This is a very naive analysis. It defines toxicity as propensity to be fired, only looks at big corporations, and uses self-reports in interview tests (like paper tests) as the independent variable.
I suppose the reason this piece strikes this community as so silly is that the general philosophy of the startup is that a group of 5 or 10 people can best multi-million dollar corporations because corporations are so inefficient. To turn to such a company's firing decisions as the source of truth of worker quality feels profoundly backward.
I guess there's a place for this type of research, but it seems like a fool's errand to try to make any kind of objective analysis about such things.
In any business you have a fairly unique problem and a unique set of people, the only way to succeed is to be aware of the specifics you're dealing with and react appropriately. Management trying to apply insight from academic research so often ends up looking clueless and alienating the rank and file.
Also, their analysis that avoiding toxic workers is twice as good as hiring a superstar makes no sense, because it doesn't account for the fact that some jobs have vastly more leverage than others. The obvious example is software developers, whose work easily scales at least 100x as much as emissions testers. But even basketball coaches are a good example. (Mark Cuban anyone?)
I'm not claiming that we hiring toxic workers is a good bet by any means, but the analysis also doesn't make any sense.
For what it's worth, I've found that quantitative economics research coming out of Harvard tends to be complete nonsense. I have no idea why this would be the case, it's just something that I've observed.
Harvard trades on a reputation earned centuries ago. Since people don't really care about the accuracy of the research, there is nothing to challenge in the content. People who agree with the vague conclusions get to say "Harvard is on my side", and then twist the vagueness to suit their purposes.
As common in studies of personality, the researchers perform a bait-and-switch in defining key terms. To ensure mass appeal, like a skillful astrologist crafting this week's horoscope, they introduce the term "toxic worker" as an employee who is "harmful to an organization", leaving the authors plenty of room to maneuver.
Unfortunately for the authors, today's scientific paradigm requires at least the semblance of a statistical analysis. Finding a numeric measurement for an aspect of human consciousness, such as "toxicity" (or how much of a jerk you are), can be quite difficult. For the empirical portion of the paper, the authors choose to define a "toxic worker" as one who has been fired for "an egregious violation of company policy" such as "sexual harassment, workplace violence, falsifying documents, fraud, and general workplace misconduct". In their data, 1 in 20 employees was eventually deemed "toxic".
While I'm sure most people can identify a jerk at the office, they might also tell the authors, "You keep using that word [toxic]. I do not think it means what you think it means."
This is a pretty embarrassing paper, even by the low standards of the Harvard Business Review. A toxic worker is defined as anyone who takes too many pens, or someone who embezzles huge sums of money, or a sexual harasser, or a physically violent person, or a jerk, or... anyone retroactively deemed toxic by the manager who fired them.
All this time we thought Wal-Mart was firing people who were trying to unionize, turns out it was just far-sighted managers keeping a deadly contagion at bay. The Harvard Business Review inches ever-closer to the methodology of soviet psychiatry.
The definition surprised me. I mean the actual definition, as it is used later, not the one presented in the beginning. First they say someone who does more bad than good is a toxic worker, but then they have some quite strange attributes to that, like someone being fired for toxicity (which often means he's annoying to the boss, not that he's toxic), or someone who steals office supplies (I don't know anybody who would even complain about that, as long as it's not in exceptional ammounts).
And finally they conclude you should always try to cut losers instead of fighting for winners. There is no game (in the sense of game theory, not child's play) in which "always" really makes you better than average. Sometimes everybody knows that someone is toxic but you can't cut him lose, because he is in a very specific position (e.g., the brother of your most important customer). Another time having that one super-star will put you forward by a dimension, not just a few bucks. "Instead of always do A better always do B"? No, both is wrong. Don't "always".
HBS's audience isn't people building huge accretive value, this publication doesn't gain anything by giving advice that let's people "put you forward by a dimension" (meaning go from x^2 to x^3 in sales or reach, let's say.)
Its audience is middle managers who want to CYA. (cover your ass.) That's why a superstar programmer who brings home ('steals') a reams of printing paper after staying for three hours of crucial unpaid overtime is a "toxic worker" - since those kinds of shenanigans could well end up reflecting poorly on the manager. even if the guy will just use it to scribble all over his kitchen table, since even after 3 hours he hasn't been able to wrap his head around the logic bug.
and people wonder why it's so easy for three programmers to compete with these huge lumbering organizations.
I was really surprised how that CYA stuff is applied in the real world literally. And in their world it really makes sense, because they will not just cover their own ass, they will gut you the moment you leave yours blank. So the first thing a manager may tell you is to cover your ass as quickly as possible and as good as you can, and he thinks he's doing you a live saving favour.
In terms of stealing office supplies, it boils down to intent. There's a big difference between taking pens home as part of work and not bothering to bring them back -vs- thinking "my kids need more pens for school" and then going out of your way to take them home. The former is mostly absentmindedness and isn't really stealing; the latter is an ethical failure and is stealing.
1. The loss of a certain % of pens is trivial to any non-tiny business.
2. Most people dont like work. They do it because they have to in order to pay rent, bills and buy food and whatnot, and deep down they resent their jobs to some degree.
3. 'Stealing' s few pens can relieve some of this resentment as even though it is completely trivial you at least feel as though you are breaking the rules and getting one over to some degree, or at least just getting some free pens.
4. I would posit that the benefits of this pressure release greatly outweigh the cost of the pens.
I don't think that's okay. Any reasonable expenses you incur when asked to work from home should be submitted as a paid expense. Others have mentioned overtime, which you are entitled to if you are paid hourly. If paid as a salary, then in my experience the general arrangement is time off for extra hours worked, whether that's a formal process or simply cutting out a couple hours early on a Friday if you put in a couple hours extra during the week.
the corruption runs both ways though: the stealing of employee overtime hours is also stealing yet really is only frowned on if it is systematic versus absentminded.
This is one of those papers where the title sounds vaguely legitimate, but the underlying analysis is only tangentially related to the topic. Also, consider
"less attention has been paid to the question of how to manage those workers on the opposite side of the spectrum: those who are harmful to organizational
performance."
Which is complete rubbish--GE placed a huge focus on doing performance reviews and culling the "bottom 10%". A practcice which started a trend (called forced ranking) that grew to include many of the fortune 500, in cluding perhaps famously Microsoft.
Its either a bafflingly careless omission or a cynical gesture to pretend this topic is somehow "under the radar" in business school or corporate HR departments.
The GE-style ranking is looking at something else, people with poor productivity. This is looking at people who are productive in their jobs by the usual metrics you'd use, but are a liability for the company, due to a propensity to cut corners, break laws, embezzle, etc.
I read this study via the article http://qz.com/563683/turns-out-toxic-coworkers-are-more-expe... which pointed out the net savings between hiring a superstar vs firing a toxic employee, and I was more shocked that the numbers were so low for superstars. That a Top 1% person in their field can only deliver $ 5303 worth of extra value to their company over their whole period of employment? I mean I'm no superstar, but that seemed pitifully low just from my own experience of the right action at the right time by the right person being able to secure deals worth a lot for the company.
As per the paper:
"To generate a straightforward comparison of the value of each of these focuses, we quantify the value of a star performer by identifying the cost savings from her increased output level. That is, a superstar is a worker that adds so much value that without her a firm would have to hire additional workers (or pay for additional hours from existing workers) to achieve the same level of output as when they have that single superstar worker."
"We calculate the percent in increased performance for each of these performance levels and multiply it by the average worker salary, based on company records."
"For comparison, we report in the "Avoid a Toxic Worker" column the induced turnover cost of a toxic worker, based on company figures. Induced turnover cost captures the expense of replacing additional workers lost in response to the presence of a toxic worker on a team."
"The data were obtained from a company that builds and deploys job-testing software to large employers. Many of these companies are business-process outsourcers (BPOs) that themselves provide a variety of business services (e.g., customer care, outbound sales, etc.) to their clients. The employees included in the dataset are all engaged in frontline service positions and paid on an hourly basis."
So these stats are coming from call center staff and telemarketing where there's a lot more priority on people following the script/process than showing flashes of brilliance. But all of that gets lost when it gets breathlessly repeated in the abstract or in articles referencing it.
i was immediately thinking the same thing: was john carmack's genius practically negligible for id software? how about fabrice bellard's worth (for telekom paris)? could have ten other average ("non-toxic") programmers achieved what they did?
HBS always writes about obvious things. This is known to anybody with more than 20 years of experience: toxic workers ("jerks") exists and they are mainly super stars. And they will raise in ranks (they will become managers, directors, etc.). But not necessary by growing the business.
Earlier in my ex-companies life, they had a real low threshold for terminating someone who was toxic. As the other CEO's took over, company lost its customers, and became more financially stressed, the tolerance for wrongdoing increased until it was basically as the chart in the article stated with offences having to be pretty severe (bullying, sexual harrassment) to cause termination. It was almost like the company needed the toxic people to keep the lights on.
1. Why are workers toxic? Is it something you're born with, or something that you have when you are hired? Is it unchanging, black and white, true or false—or can it change over time or be impacted by the environment?
2. Given the answer to the above, what types of things influence workers' toxicity? Are you able to improve any of those things?
3. Given that the time and focus of any team is limited, and assuming most of the improvements from question 2 actually impact everyone in the company positively, do you truly believe that handling toxic people is the best use of time for a human management team?
Fire when you need to, and ensure you fix mistakes quickly and with intelligence, but do not focus efforts on the bottom tier. Work to improve the environment that makes people "go toxic." Ignoring that environmental effect is the type of behavior that makes toxic companies, which is the far greater problem, and should be the far greater concern.
We all know toxicity exists, and we've worked with people who, on the surface, don't belong anywhere near the company. But often the medium that breeds toxicity—the politics, fear, competition, in-fighting and more—is more important to tackle than the individuals themselves. There's a reason companies foster and hire people who tend toward toxic behavior; and there's a reason they act the way they do.
As W. Edwards Deming said: “Give the work force a chance to work with pride, and the 3 per cent that apparently don’t care will erode itself by peer pressure.”
Not sure if serious. I have no idea if this will engender downvotes for myself, but personally, I've always found HBS to do almost NOTHING BUT lend its name to pseudoscience twaddle and call it "business" or "management". Rest assured, in corporate life whenever you get some bullshit speaker spruiking some bullshit idea, a HBS paper is always eerily near.
Frankly I'm shocked they have the cachet they do...humans and their authority/prestige structures continue to fascinated me.
In my B-school program we did read lots of HBS cases and most of those were about objective things like "could the company have adopted a different marketing strategy" not "how can the company legitimize its political machinations in a totalitarian labor system" So I was giving them the benefit.
Both. Readers should stop reading when they get to "a toxic worker is defined as a worker that engages in behavior that is harmful to an organization." Could mean literally anything. Their data is equally ridiculous. It consists of a) lists of people that were fired because management deemed them "Toxic" and b) "daily performance data that represent productivity by measuring the average amount of time an employee required to handle a transaction and customer satisfaction scores." Their conclusion is that people that have self respect are more likely to be fired for being toxic. Ha! My only question is, how the hell do I get paid to write this stuff?
I've been in people management roles and certainly dealt with a rare few who just couldn't get along with anyone and also a disruption to the productivity of others. In hindsight, what I found is these individuals will tend to have a history of short stays at many companies. Because of the litigious nature of US, the former employers will never disclose anything wrong in a reference check plus they just want to get that person out the door.
Excellent analysis. This simplifies management down to a black and white, rigid, mechanical chopping block. This isn't fit for anyone doing business in today's world.
How to get paid? Become a social scientist. The data is boring and irrelevant to reality, so write any narrative you want, and be judged how well you fit your audience's preconceived notions (double credit for publishing opposite conclusions on the same data)
Very interesting, currently there is this post and " We all don't know what we are doing (cyberomin.github.io)" next to each other on the front page.
The whole idea that good leaders wing it and do their best, but that they can and should detect and weed out toxic workers and end their career with absolute certainty (because being declared toxic will be very hard to recover from) is a really interesting cognitive dissonance. Allow margin for error to good role model, but bad role model are intrinsically bad and unredeemable. I'm sensing straw man, posturing and feel good speech all the way down.
This looks like nonsense, especially the hilarious costs/benefits calculated to the nearest dollar.
And yes, of course toxic management is more likely to be a problem, even though it remains curiously unstudied by the HBS.
Even so, I think the takeaway - don't hire self-aggrandising control freaks, no matter how talented they seem to be - is probably sound at every level.
Except possibly the CEO. (Was Steve Jobs toxic? Bill Gates? Haven't most toxic CEOs been disastrous for their companies?)
What struck me is the fine line between superstars and toxic workers. More or less, they look much alike. The main difference seems to be how they're perceived by coworkers and management. And how that affects their attitude toward the organization. I would have liked to see more about the impact of management style on forking superstars and toxic workers.
I've always viewed superstars who are toxic as flawed. Superstars without toxic traits are rare. As the article states, you are most likely better off without the flawed superstars, the problem is, how do you come up with a consistent and effective way of getting rid of them?
A far better question is, how do you create an environment that helps flawed people improve?
Most people have flaws. It's 'consistent and effective ways for getting rid of them' that drives them crazy in the first place. One of many things that creates a negative feedback loop that drives a toxic culture, which breeds more "toxic workers."
1. I believe that termination for toxic behaviour is the only way to adjust the behaviour of these types of people. There's the PIP, but these individuals would quit if they were put on a PIP. These individuals aren't dumb, they will have F.U. money, and some of them would be financially independent.
2. In electronics, that would be a positive feedback loop as it is reinforcing the behaviour. A negative feedback loop would reduce the undesired behaviour.
I agree that superstars without toxic traits are rare. However, I believe that both superstar and toxic qualities are learned, and can change. I believe that it's toxic organizations, or at least toxic managers, that shape potential superstars into toxic workers.
1. You are correct, once a toxic CEO gets put in by the board, he fills senior management with people similar to himself. This is the primary way management becomes toxic.
2. Some of them could be coached and become excellent leaders, but most depend on their toxic traits to get ahead, and their performance would plummet if they could not use their toxic traits. Most need to be identified and rooted out.
I suppose the reason this piece strikes this community as so silly is that the general philosophy of the startup is that a group of 5 or 10 people can best multi-million dollar corporations because corporations are so inefficient. To turn to such a company's firing decisions as the source of truth of worker quality feels profoundly backward.