Introduction
An increasing number of scholars contend that political behavior plays a critical role in organizational decision making and that organizations are hotbeds of political intrigue (Sharfman, Dean, Mitchell, & Bartkoski, Reference Sharfman, Dean, Mitchell and Bartkoski2009; Child, Elbanna, & Rodrigues, Reference Child, Elbanna and Rodrigues2010). They have shown that managers engage in political behavior to develop or maintain their power bases and amass support for a particular choice. Many research studies also find that successful decisions require an efficient decision-making process, meaning one in which the process unfolds smoothly and which has been selected in a timely way (Mintzberg, Raisinghani, & Theoret, Reference Mintzberg, Raisinghani and Theoret1976; Hickson, Butler, Cray, Mallory, & Wilson, Reference Hickson, Butler, Cray, Mallory and Wilson1986). Nevertheless, political behavior has been shown to reduce efficiency (Roberto, Reference Roberto2004) and hence negatively affects decision success (Dean & Sharfman, Reference Dean and Sharfman1996; Elbanna & Child, Reference Elbanna and Child2007b). We cannot, however, generalize these results without an empirical examination in non-Western settings. For example, unexpectedly, Dayan, Elbanna, and Benedetto (Reference Dayan, Elbanna and Benedetto2012) indicated that political behavior positively influences decision speed. The authors argue that a possible explanation of the conflict between their results and previous research is their new study setting, Turkey. Hence, it is important to be sensitive to the role of national contexts and examine whether the results of research conducted in the Western settings apply in other countries (such as Tunisia, which is studied in this research), as little is currently known about the influence of national culture on political behavior (Child, Elbanna, & Rodrigues, Reference Child, Elbanna and Rodrigues2010).
Prior research mainly focused on the main effects, while the interactions between decision processes and contingencies are rarely examined. For example, an important issue still lacking research attention is the role of environmental factors, and how they affect the impact of political behavior on successful decision making. Contingency theories suggest that the decision process–success relationship is shaped by the interrelationships between a range of internal organizational characteristics and external environmental features and therefore closer scrutiny of the environment reveals insights that can reconcile contradictory results of prior research (Shepherd & Rudd, Reference Shepherd and Rudd2014). Hence, we seek to extend research on the strategic decision-making processes by focusing on the possible moderating role of key environmental factors on the relationship between political behavior and decision success. The selected factors represent two levels of environment, namely, industry level (product and competition uncertainty) and society/nation level (global financial crisis) (Miller, Reference Miller1993; Hill & Jones, Reference Hill and Jones2012), which can help to provide a more complete analysis, and thus a better understanding the impact of environment on politics–success relationship. These two levels are highly related as the society/nation environment influences the strength of forces in industry level and ultimately, the attractiveness of the industry (Hill & Jones, Reference Hill and Jones2012). To broaden our understanding of the link we examine the relationship between political behavior and decision success across a wide variety of firms and industries.
Moreover, enhancing our understanding of the decision process–success relationship requires greater attention to the interactions between different modes or dimensions of decision process (Elbanna, Reference Elbanna2006), as it is doubtful if managers depend only on one mode to make important decisions. Rather, it is more likely that there is a continuum of process modes involving an intimate combination of these modes (Hart, Reference Hart1992). However, it is still not clear what really occurs in the strategic decision-making process (Maitlis & Lawrence, Reference Maitlis and Lawrence2003). For example, the mediating decision processes are generally still not well understood in the decision-making literature (Lawrence, Reference Lawrence1997).
A form of modeling that we can use it here to offer a way forward to understand political processes of strategic decisions is ‘process analysis’ (Mohr, Reference Mohr1982), which refers to the sequence of events or processes leading up to a decision (Child, Elbanna, & Rodrigues, Reference Child, Elbanna and Rodrigues2010). Executives, for example, can combine different processes or modes when making decisions (Miller & Friesen, Reference Miller and Friesen1984) either sequentially (Allison, Reference Allison1971) or simultaneously (Mintzberg, Reference Mintzberg1973). Using different modes of decision making will have implications for decision success, which varies according to which modes decision makers use. Hart (Reference Hart1992), for example, argues that configurations of specific modes of decision making may be associated with lower performance. Although theories of decision making propose the existence of interrelationships among decision-making modes, few studies have addressed this deceptively simple proposition in the decision-making literature and no study has yet done so for the interrelationships between politics and intuition. This is surprising in that political behavior is recognized as a critical determinant of decision success. Any modeling of the politics in strategic decision making will thus be suggestive rather than definitive, unless we examine these under-researched yet important relationships. In this study, we begin this process by exploring the mediating effects of intuition on the political behavior–decision success relationship.
In sum, the study sets out to examine the role of environment and intuition in the relationship between political behavior and decision success, as depicted in Figure 1, aiming to answer three pressing research questions. First, to what extent does political behavior influence the success of strategic decisions in our new setting? Second, does political behavior have an even bigger impact on decision success under conditions of industry and society/nation environment effects, namely, product and competition uncertainties, and global financial crisis? Third, might the link between political behavior and decision success be attributable to the intuitive process of decision making? The last two questions are of special importance in advancing the theories of strategic decision making as discussed below.
Political behavior and successful decisions: the role of environment and intuition
Theory and Hypotheses
Political behavior among the actors concerned is recognized as an important aspect of decision making and has received considerable research attention. Recent work in this research stream has shown that managers use political and intuitive processes as well as rational procedures in decision making (Dean & Sharfman, Reference Dean and Sharfman1996; Elbanna & Child, Reference Elbanna and Child2007b). Our study uses a process model of political behavior in strategic decision making. This model suggests that strategic decisions are arenas where decision makers compete to satisfy their individual/subunit interests rather than organizational ones (Sharfman et al., Reference Sharfman, Dean, Mitchell and Bartkoski2009).
The inconsistent findings of previous research on political behavior may be the result of applying over simplified models to complex phenomena. To overcome these inconsistencies, we advocate including key contextual variables as moderators and a mediator of the relationship between political behavior and decision success. To do so, we consulted previous research on strategic decision making in order to develop our conceptual framework. We selected two moderating factors (product and competition uncertainties), representing industry environment as it has been frequently used in the literature as a key moderator of decision success (e.g., Calantone, Garcia, & Dröge, Reference Calantone, Garcia and Dröge2003; Shepherd & Rudd, Reference Shepherd and Rudd2014). Moreover, we examine the role of an additional moderator variable at the society/nation environment level, namely, the global financial crisis, as this will be particularly relevant to the study of the use of political behavior in decision making (e.g., Ashmos, Duchon, & Bodensteiner, Reference Ashmos, Duchon and Bodensteiner1997; Vaaler & McNamara, Reference Vaaler and McNamara2004) and aids our understanding of organizational behavior under such conditions (Tushman, Reference Tushman1977). For example, decision making in a context of high uncertainty, which is typical for global financial crisis, is particularly susceptible to political behavior (cf. Papadakis, Lioukas, & Chambers, Reference Papadakis, Lioukas and Chambers1998). In such situations of high levels of uncertainty and risk, decision makers are faced with ambiguity as to what behaviors are effective; therefore, they are likely to develop their own, possibly self-serving rules (Walter, Kellermanns, & Lechner, Reference Walter, Kellermanns and Lechner2012).
Finally, the literature on decision process generally remains fragmented and dispersed throughout the examination of individual decision modes, with little systematic considerations of the interrelationships between different modes (see, e.g., Child & Hsieh, Reference Child and Hsieh2014). In this study, we seek to cast light on this critical issue running through scholarly debates by examining the impact of intuition on the relationship between political behavior and decision success. In so doing, our study attempts to integrate the fragmented literature on the topic. Indeed, much scholarly attention has been devoted to the phenomenon of decision making over the past 60 years. Many classic works have tackled the question of whether decision making is simple or complex (Snyman & Drew, Reference Snyman and Drew2003). However, most prior literature has focused on a particular mode to the exclusion of others (Hart, Reference Hart1992) without examining the interrelationships between these modes (Elbanna, Reference Elbanna2006). Resulting typologies have therefore tended to compete or overlap, but few, if any, capture the full range of factors influencing decision making. Much work remains needed, therefore, to operationalize and apply different integrative frameworks by examining interrelationships between decision process modes such as ours, politics and intuition.
Political behavior
Political behavior in decision making seeks to ‘get others to do what we want, when they might not elect to do so’ (MacMillan & Jones, Reference MacMillan and Jones1986: 1). Such organizational behavior is sometimes depicted as game-playing (Allison, Reference Allison1971). Game-playing constitutes normal behavior for ambitious executives, because it gives them greater ability to influence events according to their interests and to enhance their power and position. For this reason, they try to satisfy their personal needs by using political tactics. It is uncommon to find political behavior limited to one tactic alone. In addition to the use of power, decision makers may employ other tactics such as forming a coalition (for more information on political tactics see Child, Elbanna, & Rodrigues, Reference Child, Elbanna and Rodrigues2010). Given the above, we define political behavior as intentional attempts to enhance or protect the self-interest of individuals or groups (Hickson et al., Reference Hickson, Butler, Cray, Mallory and Wilson1986). This definition captures political behavior as it is broadly presented in the literature, and encompasses the use of power (Pfeffer, Reference Pfeffer1992; Duarte, Reference Duarte2010), cooptation (Eisenhardt & Bourgeois, Reference Eisenhardt and Bourgeois1988), limiting the alternatives considered (Cyert & March, Reference Cyert and March1963), distorting information (Pettigrew, Reference Pettigrew1973) and timing tactics (Elbanna, Reference Elbanna2006). We note here that our definition of political behavior assumes that its impact on decision success is largely negative or dysfunctional, which is consistent with much of the previous literature, but which has been criticized for ignoring possible positive or functional aspects (Fedor, Maslyn, Farmer, & Bettenhausen, Reference Fedor, Maslyn, Farmer and Bettenhausen2008). We take up this potential shortcoming in our concluding section.
As interaction between people, both within organizations and across their formal boundaries, is one of the underlying features of the strategic decision-making process, forms of political behavior affecting this process can be categorized according to their source and the relations involved (Child, Elbanna, & Rodrigues, Reference Child, Elbanna and Rodrigues2010). Some political actions arise primarily from the exercise of formal authority by those in superior hierarchical positions. Others amount to politicking between the people or organizational units occupying similar hierarchical levels, but often oriented toward persuading strategic decision makers higher up. There are also instances in which political behavior amounts primarily to the exercise of upward influence.
The impact of political behavior on decision success
Political behavior by senior management can have a significant impact on the outcomes of managerial decisions. Table 1 presents three selected real-life examples, illustrating that the impact of political behavior on decision-making quality can be positive or negative. Most previous research on the subject, however, has concluded that political behavior adversely affects the quality of decision-making and/or organizational performance. The following reasons may help to account for this negative relationship. First, many of the political tactics, which are successful in exerting influence over strategic decision making, such as manipulation and secret communication, lead to the selective and biased disclosure of relevant information (Pettigrew, Reference Pettigrew1973). Organizational politics are therefore likely to distort the information required for effective decision making (Cyert & March, Reference Cyert and March1963; Pfeffer, Reference Pfeffer1992). This behavior contrasts with open and straightforward methods of eliciting opinion in favor of the best outcome, which rely on open discussion and the sharing of information among decision makers (Eisenhardt & Bourgeois, Reference Eisenhardt and Bourgeois1988). The consequence is that organizational politics may lead managers to make decisions based on incomplete information, giving rise to disappointing outcomes (Dean & Sharfman, Reference Dean and Sharfman1996).
Political behavior in managerial decision making: three real examples
Second, political decision processes are divisive. This divisiveness may inhibit agreement on key strategic concepts and on the allocation of responsibilities for carrying out strategies effectively. Moreover, they are time-consuming. They may delay decisions, with the possible loss of opportunities and profits (Pfeffer, Reference Pfeffer1992). Some therefore regard political behavior as a waste of organizational resources that could instead be directed toward achieving the organization’s objectives (Eisenhardt & Bourgeois, Reference Eisenhardt and Bourgeois1988). Third, as argued by Dean and Sharfman (Reference Dean and Sharfman1996), political behavior may lead to incomplete understanding of the environmental constraints, resulting in the undermining of strategic decision quality, in two ways. One is that political tactics are directed toward serving the interests, power bases and positions inside the organization, rather than on what is feasible, given the prevailing environmental forces, making the decisions that result from such processes less likely to consider environmental constraints. The other is that political processes may exclude some feasible alternatives, because they conflict with the interests of powerful individuals.
In conclusion, political processes are unlikely to encourage a complete and accurate analysis of strategic decisions; consequently, they increase the possibility of poor performance and unsuccessful decisions. Empirically, previous studies suggest a negative link between political behavior and decision outcomes (e.g., Dean & Sharfman, Reference Dean and Sharfman1996; Elbanna & Child, Reference Elbanna and Child2007b). Thus, we hypothesize that:
Hypothesis 1: Political behavior will be negatively related to decision success.
Environmental uncertainty as a moderator
Contingency theory argues that environmental factors should have some kind of impact on the relationship between strategic decision processes and outcomes (Goll & Rasheed, Reference Goll and Rasheed1997), and suggests that its potential moderation of the political behavior–decision success relationship should not be overlooked. In the product innovation literature, there is evidence that environmental turbulence moderates the effect of key decision-making processes such as team reflectivity and work group quality on new product success (Dayan & Elbanna, Reference Dayan and Elbanna2011).
In an uncertain environment, decision makers have to respond to intense changes, such as rapidly changing customer needs, wants and competitors’ strategies. Major organizational changes may cause fear that benefits or even continued employment may be at risk. In such an environment, political tactics may be even more evident, as decision makers may resort to politics to enhance their power, obtain greater benefits and secure their organizational position. In cases where these personal goals conflict with organizational ones, they are likely to put their own interests ahead of those of the organization (Child, Elbanna, & Rodrigues, Reference Child, Elbanna and Rodrigues2010). This can slow down decision making, which is a particular problem as environmental uncertainty usually calls for quicker decisions (Pfeffer, Reference Pfeffer1992). Generally, studies of environmental uncertainty find that the relationships between decision making and its antecedents are stronger as uncertainty increases (Judge & Miller, Reference Judge and Miller1991). This suggests that environmental uncertainty acts as a homologizer moderator that does not directly influence political behavior, but affects the magnitude of the political behavior–decision success relationships.
The literature distinguishes between several forms of environmental uncertainty (Calantone, Garcia, & Dröge, Reference Calantone, Garcia and Dröge2003). For instance, product uncertainty exists when one faces unpredictable changes in product demand; while competitive uncertainty exists where the market is still emerging (Moorman & Miner, Reference Moorman and Miner1997; Song, Di Benedetto, & Parry, Reference Song, Di Benedetto and Parry2009) and covers the uncertainties regarding strategic moves by existing rivals and potential ones (Miller, Reference Miller1993). Both forms of uncertainty are related to higher organizational risk and a lower likelihood of decision-making success (Song, Di Benedetto, & Parry, Reference Song, Di Benedetto and Parry2009). We therefore state two moderating effect hypotheses for environmental uncertainty to test the different forms of uncertainty separately:
Hypothesis 2a: The negative relationship between political behavior and decision success will be stronger when there is a higher level of product uncertainty.
Hypothesis 2b: The negative relationship between political behavior and decision success will be stronger when there is a higher level of competition uncertainty.
The role of the global financial crisis
The economic crises have a significant impact on managerial process and hence organizational outcomes. For example, it was found that there was a shift in the dynamic of Thai multinationals’ international expansion after the Asian economic crisis (Pananond, Reference Pananond2007). Decision-making processes designed for generally stable conditions give way occasionally to different processes, prompted by unplanned environmental discontinuities, such as unexpected events at the national and even global level (Watkins & Bazerman, Reference Watkins and Bazerman2003). Such discontinuities influence the decision-making process through the higher perceived levels of uncertainty, risk, urgency to act, narrowing options and high-stakes implications for organizational survival (Vaaler & McNamara, Reference Vaaler and McNamara2004; Mousavi & Gigerenzer, Reference Mousavi and Gigerenzer2014).
An example of environmental discontinuities is the late 2000s financial crisis and subsequent economic downturn, which are among the most important global economic events since the Great Depression of the 1930s. They contributed to the failure of key businesses, declines in consumer wealth and a significant decline in economic activity, leading to a severe global economic recession in 2008 and later years. Since 2008, the world has witnessed a sharp change from the previous decade of stability and growth. This crisis was a severe shock for many countries. Hence, it is understandable that the global financial crisis might affect managerial practices, such as decision making, and it is necessary for academics to study the impact of this event on the quality of decision making by managers.
While all countries are affected by the global financial crisis, it should be clear that the impact varies greatly from one country to another. As its economy is open and dependent on international markets, Tunisia has suffered greatly from the global financial crisis. This influence was made gradually via a number of mechanisms and transmission channels. Even if the impact on the financial (asset prices, external financing, interest rate spreads, volatility of financial markets, etc.) and monetary levels (currency and bank credit) was limited, it was large on the level of the real economy including trade, remittances and debt services (Mouley & Baccouche, Reference Mouley and Baccouche2012; Mouley, Reference Mouley2013).
For example, Europe, which suffered severely from the financial crisis, is the principal trade partner of Tunisia in terms of exports (76.3%), imports (71.5%) and tourism revenues (83.2%) (The Central Bank of Tunisia, 2009). Because of the global financial crisis, the Tunisian economy suffered from many problems in 2008–2009, such as an increasing rate of unemployment (e.g., a decline by 60% in the new jobs directly created in the tourism sector in 2008) and a decline by 1.5% in the growth rate of the Tunisian economy in the same year (The Central Bank of Tunisia, 2009).
A decline in foreign demand in the main export sectors since October 2008 has led to a fall in the growth of exports from 12% in 2007 to 1% in 2008 and −22% in the first semester of 2009, especially in textile and clothing sectors (−16.6%) and in the mechanical and electrical sectors (−12.6%) (Mouley, Reference Mouley2013). At a microeconomic level, these problems increased difficulty of access of small- and medium-sized firms to financing and diminished their export performance (Mouley & Baccouche, Reference Mouley and Baccouche2012). These problems forced many firms to withdraw and liquidate their business and left many others struggling to stay solvent. In such circumstances, firms have to respond to intense pressures.
Here, decision makers may perceive that the survival of the organization is at stake (Ashmos, Duchon, & Bodensteiner, Reference Ashmos, Duchon and Bodensteiner1997) and that immediate action should be taken, for instance, seeking a merger to stave off bankruptcy (Mintzberg, Raisinghani, & Theoret, Reference Mintzberg, Raisinghani and Theoret1976). In this environment, some prospects may negatively influence decision makers themselves, for example, of a drop in income and of being laid off. Therefore, political tactics may be much more important in such situations. In coping with the possibility of losing their benefits, decision makers will have more desire to use political tactics, not only to enhance their power or get more benefits but also, and most importantly, to secure their current positions and the benefits they still retain. This magnifies the negative impact of political behavior on decision success in such circumstances. Formally:
Hypothesis 3: The negative relationship between political behavior and decision success will be stronger during the period of global financial crisis than before the global financial crisis breaks out.
The mediating role of intuition
Political behavior is likely to affect other decision processes, for example, by deciding which alternatives to be considered, and the information gathered about those alternatives (Pfeffer, Reference Pfeffer1992). Further, previous research on the topic has largely focused on the direct effects of political behavior on decision success and the moderating impact of environment on this relationship (e.g., Dean & Sharfman, Reference Dean and Sharfman1996), rather than the possible mediating impact of other decision modes such as intuition. As argued by Hart (Reference Hart1992), the politically motivated behavior among decision makers results in a disjointed and incremental decision process (Cyert & March, Reference Cyert and March1963; MacMillan & Jones, Reference MacMillan and Jones1986). Similarly, Finkelstein (Reference Finkelstein1992) suggested that power (a well-known political tactic) has an impact on strategic actions of managers. For example, managers may use their power to reduce information processing and sharing, and enhance intuitive decisions instead. In other words, when decision making is highly political, the decision-making process can be characterized as more intuitive.
Recently, knowledge of intuition has significantly advanced and much attention has been given to its relationship with rationality (such as dual process theory) compared with other decision modes such as political behavior. Decision intuition can be defined as ‘a mental process based on a “gut feeling” as opposed to explicit, systematic analysis, which yields an intuitive insight or judgment that is used as a basis for decision making’ (Elbanna, Child, & Dayan, Reference Elbanna, Child and Dayan2013: 150). Intuition is not a conscious analytical step-by-step way of thinking, it is rather incremental adaptations based on deep and intimate knowledge of the situation faced by decision makers (Eisenhardt & Zbaracki, Reference Eisenhardt and Zbaracki1992). It can be understood as a composite phenomenon involving knowing and sensing (Sadler-Smith & Shefy, Reference Sadler-Smith and Shefy2004) and depend on a low level of formalization. Decision makers can use intuition when they cannot access rational processes (Parikh, Neubauer, & Lank, Reference Parikh, Neubauer and Lank1994) or they do not want to do so.
There are several reasons why political behavior in strategic decision making may increase decision intuition. First, high levels of political behavior are characterized by informal discussions over strategic issues (Dayan, Elbanna, & Benedetto, Reference Dayan, Elbanna and Benedetto2012) and restricting the ability to exchange expertise and knowledge (Emerson, Reference Emerson1962), which is likely to translate into intuitive decision making. Even when some decision makers have important information available, they may not get the chance to contribute their perspectives as the powerful decision makers may not attend to these insights fearing that they would undermine their own preferences. Similarly, strong power variation, which is an indication of political behavior, seems to place a blind spot on considering the viewpoints of others and rational analyses that locks individuals into their own myopic perspectives and judgments (Galinsky, Rus, & Lammers, Reference Galinsky, Rus and Lammers2011). In contrast, when political behavior is low, decision makers have a say in the decision process and therefore, more decision makers are expressing their views and actively participating in a dialogue that leads to a stronger information exchange (March & Olsen, Reference March and Olsen1982).
Second, political behavior is an important indicator of top management’s attitudes to consider their own preferences rather than those of the organization. The more political activities that is devoted during strategic decision making, the more that managers tend to practice intuition, which may be easier to control its outcomes compared with rational analysis. Third, it was argued that the use of intuition is favored by managers when there is no objective criterion for success (Hodgkinson, Sadler-Smith, Burke, Claxton, & Sparrow, Reference Hodgkinson, Sadler-Smith, Burke, Claxton and Sparrow2009). Therefore, organizational politics may be underlying element of intuitive decision making. Together, these arguments suggest that political behavior is likely to have direct implications for the decision intuition. As depicted in Figure 2, it is rational to use intuition in making important decisions, directed by self-interests of decision makers, because this can vastly consider these interests, which they cannot express the real rationale behind them.
Assumptions underlying the mediating role of intuition in the decisions politics–success relationship. The idea of this figure was adapted from Dean and Sharfman (Reference Dean and Sharfman1996)
When strategic or ill-structured problems are driven by political behavior, decision makers may use judgment rather than objective or rational criteria in making such decisions and hence the likelihood of decision success may well be decreased by the political behavior of decision makers. That is, it is not only political behavior that matter for decision success, but also the intuitive processes that accompany political processes of decision making. We, therefore, propose that:
Hypothesis 4: Intuition will mediate the relationship between political behavior and decision success.
Methods
Questionnaire development
We developed our questionnaire in five phases. First, the first author developed a draft questionnaire in English. Second, three academics reviewed the draft questionnaire. Third, the final version of the questionnaire was translated into French by the third author, because French is the language most often used in the business community in Tunisia. Fourth, three bilingual academics reviewed the two versions of the questionnaire to ensure that the translation was equivalent. Finally, the modified French version was administered to several Tunisian executives to elicit feedback.
Sampling and data collection procedures
The study was conducted among private Tunisian firms employing from 50 to 250 employees and located in the three biggest cities in Tunisia, namely, Tunis, Sfax and Sousse. Considering the difficulty of conducting a research survey on the basis of a random sample in Arab countries, for example, the absence of accessible sampling frames (Zahra, Reference Zahra2011) and the suspicion aroused by academic research in Tunisia before the recent Tunisian revolution, a convenience sample was used as recommended by previous research conducted in Arab countries (Elbanna & Child, Reference Elbanna and Child2007b).
We examined non-response bias by separating our sample into two subsamples, early and late respondents, and performing t-tests on the responses of each subsample based on the number of employees, percentage of export to total sales and industry type. The two sample t-tests were found to be insignificant, showing that there were no significant differences between the two samples at the p<.05 level.
A successful technique (personal delivery and pick-up) in collecting data from Arab countries was used in this study (Elbanna, Reference Elbanna2010). In order to reduce bias, the Tunisian author first held a training workshop for the data collection team of MBA students to explain the questionnaire and data collection procedures. She also closely supervised the team during the period of data collection. In mid-2009, the team collected data on the decisions made and completed before the financial crisis (2006–2007). In mid-2010, the team collected data on decisions made and completed during the financial crisis (2008–2009). All of the 250 firms receiving the questionnaire were contacted at least once. Two reminders were used to encourage late respondents to reply. All persons involved in the study were closely involved in making the chosen decisions and gave their informed consent before their inclusion in the study.
From the 250 questionnaires distributed to our sampled firms, 131 usable questionnaires were collected (a response rate of 52.4%), of which 49% represented decisions made during 2006–2007 and 51% represented decisions made during 2008–2009. Moreover, three respondents completed the final format of the questionnaire twice for purposes of test–retest reliability; and five firms were asked for two responses each, to check interrater reliability.
The respondent sample comprised presidents (18%), general managers (8%), managers of sectors or departments (44%), heads of sections (25%) and other positions (5%). The firms sampled belong to a wide range of manufacturing (70%) and service (30%) industries. The decisions examined vary widely. The average size of the firms is 131 employees. Non-exporting firms represent 49% of the sample; while the remaining (51%) are exporting firms.
Operationalization
We developed a decision success measure based on scale items used in related research (Rodrigues & Hickson, Reference Rodrigues and Hickson1995; Elbanna & Child, Reference Elbanna and Child2007b). The scale (α=0.82) comprised three items measuring whether the firm was successful in achieving desired objectives, the extent to which there were major negative unexpected outcomes of the decision (reversed), and whether the decision was the right choiceFootnote 1.
To measure political behavior, we used a 4-item scale (α=0.70) adapted from Elbanna (Reference Elbanna2010). This scale measured whether decision makers use their power to defend their preferences (e.g., Pfeffer, Reference Pfeffer1992; Dean & Sharfman, Reference Dean and Sharfman1996), managers form alliances with each other to promote their points of view (e.g., Eisenhardt & Bourgeois, Reference Eisenhardt and Bourgeois1988; Walter, Kellermanns, & Lechner, Reference Walter, Kellermanns and Lechner2012), decision makers hide or distort information (e.g., Pettigrew, Reference Pettigrew1973; Eisenhardt & Zbaracki, Reference Eisenhardt and Zbaracki1992), or variations exist between formal and informal discussions (e.g., Dayan, Elbanna, & Benedetto, Reference Dayan, Elbanna and Benedetto2012). Similarly to related research (e.g., Dean & Sharfman, Reference Dean and Sharfman1996; Elbanna & Child, Reference Elbanna and Child2007b; Dayan, Elbanna, & Benedetto, Reference Dayan, Elbanna and Benedetto2012), the above scale incorporates both leading (the use of power, forming alliances and control of information) and lagging (variations in discussions) indicators of political behavior. Leading indicators measure political tactics and activities, while lagging indicators are caused by the level of leading indicators and hence cannot be directly managed. Thus, complementing both leading and lagging measures, as we did in this study, provides a more balanced view of political behavior.
A measure of intuition that was originally developed by Khatri and Ng (Reference Khatri and Ng2000), and validated against an established measure of intuitive decision-making style introduced by Scott and Bruce (Reference Scott and Bruce1995) and examined in a neighboring Arab country (Elbanna, Child, & Dayan, Reference Elbanna, Child and Dayan2013) was adopted in this study (α=0.68). This measure captures the main indicators of intuitive decision making, namely, knowing (judgment) and sensing (gut feeling).
Both product uncertainty (α=0.76) and competition uncertainty (α=0.87) were adopted from Miller (Reference Miller1993). The product uncertainty scale measured uncertainties in client preferences, product demand, changes in product components and changes in product quality; the competition uncertainty scale measured uncertainties in competitive prices, markets and strategies, as well as uncertainty about new competitive entrants. A dummy variable was used to distinguish between decisions made and completed before the financial crisis (coded 0) and decisions made and completed during the financial crisis (coded 1).
Control variables
We controlled for six contextual variables, representing four levels of analysis: decision level (decision importance), firm level (firm performance, firm size and exporting experience), industry level (industry type) and environmental level (environmental munificence). We selected these variables because of the expectation that decision processes and outcomes might vary systematically with them as discussed in related research, decision importance (e.g., Dayan, Elbanna, & Benedetto, Reference Dayan, Elbanna and Benedetto2012), firm performance (e.g., Rodrigues & Hickson, Reference Rodrigues and Hickson1995), firm size (e.g., Miller, Burke, & Glick, Reference Miller, Burke and Glick1998), exporting experience (e.g., Collinson & Houlden, Reference Collinson and Houlden2005), industry type (e.g., Elbanna, Reference Elbanna2012) and environmental munificence (e.g., Castrogiovanni, Reference Castrogiovanni1991). For example, while exporting Tunisian firms have learned to face international competition and to modernize their management style, domestic firms are less modernized and have many shortcomings in their management style (Madani & Ayoub-Jedidi, Reference Madani and Ayoub-Jedidi2005).
We measured decision importance with a 5-item scale adapted from related research (Elbanna & Child, Reference Elbanna and Child2007a; Dayan & Elbanna, Reference Dayan and Elbanna2011). This scale measures whether the decision is perceived to be a base for future decisions, whether it will result in substantive changes in the firm, whether it is viewed as important, and whether there would be serious consequences if the decision were delayed or a bad decision were made. Financial and business performance was measured in terms of return on assets, operating profits, market share and sales revenue growth, while organizational effectiveness was measured in terms of employee satisfaction, operational efficiency and social responsibility (adapted from Hart & Banbury, Reference Hart and Banbury1994). The number of full-time employees was used to indicate firm size, and the logarithm of firm size was used in the analysis. A dummy variable was used to distinguish between non-exporting firms (coded 0) and exporting firms (coded 1).
A 3-item scale of environmental munificence was adopted in this study (Khandwalla, Reference Khandwalla1977). This scale measured whether the environment was perceived to be safe and non-threatening, rich in investment opportunities and controllable. Higher scores reflected more munificent environments. Finally, a dummy variable was used to distinguish between manufacturing industries (coded 0) and service industries (coded 1).
Internal consistency
As indicated in Table 2, the results of α coefficients range between 0.70 and 0.87 for all scales, except intuition (0.68), indicating a satisfactory degree of internal consistency.
Descriptive statistics and correlations
Note. N=131.
†p<.10; *p<.05; **p<.01.
Test–retest reliability (stability)
The average of Pearson correlation coefficients is 0.82 (p<.01 in the three cases) between the answers of the three respondents who completed the identical questionnaire on the same decision on two different occasions (with a time gap of 4 months). This result suggests a high degree of stability in the measures used in this study.
Interrater reliability
The average of Pearson correlation coefficients is 0.91 (p<.01 in the five cases) between the answers of the multiple respondents in the same organization, who completed the identical questionnaire on the same decision. This result shows a high level of interrater reliability.
Criterion-related validity
Recalling the relevant research, we examined the criterion-related validity of our independent and dependent variables. Evidence from related research shows a positive relationship between strategic decisions that are considered as a threat/crisis and political behavior in strategic decision making (Elbanna, Reference Elbanna2010). Hence, the concurrent validity of political behavior, an indicator of criterion-related validity, was assessed by examining the correlation of the scale with a measure of decision motive. The results show a positive relationship between political behavior and decision motive, when a decision is considered as a threat/crisis (r=0.29, p<.01). Similarly, our measure of decision success is significantly correlated with a well-established measure of decision effectiveness developed by Dean and Sharfman (Reference Dean and Sharfman1996) (r=0.58, p<.01).
Construct validityFootnote 2
Because of our small sample size (131 observations) and the 32 items included in our analysis, our measures were divided into two subsets for conducting confirmatory factor analysis (CFA). The first subset includes our main variables (decision success, politics, product uncertainty, competition uncertainty and intuition); while the second one incorporates control variables (decision importance, financial and business performance, organizational effectiveness and environmental munificence). The CFA results for the main variables show that, with one exception (0.42) all item measures had standardized factor loadings of >0.50 and the loadings estimates ranged from 0.52 to 0.95. Thus, adequate convergent validity for the items was established. The CFA results indicate a good fit of the confirmatory measurement model by different indices (χ2=180, df=109, p=.00, χ2/df=1.65, IFI=0.91, CFI=0.92, RMSEA=0.07, confidence interval [CI]=0.05–0.09, PClose=0.04). Conducting CFA for control variables shows that the loadings estimates ranged from 0.55 to 0.83 with two exceptions (0.33 and 0.38) and indicate a good fit by different indices (χ2=145, df=84, p=.00, χ2/df=1.73, IFI=0.91, CFI=0.90, RMSEA=0.08, CI=0.05–0.095, PClose=0.03). The above results showed that the two CFA models fit adequately.
Biases
We attempted to reduce the possibility of incomplete recall and retrospective rationalization in several ways. (1) Scale anchors were reversed in several places. (2) The survey was arranged so that the dependent variable followed the independent variables. (3) The items which constituted a specific construct were separated from the others to limit the consistency bias and reduce repetitiveness. (4) Complete anonymity and confidentiality were assured. (5) Objective data were used to measure four of our variables: firm size, industry type, the global financial crisis and exporting experience.
Harman’s one-factor test was used to test the possibility of a common method bias by performing an exploratory factor analysis (Podsakoff & Organ, Reference Podsakoff and Organ1986). We found nine factors, rather than one single factor. The first factor captured only 18% of the variance in the data. Hence, little common method variance appears to be present. Common method concerns are further mitigated by the complex relationships examined in our study, because it is unlikely that survey respondents were able to guess our moderation and mediation hypotheses. Moreover, finding empirical support for predicted interaction effects shows that the results are not inflated by a common method (Evans, Reference Evans1985).
Results
Table 2 provides descriptive statistics. To test our hypothesized relationships, we performed regression analysis and entered the variables in seven steps as shown in Table 3. Model 1 includes control, moderator and mediator variables and explains 28% (p<.01) of the variance in decision success. In Model 1, decision importance (β=0.30, p<.01) and financial and business performance (β=0.35, p<.01) are positively associated with decision success. Adding political behavior to the control variables explains significantly more variance in decision success (∆R 2=0.05, p<.01), and political behavior is negatively related to decision success (β=−0.24, p<.01). These results provide full support for our first hypothesis.
Regression modelsFootnote a
Note. N=131.
a In Model 6, the mediator (intuition) regressed on predictors; in all remaining models, decision success regressed on predictors.
*p<.05; **p<.01.
The moderating effects of both environmental uncertainty and global financial crisis were tested using hierarchical moderated regression analysis. The procedures of this analysis, as used in this study, consist of two steps. The first step was to enter the main effects of political behavior and the control and moderating factors as one block into the equation. This step represents the base model (Model 2). The second step involved three further models that each entered interaction term, which were calculated as the product of political behavior and a moderating variable (Models 3–5).
Model 3 in Table 3 shows that the interaction between product uncertainty and political behavior significantly contributed to Model 2 (∆R 2=0.03, p<.05) and the interaction coefficient between product uncertainty and political behavior is negative and significant (β=−0.70, p<.05). Following Aiken and West (Reference Aiken and West1991), we plotted the equation at low and high values to illustrate the nature of the interaction. As shown in Figure 3, product uncertainty strengthens the negative relationship between political behavior and decision success. At high levels of product uncertainty the relationship between political behavior and decision success becomes stronger (t=−4.25, p<.01), whereas at low levels, the relationship remains largely unchanged (t=−1.14, ns). These findings provide support for Hypothesis 2a.
The interactive effects of product uncertainty and political behavior on decision success
Model 4 in Table 3 shows that the interaction between competition uncertainty and political behavior did not significantly contribute to Model 2 (∆R 2=0.00, ns) and that the interaction coefficient between competition uncertainty and political behavior is not significant (β=−0.18, ns). These findings provide no support for Hypothesis 2b.
Model 5 shows that the interaction between global financial crisis and political behavior did not significantly contribute to Model 2 (∆R 2=0.00, ns); and that the interaction coefficient between global financial crisis and political behavior is not significant (β=−0.08, ns). These findings provide no support for Hypothesis 3Footnote 3.
The mediating effect of intuition was examined following the seminal work of Baron and Kenny (Reference Baron and Kenny1986), which suggests that a variable acts as a mediator if it meets four conditions. These are (1) the independent variable (political behavior) significantly affects the dependent variable (decision success); (2) the independent variable significantly affects the mediator variable (intuition); (3) the mediator variable significantly affects the dependent variable; and (4) when both the independent and mediator variables are considered together, the impact of the independent variable on the dependent variable is reduced. Full mediation is present if the independent variable has no effect when the mediator is controlled and partial mediation is present if the independent variable is significant but the strength of the relationship is reduced when the mediator is controlled.
As discussed earlier, political behavior significantly affects decision success, the dependent variable (β=−0.24, p<.01) and hence achieving the first condition for the existence of a mediating effect (Model 2). Model 6 shows that political behavior significantly affects the mediator variable, intuition (β=0.35, p<.01), which achieves the second condition. The regression shows that when both political behavior and intuition are incorporated in the same model (Model 7), the explained variance in decision success increased significantly (∆R 2=0.03, p<.05). We found that both variables, political behavior (β=−0.17, p<.05) and intuition (β=−0.20, p<.05), were significantly related to decision success. However, the explanatory value of political behavior decreased when the mediating variable (intuition) was included in the regression as shown in Model 7 (β=−0.17, p<.05) compared with its explanatory value in Model 2 (β=−0.24, p<.01). This supports the third and fourth conditions for the existence of a mediating effect. In conclusion, the above results show a partial mediation effect of intuition as the impact of political behavior on decision success was reduced when the mediator, intuition, was controlled for in Model 7. Hence, Hypothesis 4 is supported.
Discussion
Though there are many studies of political behavior in the existing literature, very few studies have considered the moderating effect of different characteristics and levels of environment and no study examines the mediating impact of intuition in decision politics–success relationship. The current study attempted to fill this gap by developing a conceptual model that examines the impact of political behavior on decision success in a new setting, while considering the influence of key environmental factors (two aspects of environmental uncertainty and the global financial crisis) and an important dimension of decision process (intuition). The empirical findings supported the following hypotheses derived from our conceptual model:
∙ A higher level of political behavior was related to less decision success (Hypothesis 1).
∙ The negative relationship between political behavior and decision success was stronger when there was a higher level of product uncertainty (Hypothesis 2a).
∙ Intuition mediated the relationship between political behavior and decision success (Hypothesis 4).
Our hypotheses on the moderating impacts of competition uncertainty (Hypothesis 2b) and the global financial crisis (Hypothesis 3) were not supported, however, and these surprising observations are more fully discussed below. The explanatory power of our regression models in Table 3 is acceptable (ranging from 27 to 36%).
The above results suggest that political behavior has a detrimental impact on organizational outcomes. This is consistent with previous research conducted both in the West (e.g., Dean & Sharfman, Reference Dean and Sharfman1996; Walter, Kellermanns, & Lechner, Reference Walter, Kellermanns and Lechner2012) and in the Arab world (e.g., Elbanna & Child, Reference Elbanna and Child2007b), and hence supports the culture-free argument. It lends additional support to the deep-rooted arguments in the strategic decision-making literature on the potential negative impact of political behavior on decision success and other aspects of decision quality. This is not to say that political behavior is always directed toward self-interest or that it has no positive impact on decision outcomes. Some authors have emphasized the potential positive effects of political behavior (e.g., Gotsis & Kortezi, Reference Gotsis and Kortezi2010). It has been argued that decision makers need to be aware of political tactics and sometimes to use them in order to defend their firms’ interests and overcome the negative effects of other people’s political tactics (Child, Elbanna, & Rodrigues, Reference Child, Elbanna and Rodrigues2010). These contributions are, however, mainly theoretical in nature and empirical evidence is almost entirely absent as discussed below.
Although product uncertainty was found to be a significant moderator of the relationship between political behavior and decision success, no support was found for the moderating role of competition uncertainty on this relationship. These results show that executives are capable of differentiating between the uncertainties associated with different subenvironments and thus lend further support to the conceptualization of environmental uncertainty as a multi-dimensional construct (Miller, Reference Miller1993). Moreover, they show that the subscales of environmental uncertainty play different moderating roles in the link between political behavior and success, which may help to explain the contradictory results of previous research by providing an empirical support to the argument of several scholars (e.g., Werner, Brouthers, & Brouthers, Reference Werner, Brouthers and Brouthers1996; Elbanna, Child, & Dayan, Reference Elbanna, Child and Dayan2013) that variations in different dimensions of environmental uncertainty may reflect different types of risk and implications for decision making. This is particularly important in some settings where researchers show surprising results on the role of environmental uncertainty in strategic decision making (Papadakis, Lioukas, & Chambers, Reference Papadakis, Lioukas and Chambers1998). A possible explanation of these results is that different aspects of environmental uncertainty, influenced to some extent by the country in which the research is conducted, play different roles in strategic decision making.
Another surprising result of our study is the absence of a significant role of the global financial crisis in the relationship between political behavior and decision success. This finding may be owing to several reasons. The first explanation is our small sample size and its possible impact on the results of hierarchical moderated regression analysis. Second, this may be because of the way in which we examined this relationship. We used a dummy variable to distinguish between decisions made and completed before the financial crisis (coded 0) and decisions made and completed during the financial crisis (coded 1), which reduces the variety of interaction terms for decisions made before financial crisis even if they are highly political. Third, perhaps, the time gap between the start of the financial crisis and collecting data for this study was not enough for decision makers to perceive the detrimental effects of financial crisis and therefore that the variation in the perception of the impact of financial crisis among respondents was not sufficient significantly to affect the relationship between decision politics and success. An improved scenario for future research would be to examine the impact of financial crisis by including a third group of decisions, which were made at a substantially later time after the financial crisis. In sum, the insignificant role of financial crisis in this study is suggestive rather that definitive and we should consider the above concerns and also examine this impact in different national settings and industries to determine the generalizability of resultsFootnote 4.
A unique contribution of this study is the analysis of the mediating role of intuition in decision politics–success relationship. Our results show that intuition partially mediates the relationship between political behavior and decision success. This result lends empirical support to theorists who argue that decision makers may use ‘complex’ strategic decision processes (Snyman & Drew, Reference Snyman and Drew2003); consequently, enhancing our understanding of decision making would require greater attention to the role of process capability, the use of multiple decision-making process modes (Hart & Banbury, Reference Hart and Banbury1994). Studies that investigate process capability or complex decision processes are still rare in the literature (Rajagopalan, Rasheed, Datta, & Spreitzer, Reference Rajagopalan, Rasheed, Datta and Spreitzer1997; Elbanna, Child, & Dayan, Reference Elbanna, Child and Dayan2013). This was a gap, and the most important one, we tried to fill in this study.
Limitations and future research
Our findings should be evaluated in the light of the limitations of the current research. First, our data are cross-sectional, which impedes the precise understanding of the causal relationship between political behavior and decision success. Second, we used conceptual measures for most variables. Self-reported data may pose problems, such as limited recall by respondents. We, however, do not believe that recall lapses were a significant problem in this study, because, as reported in the ‘Methodology’ section, we took measures to improve the reliability and validity of retrospective reporting. Examining decisions in real time, using other methods such as interview and observation (e.g., Gilbert, Reference Gilbert2003), can avoid recollection biases, but pose a major challenge for researchers, owing to the time required to observe ongoing decision processes (Roberto, Reference Roberto2004). Third, we developed new measures for decision success, financial crisis and exporting experience that should be taken into account when interpreting the results of this study. Although the new measure of decision success is based on related research (Rodrigues & Hickson, Reference Rodrigues and Hickson1995; Elbanna & Child, Reference Elbanna and Child2007b) and shows good levels of reliability and validity, we recommend future research to further examine it in other settings and develop new measures for both financial crisis and exporting experience to better capture the essence of these variables.
Future research can also build on the remaining limitations of this study to explore additional research questions. First, political behavior is a complex construct and is perhaps not sufficiently well captured by a single scale. However, the current research, like most previous research, operationalized political behavior in terms of its dysfunctional role. Thus, the potentially positive, functional role of political behavior is not captured in this simple operationalization. As discussed above and shown in the cases of Table 1, political behavior can be functional or dysfunctional, that it, it can result in positive as well as negative outcomes. Functional political behavior is assumed to encourage the attainment of organizational goals and does not harm the organization. An example of this behavior would be forming a coalition with other decision makers who have similar thoughts on a new and constructive idea in order to lobby to pursue it. Moreover, political behavior can help in building commitment and understanding, by generating support for projects and persuading others of their merits (Roberto, Reference Roberto2004). This behavior was illustrated, for example, in the family-business paint company, previously seen as Case 3 in Table 1. Over-reliance on negative definitions of political behavior is a shortcoming of the research on political behavior in organizations, insofar as it ignores functional or positive aspects for individuals, groups and organizations.
Second, to examine the relationship between political behavior and decision success, more elaborate models may be required, as simple models cannot properly explain this complex phenomenon. Further research can examine the moderating role of other aspects of environmental uncertainty, such as economic and governmental uncertainty. Other organizational behavior variables such as organizational culture, trust, power effects, conflict resolution strategies and characteristics of decision makers (such as tenure and propensity to risk) are likely to interact with political behavior (Shepherd & Rudd, Reference Shepherd and Rudd2014). These interactions were not included in the present study but merit further investigation. Because the answers of respondents may be affected by the degree to which they agreed with the decision process during the period when it was made, it is recommended that future investigations ask how far respondents agreed with the decision under investigation, as a potential control variable. Furthermore, it would be desirable to assess how well decisions are implemented because of the potentially significant impact of implementation on decision success (Dean & Sharfman, Reference Dean and Sharfman1996).
Third, considering the multi-dimensionality of decision outcomes, we recommend future research to investigate the links between political behavior and other decision outcomes, such as decision speed, creativity, propitiousness, disturbance, efficiency, commitment, implementation and learning, which few, if any, studies have examined. Moreover, many scholars contend that managers must make painful trade-offs between competing outcomes (Janis, Reference Janis1989). Hence, future studies can also examine the role of political behavior in identifying the trade-offs between competing outcomes, which they must make (Roberto, Reference Roberto2004). Fourth, we do not include the effects of rationality in decision making in this study. Rational behavior in the context of imperfect or incomplete knowledge, or the possibility of bluffing or similar behavior by participants, is complicated and merits further investigation (cf. Sinclair, Ashkanasy, & Chattopadhyay, Reference Sinclair, Ashkanasy and Chattopadhyay2010). We recommend incorporating the effects of rationality in decision making in future studies. Fifth, this research took place within firms operating in a specific, single society/nation setting (Tunisia) and conditions of particular environmental contingency. Future cross-cultural studies can examine the generalizability of our findings for different societies/nations and other environmental contingencies.
Sixth, although it is argued that decision processes are interdependent (Pillemer & Racioppo, Reference Pillemer and Racioppo2003), scant systematic empirical research has investigated the interrelationships of different modes of decision-making process and their impact on decision outcomes, and searched for the optimal approach to blending them in practice, which hinder further develop decision-making theory. For example, rational mode that provides informative frameworks can provide a good foundation for intuitive mode of decision making; while intuition that is driven by political behavior negatively influences decision success. Similarly, the examination of the relationships between intuition and other decision processes such as reflexivity and improvisation is necessary to better understand the dynamics of intuitive decision making and its impact on decision outcomes.
From the standpoint of practice, this study helps decision makers by pointing out the negative influence of political behavior on decision success. A better understanding of the contextual factors that affect this influence can help practitioners to understand the contributions that political behavior may make toward the failure of strategic decisions with different sets of environmental variables and other decision modes. This understanding may ultimately lead to better strategic decision making in organizations.
Acknowledgements
This research was completed while the first author was a visiting scholar at Carleton University in Ottawa, Canada, and was partially supported by the National Research Foundation in the UAE (research grant no. RSA-1108-00788) and the UAE University (research grant no. 01-03-4-11/09).
