Two Famous Management Statistics That Do Not Hold Up
Managers drive 70% of engagement variance. Firms pick the wrong manager 82% of the time. Neither claim has a published derivation.
Two statistics dominate writing about management. Both come from the same consultancy. Neither has a derivation you can check.
"Managers account for 70% of the variance in engagement"
The source is a 2015 article stating that managers "account for at least 70% of the variance in employee engagement scores across business units". The article contains no methodology section at all.
A related page discloses this much: a multiple regression across 11,781 work teams, with predictors being "team members’ perceptions of their manager, the managers’ engagement, and manager talent", against team engagement measured by the publisher’s own instrument. No coefficient of determination, no coefficients, no sampling frame, no confidence interval, no published paper.
The deeper problem is circularity. Two of the three predictors are employees’ perceptions of their manager, collected from the same people, with the same instrument, at the same moment as the outcome. The engagement instrument contains items such as whether a supervisor cares about the respondent as a person. Regressing team engagement on team members’ ratings of their manager is largely regressing something on part of itself.
It is also frequently miscited as coming from the same publisher’s large engagement meta-analysis, which covers 112,312 business units and over 2.7 million employees and is genuinely well documented. The phrase "70%" does not appear in it.
"Firms pick the wrong manager 82% of the time"
The second claim is that companies "fail to choose the candidate with the right talent for the job 82% of the time".
Reading the article, the 82% is 100% minus 18%, where 18% is the share of sitting managers scoring above a cut point on the publisher’s own unpublished assessment. There is no independent measure of hiring success, no counterfactual, no observation of the hiring decision, and no comparison with candidates who were not hired. The threshold was set, in the publisher’s words, to optimise the probability of selecting high performers, which is a chosen line rather than a measured failure rate.
The article also contains a tension worth noticing. It says about one in ten people in the general population have the talent, while 18% of current managers do. By its own numbers, firms select roughly twice as well as chance.
The coincidence that causes most of the confusion
A UK management body reported in 2023 that 82% of workers entering management positions had received no formal management or leadership training. That survey has a disclosed method: online fieldwork with 2,524 employees with management experience.
It is a self-selecting panel from an organisation that sells management qualifications, so it is advocacy research, but the method is stated and the claim is falsifiable. It is also a completely different statistic that happens to share a number, and the two are routinely merged into a single sentence.
Why we are being pedantic
Because management genuinely matters, and there is real evidence that it does. Building the case on unfalsifiable vendor numbers makes it easy to dismiss when someone checks. That is the same argument we make about Botswana graduate statistics and about promotion decisions.
Frequently asked
Do managers really account for 70% of engagement?
The claim comes from a 2015 consultancy article with no methodology section. The disclosed model regresses team engagement on employees’ perceptions of their manager, measured with the same instrument at the same time, which makes it substantially circular rather than a variance decomposition.
Is it true that 82% of managers are the wrong choice?
That figure is the complement of the share of sitting managers who score above a threshold on the publisher’s own unpublished assessment. It never observes hiring decisions or outcomes, so it is not a measured failure rate and cannot be checked externally.
Are there any reliable statistics about management quality?
Yes. A randomised trial in Indian textile plants and the World Management Survey covering over 13,000 firms are both properly documented, and a study of 38,843 sales workers measured how promotion decisions affect managerial quality. All three publish their methods.