Friday, April 24, 2020

Data (Analytics) on COVID-19: Lessons for People Analytics

Data visualization, dashboards, and statistical modeling have been thrust into the spotlight because of the COVID-19 pandemic. I am not a biostatistician or an epidemiologist (not even an armchair one!) so I am not in a position to evaluate or criticize these visualizations and models. But I’m currently teaching a course on data and metrics for human resources, so there is an educational opportunity to consider lessons the spotlighted data (analytics) on COVID-19 might have for people analytics.

Let’s start with dashboards, which are common in people analytics. Here is a COVID-19 dashboard from Johns Hopkins:

It’s impressive in the amount of data it brings together and in the ability for the user to change views. You can certainly easily grasp the major metrics, numerically as well as graphically—which is the purpose of a dashboard, whether pertaining to COVID-19 or HR metrics such as employee headcounts. But as with all dashboards, there are at least two major questions. 
  1. Are these the right metrics for what you are trying to understand? It’s easy, for example, to find Twitter threads debating whether total deaths or deaths adjusted for country population is the better measure. But like many debates over metrics, rather than seeing this as a competition over which metric is better, it would be more productive to see various measures as complements that measure different aspects (e.g., total cases reflects the pace at which an outbreak is growing; per capita cases indicates strain on a health care system). 
  2. Are the data accurate or comparable, especially when collected from diverse sources? Do individuals within an organization have a self-interest to report data in certain ways? Or are there different capabilities that produce different measures. As Ryan Lamare reminds me, dashboards and data visualizations work best when there is a common baseline. Otherwise, users need to think carefully about what they're actually seeing and how they're interpreting it. In the COVID-19 case, for example, how should we interpret national comparisons of total tests when testing capacity differs? A similar example in HR might be a comparison of training numbers across units with different training capacities. 
  3. Beyond seeing the scope of a current situation, what actions can you take from metrics that are largely descriptive? This dashboard, for example, shows which areas have the most cases by COVID-19 but how do we act upon that information? An HR dashboard might reveal areas of an organization with low employee engagement, but is unlikely to help reveal why.  

Next, here is a visualization from John Burn-Murdoch of the Financial Times that has also frequently been spotlighted:

This is a great visualization for seeing trends within countries, and across them, too, if you carefully remember what's being compared. In a people analytics context, this could be seen almost as a scorecard to see how your organization stacks up against others, or how areas within your organization compare to each other. But there are at least three things to be cautious about. First, there are the same concerns as with a dashboard—are these the right measures, the right comparisons, is there a common baseline, etc. (in fact, the source of the data for this visualization is the John Hopkins dashboard data, so the same concerns apply). Second, the nature of visualization tempts you to forecast into the future. But what’s the basis for that forecast?

For example, let’s go back to the March 15 version of the same visualization:
Based on this visualization, we might have projected that the U.S. would look more like South Korea, and that Spain was on the worst trajectory of all. Unfortunately, Spain has indeed been hard hit, but it’s been exceeded by the United States in terms of total cases. Moreover, I think our minds are tempted to draw single lines that project out from each trend line. Even if these lines grasp the complicated curvature reflected in the trends to-date (so you do a complicated rather than simplistic projection), there is still a major problem. Namely, this ignores forecast error—instead, we should also be trying to ascertain how much variability and uncertainty there is in any forecast, including HR-related projections. More broadly, in making any statistical inference, we should understand whether the sampling error is large or small, and thus the magnitude of the margin of error and the soundness of concluding that there is a meaningful relationship or result.

A third caution for people analytics that we can take away from this visualization is a reminder that this metrics-focused approach doesn’t inquire as to what factors influence the trends portrayed. Note that is doesn’t claim to, so this isn’t a criticism per se. Rather, it’s a reminder that if you want to act upon information by, for example, implementing new HR initiatives, you should always be asking what’s influencing the metrics. What levers can you nudge that will change the metrics in the desired ways? Even if you can’t estimate an actual regression, it can be helpful to approach problems with that mindset—what variables would you like to include in a regression to explain the metric? In the absence of a regression, is there other evidence to support the importance of these factors? What’s missing from your (mental) model? 

Thinking about factors that influence a trend or a metric represents a shift from a metrics approach to more of a predictive analytics approach. In the COVID-19 pandemic, this is reflected in the importance of statistical models for policy-making—for example, using predictions from models for implementing stay-at-home orders. Let's consider two broad approaches.

One approach to modeling the spread of COVID-19 essentially tries to figure out the shape of the curves in the above visualizations by fitting statistical parameters to the curves that are the most complete (e.g., China, Italy). If you then assume that the lagging counties (or other geographical units) are on an earlier part of that same curve, then you can predict where those countries are headed. This is the approach of the Institute for Health Metrics and Evaluation (IHME):

Importantly, note the shaded area which reflects a 95% confidence interval. And note that it’s quite large for the immediate future. This is a good reminder for people analytics that estimates are just estimates. There is always uncertainty, and it’s important to understand the magnitude of that uncertainty before making decisions.

But note that this curve-fitting approach is akin to a data mining exercise. There is no epidemiological model that underlies these forecasts. In HR, this would be like observing the retirement ages of previous workers, and predicting a particular worker’s retirement probability based solely on their age. There’s no accounting for that person’s particular characteristics or changes in the environment particular to that person.

As an alternative modeling strategy, a long-standing epidemiological approach is the susceptible (S)-exposed (E)-infected (I)-resistant (R) model (SEIR, for short) (or alternatively, a SIR model with three classes: susceptible, infected, and recovered individuals). A SEIR model starts with the number of susceptible, exposed, infected, and resistant individuals, and then sets up a formulaic relationship across the categories based on estimates of incubation periods, frequency of contact across individuals, the probability of being infected after exposure, and the like. The spread of COVID-19, hospitalization usage, and other outcomes can then be simulated by projecting out what happens as exposure and infection increases. And by changing key parameters, you can also forecast alternative scenarios, such as the impact of various social distancing measures. This type of model is being used to guide public policy in Minnesota. 

An analogous people analytics example would be a workforce planning model where you start with the current number of employees and make assumptions about retention rates, mobility, hiring rates, and future needs. This creates forecasts into the future, and by changing different assumptions, you can model alternative scenarios, forecast shortfalls, and infer needed responses.

Note that there is expert judgement or past empirical trends built into this model—it’s not just curve fitting. And a realistic recognition of the range of uncertainty around the underlying assumptions yields confidence intervals that help inform how strongly you should interpret the results. These confidence intervals, or estimates of uncertainty, can be seen here (in red) for the Minnesota modeling of COVID-19, and at the same time, note the modeling of different scenarios (rows) and the estimated impact on different metrics (columns):

But important questions can always be asked, such as where to the assumptions and parameters come from (especially when trying to model a new issue), how much do they vary by different groups (e.g., age groups in the COVID-19 case; occupations in a workforce planning model), how fully-specified are the relationships, and are there important things that are missing? It’s also important to consider the decision-making criteria. In social science research and people analytics, we might be looking for results that characterize a typical (i.e., average) situation; in a public health crisis, it’s likely more important to identify how to avoid worst case scenarios.

Unfortunately, the IHME's curve-fitting model and Minnesota's SEIR model give very different predictions of where we're headed. Both approaches contain significant unknowns, such as how well (or not) states or countries fit the earlier experiences of China (which had much stricter social distancing) and Italy because there are so many variables that presumably affect how the outbreak spreads, or in the SEIR approach, whether key parameters are accurate because COVID-19 is a new virus. This highlights the importance of understanding the nature and limitations of any kind of statistical model, and paying attention to the sensitivity of the results. The starkly-different projections of these particular models are also a reminder that actions based on statistical models will only be as good as the explanatory power or fit of those models. Ideally, imprecision in the degree of fit will translate into margins of errors and confidence intervals, but if a model is being applied to a new situation, then purely statistical margins of error maybe too conservative. The onus is always on the decision-maker to use their subject-matter expertise when interpreting and applying statistical results. But what to do when you have to make a decision? Explicitly recognize the decision rules and include the costs of making different types of inferential errors in any decision calculus.   

Putting all of this together, then, a good people analytics person is always skeptical—or at least probing…where did the data and assumptions come from, how do we know they are accurate, how sensitive are the results to particular assumptions, how much uncertainty is there, what’s the decision-making criteria, what’s missing? And notice that this is as much about subject-matter expertise—whether that's infectious diseases or human resources—as it is about statistical sophistication. It's not just data mining.

It might also be useful to note that neither of these modeling strategies (curve-fitting or simulation based on parameterized flow models) match the dominant predictive approach in HR, especially in HR research (I don’t say this as a critique, just as another point of comparison). From a social sciences perspective, it’s much more common to predict outcomes in a regression framework where an outcome variable is modeled as a statistical function of a set of explanatory variables. For example, if employees’ level of engagement with their supervisor (inversely) predicts an intention to quit, then if an organization can increase engagement, we’d expect that quit probabilities would decrease, albeit imperfectly and with variation. This is a reminder that analytically, some issues are best modeled as societal phenomena, some modeled at an organizational level, and some at an individual-level. They each involve unique measures, and their own analytical challenges. A good people analytics person matches the methods and data to the problem—while still being probing as defined in the previous paragraph.    

Lastly, COVID-19 dashboards and modeling raise challenging ethical questions. What data are being collected and how are they being used? Are metrics and results being presented in sensationalized or inaccurate ways? What’s the role of modeling in determining public policy decisions? There are no easy answers to these and other ethical challenges, but they are a good reminder that people analytics also involves important ethical challenges. How is employee data being used? What kind of consent should be required? How transparent is the decision-making? Are implicit biases embedded in modeling decisions furthering rather than redressing historical inequalities? Throughout the people analytics process, it’s essential to remember that most data, and certainly most decisions, pertain to real people, not data points in a database or costs on an income statement. The science of people analytics is important, but so is the humanity. And in terms of presenting data in skewed ways, this has long been recognized as a danger with statistics, and perhaps the best defense is to be a wise consumer of statistics who doesn't naively take everything at face value (see "probing" above). 

In closing, it’s nice to see data visualization, dashboards, and statistical modeling getting such public attention, but it’s obviously unfortunate that this is because of a global pandemic that has harmed so many people and communities. While not losing sight of what’s most important, there are also lessons here for people analytics.

Saturday, February 15, 2020

Managing Conflict at its Sources

In Director Bong Joon-ho’s highly-acclaimed movie Parasite (2019), the wealthy Park family believes that they have a win-win relationship with the lower-class Kims. The Kims, however, view this relationship very differently, allowing them to prioritize their own interests in this relationship. If we were to step into this in the middle of the movie, we’d need to get the Parks to see the actual nature of their relationship, while also addressing the perceptions and emotions that are fueling a simmering conflict between the two families. Moreover, on an appropriately dark and stormy night, the xxxxxxx’s have an unexpected encounter with xxxxxxx and xxxxxxx (redacted to avoid a spoiler). Emotions flare up (especially fear), and magnify the fixed pie cognitive bias that pushes us to assume sharp conflicts of interest, leading all involved to treat this as a win-lose battle for self-preservation. Again, if we were to step in and try to resolve this before it spirals downward and reaches lower depths (an inside reference for those who have seen the movie), we’d need to re-frame the nature of their relationship (they have some common goals), address their decision-making, and help them cool their emotions.

Alex Colvin (Cornell), Dionne Pohler (Toronto), and I call this “managing conflict at its sources.” In other words, to successfully resolve a conflict or dispute, you must first understand its roots or sources, and then appropriately match a dispute resolution method. So we’ve created a three-part typology of the roots of conflict—specifically, structural, cognitive, and psychogenic sources of conflict—to facilitate the identification of effective dispute resolution methods tailored to the particular sources of a given dispute. These are described in my earlier blog posting, but brief definitions are useful here. Structural sources pertain to nature of the parties’ relationship, including their power, rights, and interlinked interests or goals. Cognitive sources relate to mental functioning, including interpretation, perception, information processing, decision-making, and (mis)communication. Psychogenic sources arise from the psychology of feelings, especially emotions, moods, and personality.

We believe that it’s important to diagnose a conflict by looking for these sources because they require different approaches to resolve them. Resolving structural conflicts requires diagnosing the nature of the parties’ relationship. Key alternatives include (i) a self-interested exchange with accessible alternatives (egoist); (ii) lasting interdependence with a mutual gains structure (unitarist); (iii) lasting interdependence with a mixed-motive structure (pluralist); or (iv) lasting interdependence with a win-lose structure (antagonistic). Recognizing these structural forms is important for factoring in issues of power. In an egoist relationship, power is less important than self-interest. If someone gives you a good deal, take it; if not, take your next best alternative. In a unitarist relationship, a focus on power likely interferes with finding interest-aligning policies. In contrast, power differences are likely a significant aspect of an antagonistic relationship, and distributive negotiations would be fully consistent in this structure. Integrative bargaining is very difficult in an antagonistic structure. In a pluralist relationship, both distributive and integrative negotiations are likely, and the parties or third party dispute resolution actors would likely need to ensure that power is not exercised in an overly aggressive way that undermines the shared interests and enduring nature of the relationship.

The effectiveness of third party interventions also varies across these relationship types. In an egoist relationship, the main need for third party intervention is to adjudicate alleged violations of contractual terms, which points toward arbitration-type procedures that provide a clear determination. In a unitarist relationship, in contrast, the importance of mutuality means that the arbitration of conflicts could be counter-productive; rather, mediation-type interventions are most useful in helping the parties recognize their mutual interests and resolve any coordination problems or barriers to achieving the integrative potential inherent in their relationship. But in antagonistic relationships, mediation efforts that search for common interests are incompatible with the fundamental oppositions of interests that drive conflict in this structural form, and thus would likely be futile. By contrast, pluralist relationships are most open to a range of interventions, including mediation- and arbitration-type third party interventions, reflecting the diverse nature of distribution and integrative issues inherent in this type of relationship.

Turning to the cognitive dimension, there are various techniques to address perceptual differences rooted in contrasting cognitive frames, such as a process of unfreezing, changing, and refreezing frames, either with or without mediator assistance. Other interventions can explicitly address cultural differences (more generally, in-group versus out-group conflicts). Regarding conflicts that have an aspect of limited information processing, people can more easily identify cognitive errors made by others than themselves. Providing individuals training in decision-making biases and teaching them critical thinking and self-awareness can help them become aware of decision-making blind spots to work through this type of cognitive conflict. Similarly, recognizing when miscommunication causes or contributes to a conflict also points to specific conflict resolution strategies. This can include avoiding communication channels with low signal-to-noise ratios, listening for the intended meanings of what’s being said, communicating in ways that the listener will understand your intent and that reflects the listener’s perspective, and establishing conditions under which an effective dialogue can occur.  

Psychogenic conflict is perhaps the most difficult type of conflict to tackle, and again requires tailored dispute resolution strategies. This aspect of conflict is not easily resolved through negotiation, nor is it likely to be truly resolved by the imposition of a solution by a third party such as a manager or an arbitrator. Indeed, the most accessible strategy is to give people tools to work through their own emotions, or to control their moods in different situations, either in advance of a conflict or during it. When dealing with hot emotions, cooling strategies such as taking a time-out or a break and trying to re-orient an individual’s attention to be more reflective and self-distanced rather than self-immersed can facilitate problem solving. If hot emotions like anger or humiliation  are contributing to a conflict, then facilitators can lessen these emotions by acknowledging them. An understanding of how different personality types approach not only conflict, but feeling, thinking, and behavior more generally also can be useful to understand how to engage with others constructively with others.

Lastly, not only might a dispute be complex (so don't stop after identifying the first cause), conflict can be dynamic and evolve around over time. As such, the source(s) of the conflict can change in the midst of attempts to resolve the initial source(s) of the dispute. This reinforces the need for those trying to resolve disputes to understand the range of possible sources of conflict, so that changes in the nature or sources of a particular dispute can be identified and appropriately addressed, rather than inadvertently contributing to compounding the conflict. In Parasite, the initial conflict between the wealthy and poor families appeared economic in nature, but with greater personal contact came new challenges that were more cognitive and especially psychogenic in nature. To continue to treat this conflict as purely economic (structural) and to ignore other smelly issues (another inside reference) would not produce a lasting resolution to this conflict. To effectively manage conflict at its sources is to recognize that dispute resolution needs to be tailored to the specifics of each conflict based on a careful diagnosis of the possible overlapping and changing structural, cognitive, and psychogenic dimensions.

Source: John W. Budd, Alexander J.S. Colvin, and Dionne Pohler (2020) "Advancing Dispute Resolution by Understanding the Sources of Conflict: Toward an Integrated Framework," ILR Review 73(2): 254-80. [free access to the pre-publication version here]

Monday, January 6, 2020

A New Culprit in the Decline of American Labor? Robert F. Kennedy and the Long Cast of Hoffa's Shadow

I just finished reading Jack Goldsmith’s In Hoffa’s Shadow: A Stepfather, a Disappearance in Detroit, and My Search for the Truth (Farrar, Straus and Giroux, 2019) which I highly recommend. Who needs fiction when real-life history produces stories like these? The author is a Harvard law professor whose mother married Chuckie O’Brien on June 16, 1975 when the author was 12 years old. In the author’s own words, Chuckie was “a great father” who “smothered me in love” (p. 5). But on July 30, 1975, former Teamsters President Jimmy Hoffa disappeared and Chuckie—Hoffa’s longtime friend and aide in the Teamsters—quickly became a leading suspect in this extremely high-profile case.

In Hoffa’s Shadow chronicles Hoffa’s rise and fall—often with Chuckie at his side—and his disappearance—where the FBI long thought Chuckie was also at his side, unwittingly delivering him to mob hitmen (a fiction often repeated in popular culture, including most recently in Netflix’s The Irishman). The focus is uniquely on Chuckie—his life, his ties to the Teamsters and the mafia, his personal values, his decades-long public mistreatment at the hands of the FBI, and the sheer improbability of any culpability in Hoffa’s disappearance. All of this is quite interesting, but what really makes this book such a compelling read is how deeply personal it is. Goldsmith is exceptionally candid in describing how he idolized Chuckie in high school but at age 21, renounced him and changed his name from Jack O’Brien to Jack Goldsmith out of fear that “the association with Chuckie might jeopardize my legal career” (p. 26). After 20 years, Goldsmith reconciled with Chuckie, who accepted Goldsmith “back into his life without qualification, rancor, or drama” (p. 41). The author eventually convinced Chuckie to let him tell his story, in the author’s hope that it would solve the 45 year-old mystery of Hoffa’s disappearance. Alas, the author ultimately fails on this last account, but in the end that seems like a minor footnote given the depth of insight we get into Hoffa’s leadership of the Teamsters, the relationship between the mafia and the Teamsters, the likely reasons for his disappearance, the troubling extent of the federal government’s use of its own power, and at a personal level, the complex character of Chuckie.

From a labor relations perspective, one thing that jumped out to me is the provocative claim that the field has overlooked “the most fundamental” reason for the decades-long decline in labor union membership. It is well-recognized that the fraction of workers represented by a union (“union density”) peaked in the private sector in the mid-1950s, and since that time has fallen from around 35 percent to 6 percent. Many explanations have been proposed, including structural change (e.g., the decline of manufacturing, demographic shifts, globalization), decreased demand for union representation (e.g., laws and paternalistic human resource management provide some of the protections that unions provide, or unions have failed to keep up with what workers want), and legal and illegal employer opposition facilitated by hostile legal rulings. But Goldsmith argues that “the most fundamental reason [that membership fell] was the identification of the entire labor movement with corruption, violence, and bossism—an identification that crystallized with Bobby Kennedy’s singular crusade” (p. 108). Wow!

What was this singular crusade? Senator Estes Kefauver led a special Senate investigation into organized crime in the early 1950s, and the resulting public attention on the sensational hearings helped propel Kefauver to national prominence (including being selected as the Democratic Vice Presidential candidate in 1956). According to Goldsmith, Robert F. (Bobby) Kennedy saw this as a model for elevating the profile of the Kennedys (which included his older brother John F. Kennedy), and perhaps, too, for Bobby Kennedy to prove his worth within the Kennedy clan. So in 1957, the United States Senate Select Committee on Improper Activities in Labor and Management (“the McClellan Committee”) was created to investigate labor racketeering (the corruption of labor unions by organized crime), with Bobby Kennedy as its chief counsel. Enter Jimmy Hoffa and the Teamsters. Goldsmith quotes historian Arthur Schlesinger as saying that before the hearings even started, Kennedy had already concluded that Hoffa was corrupt and ran the Teamsters solely for his own benefit. As such, Hoffa was “the enemy [Bobby Kennedy] had been seeking” (p. 99).

The reality of Hoffa is seemingly much more complex. Hoffa seemed to genuinely care for the economic well-being of truck drivers and other workers, and fought hard on their behalf—albeit often too hard in terms of taking an extreme ends-justifies-the-means approach, even if this meant hiring mob goons to literally fight employers and giving kickbacks to the mafia to maintain his own power. So of course Hoffa was no angel, but In Hoffa’s Shadow shows the extent to which Kennedy became obsessed with publicly vilifying Hoffa. And each time this failed, “Kennedy got angrier, become more vindictive, and invariably cut more corners” (p. 102). This included sending the IRS on a fishing expedition looking for evidence of criminality in over 3,500 tax returns, and then illegally entering confidential IRS information into the public record.

Students of labor relations know that these hearings resulted in the Landrum-Griffin Act in 1959 which sought to make unions more democratic while also placing a few additional restrictions on union activities (especially banning secondary boycotts). But Goldsmith interestingly argues that the larger effect was that the hearings led by Bobby Kennedy “embedded in the public mind, including the minds of many workers, the idea that unions were flawed institutions exercising illegitimate power” (p. 106). And thus we have Goldsmith’s provocative claim that “the most fundamental reason [for declining union power] was the identification of the entire labor movement with corruption, violence, and bossism—an identification that crystallized with Bobby Kennedy’s singular crusade.” Whether we can trace 65 years of union decline to this one moment is debatable and would represent an influence with remarkable staying power, but it is certainly stimulating to consider its role among other factors.

Goldsmith doesn’t let Hoffa off the hook: “his defiant embrace of criminal tactics and associations [even if done in with the sincere belief that this was to help the rank and file] allowed Kennedy [and others] to paint him as a subversive force…and his performance tarnished the entire labor movement” (p. 107). But Kennedy was anything but balanced, and ignored, for example, the role of employers in fighting workers. Kennedy’s campaign against Hoffa continued in the 1960s with Kennedy’s appointment (by his then-president brother) as U.S. attorney general. In the end, according to Goldsmith, Kennedy “neglected, elided, or interpreted away ethical and legal restrictions that are supposed to channel and constrain the federal government’s colossal power to destroy one’s reputation and liberty” (p. 121). This included a sharp rise in the government surveillance of individuals, including breaking into homes and businesses to plant listening devices, typically without any warrants or legal oversight.

The extent to which this rise in illegal government surveillance connects to Goldsmith’s own work in government is another unique aspect of In Hoffa’s Shadow making for a compelling read. But a larger take-away, in my eyes, is that these revelations implicitly highlight the need for democracy, transparency, and institutional balance. When the government holds all the cards, where are the checks on its power? Or to what end is government power being exercised? These questions are as important as ever when legislation and judicial rulings are seemingly weakening organized labor for political gain, and we seem to have forgotten the importance of the labor movement and other groups for a vibrant democracy. Hidden in In Hoffa’s Shadow, then, is a strong conservative case for labor unions, even if the focal union in this book has historically struggled with democracy and corruption.

So in the end, this book is about much more than Hoffa’s disappearance. Indeed, I assume that “in Hoffa’s shadow” refers to the personal experiences of Chuckie O’Brien. But as we continue to confront questions of power, democracy, and surveillance, it seems that we’re all living in the shadows of Hoffa and Bobby Kennedy, with their lasting implications for labor unions and democracy.