Emotional Intelligence in Leadership: What the Evidence Supports

Emotional intelligence, or EI, is one of the most popular ideas in leadership development. It appears in competency models, hiring tools and coaching plans, often with big claims attached.

The research is more mixed than the marketing, and more useful. Some forms of EI are well defined and measurable. Others overlap heavily with traits we already measure. Knowing the difference helps you invest in skills that matter and avoid paying for relabeled personality tests.

Two different ideas share one name

In a 2008 American Psychologist article, Mayer, Salovey and Caruso, who use the term for a set of abilities, described EI as the capacity to carry out sophisticated information processing about emotions and to use that information to guide thinking and behavior. They also described a split in the field: some researchers study EI as distinct mental abilities, while others study what the authors called "an eclectic mix of positive traits such as happiness, self-esteem, and optimism."

That split shows up in how EI is measured:

• Ability tests present emotional problems and score answers as more or less correct, using expert or consensus scoring. The U.S. Office of Personnel Management (OPM) describes this approach in its assessment guidance.
• Self-report and "mixed" measures ask people to rate their own emotional skills and related traits. OPM notes that employers concerned about faking on self-reports may find ability measures more acceptable.

So when someone says "EI predicts leadership," the first question is: which EI, measured how?

What EI predicts at work

Large meta-analyses give a nuanced picture.

Joseph and Newman's 2010 meta-analysis in the Journal of Applied Psychology proposed a "cascading" model in which emotion perception comes first, then emotion understanding, then emotion regulation, which in turn relates to job performance. The data supported the model, but the link between ability EI and job performance was inconsistent: positive in jobs with high emotional labor and negative in jobs with low emotional labor.

A 2011 meta-analysis by O'Boyle and colleagues in the Journal of Organizational Behavior grouped EI research into three streams (ability tests, self- or peer-report measures based on the ability model, and mixed models) and found corrected correlations with job performance ranging from 0.24 to 0.30.

Then the picture got more complicated. A 2015 follow-up by Joseph, Jin, Newman and O'Boyle examined self-reported, mixed EI measures. The content of those measures strongly overlapped with ability EI, self-efficacy, self-rated performance, conscientiousness, emotional stability, extraversion and general mental ability. Mixed EI's correlation with supervisor-rated performance was estimated at .29, but the relationship became nil after controlling for those overlapping factors.

Put plainly, many popular EI questionnaires predict performance largely because they repackage things already known to matter.

EI and leadership specifically

A 2010 meta-analysis by Harms and Credé tested claims that EI relates to transformational leadership. When the same source rated both EI and leadership behavior, the validity estimate was .59. When the ratings came from different sources, it dropped to .12. Trait measures of EI showed higher validities than ability measures, and different raters agreed only weakly about a leader's EI (.16) or transformational leadership (.14).

This is a common trap. If a leader rates both their own EI and their own leadership, or one direct report rates both, the two will look strongly related, because one person's overall impression drives both scores.

Can EI be developed?

There is some encouraging evidence. A meta-analysis by Hodzic and colleagues, published online in Emotion Review in 2017, pooled 24 studies (28 samples) of EI training for healthy adults. It found a moderate improvement from before to after training, with effects that held at follow-up. The type of EI model and the length of training made a difference.

Two caveats are worth keeping in mind. These studies measured change in EI scores, not necessarily change in how a leader's team experiences them. And a gain on a self-report questionnaire is not the same as a skill others can see.

What this means for leaders

Treat EI as a set of specific, observable skills rather than a personality score. The three facets in Joseph and Newman's model make a practical checklist:

1. Perceiving emotion. Notice your own state before high-stakes conversations, and watch for signals in others, such as a quiet team member, a sharp tone or a sudden drop in engagement.
2. Understanding emotion. Ask what is driving a reaction before you respond to it. Frustration about a deadline may really be about fairness or workload.
3. Regulating emotion. Build habits that create a pause, such as sleeping on a heated email or saying, "Let me think about that and come back to you this afternoon."

For example, imagine a plant manager who learns that a long-planned equipment upgrade has been cut. Before the shift meeting, she notices she is angry and decides not to share her own reaction first. She announces the decision, asks what worries people most, hears that overtime is the real fear, and commits to a staffing plan by Friday. Nothing about her personality changed; she perceived, understood and regulated in sequence.

Because self-ratings and others' ratings agree only weakly, get outside data. A well-run 360 is one way to do that; see our guide 360-Degree Feedback: How It Works and How to Use It Well.

A note for anyone buying EI assessments

If your organization is considering an EI tool for hiring, promotion or development:

• Ask whether it is an ability test or a self-report measure, and what it adds beyond personality and cognitive ability measures you may already use.
• Ask for evidence from studies in which EI and outcomes were rated by different people.
• Avoid using any single EI score as the basis for a selection or promotion decision.
• Be cautious with sweeping claims that EI matters more than intelligence or expertise. Ask what was measured, who rated it and what was controlled for.

Key takeaways

• "EI" refers both to a defined set of emotional abilities and to a broader mix of self-reported traits, and the evidence differs for each.
• Ability EI's link to job performance depends on the job, with a positive link where emotional labor is high.
• Self-report EI predicts performance largely through its overlap with personality, ability and self-perceptions.
• Apparent links between EI and leadership shrink sharply when different people rate each.
• Training can raise EI scores; focus development on observable perceiving, understanding and regulating behaviors.

Sources

• Mayer, Salovey and Caruso, Emotional Intelligence: New Ability or Eclectic Traits?, American Psychologist (September 2008)
• U.S. Office of Personnel Management, Emotional Intelligence Tests
• Joseph and Newman, Emotional Intelligence: An Integrative Meta-Analysis and Cascading Model, Journal of Applied Psychology (January 2010)
• O'Boyle, Humphrey, Pollack, Hawver and Story, The Relation Between Emotional Intelligence and Job Performance: A Meta-Analysis, Journal of Organizational Behavior (2011)
• Joseph, Jin, Newman and O'Boyle, Why Does Self-Reported Emotional Intelligence Predict Job Performance?, Journal of Applied Psychology (March 2015)
• Harms and Credé, Emotional Intelligence and Transformational and Transactional Leadership: A Meta-Analysis, Journal of Leadership and Organizational Studies (2010)
• Hodzic, Scharfen, Ripoll, Holling and Zenasni, How Efficient Are Emotional Intelligence Trainings: A Meta-Analysis, Emotion Review (2017)