KEY CONCEPT

Avoiding a toxic worker enhances performance and costs less than replacing an average worker with a superstar — even if the superstar performs in the top 1% of employees.

---

IDEA SUMMARY

While a great deal of attention has been paid to the performance and cost advantages of hiring superstars, less attention is paid to employees at the other end of the spectrum: toxic workers who undermine the organization’s productivity and incur significant costs regulatory and legal liabilities.

New research shows that avoiding toxic workers (or converting them into average workers) increases an organization’s productivity and performance more than hiring superstars. Using assessment, job history and performance data from more than 50,000 employees in 11 firms, the researchers first identified toxic workers as workers who had been terminated for sexual harassment, workplace violence, falsifying documents, and fraud and general workplace misconduct. 

They then identified from the data the antecedents of toxicity. Specifically, they found that workers mostly like to be fired in the future for toxic behaviour were:

• Overconfident workers — those who vastly overestimated their skills (the research showed that they were 15% more likely to be toxic)
• self-regard workers — those who only cared about their own self-interests (25% more likely to be toxic) 
• workers who declared in job interviews that rules should always be followed (22% more likely to be toxic — and ironically, workers who declared that rules might sometimes need to be ignored were less likely to be toxic)
• workers exposed to other toxic workers (46% more likely to be toxic).

This result highlights the potential damage of toxic workers: their behaviour not only impacts their own performance but can significantly impact the performance of others.

In addition to these personal characteristics antecedents of toxic behaviour, the research also looked at environmental factors, notably 1) exposure to other toxic workers, and 2) the extent to which the workers were monitored by managers or superiors. For example, the data showed that workers exposed to other toxic workers were 46% more likely to be toxic. Generally, the researchers found that 70% of the toxic predictability of a worker was based on his or her personal characteristics, and 30% on job environment.

The performance consequences of toxicity, as revealed by the data, were mixed. For example, toxic workers tended to work faster than non-toxic workers (were thus more productive), but also produced less-quality work.

A quantitative analysis of the data, comparing the costs of a toxic worker vs. the benefits of hiring superstar performers, revealed the full extent of the damage a toxic worker can inflict on an organization. The quantitative analysis calculated

• the value of a star performer based on his or her increased output (that is, the cost of hiring other workers to make up the output of one star performer)
• the induced turnover cost of a toxic worker (that is, the cost of replacing workers who left because of the presence of a toxic worker on a team).

A comparison of these two calculations showed that if a company hires a superstar who performs in the top 1% of employees, the costs savings to the company would be $5,303. However, if the company avoids hiring a toxic worker, the cost savings is $12,489. In short, the benefit of avoiding a toxic worker is more than twice the benefit of hiring a superstar (or replacing an average worker with a superstar).

The quantitative analysis allowed the researchers to drill even deeper, not only looking at toxic workers in general, but at specific characteristics of toxic workers. For example, a one standard deviation increase in confidence led to $122 in saved wages to the company due to increased productivity; however, that same one standard deviation increases led to $1,327 in increased induced turnover costs. Productive or not, the toxicity of the confident workers costs the company a net $1,000. 

The research was based on data from a company that builds and deploys job-testing software to large employers, and included:

• Assessment data from a proprietary job test (a test to determine an applicant’s fit for the position for which they were applying). The researchers used responses to selected questions from this test to gather information on the antecedents of toxicity.
• Attrition data, which included hiring and termination dates and reasons for the termination.
• Performance data, which measured employees’ time to produce one unit of quantity, as well as quality of work; the performance data also allowed researchers to calculate the cost savings of superstars and turnover costs of replacing terminated toxic workers.

---

BUSINESS APPLICATION

Major takeaways from the research included the following:

1. Workers who are overconfident, self-regarding and who claim rules should always be followed are more likely to be toxic workers. Managers should carefully screen job applicants for these traits.
2. Toxic workers are often more productive (they work faster than other workers and generate greater results). Overconfidence in a worker is particularly indicative of the productive worker who is also more likely to be toxic. Because of their increased productivity, many leaders and organizations often allow toxic workers to stay in the organization: it is difficult to fire workers who are bringing money into the organization. However, these revenues are deceptive: the costs of the workers’ toxicity far outweigh any benefits from their productivity. Multiple scandals, for example in the financial industry, indicate how easy it is to overlook toxicity in favour of productivity, with disastrous results.
3. In addition to personal characteristics, the research indicates that on average one-third of a workers’ toxic behaviour can be due to the workplace environment. This is significant, and highlights the need for leaders and organizations to not only avoid the type of worker most likely to engage in toxic behaviours, but also to carefully ensure that the work environment does not encourage this behaviour. One major step, according to this research, is to ensure that workers are sufficiently monitored.

---

---

FURTHER READING

  Michael Housman’s profile at Cornerstone OnDemand
  Dylan Minor’s LinkedIn profile

---

REFERENCES

Toxic Workers. Michael Housman, Dylan Minor. HBS Working Paper 16-057 (November 2015). 

•   DOWNLOAD IDEA
•   PIN TO MY IDEAS
•   PRINT IDEA
•   LIKE

---