In their informative article, Speer, Dutta, Chen, and Trussell (Reference Speer, Dutta, Chen and Trussell2019) have created a resource for predicting turnover by using available data to help organizations know if employees are “here to stay or go.” In an article with colleagues, we explored the other side of this coin—how employees decide, “Should I stay or should I go?” (Rothausen, Henderson, Arnold, & Malshe, Reference Rothausen, Henderson, Arnold and Malshe2017). We have two broad cautions for consultants in response to Speer et al. from our findings in this article and related work (Rothausen & Henderson, Reference Rothausen and Henderson2019). First, the “detrimental effects to company productivity, financial performance, … and morale …” referred to by Speer et al. (Reference Speer, Dutta, Chen and Trussell2019, p. 272) are not only, or perhaps even primarily, caused by losing employees but rather by the impact of the job, as delivered by the organization, on employees’ lives. Our data show that the organizational treatment of employees not only leads some to leave but also impacts “stayers” directly rather than primarily through their colleagues’ leaving. Second, though quantitative data are good at answering certain questions, there are broader questions about turnover quantitative data cannot answer, of which consultants and researchers should not lose sight.
In our study, we took a different direction from the dominant turnover models reviewed by Speer et al. (Reference Speer, Dutta, Chen and Trussell2019), in order to broaden attrition modeling, because reviews indicate that researchers are not content with the predictive power of current models of turnover (e.g., Hom, Mitchell, Lee, & Griffeth, Reference Hom, Mitchell, Lee and Griffeth2012). The bounds of the extant models Speer et al. review may be due in part to over-reliance on limited frameworks, types of data, and methodologies in turnover research (Russell, Reference Russell2013). There have been calls in the literature to expand understanding of turnover by moving beyond reductionist measures from employees, for example by exploring leavers’ self-reports and in-depth exit interviews (Bergman, Payne, & Boswell, Reference Bergman, Payne and Boswell2012; Maertz, Reference Maertz2012), which is what we did. We began by interviewing leavers across four industries and later added leavers from other industries as well. Our findings demonstrate that the process of turnover is more complex, and goes deeper, than extant models suggest. Our findings also demonstrate that the process of turnover begins well before the timeframes common in extant research. Once we learned this, we also interviewed “stayers” to make sure the process we uncovered was actually happening pre-leaving and was not post-hoc rationalizing. Our findings suggest that how current employees are—and how former employees were—treated by the organization has the potential to cost the organization much more than the cost of replacing and training new employees.
Another way of saying this is that we offer an alternative attrition model based on what employees told us about why they leave or stay in organizations, that is not limited to what happens inside the organizational walls but carries over into the impact of these experiences on people’s whole lives. In line with others’ suggestions (e.g., Picoult, Reference Picoult2010), we show that frustrated desire to turnover may have even more deleterious effects on organizations and individuals than does turnover itself and should be of perhaps even more concern to organizations. As Bergman et al. (Reference Bergman, Payne and Boswell2012) predicted, we found that, “we would be better off predicting turnover intentions rather than turnover behavior. … [O]rganizations have a better chance of changing behavior if they can intervene before the intentions manifest themselves. And, [turnover intentions have] implications for other workplace behavior” (Bergman et al., Reference Bergman, Payne and Boswell2012, p. 867).
To illustrate this, we briefly review the story with which we opened our article. On Monday, August 9, 2010, JetBlue flight attendant Steven Slater reacted strongly to a belligerent customer during the deplaning process. This passenger started to retrieve her luggage from an overhead bin before it was safe to do so, and when Mr. Slater asked her to desist, she cursed him.
He said, “I’ve had it! To the passenger who called me a mother******, **** you! I’ve been in this business for 28 years and that’s it. I’m done. … [He then] grabbed his bags—and two cans of beer from the galley—and popped the lever for the plane’s inflatable emergency chute. He threw the bags on to it before sliding down to the tarmac himself. (Gardner, Reference Gardner2010)
This widely reported incident of voluntary turnover is not typical, and yet it highlights elements of turnover that may be more impactful for employers than the turnover event itself. As we also note in our article, the widespread response to the memorable exit of Mr. Slater from JetBlue illustrates this potential impact and is perhaps more alarming than the turnover incident itself, which could be viewed as idiosyncratic. Later in the month he quit: “Slater [became] an unlikely folk hero in the U.S. … a string of pages had been set up in tribute to him on the Facebook website, with many social networkers admiring his grand exit” (Carey, Reference Carey2010).
The glee with which other workers embraced Mr. Slater’s spectacular exit suggests that quite short of turnover, there is much to be gained from understanding why people stay or go, attitudes toward turnover that exist in them prior to actual turnover, and conversely, why people remain in organizations.
Given that one impetus for Speer et al. (2018) is mining the data that are available in order to manage the impact of turnover on organizations (as well as sometimes to collect additional data), our findings suggest that another way to measure the impact of turnover is to explore not only pre-turnover withdrawal behaviors but also what former employees say about the company to potential customers and to other workers. We found that employees identify deeply and often passionately with their work roles and organizations, for the roles themselves, and also because of the broad impact jobs have on lives and on workers’ sense of purpose, trajectory in life, relatedness, expression, acceptance, and differentiation (PTREAD) both at work and outside of work. Jobs in which these values are supported are appreciated, and jobs in which these values are threatened are not.
The elements of PTREAD suggest that in exploring and predicting the impact of turnover on organizations, we should also include measuring the impact former employees may have on company productivity and financial performance, as well as the morale of existing employees. Especially in today’s world of social media that Speer et al. (Reference Speer, Dutta, Chen and Trussell2019) reference frequently, data on how former employees talk about the company are likely available from similar sources that Speer et al. mention.
In sum, the costs of turnover are not just the costs of rehiring and getting new employees to the same productivity levels as former employees, but also the result of the number of former employees a company has and how they feel when they leave, as well as how current employees feel for months or years before they leave. Similarly, the opportunity in turnover is not only getting rid of poor performers but the opportunity to have people leave feeling good about the way they were treated, as well as about the organization’s products or services. If people feel justly treated, they tend to treat the organization justly (Greenberg, Reference Greenberg and Cropanzano1993). Negative reviews on social media sites in general, as well as on sites related to job searches, can quash sales and pools of potential recruits; positive reviews can bring in more business and increase labor pools (Etter, Ravasi, & Colleoni, Reference Etter, Ravasi and Colleoni2019; McFarland & Ployhart, Reference McFarland and Ployhart2015)
Second, though Speer et al. (Reference Speer, Dutta, Chen and Trussell2019) make no claim that attrition modeling as they define it is the only important process to follow when consultants are trying to study turnover in organizations, nor do they suggest vital complementary practices. In addition to collecting quantitative data and “mining” for more numbers, qualitative data in the form of people’s stories and experiences in the employment and turnover process can be a rich resource for understanding the meaning of the numbers and the interrelatedness of many factors. Data such as leavers’ stories, which we collected for our study, and other narrative data, provide not only further explanations of turnover but flesh out the picture of what the human community within an organization is like. This provides rich and thick data suggesting potentially impactful changes the organization could make to both prevent unwanted turnover and to impact the engagement of current employees. The meaning of the numbers is what really counts to those turning over or staying (Rothausen & Henderson, Reference Rothausen and Henderson2019).
Foreseeing the impact of the direction science was taking toward an overfocus on quantitative measurement (which the first author has also written about; see Rothausen, Reference Rothausen2016), Ralph Waldo Emerson noted that when one lacks a deeply felt sympathy with one’s subject, for example an ornithologist’s work results in a dead bird measured in ounces and inches. A full appreciation of the beauty and wholeness of the bird alive in all of its interrelations in nature is absent when we take this approach. As Emerson said: “The skin or skeleton you show me, is no more a heron, than a heap of ashes or a bottle of ashes into which his body has been reduced, is Dante [Alighieri] or [George] Washington.” (Reference Emerson1860, p. 1099; parenthetical material added for clarity).
We caution that a set of statistically significant predictors of the binary “stay or go,” although useful, really does no more to explain the richness of reasons why people stay or go than that bottle of ashes explains any person.
There is a shibboleth right now in management that only what can be measured gets managed. This is, of course, not even close to the truth. Many things that cannot—easily anyway—get measured are managed every day and make a difference to human connections and identity ties to work organizations. It does seem true that our attention is drawn to the measureable, often in dysfunctional ways (e.g., Kerr, Reference Kerr1975). Based on our findings, for example, human connections at work are vitally important and are only partially captured by embeddedness. Similarly, the dignity with which employees have been treated both while at the organization and as they are leaving impact their memories of it, how they feel and think about it thereafter, and what they tell others about it.
It is in the significance to people of the immeasurable that we find the meaning of the measurements. For example, in our study we found that the meaning of the dichotomous variable “leaving/not leaving” had to do with agonizing trade-offs related to the impact of jobs on people’s sense of PTREAD—purpose, trajectory, relatedness, and self-expression, -acceptance, and -differentiation. To suggest that important findings in turnover research should all be made measurable implies that prediction is what is most important about turnover. Although it is important, of course, we believe the impact of organizations on their current and former employees’ lives in both measurable and immeasurable ways is, at the end of our day, what makes work life meaningful and is a rich source of ideas and “strategies” for managing retention and turnover.
You have
Access
In their informative article, Speer, Dutta, Chen, and Trussell (Reference Speer, Dutta, Chen and Trussell2019) have created a resource for predicting turnover by using available data to help organizations know if employees are “here to stay or go.” In an article with colleagues, we explored the other side of this coin—how employees decide, “Should I stay or should I go?” (Rothausen, Henderson, Arnold, & Malshe, Reference Rothausen, Henderson, Arnold and Malshe2017). We have two broad cautions for consultants in response to Speer et al. from our findings in this article and related work (Rothausen & Henderson, Reference Rothausen and Henderson2019). First, the “detrimental effects to company productivity, financial performance, … and morale …” referred to by Speer et al. (Reference Speer, Dutta, Chen and Trussell2019, p. 272) are not only, or perhaps even primarily, caused by losing employees but rather by the impact of the job, as delivered by the organization, on employees’ lives. Our data show that the organizational treatment of employees not only leads some to leave but also impacts “stayers” directly rather than primarily through their colleagues’ leaving. Second, though quantitative data are good at answering certain questions, there are broader questions about turnover quantitative data cannot answer, of which consultants and researchers should not lose sight.
In our study, we took a different direction from the dominant turnover models reviewed by Speer et al. (Reference Speer, Dutta, Chen and Trussell2019), in order to broaden attrition modeling, because reviews indicate that researchers are not content with the predictive power of current models of turnover (e.g., Hom, Mitchell, Lee, & Griffeth, Reference Hom, Mitchell, Lee and Griffeth2012). The bounds of the extant models Speer et al. review may be due in part to over-reliance on limited frameworks, types of data, and methodologies in turnover research (Russell, Reference Russell2013). There have been calls in the literature to expand understanding of turnover by moving beyond reductionist measures from employees, for example by exploring leavers’ self-reports and in-depth exit interviews (Bergman, Payne, & Boswell, Reference Bergman, Payne and Boswell2012; Maertz, Reference Maertz2012), which is what we did. We began by interviewing leavers across four industries and later added leavers from other industries as well. Our findings demonstrate that the process of turnover is more complex, and goes deeper, than extant models suggest. Our findings also demonstrate that the process of turnover begins well before the timeframes common in extant research. Once we learned this, we also interviewed “stayers” to make sure the process we uncovered was actually happening pre-leaving and was not post-hoc rationalizing. Our findings suggest that how current employees are—and how former employees were—treated by the organization has the potential to cost the organization much more than the cost of replacing and training new employees.
Another way of saying this is that we offer an alternative attrition model based on what employees told us about why they leave or stay in organizations, that is not limited to what happens inside the organizational walls but carries over into the impact of these experiences on people’s whole lives. In line with others’ suggestions (e.g., Picoult, Reference Picoult2010), we show that frustrated desire to turnover may have even more deleterious effects on organizations and individuals than does turnover itself and should be of perhaps even more concern to organizations. As Bergman et al. (Reference Bergman, Payne and Boswell2012) predicted, we found that, “we would be better off predicting turnover intentions rather than turnover behavior. … [O]rganizations have a better chance of changing behavior if they can intervene before the intentions manifest themselves. And, [turnover intentions have] implications for other workplace behavior” (Bergman et al., Reference Bergman, Payne and Boswell2012, p. 867).
To illustrate this, we briefly review the story with which we opened our article. On Monday, August 9, 2010, JetBlue flight attendant Steven Slater reacted strongly to a belligerent customer during the deplaning process. This passenger started to retrieve her luggage from an overhead bin before it was safe to do so, and when Mr. Slater asked her to desist, she cursed him.
He said, “I’ve had it! To the passenger who called me a mother******, **** you! I’ve been in this business for 28 years and that’s it. I’m done. … [He then] grabbed his bags—and two cans of beer from the galley—and popped the lever for the plane’s inflatable emergency chute. He threw the bags on to it before sliding down to the tarmac himself. (Gardner, Reference Gardner2010)
This widely reported incident of voluntary turnover is not typical, and yet it highlights elements of turnover that may be more impactful for employers than the turnover event itself. As we also note in our article, the widespread response to the memorable exit of Mr. Slater from JetBlue illustrates this potential impact and is perhaps more alarming than the turnover incident itself, which could be viewed as idiosyncratic. Later in the month he quit: “Slater [became] an unlikely folk hero in the U.S. … a string of pages had been set up in tribute to him on the Facebook website, with many social networkers admiring his grand exit” (Carey, Reference Carey2010).
The glee with which other workers embraced Mr. Slater’s spectacular exit suggests that quite short of turnover, there is much to be gained from understanding why people stay or go, attitudes toward turnover that exist in them prior to actual turnover, and conversely, why people remain in organizations.
Given that one impetus for Speer et al. (2018) is mining the data that are available in order to manage the impact of turnover on organizations (as well as sometimes to collect additional data), our findings suggest that another way to measure the impact of turnover is to explore not only pre-turnover withdrawal behaviors but also what former employees say about the company to potential customers and to other workers. We found that employees identify deeply and often passionately with their work roles and organizations, for the roles themselves, and also because of the broad impact jobs have on lives and on workers’ sense of purpose, trajectory in life, relatedness, expression, acceptance, and differentiation (PTREAD) both at work and outside of work. Jobs in which these values are supported are appreciated, and jobs in which these values are threatened are not.
The elements of PTREAD suggest that in exploring and predicting the impact of turnover on organizations, we should also include measuring the impact former employees may have on company productivity and financial performance, as well as the morale of existing employees. Especially in today’s world of social media that Speer et al. (Reference Speer, Dutta, Chen and Trussell2019) reference frequently, data on how former employees talk about the company are likely available from similar sources that Speer et al. mention.
In sum, the costs of turnover are not just the costs of rehiring and getting new employees to the same productivity levels as former employees, but also the result of the number of former employees a company has and how they feel when they leave, as well as how current employees feel for months or years before they leave. Similarly, the opportunity in turnover is not only getting rid of poor performers but the opportunity to have people leave feeling good about the way they were treated, as well as about the organization’s products or services. If people feel justly treated, they tend to treat the organization justly (Greenberg, Reference Greenberg and Cropanzano1993). Negative reviews on social media sites in general, as well as on sites related to job searches, can quash sales and pools of potential recruits; positive reviews can bring in more business and increase labor pools (Etter, Ravasi, & Colleoni, Reference Etter, Ravasi and Colleoni2019; McFarland & Ployhart, Reference McFarland and Ployhart2015)
Second, though Speer et al. (Reference Speer, Dutta, Chen and Trussell2019) make no claim that attrition modeling as they define it is the only important process to follow when consultants are trying to study turnover in organizations, nor do they suggest vital complementary practices. In addition to collecting quantitative data and “mining” for more numbers, qualitative data in the form of people’s stories and experiences in the employment and turnover process can be a rich resource for understanding the meaning of the numbers and the interrelatedness of many factors. Data such as leavers’ stories, which we collected for our study, and other narrative data, provide not only further explanations of turnover but flesh out the picture of what the human community within an organization is like. This provides rich and thick data suggesting potentially impactful changes the organization could make to both prevent unwanted turnover and to impact the engagement of current employees. The meaning of the numbers is what really counts to those turning over or staying (Rothausen & Henderson, Reference Rothausen and Henderson2019).
Foreseeing the impact of the direction science was taking toward an overfocus on quantitative measurement (which the first author has also written about; see Rothausen, Reference Rothausen2016), Ralph Waldo Emerson noted that when one lacks a deeply felt sympathy with one’s subject, for example an ornithologist’s work results in a dead bird measured in ounces and inches. A full appreciation of the beauty and wholeness of the bird alive in all of its interrelations in nature is absent when we take this approach. As Emerson said: “The skin or skeleton you show me, is no more a heron, than a heap of ashes or a bottle of ashes into which his body has been reduced, is Dante [Alighieri] or [George] Washington.” (Reference Emerson1860, p. 1099; parenthetical material added for clarity).
We caution that a set of statistically significant predictors of the binary “stay or go,” although useful, really does no more to explain the richness of reasons why people stay or go than that bottle of ashes explains any person.
There is a shibboleth right now in management that only what can be measured gets managed. This is, of course, not even close to the truth. Many things that cannot—easily anyway—get measured are managed every day and make a difference to human connections and identity ties to work organizations. It does seem true that our attention is drawn to the measureable, often in dysfunctional ways (e.g., Kerr, Reference Kerr1975). Based on our findings, for example, human connections at work are vitally important and are only partially captured by embeddedness. Similarly, the dignity with which employees have been treated both while at the organization and as they are leaving impact their memories of it, how they feel and think about it thereafter, and what they tell others about it.
It is in the significance to people of the immeasurable that we find the meaning of the measurements. For example, in our study we found that the meaning of the dichotomous variable “leaving/not leaving” had to do with agonizing trade-offs related to the impact of jobs on people’s sense of PTREAD—purpose, trajectory, relatedness, and self-expression, -acceptance, and -differentiation. To suggest that important findings in turnover research should all be made measurable implies that prediction is what is most important about turnover. Although it is important, of course, we believe the impact of organizations on their current and former employees’ lives in both measurable and immeasurable ways is, at the end of our day, what makes work life meaningful and is a rich source of ideas and “strategies” for managing retention and turnover.