure survival and the achievement of positive cash flow (Robbins and Pearce, 1992). Conversely, during recovery, firms shift objectives to growth and development (Robbins and Pearce, 1992). The main source of confusion in the turnaround literature that prevents research progress is the role of strategic action in the turnaround stages (Arogyaswamy et al., 1995; Barker III and Duhaime, 1997). Theorists have adopted two perspectives on this issue. First, some theorists categorize retrenchment as operating actions rather than strategic actions (Ndofor et al., 2013; Schmitt and Raisch, 2013). This perspective defines retrenchment as the elimination of assets and cost reductions as a means of increasing 40 firm efficiency (Dewitt, 1998; Hambrick and Schecter, 1983; Schmitt and Raisch, 2013) and posits that strategic actions do not have a relevant role in the retrenchment stage. Second, other theorists define retrenchment as activities taken in lieu of not only of efficiency (operating actions), but also, and most importantly, the refocus of strategy (Arogyaswamy et al., 1995; Barker III and Duhaime, 1997; Boyne and Meier, 2009; Dawley et al., 2002; Johnson, 1996; Pearce II and Robbins, 2008). From this perspective, retrenchment is not confined to operating actions but also entails strategic actions. Retrenchment is, therefore, defined as a reduction in the size and scope of a business (Boyne and Meier, 2009). This perspective considers strategic actions such as exiting difficult markets, deleting unprofitable product lines, divesting lines of business that do not fit the core, discontinuing unpromising products, and changing the mix of products and services controlled by a strategic unit as part of retrenchment (Barker III and Duhaime, 1997; Boyne and Meier, 2009; Pearce II and Robbins, 2008; Sudarsanam and Lai, 2001). Therefore, from this perspective, firms do engage in strategic actions during retrenchment. Actions taken during retrenchment are unanimously considered operational in nature; however, the specific content of those actions is also strategic. Nonetheless, consistent with the definition of retrenchment (Barker III and Duhaime, 1997; Bibeault, 1982; Robbins and Pearce, 1992), the type of 41 strategic action taken during this stage is survival and positive cash floworiented strategic action rather than growth-oriented strategic action. A turnaround involves several years of declining profitability and a final year of straight loss (Bruton et al., 2003; Morrow et al., 2004). Firms experiencing losses or lack of cash find it difficult to implement growth-oriented strategic action. A strategic approach based on investments would further deteriorate the firm´s position. Only when the company can generate positive cash flow is growth-oriented strategic action an option (Platt, 2004). This situation occurs after the retrenchment stage and during the recovery stage (Robbins and Pearce, 1992). However, as Porter (1996: p. 18) clarifies “strategy renders choices about what not to do as important as choices about what to do.” Those retrenchment-related decisions and actions are purely strategic because they represent adjustments in a firm´s domain (Ndofor et al., 2013) and long-term strategic orientation is necessary (Dewitt, 1998; Lim et al., 2013). The next subsection describes retrenchment as a process by which firms, through sensing, shaping, and reconfiguration, readapt to the environment after years of neglect to respond to fast-changing markets. Retrenchment involves a set of competencies that have been categorized as dynamic capabilities by research. 48 configurations (Katkalo et al., 2010; Teece, 2012) and 2) the mobilization of resources to address threats and opportunities to capture value (Katkalo et al., 2010; Teece, 2012). First, with respect to the evaluation and choice of resources, seizing involves the when, where, and how much of the reconfiguration process and the selection and creation of a business model (Teece, 2007). This activity implies an adjustment in the firm´s domain, which, as argued earlier, represents an activity component of retrenchment. Theorists have acknowledged that these activities are also dynamic capabilities-related (Bowman and Ambrosini, 2003; Katkalo et al., 2010; Moliterno and Wiersema, 2007). In evaluating resources to change asset configurations, retrenching firms lack resources to invest because the decline has eroded the asset base. Thus, the orchestration of assets – identification, prioritization, and the selection of projects (Teece, 2012) – is linked to asset reconfiguration (Teece, 2007). Reconfiguration describes the methods used by firms to modify existing capabilities (Lavie, 2006) and is a core element of dynamic capabilities (Teece, 2007). By reconfiguring, the firm tries to escape unfavorable path dependencies, inertia, and other rigidities (Helfat et al., 2007; Teece, 2007). Lavie (2006) describes three reconfiguration mechanisms: substitution, evolution, and transformation. During retrenchment, firms are more likely to pursue asset reconfigurations using a substitution or transformation mechanism. Because successful evolution 49 requires substantial time (Lavie, 2006), it is less applicable during the timeconstrained retrenchment stage. As argued, during retrenchment, firms decide which activities, market segments, and products the company will stop producing to survive (Barker III and Duhaime, 1997). Substitution and transformation are linked to retrenchment. Second, with respect to resource mobilization, retrenchment requires change capabilities to execute asset reconfigurations. At the core of dynamic capabilities is the dynamic ability to change or reconfigure operating capabilities (Winter, 2003; Zahra et al., 2006; Zollo and Winter, 2002). The required change capacities may exist within the organization; however, it is impossible to retain full transformational capacities within the firm. Consequently, these capacities may be external to the organization (Teece, 2012). Transformation capacity that has been absent can be provided by enhancing and reconfiguring routines and resources and combining and protecting resources (Teece, 2007). A successful turnaround depends on the efficient and effective management of the retrenchment process (Bruton et al., 2003; Dewitt, 1998). The role of leadership is integral to successful turnarounds because dynamic capabilities require superior management entrepreneurial and leadership skills to effect aggressive decisions concerning what and where to retrench (Teece, 2012). 50 In summary, seizing at the retrenchment stage involves dynamic capabilities that have been recognized by the literature, such as strategic decisionmaking capabilities (Bowman and Ambrosini, 2003; Eisenhardt and Martin, 2000; Katkalo et al., 2010; Moliterno and Wiersema, 2007; Slater et al., 2006), product, production process, scale of markets served, change capabilities (Winter, 2003), operations alteration capabilities (Helfat and Winter, 2011; Zahra et al., 2006), new business processes implementation capabilities (Drnevich and Kriauciunas, 2011), capabilities to change ways of doing business (Drnevich and Kriauciunas, 2011; Teece, 2007), renewal capabilities (Shamsie et al., 2009), and resource orchestration capabilities (Katkalo et al., 2010). Transforming. Transforming entails the continuous renewal of the firm to sustain competitive advantage (Katkalo et al., 2010; Teece, 2007, 2012). It refers to the reconfiguration of assets and structures to maintain competitiveness, given that firms have a continuous need to modify product offering, business models, enterprise boundaries, and organizational structures (Teece, 2007). Continuous renewal softens the rigidities derived over time from asset accumulation, standard operating procedures, and insider misappropriation of rent streams (Teece, 2014b). 51 Transforming is related more to the second turnaround stage, recovery 1 . Recovery is the stage in which competitive advantage is pursued. However, retrenchment is relevant to transforming, as during this turnaround stage, the firm prepares for transforming by modifying organizational traits. The result of retrenchment is an organization that is better able to achieve continuous renewal. Retrenchment involves the reduction of organizational assets and costs. The resulting firm will be a smaller organization. By reducing the size, the firm reduces its complexity, reduces the need for organization and control, and becomes less structurally mechanistic (Sutton and D´Aunno, 1989). Less structurally mechanistic firms are more flexible and thus, are more apt to adapt to a changing environments (Davis et al., 2009). In addition, retrenchment devolves the organization to financial stability and the firm will become cash flow positive (Pearce II and Robbins, 1993; Robbins and Pearce, 1992). Beyond those internal sources of financing, the improvement of the financial health will provide the organization with access to external finance. Thus, the new organization will be able to become involved in continuous renewal projects requiring financial support. In summary, after retrenchment, the declining organization will possess traits that allow it to 1 In line with Arogyaswamy et al. (1995), we do not intend to suggest that the retrenchment and recovery stages should necessarily be sequentially addressed. Our opinion is, they could be simultaneously dealt with, as recent research has suggested (Schmitt and Raisch, 2013) . Similarly, Teece (2007) sensing-seizing-transforming model does not seem to be conceived as a sequential one. The author has suggested in numerous instances how the model is often simultaneously applied (Teece, 2007, 2014a, 2014b). 52 achieve a better fit between the strategy, structure, and the processes. Hence, retrenchment is preparatory for the transformation of a firm, which prepares the firm for continuous renewal. Specifically, the retrenched firm will be better able to meet the four microfoundations of transforming described in the model of Teece (2007). We describe next how the retrenched organization is better able to cope with the microfoundations of transforming. First, with regard to decentralization and near decomposability, during retrenchment, the organization needs to become more centralized in order to gain tighter control because survival is at stake (D´Aveni, 1989; Schmitt and Raisch, 2013; Whetten, 1987). Subsequently, retrenchment reduces personnel and delayers the organization to produce a more agile, simplified, and smaller firm that is better able to communicate. Delayering and simplification will help the organization to achieve the decentralization and near decomposability necessary for transformation (Teece, 2007). A more agile organization will be able to speed up the decision-making process. Faster decision making allows firms to adopt faster new products, processes, business models, and economies of learning faster (Baum and Wally, 2003; Eisenhardt, 1989). Second, with regard to managing cospecialization (Teece, 2007), cospecialized assets are a class of complementary assets whose value depends on their joint use with other assets. Asset cospecialization allows organizational renewal in turnaround processes 53 (Ruiz-Navarro, 1998). Cospecialization requires investment or access to external resources (Teece, 2007). Retrenchment enables the firm to devolve to financial stability and positive cash flow (Pearce II and Robbins, 1993; Robbins and Pearce, 1992). The retrenched firm will be more able to obtain financing to build or buy cospecialized assets. Third, with regard to learning and knowledge management, the retrenched organization is a more simplified less complex organization. These traits will improve organizational learning and knowledge management. In addition, retrenched organizations are likely to reduce routine learning and groupthink, and open opportunities for the creation of new network connections (Fisher and White, 2000). Finally, with regard to corporate governance, through retrenchment, firms replace the top management team demonstrating weak sensing, seizing, and transforming capabilities (Teece, 2007). Retrenchment provides the organization with a team that will help the organization achieve continuous renewal. An issue critical to dynamic capabilities is the need for repeated or patterned behavior that renders activities a capability (Helfat and Winter, 2011; Helfat et al., 2007). There are two arguments that lead to the consideration of retrenchment strategies as routines. First, literature has supported the theory that some of the measures regularly taken during asset retrenchment and cost retrenchment, such as sell-offs or spin-offs, although occasionally practiced, are practiced by firms with sufficient 54 frequency to be considered a capability rather than an ad hoc activity (Bergh and Lim, 2008). Firm decline – and thus retrenchment – is a normal firm process (Bergh and Lim, 2008). Time spans between periods of decline can be long and the degree of retrenchment intensity can be less than that of an actual turnaround. However, because decline is an event that repeats itself in firms, retrenchment is a routine activity. Third, asset and cost retrenchment are clusters of dynamic capabilities. Even if a firm does not embark in a turnaround every year, those capabilities that compose asset and cost retrenchment – analytical capabilities, absorptive capacity, knowledge transfer capabilities, strategic decision-making capabilities, and resource orchestration capabilities – are activities routinely practiced by firms. In summary, retrenchment involves sensing, seizing and transforming to address and shape a rapidly changing business environment (Danneels, 2011; Teece et al., 1997). By using the dynamic capabilities framework proposed by Teece (2007), we have described and detailed the specific dynamic capabilities used during the retrenchment stage of a turnaround. 3. Hypoheses This section first hypothesizes that the more dynamic the environment, the greater the (positive) impact that retrenchment, as a set of dynamic 55 capabilities, has on firm performance. Second, we hypothesize that the extent of the environmental dynamism´s impact on performance depends on the type of retrenchment strategy selected – asset retrenchment or cost retrenchment. Asset retrenchment is more dynamic capability-oriented than cost retrenchment; thus, we argue that, under dynamic environments, firm performance of asset retrenchment will be superior to that of cost retrenchment. Abundant literature has supported that the environment affects organizations (Baum et al., 2001; Lumpkin and Dess, 1996). Dynamism is defined as the level of turbulence or instability facing an environment (Boyd, 1990). Dynamism is an environmental condition and there is substantial evidence of its effects on firm performance (Ensley et al., 2006; Priem et al., 1995; Wang and Li, 2008; Wiklund and Shepherd, 2003). Similarly, the most recent research on turnaround provides evidence of environmental and, more particularly, dynamism effects on the processes and outcomes of declining firms (Boyne and Meier, 2009; Lim et al., 2013; Morrow et al., 2004; Ndofor et al., 2013). A dynamic environment may trigger a firm to change and to use dynamic capabilities. Dynamism has two effects on firms. First, dynamism raises the awareness of the need to change and, second, it drives organizations to effect change. 56 Dynamism can facilitate firm awareness of the need to change. An unusual level of environmental dynamism, or a level that differs from management expectations, attracts the attention of the executive team (Daft et al., 1988). The environment alerts management of the inadequacies in proprietary ordinary capabilities. Management will, upon a surge in dynamism, start to consider the need to use proprietary dynamic capabilities to reconfigure ordinary capabilities (Zahra et al., 2006). Management awareness of the need to change is also raised by the speed of firm decline. A slow firm decline results in less awareness of the situation by management (Barr et al., 1992). Conversely, awareness of the need to change is more salient in firms subject to rapid decline. Hence, firm awareness of the need to change is raised by both, a dynamic environment and a high level of distress. Even if the declining firm is aware of the need to change, research suggest that poor performance is not sufficient to move towards change (Barker III et al., 2001). The firm will not act until the level of stress peaks. This happens when poor results are combined with a precipitating event. Additionally, Barker and Patterson (2001) suggest that this precipitating event can be caused by an environmental factor (Barker III, Patterson Jr., and Mueller 2001). High dynamism could be the trigger for the initiation of retrenchment – composed of a set of dynamic capabilities – as the first stage in a 57 turnaround. Thus, awareness of the need to change can be initiated by the joint action of both environmental dynamism and a high level of distress. Second, once dynamism has raised the awareness of the need to change and a precipitating event has triggered the initiation of retrenchment, firms will retrench by deploying the dynamic capabilities retrenchment represents. A dynamic environment is one of the precedents of organizational and strategic change because it stimulates managerial action. Managers, under such conditions, are more likely to invoke change (Gordon et al., 2000). Similarly, in high velocity industries – an industry characterized by rapid and unpredictable changes – the speed of organizational response and change is faster than in low velocity industries (Nadkarni and Barr, 2008). In the context of declining firms, the lower the level of satisfaction with past performance, the higher the incentive to seek change (Audia et al., 2000). Distressed firms have a higher incentive to change. Research on underperforming firms confirms the recognition of the need to retrench (Datta et al., 2010; Filatotchev et al., 2000) or embark on strategic change (Barker III and Duhaime, 1997; Haveman, 1992; Ndofor et al., 2013). Not only does the need to change under dynamic environments stimulate firms to use dynamic capabilities, but the opportunity to use those capabilities provides firms with an incentive to use them. Dynamic 64 retrenchment measures requires less capital, which is critical to be conserved in such a situation of cash starvation. As an alternative to cost retrenchment, firms can resort to asset retrenchment to generate cash flow. Consequently, asset retrenchment involves the collection of cash through the disposal of the firm assets, such as closing plants, divesting equity (sell offs, spin offs, and carve-outs), and reducing stocks of property, equipment, and inventory (Lim et al., 2013; Morrow et al., 2004). Research has found numerous difficulties in the process of asset divestments by declining firms. The difficulties relate to the need for creditor discretion of the disposal, exit barriers, industry liquidity problems, low resale prices derived from fire sale, and asset specificity (Filatotchev and Toms, 2006; Harrigan, 1981). All these difficulties make asset retrenchment a complex and very time-consuming process. For these reasons, asset retrenchment becomes the least used retrenchment alternative to generate cash flow given cost retrenchment is quicker, less capital consuming, and thereby more effective than asset retrenchment. In summary, a critical function of retrenchment is the generation of positive cash flow to avoid firm collapse. Cash flow is then generated to support current operations. Those capabilities involved in the support of current operations are ordinary and not dynamic (Teece, 2012; Winter, 2003), and hence, the function of cash flow generation is related to ordinary capability. Declining firms can improve cash flow generation through asset 65 or cost retrenchment. Cash flow can be generated more optimally through cost retrenchment than through asset retrenchment. Cost retrenchment allows a firm to generate cash flow faster and with a lower use of capital. Declining firms will use asset retrenchment only as a second-best option to cost retrenchment oriented to generate cash. Cost retrenchment is more suited to accomplish this ordinary capability-related function than asset retrenchment. Function 2: Retrenchment as a mean to change a firm’s strategy. We know that “the average turnaround involves significant” strategic change (Barker III and Duhaime, 1997: p. 33). The change of a firm strategy is critical to turnaround success (Zúñiga-Vicente and Vicente-Lorente, 2006). Research on strategic content is concerned with two main matters: the scope of the firm and competition within individual markets 4 (Helfat et al., 2007; Montgomery et al., 1989). We discuss each of these two matters in the context of declining firms and retrenchment. However, it is important to stress that turnaround literature clearly indicates that the most important strategic content element during retrenchment is the delineation of firm scope (Boyne and Meier, 2009; Dawley et al., 2002; Sirmon et al., 2011). 4 Turnaround literature also discusses strategic content. Barker III and Duhaime (1997) concur these two elements are also the relevant ones to firms under turnaround. 66 First, through strategy, a firm delineates the firm scope. The choice of scope (“breadth”) is a key task of the resource orchestration for declining firms (Sirmon et al., 2011: p. 1392). Delineating the firm scope rather than being important for strategy may constitute the very essence of strategy for many firms (Bingham et al., 2007). Delineating the firm scope is part of the strategy of the firm and part of the dynamic capabilities of a firm (Helfat et al., 2007; Sirmon et al., 2011; Teece, 2007). Delineating the firm scope is the principal strategic content of retrenchment (Boyne and Meier, 2009; Dawley et al., 2002). During the early stages of decline, the delineation of a firm boundaries consists of narrowing those boundaries rather than broadening them. Resource orchestration of declining firms will primarily require resource divestments (Sirmon et al., 2011), especially during the early stages of decline. Firms then refocus the scope of the firm to return to profitable core businesses rather than diversifying away from it (Bibeault, 1982; Dawley et al., 2002; Johnson, 1996; Robbins and Pearce, 1992). There are two reasons why declining firms resort to refocusing for boundary delineation. First, diversifying requires the availability of financing and declining firms lack financing or access to it (Gilson, 1990; Hambrick and D´Aveni, 1988). Refocusing is less capital intensive, and hence, is more feasible for cash strapped firms. Second, evidence shows that in situations of decline, to refocus rather than to diversify is a better alternative given that the risks of diversification may be too high (Anand and Singh, 1997; Blatz et al., 2006; 67 Helfat and Eisenhardt, 2004). Firms avoid further risks in a situation in which risks abound, such as decline. Then, declining firms will use retrenchment to delineate a firm´s scope and refocus the firm towards the profitable core. Danneels (2008) provides a good example of refocusing through retrenchment involving dynamic capabilities. Smith Corona, the typewriter firm, retrenched firm assets by divesting its manufacturing facilities because of decreasing demand for the product. The divestment allowed the firm to obtain cash and to concentrate on the typewriter ribbons, a high-margin product for the firm. Declining firms adjust their scope through retrenchment. The choice of scope can range through a continuum from narrow to broad. Thus, using retrenchment, firms adjust the scope by discontinuing SKUs, products, segments, lines of products, lines of business stock keeping units (SKU´s), divisions, domestic geographic areas, countries and entire geographic areas or by backward or forward vertical deintegration. Depending on the choice of scope adjustment, firms will use a different type of retrenchment. Cost retrenchment is used for smaller scope adjustments and thus, for less important strategic moves. Firms requiring less important strategic moves will eliminate SKUs, products or segments. To achieve this, firms will then reduce marketing expenses, salaries and employee expenses, R&D costs, etc. These cost reductions are part of cost retrenchment (Lim et al., 2013). In smaller scope adjustments, asset retrenchment becomes less important, and will 68 typically be limited to inventory or equipment reductions. By contrast, asset retrenchment is required for larger scope adjustments, and thus, it is required for more important strategic moves. Large scope adjustments in turnarounds are implemented through backward or forward deintegration, or the discontinuation of lines of products, lines of business, divisions, domestic geographic areas, countries, or entire geographic areas. These types of large strategic changes can be implemented mainly only through large asset reductions. Asset retrenchment actions, such as plant closing, selling off, spinning off, or carving out full portions of the business (Lim et al., 2013), are required to accomplish those important strategic changes. Furthermore, asset retrenchment has been found to work more successfully in more severe turnaround situations, that is, those situations in which a greater type of strategic adjustment is required (Hofer, 1980; Robbins and Pearce, 1992). Thus, more important strategic changes for declining firms require asset reductions. Conversely, less important strategic changes are pursued through cost retrenchment. This seems to be the view by financial markets as concluded by Morrow et al. (2004). Given that changes in the firm scope are driven by dynamic capability, asset retrenchment will have a greater dynamic capability nature as they are used for more important strategic changes. Second, regarding competition within markets, strategy is concerned with pursuing superior profits at the individual business level from rent-producing 69 resources. Dynamic capabilities do not just rely on any resources but on valuable, rare, imperfectly imitable, non-substitutable (VRIN) resources (Teece, 2014a). We discuss now whether firms can reconfigure resources to generate VRIN resources only by the means of retrenchment. The RBV perspective implies that a firm can obtain superior profits either by endowment or investment of resources (Teece, 2014b). A turnaround attempt is triggered a few years after the initiation of decline (Barker III et al., 2001; Tangpong et al., 2015). By then, the resource base tends to be depleted due to inertial constant forces acting over time (Arogyaswamy et al., 1995). The type of resources most eroded are reputational, human, and financial (Barker III and Duhaime, 1997), with the first two categories being the most important to produce a competitive advantage (Teece, 2014b). Thus, by the start of a turnaround, resource endowment of declining firms is low, given the resource base has been eroded and the stock of resources possessed are second tier. The question is, then, whether a firm with depleted resources can modify existing individual business resource base to generate VRIN resources. This view is not supported by empirical (although scarce) evidence. Sirmon et al. (2007) argues that to alter current capabilities, firms require slack. However, depletion of resources during the decline process has eroded the firm slack. Morrow Jr. et al. (2007) showed that declining firms find it difficult to improve the resource base significantly only by reconfiguring the firm’s existing resources. They indicate that 70 declining firms will require new resources to reconfigure existing resources into new VRIN resources. Furthermore, minor acquisitions of additional resources (e.g., computer equipment and staff members) may not suffice. Declining firms will need to obtain significant levels of new resources from external sources (Morrow Jr. et al., 2007). In summary, reconfiguring the resource base to generate VRIN resources at the individual declining business is hardly feasible only through retrenchment, given that the firm will require acquiring significant levels of external resources after a period of resource erosion. Thus, using retrenchment, the feasible changes to the individual business resource base mainly will be operational and not strategic. Then, the principal strategic alternative during retrenchment is reducing assets and costs to refocus the firm scope (Boyne and Meier, 2009; Sirmon et al., 2011). Scope refocusing is a strategic activity and involves dynamic capabilities. Broader scope with more important strategic changes requires asset reallocations through asset retrenchment, given that the potential of cost retrenchment for strategic change is narrower. Thus, asset retrenchment has a more dynamic capability than cost retrenchment. Finally, when declining firms adjust the firm scope or reconfigure resources through retrenchment, they will use some of the dynamic capabilities described earlier, which are part of the sensing and seizing clusters of 71 dynamic capabilities. Among the dynamic capabilities most used are analytical methodologies (Teece, 2012), learning (Zollo and Winter, 2002), strategic decision-making (Katkalo et al., 2010; Slater et al., 2006), operations alteration capabilities (Helfat and Winter, 2011), divestment (Mathews, 2003; Moliterno and Wiersema, 2007; Teece, 2014b), and unlearning (Lavie, 2006; Schilke, 2014). Function 3: Retrenchment as a means to improve operations. Declining cash-strapped firms use asset and cost retrenchment to become more profitable and competitive by analyzing the efficiency of their operations and by benchmarking them with the industry. First, declining firms analyze their internal structure to find surplus expenses and inefficient operations. Once they find surplus expenses and redundant assets, they use asset retrenchment and cost retrenchment to become more operationally efficient. Even though activities oriented to raise operational efficiency are necessary to achieve superior profits, they cannot be considered strategic but operational (Grant, 2013; Porter, 1996). Strategy involves “doing things differently,” whereas efficiency measures involve “doing the same things on a smaller or more efficient way” (Pearce II and Robbins, 1993: p. 621). Thus, when firms implement measures whose effect is to become more efficient, they involve ordinary capabilities (Helfat and Peteraf, 2003; Zahra et al., 72 2006). Retrenchment as a means to improve firm efficiency is then intrinsically more related to ordinary capabilities. Second, through industry benchmarking, declining firms align their structure to the industry average and thus, become more competitive (Grinyer and Spender, 1988). When firms benchmark they are aligning themselves to industry best practices. Teece (2014) indicates that ordinary capabilities are considered strong when the firm has achieved best practices. However, benchmarking best practices only involve ordinary capabilities and do not involve dynamic capabilities (Teece, 2014a). The reason is that benchmarking provides operational efficiency and capabilities possessing an operational nature are considered to be dynamic (Teece, 2007; Zollo and Winter, 2002). Firms can become efficient through asset or cost retrenchment. By means of asset retrenchment, firms become more efficient by streamlining the balance sheet through the sale of inventories, raw materials, etc. In addition, firms can also become more efficient through cost retrenchment by cutting costs such as selling, general, and administrative (SGA) expenses, interest, and pensions. Cost retrenchment is the preferred alternative by managers to improve operations. Cost retrenchment allows firms to become profitable faster and requires less capital than asset retrenchment (Robbins and Pearce, 1992; Schoenberg et al., 2013). Hence, when pursuing 73 efficiency, firms find cost retrenchment “so pervasive as to be considered indispensable” (Robbins and Pearce, 1992: p. 303). Finally, given this function is realted to ordinary capability, cost retrenchment will have a more ordinary capability nature. In summary, although cost retrenchment can be used for strategic purposes, firms mainly use it for operational improvement and cash generation purposes. That is for two reasons. First, cost retrenchment allows a faster and less capital-consuming introduction of efficiency measures and cash generation. Second, the amount of strategic change that can be implemented through cost retrenchment is limited given larger strategic changes require asset reallocations. Given, the higher efficiency seeking nature of cost retrenchment, the measure will possess a more ordinary capability nature. Conversely, asset retrenchment can be used to generate cash and improve operations. However, firms mainly use it to implement strategic change. First, asset retrenchment requires the transfer of assets which is a slow and difficult process, whereas cash-strapped firms require fast liquidity generation. Second, asset retrenchment is a more optimal choice to modify a firm’s strategy given that it allows more important strategic moves. Given the larger strategic nature of asset retrenchment, it will be more related to dynamic capability. 80 the legally incorporated for-profit entities in Spain at the sector’s four-digit level. Control variables. This study included seven control variables to mitigate potential misinterpretations of our findings. The initial state of the firm can impact the firm´s retrenchment process (Lim et al., 2013; Love and Nohria, 2005; Morrow et al., 2004). We included the variable severity to control for this impact, which was calculated as the sector-adjusted ROS the year prior to retrenchment. Organizational performance is likely to vary with the size and age of a firm. Prior studies have shown that firm size negatively affects the ability of managers to execute changes (Boyne and Meier, 2009; Hambrick and D´Aveni, 1988; Schmitt and Raisch, 2013). The number of employees was used to measure firm size (Lim et al., 2013; Morrow et al., 2004). Firm age was measured as the number of years since founding (Latham and Braun, 2009). We log-transformed these two variables to reduce highly skewed values. Four further control variables were used: slack, gearing, liquidity, and munificence. Slack represents shortterm resource accessibility. Previous research has shown evidence of the extensive influence of slack in turnaround research (Trahms et al., 2013) and in turnaround success (Hambrick and D´Aveni, 1988; Morrow et al., 2004). Slack was measured as working capital divided by total assets (Morrow Jr. 81 et al., 2007). We included gearing as a control variable because debt limits an organization´s ability to turn around (Powell, 2001; Trahms et al., 2013). Gearing was measured as the total debt to total asset ratio (Chen and Hambrick, 2012; Schmitt and Raisch, 2013). Liquidity was operationalized as the current ratio (Morrow et al., 2004). We also controlled for munificence because munificent industries have been observed to improve turnaround performance (Morrow et al., 2004). Industry munificence was operationalized as the regression coefficient of time on annual average sales for each industry, divided by the mean value of sales (Dess and Beard, 1984; Nielsen and Nielsen, 2013). Finally, because we sector-adjusted our dependent variable, and to avoid redundant analysis, we excluded industry control variables from our model (Dess and Beard, 1984; Nielsen and Nielsen, 2013). 4.3 Analysis We analyzed our hypotheses using the two-step Generalized Method of Moments (GMM) developed by Arellano and Bond (1991) and Blundell and Bond (1998). The method proposes a dynamic panel generalized method of moments estimator. We employed the Arellano-Bond estimation because it offers four advantages to our study: 1) We use a lagged dependent variable, which raises concerns about potential inconsistency in 82 estimates because of the likelihood that the lagged dependent variable is correlated with the error term (Greene, 2003). The Arellano-Bond method accounts for the dynamic nature of the dependent variable and its dependence on its lagged values. 2) Our variable severity has a time invariant nature. System GMM allows the inclusion of time invariant regressors (Roodman, 2009). 3) The method accounts for the effects of unobserved firm-specific heterogeneity, eliminating the risk of obtaining biased results arising from this heterogeneity (Hsiao, 1993). 4) The literature has demonstrated the effect of endogeneity on multiple variables typically used in turnaround studies (Datta et al., 2010). GMM allows us to avoid concerns of endogeneity that could affect the estimation results. We employ the robust two-step system GMM estimator with the Windmeijer corrected standard errors to correct for the downward bias on standard errors (Arellano and Bond, 1991; Blundell and Bond, 1998). The estimator provides estimates that are robust to the presence of autocorrelation and heteroskedasticity. Our model included one lag of the dependent variable and independent and control variables and interaction terms specified as endogenous. We used these variables lagged up to three times as instruments. We used the variables age, dynamism, munificence, and the time dummies as exogenous instruments to reduce the possibility of spurious effects. We mean-centered all the independent variables that constitute the 83 interaction terms to avoid multicollinearity (Aiken and West, 1991). Meancentering the variables also simplified coefficient interpretation. 5. Results Table 1 presents the descriptive statistics and correlation matrix for all of the variables used in the study. The values are shown prior to mean-centering and log. ------------------------------------------------------------ Insert Table 1 about here ------------------------------------------------------------- Table 2 presents the results for the determinants of performance based on system GMM estimations. The table presents four double columns corresponding to each of our four models. Each column includes the beta and the standard error for each model. All of the models in Table 2 are significant, and the stepwise addition of variables improves the fit over the preceding model based on the 𝜒2. ------------------------------------------------------------ Insert Table 2 about here ------------------------------------------------------------- 84 All reported models include year dummies to control for any idiosyncratic effects resulting from fluctuations in macroeconomic conditions. Model 1 represents the baseline model reporting the effect of the control and main effects variables. Model 1 indicates that cost retrenchment (b = 0.019, p < 0.05) is significant and positively associated with firm performance. Models 2 to 4 include the interaction terms required to test the hypotheses. In models 2 and 3, the two interaction terms are introduced one at a time. In model 4, the two interaction terms are introduced into the regression simultaneously. Hypothesis 1 posits that the higher the dynamism, the stronger the effects of retrenchment. As shown in table 2, the interaction terms in Model 2 (Asset Retr X Dynamism) and Model 3 (Cost Retr X Dynamism) are both positive (b = 0.014 and b = 0.006, respectively) and significant (p < 0.001 and p < 0.01, respectively), providing support for Hypothesis 1. Model 4 includes both interaction terms simultaneously. The coefficients in Model 4 for asset retrenchment and cost retrenchment interaction with dynamism are both positive (b = 0.012 and b = 0.003, respectively) and significant (p<0.01 and p<0.05, respectively), consistent with the predictions of Hypothesis 1. We can also use Models 2, 3, and 4 to evaluate Hypothesis 2. Hypothesis 2 predicts that the moderating effects of asset retrenchment will be stronger than the moderating effects of cost retrenchment the higher the level of 85 dynamism. The interaction of asset retrenchment and dynamism in Model 2 is greater than the interaction of cost retrenchment and dynamism in Model 3. Model 4 shows that when both interaction terms are included simultaneously, the interaction of asset retrenchment and dynamism is greater than that of cost retrenchment and dynamism, providing support for Hypothesis 2. To further demonstrate the moderating effects of dynamism on the relationship between asset and cost retrenchment and firm performance, we plotted the relationships (Figures 1 and 2). Figures 1 and 2 show that the effects of asset and cost retrenchment on firm performance are greater in highly dynamic environments, and the slope of the relationship is more pronounced and positive the greater the dynamism. However, a comparison of both types of retrenchment reveals that the effects of asset retrenchment are greater than the effects of cost retrenchment. ------------------------------------------------------------ INSERT FIGURES 1 & 2 ABOUT HERE ------------------------------------------------------------- In summary, we find that dynamism positively moderates the relationship between retrenchment and performance in turnaround firms. The conclusion is applicable to both asset and cost retrenchment. However, the 86 effects of dynamism are greater for asset retrenchment than for cost retrenchment. With respect to the control variables, the results are consistent with previous research findings (Morrow et al., 2004; Schmitt and Raisch, 2013). We find that liquidity has a significant positive effect on firm performance. Similarly, severity also has a consistent significant positive effect on performance (Morrow et al., 2004). 6. Discussion Firm decline has been linked, among other factors, to environmentally related causes in the literature (Arogyaswamy et al., 1995; Trahms et al., 2013). Studies have found interactive effects of industry type or industry munificence on the relationship between retrenchment and performance (Lim et al., 2013; Morrow et al., 2004; Ndofor et al., 2013). Such refinement has shed some light on the role of industry traits on retrenchment success. However, extant turnaround research ignores the interactive effects of dynamism on retrenchment. In parallel, turnaround research has mainly been phenomenon-driven with less theoretical development (Trahms et al., 2013). Nevertheless, the use of relevant theories advances the field because they act as a source of new discoveries (Trahms et al., 2013). The primary goal of our study is to contribute to the retrenchment literature by examining the role of dynamism on retrenchment actions using a 87 relevant framework: the dynamic capabilities framework is applicable to situations of significant change and we have used it to argue the effects of dynamism on core turnaround actions – asset and cost retrenchment. We have theorized that retrenching firms perform a series of activities, some of which are considered dynamic capabilities by many scholars. Our results support the hypothesis that the more dynamic the environment, the more effective retrenchment actions are on turning around firm performance. If retrenchment represents a set of dynamic capabilities, the use of these capabilities under a dynamic environment will enhance performance, as other scholars have found in contexts other than declining firms (Drnevich and Kriauciunas, 2011; Schilke, 2014). Although our findings are consistent with those of Schilke (2014a) and Drnevich and Kriauciunas (2011), the results are more aligned to those of Drnevich and Kriauciunas (2011) given that the positive relationship between the interactive effect of dynamic capabilities and dynamism on performance that we found is linear rather than curvilinear. The fact that some authors find a linear relationship while others find a curvilinear one could be related to the type of capability used to measure the dynamic capabilities construct. Future research might pay attention to the differences in performance by the interaction between the different types of dynamic capabilities and dynamism. The second goal of our study is to examine the differences between the performance of asset and cost retrenchment in a dynamic environment. 88 We argue that asset retrenchment is more dynamic capability-oriented than cost retrenchment because asset retrenchment is more strategic, whereas cost retrenchment is more operational and efficiency-oriented. We found that the effect of both types of retrenchment is positive under dynamic conditions. However, consistent with our predictions, we found gains generated through asset retrenchment to be greater than those generated from cost retrenchment. Our research suggests that efficiencyoriented measures, such as cost retrenchment, are less suited to highly dynamic environments than more strategic-oriented measures such as asset retrenchment. The effects of cost retrenchment, involving a set of dynamic capabilities, are positive for firm performance. Conversely, less efficiencyoriented and more strategic and flexibility-oriented measures, such as asset retrenchment, are more suited to dynamic environments. Our results differ from those of Lim et al. (2013), who found that asset retrenchment is a suboptimal choice for firms operating in a Ricardian industry or sector, such as manufacturing. However, the authors recognize that their results are not clear-cut. By contrast, our results in a manufacturing-only (Ricardian type) sample show that the effects of asset retrenchment on firm performance are higher than those by cost retrenchment. In addition, seminal studies on turnaround performance in the manufacturing industry seem to confirm the higher performing nature of asset retrenchment over that of cost retrenchment (Barker and Mone, 1994; 89 Robbins and Pearce, 1992). Although the topic is not the focus of our study, future research should revisit the relationship between retrenchment strategy and type of industry as an important topic of research. Our study represents a contribution to both the turnaround literature and the dynamic capabilities literature. We followed recent research focusing on the effects of the environment on turnarounds and retrenchment (Boyne and Meier, 2009; Lim et al., 2013; Morrow et al., 2004). These studies showed how the characteristics of a sector –munificence, the type of sector, Schumpeterian/Ricardian – impact retrenchment. Our findings extend this research and provide arguments and evidence that dynamism also has an impact on the effects of retrenchment. Our work also represents a contribution to the dynamic capabilities literature. It has been argued that more research on dynamic capabilities is needed for new types of firms (Barreto, 2010). Our work focuses on the sparse literature researching the use of dynamic capabilities by surviving firms (Danneels, 2011; Zúñiga- Vicente and Vicente-Lorente, 2006). Our results suggest that the dynamic capabilities framework can be used beyond stable firms. The framework is applicable to situations of firm radical change (Eisenhardt and Martin, 2000; Helfat and Winter, 2011; Teece, 2007; Teece et al., 1997), such as turnaround situations (Teece, 2012). Organizational decline is caused, among other factors, by environmental changes, and a firm´s set of dynamic capabilities can help reverse the situation. 96 Management Studies, 47(3), 416–436. Drnevich PL and Kriauciunas AP (2011) Clarifying the conditions and limits of the contributions of ordinary and dynamic capabilities to relative firm performance. Strategic Management Journal, 32(3), 354–379. Eisenhardt KM (1989) Making fast strategic decisions in high velocity environments. Academy of Management Journal, 32(3), 543–576. Eisenhardt KM and Brown SL (1995) Product development: past research, present findings and future directions. Academy of Management Review, 20(2), 343–378. Eisenhardt KM and Martin JA (2000) Dynamic capabilities: What are they? Strategic Management Journal, 21(10-11), 1105–1121. Elliott D and Smith D (2006) Cultural readjustment after crisis: Regulation and learning from crisis within the UK soccer industry. Journal of Management Studies, 43(2), 289–317. Ensley MD, Pearce CL and Hmieleski KM (2006) The moderating effect of environmental dynamism on the relationship between entrepreneur leadership behavior and new venture performance. Journal of Business Venturing, 21(2), 243–263. Farjoun M and Levin M (2011) A Fractal Approach to Industry Dynamism. Organization Studies, 32(6), 825–851. Filatotchev I and Toms S (2006) Corporate governance and financial constraints on strategic turnarounds. Journal of Management Studies, 43(3), 407–433. Filatotchev I, Buck T and Zhukov V (2000) Downsizing in privatized firms in Russia, Ukraine, and Belarus. Academy of Management Journal, 43(3), 286–304. Fisher SR and White M a. (2000) Downsizing in a learning organization: Are there hidden costs? Academy of Management Review, 25(1), 244–251. Gilson SC (1990) Bankruptcy, boards, banks, and blockholders. Evidence on changes in corporate ownership and control when firms default. Journal of Financial Economics, 27(2), 355–387. Gordon SS, Stewart Jr. WH, Sweo R, et al. (2000) Convergence versus strategic reorientation: The antecedents of fast-paced organizational change. Journal of Management, 26(5), 911–945. Grant R (2013) Contemporary Strategy Analysis. eigth. Wiley. Greene WH (2003) Econometric analysis. New York: Macmillan. 97 Grinyer P and Spender J (1988) Sharpbenders: The secrets of unleashing corporate potential. Oxford: Basil Blackwell. Gruber M, Heinemann F, Brettel M, et al. (2010) Configurations of resources and capabilities and their performance implications: An exploratory study on technology ventures. Strategic Management Journal, 31(12), 1337–1356. Hambrick D and D´Aveni R (1988) Large orporate failures as downward spirals. Administrative Science Quarterly, 33, 1–23. Hambrick D and D´Aveni R (1992) Top team deterioration as part of the downward spiral of large corporate bankruptcies. Management Science, 38(10), 1445–1466. Hambrick D and Schecter S (1983) Turnaround Strategies in mature industrialproduct businesss units. Academy of Management Journal, 26, 231–248. Harrigan KR (1981) Deterrents to Divestiture. Academy of Management Journal. Haveman HA (1992) Between a Rock and a Hard Place: Organizational Change and Performance under Conditions of Fundamental Environmental Transformation. Administrative Science Quarterly, ITHACA; CORNELL UNIV, JOHNSON SCHOOL, 20 THORNWOOD DR, STE 100, ITHACA, NY 14850-1265: Administrative Science Quarterly, 37(1), 48–75. Helfat C and Winter S (2011) Untangling dynamic and operational capabilities: Strategy for the (N)ever-changing world. Strategic Management Journal, 32(11), 1243–1250. Helfat C, Finkelstein S, Mitchell W, et al. (2007) Dynamic capabilities: understading strategic change in organizations. Publishing B (ed.). Helfat CE (1997) Know-how and asset complementarity and dynamic capability accumulation: The case of R&D. Strategic Management Journal, 18(5), 339– 360. Helfat CE and Eisenhardt KM (2004) Inter-temporal economies of scope, organizational modularity, and the dynamics of diversification. Strategic Management Journal, 25(13), 1217–1232. Helfat CE and Martin JA (2014) Dynamic Managerial Capabilities: Review and Assessment of Managerial Impact on Strategic Change. Journal of Management, SAGE Publications Inc., 41(5), 1281–1312. Helfat CE and Peteraf M a (2003) The dynamic resource based view: Capability lifecycles. Strategic Management Journal, 24(10), 997–1010. 98 Hermann CF (1963) Some consequences of crisis which limit the viability of organizations. Administrative Science Quarterly, 8, 82. Hine D, Parker R, Pregelj L, et al. (2013) Deconstructing and reconstructing the capability hierarchy. Industrial and Corporate Change, 23(5), 1299–1325. Hofer C (1980) Turnaround strategies. The journal of business strategy, 1(1), 19–31. Hsiao C (1993) Analysis of panel data. Cambridge, UK: Cambridge University Press. Johnson RA (1996) Antecedents and outcomes of corporate restructuring. Journal of Management, 22, 439–483. Katkalo VS, Pitelis CN and Teece DJ (2010) Introduction: On the nature and scope of dynamic capabilities. Industrial and Corporate Change, 19(4), 1175–1186. Kay NM (2010) Dynamic capabilities as context: the role of decision, system and structure. Industrial and Corporate Change, 19(4), 1205–1223. Khurana R (2002) Searching for a corporate saviour: The irrational quest for charismatic CEOs. Princeton, NJ: Princeton University Press. Klarner P and Raisch S (2013) Move to the beat-rhythms of change and firm performance. Academy of Management Journal, 56(1), 160–184. Kogut B and Zander U (1996) What Firms Do? Coordination, Identity, and Learning. Organization Science, 7(5), 502–518. Krause R, Semadeni M and Cannella Jr. AA (2013) External COO/presidents as expert directors: A new look at the service role of boards. Strategic Management Journal, 34(13), 1628–1641. Lant TK, Milliken FJ and Batra B (1992) The role of managerial learning and interpretation in strategic persistence and reorientation: An empirical exploration. Strategic Management Journal, John Wiley & Sons, Ltd., 13(8), 585–608. Latham SF and Braun M (2009) Managerial risk, innovation, and organizational decline. Journal of Management, 35(2), 258–281. Lavie D (2006) Capability reconfiguration: An analysis of incumbent responses to technological change. Academy of Management Review, 31(1), 153–174. Lim DSK, Celly N, Morse EA, et al. (2013) Rethinking the effectiveness of asset and cost retrenchment: The contingency effects of a firm’s rent creation mechanism. Strategic Management Journal, 34(1), 42–61. 99 Lohrke FT, Ahlstrom D and Bruton GD (2012) Extending Turnaround Process Research: Important Lessons From the U.S. Civil War. Journal of Management Inquiry, 21(2), 217–234. Love EG and Nohria N (2005) Reducing slack: The performance consequences of downsizing by large industrial firms, 1977-93. Strategic Management Journal, 26(12), 1087–1108. Lu JW and Beamish PW (2004) International diversification and firm performance: The S-curve hypothesis. Academy of Management Journal, 47(4), 598–609. Lumpkin GT and Dess GG (1996) Clarifying the entrepreneurial orientation construct and linking it to performance. Academy of Management Review, 21(1), 135–172. Madsen PM and Desai V (2010) Failing to learn? the effects of failure and success on organizational learning in the global orbital launch vehicle industry. Academy of Management Journal, 53(3), 451–476. Martin de Hollan PM and Phillips N (2004) Rememberance of things past? The dynamics of organizational forgetting. Management Science, 50(11), 1603– 1613. Mathews JA (2003) Strategizing by firms in the presence of markets for resources. Industrial and Corporate Change, 12(6), 1157–1193. McKinley W, Latham S and Braun M (2014) Organizational decline and innovation: Turnarounds and downward spirals. Academy of Management Review, 39(1), 88–110. Miller D, Friesen and P. (1982) Innovation in conservative and entrepreneurial firms: Two models of strategic momentum. Strategic Management Journal, 3(1), 1–25. Miner AS, Bassoff P and Moorman C (2001) Organizational improvisation and learning: A field study. Administrative Science Quarterly, 46(2), 304–337. Moliterno TP and Wiersema MF (2007) Firm performance, rent appropriation, and the strategic resource divestment capability. Strategic Management Journal, 28(11), 1065–1087. Montgomery C a, Wernerfelt B and Balakrishnan S (1989) Strategy Content and the Research Process: A Critique and Commentary. Strategic Management Journal, 10(2), 189–197. Morrow JL, Johnson RA and Busenitz LW (2004) The Effects of Cost and Asset 100 Retrenchment on Firm Performance: The Overlooked Role of a Firms Competitive Environment. Journal of Management, 30(2), 189–208. Morrow Jr. JL, Johnson RA and Busenitz LW (2004) The effects of cost and asset retrenchment on firm performance: The overlooked role of a firm’s competitive environment. Journal of Management, 30(2), 189–208. Morrow Jr. JL, Sirmon DG, Hitt MA, et al. (2007) Creating value in the face of declining performance: Firm strategies and organizational recovery. Strategic Management Journal, 28(3), 271–283. Nadkarni S and Barr PS (2008) Environmental context, managerial cognition, and strategic action: An integrated view. Strategic Management Journal, 29(13), 1395–1427. Ndofor HA, Vanevenhoven J and Barker VL (2013) Software firm turnarounds in the 1990s: An analysis of reversing decline in a growing, dynamic industry. Strategic Management Journal, 34(9), 1123–1133. Nielsen BB and Nielsen S (2013) Top management team nationality diversity and firm performance: A multilevel study. Strategic Management Journal, 34(3), 373–382. Nystrom PC and Starbuck W (1984) To avoid organizational crisis: unlearn. Organizational dynamics, Spring, 53–65. Paint LW (1991) An investigation of industry and firm structural characteristics in corporate turnarounds. Journal of Management Studies, Blackwell Publishing Ltd, 28(6), 623–643. Pajunen K (2006) Stakeholder influences in organizational survival. Journal of Management Studies, 43(6), 1261–1288. Patel PC and Cooper D (2014) Structural Power Equality between Family and Non-Family TMT Members and the Performance of Family Firms. Academy of Management Journal, Academy of Management, 57(6), 1624–1649. Pathak S, Hoskisson RE and Johnson RA (2013) Settling up in CEO compensation: The impact of divestiture intensity and contextual factors in refocusing firms. Strategic Management Journal. Pearce II JA and Robbins DK (2008) Strategic transformation as the essential last step in the process of business turnaround. Business horizons, 51(2), 121–130. Pearce II JA and Robbins K (1993) Toward improve theory and research on business turnaround. Journal of Management, 19(3), 613–636. 101 Pearce JA and Robbins DK (1994) Retrenchment remains the foundation of business turnaround. Strategic Management Journal, John Wiley & Sons, Ltd., 15(5), 407–417. Pennings JM, Lee K and Van Witteloostuijn A (1998) Human capital, social capital, and firm dissolution. Academy of Management Journal, 41(4), 425– 440. Peteraf M, Di Stefano G and Verona G (2013) The elephant in the room of dynamic capabilities: Bringing two diverging conversations together. Strategic Management Journal, 34(12), 1389–1410. Platt H (2004) Principles of Corporate Turnaround. University of Michigan Press. Porter ME (1996) What is Strategy? Harvard Business Review, Nov/Dec, 61–78. Powell TC (2001) Competitive advantage: Logical and philosophical considerations. Strategic Management Journal, 22(9), 875–888. Priem RL, Rasheed AMA and Kotulic AG (1995) Rationality in strategic decision processes, environmental dynamism and firm performance. Journal of Management, 21(5), 913–929. Robbins DK and Pearce JA (1992) Turnaround: Retrenchment and recovery. Strategic Management Journal, John Wiley & Sons, Ltd., 13(4), 287–309. Roodman D (2009) How to do xtabond2: an introduction to difference and system GMM in Stata. Stata journal, 9(1), 86–136. Ruiz-Navarro J (1998) Turnaround and renewal in a Spanish shipyard. Long Range Planning, 31(1), 51–59. Rumelt R (2011) Good strategy/Bad strategy: The difference and why it matters. New York: Crown Business. Sapienza HJ, Autio E, George G, et al. (2006) A capabilities perspective on the effects of early internationalization on firm survival and growth. Academy of Management Review, 31(4), 914–933. Schendel D, Patton GR and Riggs J (1976) Corporate Turnaround Strategies: A Study of Profit Decline and Recovery. Journal of General Management, 3, 3–11. Schilke O (2014) On the contingent value of dynamic capabilities for competitive advantage: The nonlinear moderating effect of environmental dynamism. Strategic Management Journal, 35(2), 179–203. Schmitt A and Raisch S (2013) Corporate Turnarounds: The Duality of 102 Retrenchment and Recovery. Journal of Management Studies, 50(7), 1216– 1244. Schoenberg R, Collier N and Bowman C (2013) Strategies for business turnaround and recovery: a review and synthesis. European Business Review, 25(3), 243–262. Semadeni M, Cannella Jr. AA, Fraser DR, et al. (2008) Fight or flight: Managing stigma in executive careers. Strategic Management Journal, 29(5), 557–567. Shamsie J, Martin X and Miller D (2009) Research notes and commentaries n with the old, in with the new: Capabilities, trategies, and performance among the ollywood studios. Strategic Management Journal, 30(13), 1440–1452. Shein J (2013) Reversing the slide: A strategic guide to turnarounds and corporate renewal. Jossey-Bass. Sirmon DG, Hitt MA and Ireland RD (2007) Managing firm resources in dynamic environments to create value: Looking inside the black box. Academy of Management Review, 32(1), 273–292. Sirmon DG, Hitt MA, Ireland RD, et al. (2011) Resource orchestration to create competitive advantage: Breadth, depth, and life cycle effects. Journal of Management, 37(5), 1390–1412. Slater SF, Olson EM and Hult GTM (2006) The moderating influence of strategic orientation on the strategy formation capability-performance relationship. Strategic Management Journal, 27(12), 1221–1231. Slatter S (1984) Corporate recovery. books P (ed.), Harmandsworth, Middlesex. Slatter S, Lovett D and Barlow L (2006) Leading corporate turnaround. West Sussex, England: John Wiley & Sons. Sudarsanam S and Lai J (2001) Corporate Financial Distress and Turnaround Strategies: An Empirical Analysis. British Journal of Management, 12(3), 183– 199. Sutton R and D´Aunno T (1989) Decreasing organizational size: untangling the effects of money and people. Academy of Management Review, 14, 194– 212. Tangpong C, Abebe M and Li Z (2015) A Temporal Approach to Retrenchment and Successful Turnaround in Declining Firms. Journal of Management Studies, Blackwell Publishing Ltd, 52(5), 647–677. Teece DJ (2007) Explicating dynamic capabilities: The nature and 103 microfoundations of (sustainable) enterprise performance. Strategic Management Journal, 28(13), 1319–1350. Teece DJ (2012) Dynamic Capabilities: Routines versus Entrepreneurial Action. Journal of Management Studies, 49(8), 1395–1401. Teece DJ (2014a) A dynamic capabilities-based entrepreneurial theory of the multinational enterprise. Journal of International Business Studies, 45(1), 8–37. Teece DJ (2014b) The Foundations of Enterprise Performance: Dynamic and Ordinary Capabilities in an (Economic) Theory of Firms. Academy of Management Perspectives, Academy of Management, 28(4), 328–352. Teece DJ, Pisano G and Shuen A (1997) Dynamic capabilities and strategic management. Strategic Management Journal, 18(7), 509–533. Trahms CA, Ndofor HA and Sirmon DG (2013) Organizational Decline and Turnaround: A Review and Agenda for Future Research. Journal of Management, 39(5), 1277–1307. Walsh JP and Rivera G (1991) Organizational memory. Academy of Management Review, 16(1), 421–458. Wang H and Li J (2008) Untangling the effects of overexploration and overexploitation on organizational performance: The moderating role of environmental dynamism. Journal of Management, 34(5), 925–951. Whetten D (1987) Organizational growth and decline processes. Annual review of sociology, 13, 335–358. Wiklund J and Shepherd D (2003) Aspiring for, and Achieving Growth: The Moderating Role of Resources and Opportunities. Journal of Management Studies, 40(8), 1919–1941. Winter SG (2003) Mistaken perceptions: Cases and consequences. British Journal of Management, 14(1), 39–44. Zahra SA, Sapienza HJ and Davidsson P (2006) Entrepreneurship and dynamic capabilities: A review, model and research agenda. Journal of Management Studies, 43(4), 917–955. Zhang Y (2008) Information asymmetry and the dismissal of newly appointed CEOs: An empirical investigation. Strategic Management Journal, 29(8), 859–872. Zimmerman FM (1986) Turnaround-a painful learning process. Long range planning, 19(4), 104–114. 104 Zimmerman FM (1991) The turnaround experience. McGraw-Hill (ed.), New York. Zollo M and Winter SG (2002) Deliberate learning and the evolution of dynamic capabilities. Organization Science, 13(3), 339–351. Zúñiga-Vicente JÁ and Vicente-Lorente JD (2006) Strategic moves and organizational survival in turbulent environments: The case of Spanish banks (1983-97). Journal of Management Studies, 43(3), 485–519. 105 Table 1: Descriptive statistics and correlations Variable M SD 1 2 3 4 5 6 7 8 9 10 11 1. Return on sales (adjusted) -0.04 0.40 1 2. Severity -0.07 0.24 0.18 1 3. Asset retrenchm ent -0.02 0.44 0.01 -0.01 1 4. Cost retrenchment -0.07 0.88 0.03 -0.01 0.12 1 5. Dynamism 5.57 6.26 -0.02 -0.02 0.03 0.04 1 6. Munificence 0.09 0.05 0.02 0.02 -0.02 -0.02 0.15 1 7. Firm age 36.58 19.75 0.12 0.04 0.02 0.05 -0.09 -0.02 1 8. Firm size 716.25 1398.6 9 0.06 0.06 -0.00 0.02 -0.04 -0.08 0.15 1 9. Gearing 0.95 2.33 -0.04 0.03 -0.02 -0.03 0.01 0.05 0.02 -0.02 1 10. Liquidity 0.94 0.99 0.03 0.02 -0.01 -0.03 -0.04 -0.03 -0.00 -0.01 0.05 1 11. Slack 0.09 0.24 0.06 0.00 0.00 0.03 -0.06 0.07 0.15 0.09 -0.02 0.03 1 In thousands of euros Table 2: System GMM: Effects of asset and cost retrenchment on return on sales (t + 2) 𝑎 112 if the firm slides to this stage there is the risk of organizational liquidation. In our study, we delve into the question of under what circumstances does an early CEO replacement decision helps or hinders organizational survival. In other words, are there circumstances under which we can safely predict turnaround success in a forced CEO succession? Current studies on time and turnarounds have offered evidence that early and fast action positively affect turnaround success (Barbero, J., Di Pietro, F., Chiang, 2017; Tangpong et al., 2015). As argued by the downward spiral perspective, decline is a protracted process that erodes firm resources. Thus, time (early/fast action) is critical to turnaround success. We apply this time approach to CEO replacement. We hypothesize and test whether the causes of decline (internal vs external), CEO tenure, and duality (CEO power) interact with the positive effect of a fast CEO replacement on turnaround success. Despite our build on the downward spiral perspective, consistent with prior CEO succession studies (Berns & Klarner, 2017; Finkelstein & D´Aveni, 1994) and decline studies (Barbero, J., Di Pietro, F., Chiang, 2017; Schmitt, Barker, Raisch, & Whetten, 2016), we use a pluralistic theory approach to argue our hypothesis. Our conceptual model is depicted in figure 1. 113 ------------------------------------------------------------ Insert Figure 1 about here ------------------------------------------------------------- Using data from Compustat and Execucomp, we select a matched pair sample of 80 firms in decline between 1996 and 2007. This methodology is consistent with leading latest turnaround research (Chen, 2014; Ndofor, Vanevenhoven, & Barker, 2013; Tangpong et al., 2015). We find that the positive effect between an early timing of CEO replacement and turnaround success is positively moderated by the internal causes of decline and CEO long tenure. Such relationship is negatively moderated by CEO duality. In the next section we develop the concept of CEO replacement and clarify its relationship with the downward spiral perspective. Next, we argue our hypotheses. We explain our methodology and the results in the two following sections. Finally, we discuss our results, describe the limitations of our study and propose future lines of research. 114 2. Theoretical background Timing of action and turnarounds One of the prevalent ideas of the turnaround literature is that time is critical to turnaround success, specially during the early stages of decline (Bibeault, 1982; Robbins & Pearce, 1992; Zimmerman, 1991). Only recently have scholars empirically addressed this issue (Barbero, J., Di Pietro, F., Chiang, 2017; Tangpong et al., 2015). Decline is a protracted process that erodes firm resources (Hambrick & D´Aveni, 1988, 1992). In a self-reinforcing process, poor performance depletes slack resources which further deteriorates firm performance (Hambrick & D´Aveni, 1988; Tangpong et al., 2015). On these basis and only recently, researchers have showed evidence that an early (timing) and fast (speed) of action will avoid organizational collapse (Barbero, J., Di Pietro, F., Chiang, 2017; Tangpong et al., 2015). Other CEO replacement studies have briefly referred to the timing of CEO replacement (Berns & Klarner, 2017; Chen & Hambrick, 2012). Timing is defined as “the moment an event happens or is planned to happen” (Huy, 2001: p. 613). In our research, the focal event of study is CEO replacement. 115 The timing of CEO replacement is marked by the lapse of time between the onset of the decline period and the CEO replacement. An early timing of CEO replacement occurs when a CEO is replaced early with respect to the focal moment the company enters into decline. A late timing of CEO replacement occurs when a CEO is replaced late with respect to the focal moment 5 . The study of the time dimension in turnarounds represents a substantial change to prior literature. Prior turnaround research has analyzed the effect of certain measures on turnaround success. The evidence found on the effect of some of these actions has very often been mixed (i.e.: the value of retrenchment actions (Barker & Mone, 1994; Pearce & Robbins, 1994; Pearce II & Robbins, 1993a)) and it has frustrated scholars. The change in the study focus to a temporal approach represents a substantial change in the direction of turnaround research. Under this approach, the question is not whether a specific action has value, but when does the execution of a specific action has value. In other words, the value of actions are intrinsically tied to the moment they are executed. In sum, CEO replacement is one of the actions with a greater degree of lever to turnaround performance (Berns & Klarner, 2017; Bibeault, 1982; Trahms et al., 2013). In a decline setting, characterized by resource erosion, 5 We study early/late CEO replacement rather than fast/slow replacement. The difference is conceptual but important. The former represents the study of timing which has an event nature. The latter represents the study of speed which has a procesual nature. We study the event of CEO replacement rather than the decision process of CEO replacement. 116 the timing of CEO replacement will influence the outcome of decline. Next we hypothesize three factors that are likely to interact with the positive effect of CEO replacement on turnaround success. 3. Hypotheses Timing of CEO replacement and the causes of decline Turnaround success has been heavily linked to the causes of decline (Arogyaswamy et al., 1995; Cameron, Sutton, & Whetten, 1988; Ndofor et al., 2013; Santana, M., Valle, R., Galán, 2017). Decline can be originated by internal and by external causes. The external causes of decline are environmentally related (i.e.: business cycle, availability of credit, government regulations, etc) and as a such a firm does not have direct control on them (Arogyaswamy et al., 1995; Balcaen & Ooghe, 2006; Shein, 2013). Under internal causes, decline is inflicted by the firm on itself. These causes are intrinsically managerial (Argenti, 1976; Bruno & Leidecker, 1988; Heracleous & Werres, 2016; Slatter et al., 2006; Trahms et al., 2013) given management can decide to act and remove the factor causing decline. For example, (Bibeault, 1982) found 85% of the internal causes of decline 117 are managerial in nature. Slatter (1984) traced all the causes of decline ultimately to bad management. Scholars have agreed there are several proxies representing the internal causes of decline. Munificence represents a proxy for internal causes of decline. A munificent environment is one in which there is an abundance of resources and opportunities (Boyd, 1990; Dess & Beard, 1984), and a low munificent an environment of scarcity (Wiersema, Bantel, & K:, 1993). Literature has suggested that firms in decline in a munificent environment are affected by internal causes of decline, given the environment exerts a positive effect on the firm (Barbero, J., Di Pietro, F., Chiang, 2017; Ndofor et al., 2013). Literature has also pointed out that the internal causes of decline can be best assessed by a firm´s performance compared to the industry performance: those firms whose performance is below that of the industry are affected to internal causes of decline. However, those firms still in decline, which are overperforming compared to the industry are clearly affected by external causes (Barbero, J., Di Pietro, F., Chiang, 2017; Ndofor et al., 2013). The downward spiral perspective argues that decline is a protracted process that erodes firm resources (Hambrick & D´Aveni, 1988, 1992). The process of resource erosion will continue until action is taken to remove the circumstances causing decline. The longer the period of decline is 118 extended, the more deteriorated the organization will become. Under internal causes of declne, early action will avoid excessive firm deterioration by removing the main source of underperformance (Barbero, J., Di Pietro, F., Chiang, 2017; Tangpong et al., 2015). Hence, the earlier the action on the main cause of decline (management), this is the earlier the CEO replacement the greater the probability of turnaround. Hence, we can argue: Hypothesis 1: In the presence of internal causes of decline, an early timing of CEO replacement will lead to turnaround success. Timing of CEO replacement and tenure One of the critical issues to turn a firm around is to detect the main causes of decline and to deal with them (Arogyaswamy et al., 1995; Bibeault, 1982). A mistaken diagnosis leads to further firm deterioration and finally to firm demise (Shein, 2013). Decline results from a mix of external and internal causes (Arogyaswamy et al., 1995; Schendel, Patton, & Riggs, 1976). Nevertheless, most scholars in the turnaround literature ultimately attributes failure to internal causes, being the CEO the main cause (Bibeault, 1982; Shein, 2013; Slatter, 1984). 119 It is well established in the literature that CEO tenure leads to underperformance. Long tenured CEOs are cognitively committed to the status quo (Castrogiovanni et al., 1992; Chen & Hambrick, 2012). First, they are more likely to fall prey to self serving bias, the unconscious motivation of top managers to conserve the status quo and unsolved themselves from responsibility (Barker & Patterson, 1996). Thus, they tend to attribute failure to external factors because they are less likely to attribute decline to past actions they were involved with (Barker & Patterson, 1996; Hedberg, Nystrom, & Starbuck, 1976). Short tenured management attribute failure to internal factors (Barker & Patterson, 1996). Hence, shorter tenured CEOs are more realistic with the causes of decline and more beneficial to organizational decline. Second, one of the key tools turnaround managers have to turn firms around is strategic change (Barker III & Duhaime, 1997; Boyne & Meier, 2009; Schmitt & Raisch, 2013). Evidence shows that because long tenured managers have been part of the firm strategy before decline, they feel less likely to lead the firm towards strategic change and tend to preserve the status quo (Barker & Barr, 2002). They make a lower number of changes as their tenure extends (Henderson, Miller, & Hambrick, 2006) and they retrench less when compared with substitute CEOs (Nag, Hambrick, & Chen, 2007). In contrast, short term tenured managers feel less committed to the current strategy and are more proactive in introducing strategic changes. Third, beyond cognitive commitment, CEOs with long tenures are 120 socially committed to their firms (Castrogiovanni et al., 1992; Chen & Hambrick, 2012; Simsek, Veiga, Lubatkin, & Dino, 2005). They are socially entrenched to the organization derived from long years of personal relationships with their teams (Hambrick, Geletkanycz, & Fredrickson, 1993). CEOs were an important part in the selection and promotion of those upper echelon employees, hence they are socially committed. Such social commitment causes CEO cognitive commitment, avoiding changes in their TMTs. Both CEO cognitive and social commitment drive the organization to inaction or to slow/weak action. For declining firms, based on the downward spiral perspective, during this period of lack of appropriate action, resources deteriorate and the position of the organization becomes more fragile (Barbero, J., Di Pietro, F., Chiang, 2017; Hambrick & D´Aveni, 1988, 1992). In the case of declining firms with long tenured CEO, the earlier they replace a long tenured cognitively and socially committed CEO mismanaging the firm, the lower the resource deterioration and the greater the likelihood of turnaround success. Second, a delay in CEO replacement implies an extension in the CEO tenure. A CEO tenure increase, will increase the cognitive and social 121 commitment and the organizational paralysis (Hambrick & Fukutomi, 1991). Thus, the faster the long tenured CEO of a declining firm is replaced, the lower the commitment and the earlier change is kickstarted. Accordingly, we can argue: Hypothesis 2: In the presence of a long-tenured CEO, an early timing of CEO replacement will lead to turnaround success. Timing of CEO replacement and CEO duality Duality is the practice of a single individual serving as both CEO and board chair (Krause, Semadeni, & Cannella, 2014). CEO duality has been characterized as a double edged sword (Hambrick & Fukutomi, 1991). On one hand, the agency theory argues duality promotes CEO entrenchment because CEO supervision misses independent oversight. On the other, organization theory perspectives argue duality promotes unity of command (Boyd, 1995; Donaldson & Davis, 1991) which is one of the principles in turnaround management (Bibeault, 1982; Mueller & Barker, 1997; Slatter et al., 2006). Duality has been acknowledged, as one of the most contentious issues in the governance literature (Dalton & Dalton, 2011) 128 procedure it is required to identify at least one independent variable that is associated with the dependent variable in the first-stage model, but is not related to the dependent variable in the second-stage model. This variable was size measured by the firm assets. Previous studies have documented that this variable is positively related to CEO turnover (Chen, 2014), but is not significantly associated with successful turnaround (Ndofor et al., 2013; Tangpong et al., 2015). An inversed Mill’s ratio was generated in the firststage model, and then included in the second-stage analysis as an instrumental variable (lambda) to correct for any selection bias (Heckman, 1979). We included in the second stage the results of the Probit analysis as the inverse Mills ratio variable in the main regression analyses. As shown in Table 2, the overall results are not significantly altered, suggesting that endogeneity is not a major concern. Survival bias Once selection bias was dealt with, we were left with a sample of 105 declining firms involved in CEO replacement during the three year period following the onset of decline. A total of 60 of these 105 companies achieved turnaround success, while 45 were unsuccessful. In order to be considered successful, firms should meet the following requirements: a) have at least two years of increasing ROA after the two years of declining ROA, and b) have achieved and 129 maintained positive ROA by at least the sixth year after the base year before decline started (Ndofor et al., 2013). In the group of unsuccessful firms, some of them stopped reporting results in Compustat, thus they were classified as unsuccessful (Ndofor et al., 2013). However, all these firms reported data during our six-year period of study. To test our hypotheses and avoid survival bias, we used a matched-pair sampling technique common in turnaround studies (Barker III & Duhaime, 1997; Chen & Hambrick, 2012; Tangpong et al., 2015). We used a propensity score-matching to identify two subsets, successful and unsuccessful firms. We employed Stata code psmatch2 for our analyses. The variables used to calculate the propensity score match were firm performance (measured by the ROA on the base year), industry growth (measured by the industry growth between the base year and the first year of decline), and capital expenditure, current ratio, Z-score measured at the base year. We also included industry and year dummies. The final sample consisted of 40 matched pairs of successful turnaround firms and 40 unsuccessful turnaround firms. The model had a log-likelihood of -60.61 (p<0.1) and a pseudo R-squared of 0.14. These statistics indicate the appropriateness of the choice of independent variables, as well as the overall fit of our model. Based on the propensity score, we identified those 130 firms with the closest one to one match. Our final sample size was 80 firms, given we matched 40 unsuccessful firms with 40 successful firms. This sample size is reasonable compared to prior turnaround studies using matched pair sampling (Ndofor et al., 2013; Tangpong et al., 2015). Finally as robustness and contrast test sample a t-test indicates no significant difference between the successful and unsuccessful subsamples on employees, total assets, ROA and total sales and leverage (total debt/total assets) in the base year. Measures Dependent variable The dependent variable is the likelihood of turnaround success, binary coded ‘1’ for successful turnaround and ‘0’ otherwise as in previous turnaround research (Hambrick and D’Aveni, 1992; Mueller and Barker, 1997; Ndofor, Vanevenhoven and Barker, 2013). The operationalization of this variables has been described in the “survival bias” section. Independent variables 131 The independent variable in this study is the timing of CEO replacement. We measured this variable as the time lapse (number of days) between the onset of the decline period and the date the CEO was replaced. This variable was log formed and standardized. Moderating variables Our hypotheses included three moderating effects: causes of decline, tenure and CEO duality. We used two measures to operationalize (internal/external) causes of decline. First, as argued above and consistent with the turnaround literature (Arogyaswamy et al., 1995; Barbero, J., Di Pietro, F., Chiang, 2017; Ndofor et al., 2013), we measured internal causes of decline with the variable munificence. A munificent environment indicates the industry enjoys sufficient resources and is profitable. Literature has pointed that firms in decline under a munificent environment are plagued with internal problems. We measured munificence with the coefficient of the regression slope of the set of sector sales divided by average value. The coefficients are based on a regression of time estimate for a given year is based on the previous five years, including the year of calculation analysis (Boyd, 1990; Dess & Beard, 1984). Second, we measured internal/external causes of decline based on whether the firm ROA is greater than the industry average ROA. The variable was coded as 1 for external causes 132 (Firm ROA the first year of decline>Industry ROA the first year of decline) and as 0 for internal causes (Otherwise). Tenure was calculated as the amount of time (number of days) that a CEO holds the title in the firm. This variable was log formed. CEO duality was operationalized as a dummy equaled 1 if the newly appointed CEO also held the position of chair of the board in the base year. Both the independent variable as the moderators were standardized in the calculations for analyzing the effects of joint interactions do not influence the results the different levels of each of them. Control variables We controlled for several industry, firm and agency factors that have been suggested by previous research to potentially affect the likelihood of successful turnaround and a CEO replacement. Retrenchment. Retrenchment is defined as a set of organizational activities to undertaken to achieve cost and asset reductions (Michael & Robbins, 1998; Morrow et al., 2004; Trahms et al., 2013). We measured retrenchment as the reduction in total assets. We calculated retrenchment the period before the CEO was replaced ((total assets quarter before the change of 133 management – total asset last quarter base year) / (total asset last quarter base year)), as our intention was to measure any potential effect just before the management change. We used total assets because the retrenchment strategy has been recognized as the one having a greater degree of lever on company results (Pearce II & Robbins, 1993b). Altman Z-score: The Z-score is a measure that predicts a company´s financial distress and the probability of default (Altman, 1968). Quick ratio: The quick ratio is calculated as the current assets minus inventories divided by the current liabilities. Capital intensity: We computed this variable as the ratio of the total value of assets to sales revenue generated over a given period. Board size: Board size is calculated as the number of members of the board of directors of a company in the base year. Dynamism. Industry dynamism is a variable that measures the instability and volatility in an industry (Aldrich, 1979; Boyd, 1990; Dess & Beard, 1984). We measured this variable as the standard error of the slope of the regression divided by the average value (Boyd, 1990; Dess & Beard, 1984). We measured this variable the year the CEO was replaced. Industry performance. The variable was calculated as the difference between the industry ROA and the Firm ROA the year of decline. 134 Lambda. This variable is formed with the residuals of the Heckman procedure explained in the section “Sample selection bias”. Time dummies: Finally, we included time dummies for the year decline started. 5. Results Table 1 presents descriptive statistics and correlations for the full dataset we use in our analysis. Table 2 presents six models based on the results of our regression analysis. Models 1 and 2 are base models. Models 3 to 6 test our hypotheses. Note we have four models to test three hypotheses because hypotheses 1, the one predicting a positive interactive effect of the causes of decline on the effect of timing of replacement on turnaround success is tested with two variables (munificence and causes (dummy)); thus we have two models (Models 3 and 4) for this hypothesis 1. The first model includes all the control variables except for timing of CEO replacement. Model 2 includes all the control variables and timing. Models 3 to 6 test our three hypotheses. Our hypothesis 1 posited that the internal causes of decline strengthens the effects of an early timing of CEO replacement on firm results. Our results show in Table 2 Model 3 for the 135 variable munificence is not significant, but Model 4 for the variable causes(dummy) ( = 4.40, p < 0.01) is significant and in line with our prediction. Figure 2 depicts in line with our prediction, when the causes of decline are internal, an early timing of CEO replacement leads to turnaround success. Also as predicted by our hypothesis 2 and as shown in Model 5, an early timing of CEO replacement ( = -2.31, p < 0.01) produces higher firm performance when CEO tenure is long (Figure 3). Finally, we predicted that under conditions of duality, a late timing of CEO replacement brings about a higher performance. Model 6 shows that a later timing of CEO replacement generates a higher performance under conditions of duality ( = 2.70, p < 0.05), confirming hypothesis 3. Figure 4 portrays this effect. ------------------------------------------------------ INSERT TABLES 1 AND 2, AND FIGURES 2, 3 AND 4 ABOUT HERE ------------------------------------------------------ As a robustness test we applied quadratic model for the timing of CEO replacement and had the variable interact with causes of decline, tenure and duality. These relationships were not significant. 136 6. Discussion Research has offered considerable attention to CEO replacement with mixed evidence (Chen & Hambrick, 2012). Organizational decline is a very appropriate setting for this topic because CEO dismissal is quite more common (Mueller & Barker, 1997). Moreover, empirical studies have not paid sufficient attention to CEO replacement in turnarounds despite the importance classic studies grant the topic (Bibeault, 1982; Slatter, 1984; Whitney, 1987), and the lever CEO replacement exercises on firm results (Berns & Klarner, 2017; Chen & Hambrick, 2012). In our study, we shift the question of whether a CEO should be replaced to, when should the CEO be replaced by using a temporal approach to turnaround study (Barbero, J., Di Pietro, F., Chiang, 2017; Tangpong et al., 2015). Specifically, we investigate the boundary conditions that influence the effect of the timing of CEO replacement on turnaround success. We build on the downward spiral perspective, and the agency and organizational theories. Next we discuss the implications for research and for practice of our study. 137 Implications for research In our research we study the boundary conditions of the timing of CEO replacement in declining organizations. Chen & Hambrick (2012) did not find evidence of whether CEO replacement improves (direct relationship) performance in declining organizations. Because time is critical to turnarounds (Barbero, J., Di Pietro, F., Chiang, 2017; Bibeault, 1982; Robbins & Pearce, 1992; Tangpong et al., 2015), we subtlely modify the research question and study when should CEOs be replaced. Despite we have not hypothesized a positive effect of a fast CEO removal on turnaround success, our results indicate this relationship might be significant. This incidental but obvious finding points to those classical turnaround studies attributing the ultimate cause of decline to bad management (Bibeault, 1982; Slatter et al., 2006). If management is the ultimate cause of decline, all things being equal, an early removal will lead to superior performance when decline is eroding firm resources over time. Regardless of this finding, the core of our contribution focuses on the specification of the boundary conditions on which this relationship becomes more effective. First, our research explores whether CEO replacement should be expedite when decline is caused by internal factors. Consistent with other studies, a firm underperforming when compared to the average industry firm indicates decline is firm-based (Ndofor et al., 2013). We find that for underperforming firms CEO replacement should be executed early. An 144 Boyd, B. 1990. Corporate linkages and organizational environment: A test of the resource dependence model. Strategic Management Journal, 11(6): 419–430. Boyd, B. K. 1995. CEO duality and firm performance: A contingency model. Strategic Management Journal, 16(4): 301–312. Boyne, G. A., & Meier, K. J. 2009. Environmental change, human resources and organizational turnaround. Journal of Management Studies, 46(5): 835–863. Brauer, M., & Laamanen, T. 2014. Workforce Downsizing and Firm Performance: An Organizational Routine Perspective. Journal of Management Studies, 51(8): 1311–1333. Bruno, A. V., & Leidecker, J. K. 1988. Causes of new venture failure: 1960s vs. 1980s. Business Horizons, 31(6): 51–56. Cameron, K., Kim, M., & Whetten, D. 1987. Organizational effects of decline and turbulence. Administrative Science Quarterly, 32(2): 222–240. Cameron, K., Sutton, R., & Whetten, D. 1988. Readings in organizational decline: Frameworks, research and prescriptions. Cambridge, MA: Ballinger. Castrogiovanni, G. J., Baliga, B. R., & Kidwell, R. E. 1992. Curing sick businesses: changing CEOS in turnaround efforts. Academy of 145 Management Executive, 6(3): 26–41. Chen, G. 2014. Initial compensation of new CEOs hired in turnaround situations. Strategic Management Journal, 36(12): 1895–1917. Chen, G., & Hambrick, D. C. 2012. CEO Replacement in Turnaround Situations: Executive (Mis)Fit and Its Performance Implications. Organization Science, 23(1): 225–243. Daily, C. M., & Dalton, D. R. 1995. CEO and director turnover in failing firms: An illusion of change? Strategic Management Journal, 16(5): 393–400. Dalton, D. R., & Dalton, C. M. 2011. Integration of Micro and Macro Studies in Governance Research: CEO Duality, Board Composition, and Financial Performance. Journal of Management, 37(2): 404–411. Davidson, W; Worrell, D; Dutia, D. 1993. The stock market effects of CEO succession in bankrupt firms. Journal of Management, 19(3): 517–533. Denis, D. J., & Kruse, T. A. 2000. Managerial discipline and corporate restructuring following performance declines. Journal of Financial Economics, 55(3): 391–424. Dess, G. G., & Beard, D. W. 1984. Dimensions of organizational task environments. Administrative Science Quarterly, 29(1): 52–73. Donaldson, L., & Davis, J. H. 1991. Stewardship Theory or Agency Theory: 146 Australian journal of Management, 16(June 1991): 49–66. Dowell, G. W. S., Shackell, M. B., & Stuart, N. V. 2011. Boards, CEOs, and surviving a financial crisis: Evidence from the internet shakeout. Strategic Management Journal, 32(10): 1025–1045. Finkelstein, S., & D’Aveni, R. a. 1994. CEO duality as a double-edged sword: How boards of directors balance entrenchment avoidance and unity of command. Academy of Management Journal, 37(5): 1079–1108. Finkelstein, S., & D´Aveni, R. A. 1994. CEO duality as a double-edged sword: How boards of Directors balance entrenchment avoidance and unity of command. Academy of Management Journal, 37(5): 1079–1108. Friedman, S. D., & Singh, H. 1989. CEO Succession and stockholder reaction: the influence of organizational context and event content. Academy of Management Journal, 32(4): 718–744. Guthrie, J. P., & Datta, D. K. 2008. Dumb and dumber: The impact of downsizing on firm performance as moderated by industry conditions. Organization Science, 19(1): 108–123. Hambrick, D. C., & Fukutomi, G. D. S. 1991. The seasons of a CEO ’ s tenure. The Academy of Management, 16(4): 719–742. Hambrick, D. C., Geletkanycz, M. A., & Fredrickson, J. W. 1993. Top executive commitment to the status quo: Some tests of its determinants. 147 Strategic Management Journal, 14(6): 401–418. Hambrick, D., & D´Aveni, R. 1988. Large corporate failures as downward spirals. Administrative Science Quarterly, 33: 1–23. Hambrick, D., & D´Aveni, R. 1992. Top team deterioration as part of the downward spiral of large corporate bankruptcies. Management Science, 38(10): 1445–1466. Hamilton, B. H., & Nickerson, J. A. 2003. Correcting for Endogeneity in Strategic Management Research. Strategic Organization, 1(1): 51–78. Heckman, J. J. 1979. Sample Selection Bias as a Specification Error. Econometrica, 47(1): 153–161. Hedberg, B., Nystrom, P. C., & Starbuck, W. 1976. Camping on seesaws: Prescriptions for a self-designing organization. Administrative Science Quarterly, 21(1): 41–65. Henderson, A. D., Miller, D., & Hambrick, D. C. 2006. How quickly do CEOs become obsolete? Industry dynamism, CEO tenure, and company performance. Strategic Management Journal, 27(5): 447–460. Heracleous, L., & Werres, K. 2016. On the Road to Disaster: Strategic Misalignments and Corporate Failure. Long Range Planning, 49(4): 491– 506. 148 Huy, Q. N. 2001. Time, temporal capability, and planned change. Academy of Management Review, 26(4): 601–623. Karaevli, A. 2007. Performance consequences of new CEO “outsiderness”: Moderating effects of pre- and post-succession contexts. Strategic Management Journal, 28(7): 681–706. Krause, R., Semadeni, M., & Cannella, A. A. 2014. CEO Duality: A Review and Research Agenda. Journal of Management, 40(1): 256–286. Lim, D. S. K., Celly, N., Morse, E. A., & Rowe, W. G. 2013. Rethinking the effectiveness of asset and cost retrenchment: The contingency effects of a firm’s rent creation mechanism. Strategic Management Journal, 34(1): 42–61. Michael, S. C., & Robbins, K. 1998. Retrenchment among Small Manufacturing Firms During Recession. Journal of Small Business Managment, 36(3): 35–46. Morrow, J. L., Johnson, R. A., & Busenitz, L. W. 2004. The Effects of Cost and Asset Retrenchment on Firm Performance: The Overlooked Role of a Firms Competitive Environment. Journal of Management, 30(2): 189– 208. Mueller, G. C., & Barker, V. L. 1997. Upper Echelons and Board Characteristics of Turnaround and Nonturnaround Declining Firms. 149 Journal of Business Research, 39(2): 119–134. Nag, R., Hambrick, D. C., & Chen, M. J. 2007. What is strategic management, really? Inductive derivation of a consensus definition of the field. Strategic Management Journal, 28(9): 935–955. Ndofor, H. A., Vanevenhoven, J., & Barker, V. L. 2013. Software firm turnarounds in the 1990s: An analysis of reversing decline in a growing, dynamic industry. Strategic Management Journal, 34(9): 1123–1133. Ocasio, W. 1994. Political Dynamics and the Circulation of Power: CEO Succession in U.S. Industrial Corporations, 1960-1990. Administrative Science Quarterly, 39(2): 285–312. Pearce, J. A., & Robbins, D. K. 1994. Retrenchment remains the foundation of business turnaround. Strategic Management Journal, 15(5): 407–417. Pearce II, J. A., & Robbins, K. 1993a. Toward improved theory and research on business turnaround. Journal of Management, 19(3): 613–636. Pearce II, J. A., & Robbins, K. 1993b. Toward improved theory and research on business turnaround. Journal of Management, 19(3): 613–636. Robbins, D. K., & Pearce, J. A. 1992. Turnaround: Retrenchment and recovery. Strategic Management Journal, 13(4): 287–309. Santana, M., Valle, R., Galán, J. L. 2017. Turnaround strategies for 150 companies in crisis. Business Research Quarterly, 20(3): 206–2011. Schendel, D., Patton, G. R., & Riggs, J. 1976. Corporate Turnaround Strategies: A Study of Profit Decline and Recovery. Journal of General Management, 3: 3–11. Schmitt, A., Barker, V. L., Raisch, S., & Whetten, D. 2016. Strategic Renewal in Times of Environmental Scarcity. Long Range Planning, 49(3): 361–376. Schmitt, A., & Raisch, S. 2013. Corporate Turnarounds: The Duality of Retrenchment and Recovery. Journal of Management Studies, 50(7): 1216–1244. Semadeni, M., Withers, M. C., & Trevis Certo, S. 2014. The perils of endogeneity and instrumental variables in strategy research: Understanding through simulations. Strategic Management Journal, 35(7): 1070–1079. Shein, J. 2013. Reversing the slide: A strategic guide to turnarounds and corporate renewal. San Francisco, CA: Jossey-Bass. Simsek, Z., Veiga, J. F., Lubatkin, M. H., & Dino, R. N. 2005. Modeling the multilevel determinants of top management team behavioral integration. Academy of Management Journal. Slatter, S. 1984. Corporate recovery. (P. Books, Ed.). Harmandsworth, Middlesex: Penguin Books. 151 Slatter, S., Lovett, D., & Barlow, L. 2006. Leading corporate turnaround. Chichester, West Sussex, England: John Wiley & Sons. Tang, J., & Crossan, M. 2016. Are Dominant CEOs the Saviors of Troubled Firms? Long Range Planning, In Press. Tangpong, C., Abebe, M., & Li, Z. 2015. A Temporal Approach to Retrenchment and Successful Turnaround in Declining Firms. Journal of Management Studies, 52(5): 647–677. Trahms, C. A., Ndofor, H. A., & Sirmon, D. G. 2013. Organizational Decline and Turnaround: A Review and Agenda for Future Research. Journal of Management, 39(5): 1277–1307. Vance, D. 2009. Corporate restructuring: from cause analysis to execution. London, UK: Springer-Verlag. Weitzel, W., & Jonsson, E. 1989. Decline in organizations: A literature integration and extension. Administrative Science Quarterly, 34(1): 91– 109. Whitney, J. O. 1987. Taking charge: Management guide to troubled companies and turnarounds. Homewood, IL: Dow-Jones Irwin. Wiersema, M., Bantel, & K: 1993. Top management team turnover as an adaptation mechanism: The role of the environment. Strategic management journal, 14(7): 485–504. 152 Winn, J. 1997. Asset productivity turnaround: The growth/efficiency challenge. Journal of Management Studies, 34(4): 585–600. Zhang, Y., & Rajagopalan, N. 2004. When the known devil is better than an unknown god: An empirical study of the antecedents and consequences of relay CEO successions. Academy of Management Journal, 47(4): 483–500. Zimmerman, F. M. 1991. The turnaround experience: real-world experiences in revitalizing corporations. (McGraw-Hill, Ed.). New York: McGraw-Hill. 153 Figure 1: Conceptual model TIMING OF CEO REPLACEMENT External/Internal causes Turnaround success Timing of CEO replacement (Early) Environmental variables H1 (+) Tenure Individual (CEO) variables H2 (+) Duality Board variables H3 (-)