Abstract. Three studies examining the effects of leadership humility on employee voice, continual improvement and firm performance

Transcription

1 Three studies examining the effects of leadership humility on employee voice, continual improvement and firm performance Abstract Background: This paper uses results from three separate studies to examine the impacts of leadership humility on employee voice, continual improvement, efficiency and firm performance. Many economic models incorporate or appreciate differences in efficiencies across firms, but at best, these differences lie in the error term or are controlled for using firm fixed-effects. I will argue that some of the differences in efficiencies can, in fact, be modeled and may very well be attributable to management attitudes toward employee input. Indeed, employees are often the ones on the ground and thus are very likely to be able to generate ideas that may improve productivity or customer satisfaction, which may directly impact profits and performance. If leadership humility encourages more ideas from all levels and is also associated with a greater willingness to implement new ideas, these factors may be driving some of the differences in performance that have not yet been measured in our models. Methods: Survey data were collected in all three studies, with objective performance data available in the third study. (Performance is proxied by the employees perception of the level of efficiency in the first two studies.) All studies generate complex constructs by combining multiple related survey items into one variable, much like economists often do for socioeconomic status. Simple linear regressions, fixed effects regressions and regressions on a seemingly unrelated system of equations are all employed using these created constructs to test whether humble leaders are more likely to experience greater levels of employee voice, whether they are more likely to use employee ideas and implement new ideas in general, and whether these factors are associated with greater levels of efficiency in the workplace and store performance. Summary: The data collected across all three studies support the model. Employees are more likely to voice ideas for improvements when they perceive their managers as being more humble. Humble managers are also more likely to continually improve the way things are done, and continual improvement is associated with higher sales and fewer job obstacles (things getting in the way of employees doing their jobs well). 1 P a g e

2 1. Introduction In a dynamic economy where consumer preferences are constantly changing and new competitors are constantly emerging, firms need to be willing to adapt their processes and often their products and services to be able to survive the changing market conditions they face. However, if the leadership and the decision-makers aren t open to new ideas or willing to try new things, then the firm may get pushed out of the market. Economists know and understand this, but when analyzing differences in firms profits and performance, we simply study input and output decisions and at best, mention that the error term likely includes differences in efficiency or possibly the role of management (Bloom, Genakos, Sadun, and Reenen (2011)). The purpose of this research is to try to pull the effect of management out of the error term and offer one particular management style as a potential explanation for why we see differences in firms abilities to thrive whenever adaptation is required. This particular leadership style is called humble leadership and there are several reasons to believe that this type of leadership would encourage more ideas from all levels within the organization, be associated with a higher willingness to implement new ideas and thus be associated with better performing firms. Humility has been the recent focus of several studies, but we are far from understanding its true consequences and potential benefits. It has recently been formally defined in the literature as "an interpersonal characteristic that emerges in social contexts that connotes: (a) a manifested willingness to view oneself accurately, (b) a displayed appreciation of others strengths and contributions, and (c) teachability (Owens, Johnson, & Mitchell, 2013), where "teachability" includes an openness to ideas and advice from others. It is this trait in a leader that I believe leads to a willingness to adapt and change in order to improve performance and helps a firm remain competitive in the face of constantly changing market conditions. 2 P a g e

3 This paper explores three relationships associated with leadership humility. First, leadership humility is shown to be a catalyst for employees voicing improvement ideas upward. Secondly, leadership humility is shown to describe not only a willingness to listen to other s ideas, but is also strongly correlated with implementing new ideas (including the ideas of employees). Of course, listening to ideas and acting on those ideas are two different things, but it is shown that humble leaders may be more likely to do both. Finally, this research provides support for the notion that leadership humility is positively related to performance and efficiency, where performance is measure by monthly sales in dollars, and efficiency is measured by an absence of job obstacles (things getting in the way of employees doing their jobs well). The paper is structured as follows: Section 2 details the literature behind the constructs included in the study. Section 3 details the items used to measure the constructs, and Section 4 details the three separate studies, including the results from each. Section 5 concludes and is followed by Section 6, which touches on the limitations of this work followed by ideas that may correct these limitations in the future. 2. Literature Review 2.1 Humility Leader humility is still viewed as a rare personality trait that somewhat mysteriously produces favorable organizational outcomes (Owens & Hekman, 2012). In contrast with the more conventional idea that leaders should portray themselves as superstars, there have been calls for leaders to show their humanness by being open about their limitations in knowledge and experience (Weick (2001), as quoted by Owens and Hekman (2012)), especially since the workplace is becoming increasingly more complex and adaptability is now essentially required for survival (Weick (2001), referenced by Owens and Hekman (2012)). 3 P a g e

4 However, exactly what humility is or what humble leader behaviors look like is still unclear (Owens & Hekman, 2012). In a great summary of the previous literature, Owens and Hekman (2012) synthesize the various ways scholars and others have defined humility as involving how leaders tend to view themselves more objectively, others more appreciatively and new information or ideas more openly (pg 789). Put another way, those who are humble 1) acknowledge their limitations, 2) are open to others ideas and 3) view others appreciatively without feeling threatened by their potential contributions (B. P. Owens, Rowatt, & Wilkins, 2011). These characteristics are likely to be some of the underlying attitudes and beliefs of managers that foster the kind of environment where employees feel comfortable and safe expressing ideas, as well as encouraged or appreciated for doing so. When a leader is able to acknowledge their limitations they are likely more willing to seek the ideas and opinions of others, including their subordinates. When they can spotlight followers strengths and contributions (pg. 794) then these followers may be more willing to spend the time and energy generating and communicating their ideas for improving company performance. It therefore seems likely that these types of managers would experience more ideas from below; specifically, higher levels of employee voice. 2.2 Employee Voice There has been a great deal of research done in organizational behavior which focuses on employees taking part in building a firm s competitive advantage. One such avenue to explore the potential contribution of employees can be captured in the construct coined employee voice. This construct has been defined in various ways, but most researchers follow the definition put forth by Van Dyne (see, for example, Van Dyne and LePine (1998);Botero and Van Dyne (2009); Van Dyne, Ang, and Botero (2003)), which specifies that employee voice 4 P a g e

5 occurs when an employee makes innovative suggestions for change and recommends modifications to standard procedures, utilized not to criticize, but instead to proactively express constructive suggestions (Botero & Van Dyne, 2009), pg 87). Employee voice, then, relates specifically to improvement ideas suggested by employees that may lead to a greater level of success for the organization. (To be clear, this means it does not include employee complaints or dissatisfactions that don t bring with them potential improvement solutions.) Many researchers have attempted to discover what influences employee voice, what types of people are most likely to engage in it, and under what conditions it is most likely to occur. For example, studies have found that certain managerial characteristics such as approachability and responsiveness (Saunders, 1992)or transformational leadership (Detert & Burris, 2007) encourage employees to express their ideas. In addition, employees are more likely to speak up if they perceive that speaking up is safe and/or will be effective and make a difference (Morrison, 2011), and if personal beliefs such as power distance orientation leads one to believe that authority in power should or can be challenged (Botero & Van Dyne, 2009). Additionally, gender, experience in the field and tenure with the organization have been found to be individual-level characteristics that help predict employee voice (Morrison, 2011). Organizational characteristics such as the structure and bureaucracy can also either provide or withhold opportunities for ideas to be communicated and create an environment that either makes the ideas more likely to be implemented or more likely to be ignored. Ideas are usually solutions to problems. It seems reasonable that in many cases, employees can be great sources of ideas since they are the ones "on the ground" and often deal directly with the processes in place that are supposed to ensure smooth operations and happy customers. Indeed, whenever an employee recognizes that something can be done more 5 P a g e

6 efficiently or that customers are constantly asking for a product or service that the company doesn t currently offer, that employee decides whether or not to share the news or any accompanying ideas with upper management. It therefore seems likely that management s openness to employee ideas would make a difference in whether employees decide to voice their ideas upward. If humble leadership is also associated with a greater willingness to implement the ideas from employees, then employee voice can have a positive impact on the performance of the firm and the efficiency of the processes within the workplace. Indeed, it s not simply the amount of ideas voiced by employees that makes a difference; it s the implementation of the ideas by an open-minded manager that can actually impact performance. 2.3 Dynamic Capabilities and Firm Performance This research problem focuses on a firm s desire and capability to adapt, or more specifically, improve. A firm can improve processes to increase efficiency or improve products and services to increase customer satisfaction. Either way, the desire (and ability) to continually improve may also improve both the overall performance and the likelihood of surviving the turbulent dynamic markets within which so many firms operate. This idea runs parallel to the dynamic capabilities perspective introduced over 15 years ago. The dynamic capabilities literature has been extensively developed in the strategic management field after first being introduced by Teece, Pisano, and Shuen (1997). They were extending the ideas of the resource-based view (RBV) of the firm (introduced by Wernerfelt (1984) and later popularized by Barney (1991)) which held that firms need to own resources that are valuable, rare, inimitable and non-substitutable in order to achieve competitive advantage. Teece et al. (1997) argued that owning these types of resources wasn t enough; firms also needed the internal capabilities to deploy those resources (Wang & Ahmed, 2007). Specifically, since 6 P a g e

7 most markets are dynamic and constantly changing, in order to achieve a sustainable competitive advantage, firms must be able to constantly integrate, reconfigure, renew and recreate its resources and capabilities and, most importantly, upgrade and reconstruct its core capabilities in response to the changing environment (pg. 35). It is the emphasis on upgrading and renewing the resources that make this perspective especially relevant to this research. There are three factors that Wang and Ahmed (2007) put forth that explain firms mechanisms of linking internal resource advantage to external marketplace-based competitive advantage (emphasis added, pgs. 36-7): adaptive capabilities, absorptive capabilities and innovative capabilities. What will be argued here is that the internal resource advantage may come not from patent protection or other barriers to entry that are often credited by economists as the sources of market power, but instead from having internal processes that cannot easily be imitated by the competition. Examples include specifically the constructs being studied here: namely, the amount of employee voice that occurs and the leadership style that encourages it. Managers who encourage ideas from all levels likely have the best possibility of developing these capabilities and additionally, will likely out-perform their counterparts whose adaptive or innovative capabilities are stifled by close-minded management. Humble managers are also likely to have the openness required to continually improve the way things are done and adapt whenever necessary, as they are less likely to defend practices that no longer work simply because it was their idea in the first place. Indeed, links have been found between adaptive capabilities and firm evolution, survival and success (Wang & Ahmed, 2007); see also Alvarez and Merino (2003), Camuffo and Volpato (1996) and Forrant and Flynn (1999). Other studies have found these outcomes linked specifically to innovative capabilities (see, for example, Deeds, Decarolis, and Coombs (1999); 7 P a g e

8 Delmas (1999); Lazonick and Prencipe (2005); Petroni (1998); Tripsas (1997)). The question here is whether humble leaders are more likely to improve a firm's chances for success through creating an environment where adapting and innovating (processes or products) is more widely done. 2.4 Contributions to Existing Literature Existing economic models focus mainly on inputs and outputs to describe a firm s profitmaximization problem, and describe differences in performance that we observe in data as unobservable differences in efficiencies among firms. I will argue that these unobservable characteristics occurring inside the firm can (and should) be modeled using techniques that have been employed by other fields for many years. An integrated model incorporating theories of management and psychology, for example, may shed better light on what determines performance heterogeneity across firms over a model using isolated theories from economics. Perhaps management and employees affect profits not only through productivity, but also through influencing the flow of ideas and continual improvement. If these things can be measured, I offer them as potential reasons some firms continually outlast and out-perform their rivals. While this research is only part of a first step in testing whether humble leadership is good for an organization, it can offer support for future studies that will continue the conversation. Additionally, researchers are still trying to determine the potential consequences of leadership humility. This research contributes to the existing literature by adding support to the idea that leadership humility not only encourages employee ideas, but is also strongly associated with the continual improvement of processes, products and services, and can therefore further impact performance and reduce the existence of job obstacles (things that get in the way of employees doing their job well). I also find that past sales can influence the level of employee 8 P a g e

9 voice, which hasn t yet been tested in the literature. Most researchers work under the notion that employee voice is beneficial for the firm, but I find a negative relationship with past sales. Indeed, this makes perfect sense- the lower the past sales, the higher the level of employee voice because of the greater need for improvement. The purpose of this research is therefore threefold- to analyze whether more open-minded managers experience more employees voicing their ideas, whether humble leadership is associated with a greater willingness to implement new ideas, and finally whether these same managers are associated with better performance or fewer job obstacles. Indeed, to achieve the fluidity and adaptability that may be required for market survival, firm leaders may need ideas from employees across all levels of the organization, but they also need to be willing to implement these ideas and change the processes that can be improved upon. It makes sense that managers who display greater levels of humility and openness to ideas would experience greater levels of employee voice. Additionally, humble leadership may be associated with more ideas being implemented in the workplace. Both consequences of humble leadership are tested across three separate studies, and each finds support for the notion that management humility is related to higher levels of employee voice, fewer job obstacles and better performance. 3. Constructs and Measurements There are several constructs employed across the three studies. All items used for each construct can be found in Appendix 2, but example items are listed here. Manager humility is defined as stated previously, and is measured using a 9-item scale developed by Owens et al. (2013). Example items are: This supervisor is open to ideas from 9 P a g e

10 others ; This supervisor admits it when they don t know how to do something and This supervisor acknowledges when others have more knowledge and skills than themselves. Employee Voice is measured using a mix of 3 items based on the work fromvan Dyne and LePine (1998), Detert and Burris (2007) and Van Dyne et al. (2003). Specifically, the focus here is on prosocial voice (Van Dyne et al., 2003) which specifically tries to capture expressing solutions to problems so that others in the group or organization can benefit [and] includes creative suggestion of alternatives and ideas for change (pg. 1371).The items used in these studies are as follows: I give suggestions to this supervisor about how to make our store and/or company better ; I develop and make recommendations to this supervisor that could improve the way we do things and I develop and make recommendations to this supervisor that could improve customer satisfaction. It is expected that more humble leaders experience greater levels of employee voice. Job Obstacles is a construct measuring efficiency in the workplace by measuring the extent to which things get in the way of operations or processes running smoothly. This subjective measure of efficiency is captured by responses to the following three items: Things get in the way of me doing my job well ; Things get in the way of me offering good service to my customers and Things get in the way of our customers being easily satisfied. (Responses range from Never to Always. ) Leadership humility is believed to be negatively related to job obstacles because humble leaders are more open to new ideas that may improve the way things are done. General job satisfaction is also included in the studies since not controlling for it may bias the estimate of the effect of humility on job obstacles. Specifically, people who are more satisfied with their job may report both higher levels of manager humility and fewer obstacles. 10 P a g e

11 There is a potential single-source bias since the first two studies use the subjective measure of efficiency from the same source as the humility measure. This problem can be somewhat mitigated by including and controlling for job satisfaction, measured simply by one question asking: Overall, how satisfied are you with this job? This question is also combined with supervisor satisfaction to create a construct measuring both job and supervisor satisfaction. When showing the relationship between humility and the willingness to implement new ideas or continually improve the way things are done, a 5-item construct is used which includes not only the willingness to try new things, but also a willingness to try employee ideas. Example items include: My supervisor continually improves the way things are done ; My supervisor tries new things to improve customer satisfaction ; and My supervisor uses suggestions from employees to improve how we do things. Factor analysis confirmed that the two constructs (the manager s rating of humility and the rating of continual improvement ) are in fact separate constructs. 1 Power Distance Orientation is also included and measures whether the employee believes that those in power should or should not be challenged. Three items capture this construct and include: It is better not to disagree with management decisions ; When my supervisor makes a decision with which I disagree I prefer to accept the decision rather than question it ; and I believe that it is not right to disagree with my boss. A higher score on power distance means that the employee believes that there is a larger distance between two power levels and that those in power should not be challenged. It has been found in past research that power distance beliefs are negatively related to employee voice (Botero & Van Dyne, 2009), meaning that those who don t believe they should disagree with or challenge their boss are less likely to voice ideas upward to a boss. 1 The results of the factor analyses are not provided, but are available upon request. 11 P a g e

12 4. Three Separate Studies and Their Results 4.1 Study #1: The first study includes a convenience sample of student employees, reddit.com internet users, and Facebook contacts. Requirements were simply that the respondents were employed (or had been employed in the past 6 months) and that they have (or had) at least one supervisor. Data were collected related to the respondents perceptions of their supervisors humility and their own voice behaviors to test whether employee voice is related to perceptions of the supervisor s humility. Additionally, job obstacles is used to act as somewhat of a subjective measure for performance or efficiency in the workplace. Questions related to perceptions of the manager s willingness to continually improve the way things are done and try new things (including employees ideas) were also included. A simple linear regression will first test whether employee voice relates to perceptions of the manager s humility. The results are shown in the first column of Table 3 2 in Appendix 2. INSERT STUDY 1 TABLE 3 HERE As expected, perceived leadership humility is positively related to employees reported voice behaviors. The variables are first standardized, meaning that a one standard deviation increase in manager humility ratings increases voice by 0.41 standard deviations and the effect is significant at the 1% level. The remaining columns in the table show the results after adding in controls, showing that humility ratings continue to be positively and significantly related to reported employee voice behaviors. Column (2) adds in job and supervisor satisfaction, showing that it s not simply that an employee likes his or her boss or her job that makes it more likely that they will voice ideas upward, although the effect size of the humility rating definitely decreases. Column (3) shows 2 Table 1 contains summary statistics and Table 2 is the correlation matrix of all variables analyzed in the study. 12 P a g e

13 that a manager s ability to make changes also influences the level of employee voice, but humble leadership continues to be positively and significantly associated with employee voice. The last column controls for power distance beliefs and provides support for past studies that have shown that higher power distance beliefs are associated with a smaller likelihood of employees voicing ideas to their manager. This study shows that employees are more likely to report voicing ideas on how to improve things upward to the managers when they perceive those managers to be humble, meaning these types of managers are likely to have more ideas for improvement from which to choose. However, are these types of managers also more likely to implement ideas to make the improvements or changes that could ultimately impact overall performance or efficiency? This is tested using the continual improvement score along with manager humility. The results are shown in Table 4. INSERT STUDY 1 TABLE 4 HERE The simple linear regression shows the strong positive relationship between the manager s perceived humility and the same manager s willingness to take actions and make changes to improve the way things are done. This means humble leaders may not just listen to ideas that may improve performance, but they may also be more willing to try new ideas. Of course, since these manager ratings are all from the viewpoint of the employee, it may make sense to control for employee job satisfaction, since employees with high job satisfaction may report high marks about their managers for both scales. These results are shown in column 2 of the same table, and a regression adding in the manager s ability to make changes is shown in column 3. The estimate on manager humility decreases when adding in job and supervisor satisfaction, but remains strongly significant. A one standard deviation increase in manager 13 P a g e

14 humility increases his or her rating for continually improving the way things are done by 0.42 standard deviations. Job and supervisor satisfaction also has a positive association with the continual improvement rating (as can be expected), but with this data there is no way to tell the direction of the relationship. It seems likely that working under a manager that continually improves things would lead to greater levels of job satisfaction. The reverse causality could be problematic in terms of biasing the estimate on humility, a problem I attempt to rectify in the third study when the manager s own ratings of continual improvement are available. A manager s ability to make changes is positively and significantly associated with continual improvement, which makes perfect sense- holding constant manager humility (which measures the openness to ideas), a manager who has a greater ability to make changes is perceived as making more changes. Alternatively, since all of these ratings come from the employee, it could be that when an employee observes managers making changes, they also observe that the manager is able to make changes, leading to the positive association. Either way, the association between manager humility ratings and continual improvement remains positive and significant, and actually isn t affected by the inclusion of the variable measuring the manager s ability to make changes. If managers who are humble are more likely to receive, listen to and implement new ideas (even those of lower-level employees), then these managers should also experience greater levels of efficiency and fewer job obstacles in the workplace. Job obstacles are things that get in the way of employees doing their job well, and it is expected that humble leaders who are continually improving the way things are done would reduce the number of job obstacles that exist. This relationship is tested and the results are shown in Table 5. INSERT STUDY 1 TABLE 5 HERE 14 P a g e

15 There is a significant and negative relationship between leadership humility and the reported number of job obstacles, yet this relationship is insignificant when controlling for job and supervisor satisfaction. However, since we also see evidence that humble leadership is associated with a greater willingness to continually improve things, we can test whether it is this attribute of humble leaders that has an impact on the number of reported job obstacles. The results of this regression are shown in Table 6, and the fourth column shows that once continual improvement is included, manager humility adds no significant explanation to the amount of reported job obstacles. INSERT STUDY 1 TABLE 6 HERE The effect size on continual improvement drops by almost half when including job and supervisor satisfaction, but almost remains significant at the 1% level. It makes sense that the relationship between continual improvement and job obstacles would be stronger and more robust than the relationship between manager humility and job obstacles. Manager humility captures the openness of the manager to ideas and advice, their willingness to appreciate the contributions of others and their ability to view themselves and their ideas more objectively. Continual improvement on the other hand, while strongly related to leadership humility, is taking actions on ideas to improve the workplace and customer satisfaction. When adding in the manager s ability to make changes, the direction of the relationship stays the same, but is no longer significant at the 5% level. Adding in humility shows that it offers no significant effect on job obstacles, possibly supporting the idea that it s the implementation of ideas (captured by continual improvement) that s associated with fewer job obstacles and not the openness to hearing ideas (which is captured by humility ratings). 15 P a g e

16 These results from the first study show that there may in fact be a relationship between manager humility and employee voice, where greater levels of manager humility may lead to higher levels of employee voice (and therefore more ideas flowing around about how to improve the way things are done). We also see some support for the notion that management humility impacts more than just the number of ideas flowing around; it may be associated with a greater willingness to take actions on new ideas, which may then lead to a more efficiently running workplace (as measured here by fewer job obstacles). In the face of constantly changing market conditions, the willingness to adapt and make changes may be the factor that allows an organization to remain competitive and survive. If so, the type of leadership within an organization could help determine the amount of continual improvement that actually occurs. Study #2 tests the same relationships but uses a different methodological approach to extract a more causal argument relating manager humility to employee voice and to provide further support for the other relationships tested above. 4.2 Study #2: The first study supported the idea that humble leadership is associated with greater levels of employee voice and continual improvement, which was then shown to be associated with fewer job obstacles. The second study uses the same variables as the first, but a different estimation strategy to try measure a causal effect of manager humility on employee voice and continual improvement. This study uses a different data set than that used in the first study, although obtained similarly from different students, reddit.com internet users and employees of two quick-service food restaurants. In this study, employees were asked to respond about two different managers with which they work directly, the highest level manager (here, called top manager ) and any other 16 P a g e

17 manager they choose (here, called other manager ). Using a fixed effects strategy, we can hold the person/employee constant to measure whether this same employee is more or less likely to voice ideas to their two different managers based on his or her ratings of the two different managers level of humility. The results are shown in Table 2 in Appendix 3. INSERT STUDY 2 TABLE 2 HERE This fixed effects estimation strategy shows us that while holding the employee constant, differences in the employee s perceptions of their managers humility lead to differences in that employee s reported voice behaviors toward the managers 3, implying a causal link that wasn t testable with the data used in Study #1. The rating of the manager s humility influences the level of voice toward that manager- the higher the humility rating, the more likely the employee is to voice improvement ideas to that manager. This could lead to more ideas from all levels, which, if then implemented, could improve both efficiency within the workplace and customer satisfaction (since these are the types of ideas being voiced that employee voice is measuring). However, we also see in column (2) that when controlling for supervisor satisfaction, the effect of humility is only significant at the 10% level (even though supervisor satisfaction is insignificant). This may imply that humility captures a little bit of the effect of the employee s satisfaction with the supervisor on his or her willingness to voice ideas when the impact of the manager s humility is tested alone (a positive bias caused by omitting the variable). Unfortunately, job satisfaction can t be tested in this analysis because job satisfaction doesn t vary within each employee observation. As was done in Study #1, we can also test whether differences in humility ratings are associated with differences in continual improvement ratings using the same fixed effects 3 Fixed effects estimation uses deviations from the average to estimate effects. Specifically, deviations from the average Y are regressed against deviations from the average X. Whenever there are only two observations per person, it s a regression of the difference in the two Y values on the difference in the two X values. 17 P a g e

18 strategy used above. The results are shown in Table 3, and additional variables are included in a hierarchical fashion. INSERT STUDY 2 TABLE 3 HERE We see similar results to the first study: humility ratings are positively associated with continual improvement ratings, although the effect size drops by almost half when including supervisor satisfaction. The coefficient remains significant at the 1% level across all regressions. It s important to note that we cannot necessarily infer a causal link between humility and continual improvement the way we could between manager humility and employee voice. The reason for this is that here we are still holding the employee constant, meaning the results can only be interpreted as whether differences in the employee s ratings of his or her two managers humility cause differences in the same employee s ratings of those two managers levels of continual improvement. Put another way, both ratings are about two different managers from the same employee, so what we see is that employees who perceive their manager as more humble also rate those managers higher on the continual improvement scale. For the humility-voice relationship, the results can instead be interpreted as differences in the employee s ratings of his or her two managers humility cause differences in the same employee s willingness to voice ideas. Since we hold the person whose behavior we are measuring constant, it means that an employee isn t equally likely to voice ideas to their managers no matter how they perceive the managers. Instead, the perception of the manager s humility has an influence on the employee s voice behaviors. As mentioned previously, the fixed effects approach doesn t allow us to test whether these results change when controlling for job satisfaction (as was done in Study #1), since job satisfaction takes the same value within each person. Neither can we used fixed effects to test the 18 P a g e

19 impacts of humble leadership on job obstacles, as job obstacles relate to the job (not the manager) so the value doesn t vary across the employee observation. Nonetheless, this data set can help provide further support for the relationships found in the first study if we go back to simply using linear regressions. Each employee responded about his or her highest-level manager under which they work, as well as another, lower-level manager. We can therefore again test the relationship between manager humility ratings and employee voice (for each of the two managers) by running separate regressions for each of the manager types. The results for the two simple linear regressions are provided in Table 4. We see that perceived humility still has the same positive effect on employee voice that was found in the first study, although the effect becomes significant only at the 5% level for the top-level manager once job and supervisor satisfaction are included. Additionally, as a robustness check, we can test whether the perceived humility of Manager B affects an employee s reported voice toward Manager A. (There should be no effect.) The results are shown in Table 5. INSERT STUDY 2 TABLE 5 HERE We see that voice toward the top manager is unaffected by the perceived humility of another (lower level) manager (as it shouldn t be). However, we also see an interesting result coming out of regressing voice toward a lower manager on the perceived humility of the top manager (column 2). It seems as though voice toward a lower manager may be affected by the humility of the top manager! This makes sense if we consider the possibility that a top manager creates an environment or culture where ideas are accepted from all levels. Of course, it s also possible that top managers have hiring discretion for their lower managers and may choose to hire managers with humility levels they themselves have. When controlling for and including the 19 P a g e

20 humility of the lower manager in this regression, the effect of the top manager s humility rating becomes non-significant, but not as completely insignificant as the effect of other manager s humility on voice toward the top manager (suggesting perhaps maybe the culture argument is still plausible). (See the results in the final column of Table 5.) We can now also test the effects of leadership humility on the perceived level of job obstacles present in the workplace. Here we see results similar to those found in Study #1. Manager humility ratings are negatively and significantly related to reported job obstacles. After controlling for job satisfaction, we continue to see negative and significant results- greater toplevel manager humility is related to fewer reported job obstacles. The results are provided in Table 6. INSERT STUDY 2 TABLE 6 HERE We see similar results related to the lower level manager humility ratings as well (as shown in Table 7). Table 8 shows that the results match Study #1 when controlling for supervisor satisfaction. INSERT STUDY 2 TABLE 7 HERE INSERT STUDY 2 TABLE 8 HERE The effect of the top manager s humility on reported job obstacles becomes completely insignificant, while the effect of lower-level manager s humility only barely becomes insignificant at the 10% level. The difference may be due to the lower-level manager being more closely involved with the day-to-day processes about which the employees report obstacles but this cannot be tested with this data. 20 P a g e

21 Additionally, just as we did in the Study #1, we can test the relationship between the managers actually making changes and the effect of the continual improvement score on reported job obstacles. The results are included in Table 9, and are also shown to match the results found from Study #1. INSERT STUDY 2 TABLE 9 HERE One interesting result is that including job and supervisor satisfaction has less of an impact on the lower-level manager s continual improvement coefficient. (It remains significant at the 1% level, where the top manager s continual improvement rating becomes significant only at the 5% level.) This again could perhaps be due to the fact that lower-level managers may actually be more closely related to the job obstacles that employees experience directly, whereas the top-level manager may be consumed with more broad level decisions. However, as stated above, more data and future research would have to test whether this is the case. Of course, job obstacles are still being reported by the same source that is rating manager humility and job satisfaction, potentially leading to a single-source bias. (Since it s still possible that the employee that gives positive ratings about their supervisor is more likely to say fewer obstacles exist in the workplace.) The only way to get around this problem is to use a more objective measure of performance from a separate source. This is the purpose of the third and final study. (Note: this third study is still very much in its exploratory phase- more observations are currently being sought out.) 4.3 Study #3 Data were collected from employees and managers of 2 quick-service restaurant chains in the Southeastern region of the United States. The humility and voice ratings were averaged across the employees within each unit to create a unit-level measure, since performance is measured using monthly sales in dollars at the unit level. Brand 1 has 8 different units and 21 P a g e

22 Brand 2 has 11. Using this data, we can test whether humble leadership has an impact on sales performance. The results are shown in Table 3 in Appendix 4. INSERT STUDY 3 TABLE 3 HERE We see that humble leadership, as perceived by employees, has a positive and significant effect on monthly sales. The natural log of sales is used as the dependent variable, so the estimate on the average humility rating is interpreted as follows: a one point increase in the average humility rating is associated with 28.3% higher sales. (Note that brand must be included as a control due the differences in the levels of sales for each brand. The coefficient on humility can then be thought to be measuring the effect of humility on differences in sales within each brand.) The positive and significant result remains even after controlling for past sales (but the effect size decreases). This is shown in column 2 of Table 3. We lose a few observations because some of the units didn t exist the year prior. Nonetheless, the effect of humility ratings continues to be positive and significant, and a onepoint increase in the average humility rating is associated with 23.3% higher sales. Regressing instead on Sales only two months prior gives us back the lost observations, and in Table 4 we still see a positive effect of manager humility on sales. INSERT STUDY 3 TABLE 4 HERE The relationship is almost one-to-one when using sales from only two months before- a one percentage increase in sales from May 2014 is associated with 95% higher sales in July However, humble leadership still seems to have an effect: a one-point increase in the average manager humility rating is associated with 4.6% higher sales. 22 P a g e

23 We can also test the relationship between the average level of employee voice and the average rating of the unit manager s humility, just as we did in both of the first studies. The results are shown in Table 5. INSERT STUDY 3 TABLE 5 HERE We see the positive relationship that we expect, but it s insignificant at the 5% level. The p-values are reported to provide more information since the sample is so small, and because it shows that the significance level is just above 5%. When we include past sales, we see that worse sales in the past lead to greater levels of improvement ideas being voiced. However, with such a small sample, the results are insignificant at the 10% level. INSERT STUDY 3 TABLE 6 HERE Again, the p-values are reported in parentheses. Using simple linear regression shows the coefficients are not significant at the 10% level. However, using a seemingly unrelated system of equations and estimating both the employee voice and the sales equations simultaneously achieves efficiency by exploiting the information contained within both equations and using the fact that the errors in both equations are likely correlated (Wooldridge, 2002). This yields the results found in Table 7. INSERT STUDY 3 TABLE 7 HERE Exploiting the information contained in both equations increases the efficiency of the estimates, and they all become significant at the 5% level (with estimates in the sales equation significant at the 1% level). Additionally, we still see that a one-point increase in the average humility rating increases the average level of employee voice by 0.49 points. Past sales are included in different ways to allow for different functional forms and different interpretations. Using the natural log of 23 P a g e

24 past sales makes sense in the sales equation- a one percent increase in past sales is associated with a 0.31% increase in sales this year. However, it wouldn t make as much sense to use the natural log of past sales in the employee voice equation. Instead, the standardized values of sales are used to simplify the interpretation- an increase in sales by one standard deviation leads to a decrease of approximately 0.27 standard deviations in the level of employee voice. Since the manager s average humility ratings by the employees may have embedded within them other factors that could potentially impact performance, another test using the manager s self-rating can be used to provide further (and possibly more robust) support for the idea that a willingness to take actions on ideas to improve the workplace actually has an impact on performance. Specifically, the managers were asked whether they use employee ideas to make changes and improve the way things are done. Using the manager s self-rating of using and implementing employee ideas gets away from the potential issues of using employee ratings of manager humility since the employee s ratings of their manager s humility may also include other factors that could impact sales. The results from this regression are shown in Table 8. INSERT STUDY 3 TABLE 8 HERE The results show that a manager s own rating is also positively associated with sales; a one point increase in the manager s own report of using employee ideas is associated with 17% higher sales, even after controlling for past sales. This is also interesting because it shows that managers with higher sales are more likely to report that they use employee ideas to improve the way things are done, crediting the employees with part of the success. Although this third study is only exploratory due to the current small sample size, all three studies lend support to the idea that humble leadership is associated with more improvement ideas being offered from the employees. Additionally, we see some evidence that 24 P a g e

25 humble leadership is associated with greater performance in terms of sales and fewer job obstacles as perceived by the employees. This could be due to the fact that humble leaders are not only more likely to listen to ideas (by definition), but may also be more likely to implement new ideas and try new things to increase performance. The triangulation of the three different studies allows for a more conclusive and robust argument for all of the relationships tested here, but more research always needs to be done. The limitations and possible future work are outlined in the subsequent section. 5. Conclusion, Limitations and Future Research Possibilities First, I tried to establish that a relationship exists between employee voice and manager humility. I then tested it against holding constant all time-invariant (or personal trait) characteristics to show that it is in fact the perception of the manager s humility that affects a person s decision to voice ideas upward. I also showed that average voice ratings within a store are associated with average ratings of that store manager s humility. I also tested whether humble leadership is associated with stronger performance. In the first study, I had only a subjective measure of efficiency which I called job obstacles. The data showed that at the very least, employees who perceive their managers as humble also perceive fewer job obstacles (which may be beneficial to an organization in itself). Study #3 supports this finding by showing that humble leadership (still as perceived by the employees) has an impact on monthly sales in dollars, even after controlling for past sales. Further support for the idea that managers that listen to and implement new ideas to make improvements can be shown by using the manager s own self-report of using employee voice, which is positively and significantly related to sales. This test mitigates the potential issues of employees rating their managers higher on the humility scale for other reasons that may impact sales. Using two different sources of data (the objective sales measure and the manager s own self-rating) lends 25 P a g e

26 additional support to the notion that managers openness to trying new ideas improves performance. While the results may help us argue that managers can impact performance in ways that economic models don t generally measure, more research is needed to confirm the results. Additionally, future research using other methods may be necessary to test more causal inferences. More data is needed for Study #3 to see if results remain with a larger sample, and more experimental data can help us better understand if humility can be learned, and if so, if it in fact leads to better performance. Other limitations of this research include the potential samesource bias caused by gathering all the data from the employees in Study #1 and #2. This could be rectified in future studies by attempting to gather addition data from multiple sources. This paper may just be a piece of the very first steps trying to understand what humble leadership entails and whether or not it may positively impact performance. Although there is some support found here, it s not entirely clear how or why. I posit that it is through the openness to others ideas and through trying new things to improve performance, but it s also possible that humble leaders are more willing to empower their employees, which could positively affect performance. Future research should try to more fully investigate the mechanisms through which this type of leadership affects organizational outcomes like performance. Future research should also test whether managers at different levels have varying impacts on performance. Apriori it may seem as though higher-level managers may have greater impacts on performance, but it also makes sense that lower-level managers closer to the employees processes and perhaps even closer to the customers could have a larger impact on some of the day-to-day things that impact performance and efficiency in the workplace. 26 P a g e