article » Data scientist: Avoid one ‘toxic worker’ don’t hire two ‘superstars’

Data scientist: Avoid one ‘toxic worker’ don’t hire two ‘superstars’

January 29, 2016
3 min read

Michael Housman has built a career around using big data to “decode the workforce”. This has included a wide range of statistical projects which aim to get to the nub of employee retention.

“I guess the over-arching theme is that we are, by our nature, social creatures that are influenced very much by relationships,” he tells us. “I came into my work [as Chief Analytics Officer at Evolv, since acquired by Cornerstone OnDemand and now as Workforce Scientist in Residence at HiQ Labs] thinking that if you hire the right person, everything else takes care of itself.”

“What I've found is that it's crucial that someone has the right manager, has friends at work, and is not be surrounded by toxic co-workers if they’re going to do their best work.”

This is precisely what Houseman’s latest paper [PDF], produced in conjunction with the Harvard Business School, looks at: toxic workers. It is based on research into 50,000 front-line workers employed at an hourly rate. Interestingly, it found negative ‘toxic’ workers have a far greater impact on the organisation than positive ‘superstars’.

These findings revealed that toxic workers – those who display overtly socially negative behaviour like theft, harassment and bullying – have a poisonous impact on the people around them and impact their retention.

Houseman says that neither he nor his co-writer, Dylan Minor of the Harvard Business School, were surprised by the results overall. “We had expected to see those results although the sizes of the effects were pretty staggering,” he says.

“That avoiding a toxic worker is the equivalent of hiring two superstars is pretty remarkable [though],” he adds. “We also didn't expect that toxic workers would be more productive – albeit with lower levels of quality – and would claim to follow rules more. That piece was definitely a surprise to us.”

But how does all this translate to more professional, salaried occupations? Houseman tells us, he feels part of the difficulty is that the really “egregious behaviour” like drug or alcohol violations, sexual harassment, fraud and so on tend to appear less often in more professional roles.

“My gut sense is they occur less but that there are more subversive forms of toxic behaviour – workplace bullying, intimidation, back-stabbing – that occur a lot more,” he says. “Unfortunately, we couldn't observe or measure those.”

This is very interesting point and extremely obvious in our own recent IT workplace bullying research. Toxic behaviour is definitely there in spades in professional roles but is more underhand, political and Machiavellian in nature. It also often comes from the very top.

Houseman also points out that although the cost of the toxic worker is calculated at $12,489 for an employee on an average salary of $22,000 – excluding other potential costs, such as litigation, regulatory penalty, and reduced employee morale. “If your salaried employees are getting paid a lot more, then the cost of this behaviour goes up several fold. So the stakes are certainly higher and I think the cost of a toxic worker would easily be in the tens of thousands within a professional environment.”

Through his years’ of research Houseman says the findings that surprised him most was one of the first studies he conducted which found that someone’s history as a job hopper had no impact on the likelihood they’d stay in their next job.

“I love it when I can do some research that upends some long-standing beliefs and hopefully force people to re-evaluate what they're looking for and what truly matters in the employment process,” he says.

“Managers should know they have a much bigger impact than they probably realise,” he adds. “When we throw everything possible in our models to predict attrition we find that someone's manager alone predicts as much of the variance in the models as everything else combined.” This is something he has seen replicated time and time again.

The other main thing he has noticed is that there is not a strong relationship between someone's experience doing a role and their skills managing others in that capacity. “The rock star doesn't necessarily make the best manager,” he concludes. “What that says to me is that managing is a skill unto itself that needs to be cultivated over time.”

To access the original article, click here.