Feel Less, Claim More: What 105 Studies Say About Emotional Intelligence in Negotiation

There is a slide in almost every negotiation workshop I have sat through. It has a brain on one side, a heart on the other, and a headline about emotional intelligence being one of the negotiator’s secret weapons. Read the room, manage your emotions, empathize with your counterpart, and the deal follows.

A couple of months back, an interesting paper came out that pools every empirical study on emotional intelligence in negotiations published up to June 2025. Remigiusz Smolinski and Christoph Streer at HHL Leipzig screened 1,793 records, kept 105 studies, and ran a meta-analysis on the 37 that reported usable statistics, 155 effect sizes in total.¹

The Numbers and the Dataset

Across all 155 effect sizes, emotional intelligence (EI) correlates with negotiation outcomes at r = 0.209. By textbook conventions, that is a small effect. It is also a stable one: highly significant, no sign of publication bias, and pointing the same way whether EI was measured by questionnaire or by ability test.

The interesting part is what happens when you split by outcome type. For relational outcomes such as trust, rapport, and communication quality, the correlation rises to r = 0.262. For economic outcomes such as individual gain or joint value, it falls to r = 0.140, and the difference is statistically significant. In the handful of studies that used ability-based EI tests rather than self-report, the effect on economic outcomes was not significant at all, while the effect on relational outcomes was the strongest in the whole dataset.

So, emotionally intelligent negotiators do build better relationships. Whether they take home more money is a much weaker claim, and in some settings, the sign flips.

Where It Stops Working

Two boundary conditions caught my attention in this study.

The first is structure. In negotiations where one person faces one person, EI effects were consistently stronger. In multiparty and role-play settings, they weakened and in several cases turned negative. The authors call this exploratory, but the pattern held across their data.

The second is competition. The paper lists six studies in which high EI went with worse individual results. High-EI negotiators created more value but claimed less of it, because empathy turned into concessions.² Under power asymmetry, emotionally responsive negotiators gave more away. Negotiators who had to suppress anger lost the attention they needed for strategy. In one intercultural setting, EI on its own was not enough; it needed cultural intelligence alongside it. And in competitive bargaining, counterparts who showed inconsistent emotions did better than the consistent, high-EI ones.

The authors’ own summary is the one I would keep: EI becomes a liability where empathy, regulation, or sensitivity to emotional cues outweigh assertive, goal-oriented behaviour.

What This Means for Building Trust at a Distance

My research asks how trust is built and kept in large, complex negotiations now that most of them run over Teams and Zoom rather than across a table. The sixteen senior negotiators I interviewed across Europe told me, in different words, the same thing. Trust has not disappeared; it has changed form. The affective trust of the dinner and the handshake has given way to what I call cognitive trust: punctuality, clear summaries, kept promises, a professional presence with the camera on, and consistent follow-up. I wrote about that shift in “Why Trust Is Harder to Build on a Screen”.

This paper provides more concrete evidence that connects those interview findings to the wider literature on the negotiator as a person. EI works on emotional cues. Four of the five components in the authors’ model (self-awareness, self-regulation, empathy, and social skills) run on exactly the verbal, vocal, and visual signals that a video call strips out. A gallery of head and shoulders tiles is not a room. There is procurement, legal, finance, and the business on one side and a mirror image on the other, often five or more people on the call. The two conditions under which EI effects fade in the meta-analysis, lean channels and many parties, are the normal conditions of a large complex deal in 2026.

I do not read that as emotional intelligence becoming irrelevant. I read it as EI moving in a different direction. Negotiators who can no longer read the room learn to read the record instead: the tone of the summary email, who turned the camera on and who did not, whether the counterpart did what they said they would do by Friday, and how the other side behaves when the screen share fails. The competency is the same one; different signals now feed it. I suspect this is also why the redistribution of power on screen that I described in “Everyone Gets the Same-Sized Square” has unsettled experienced negotiators so much. The emotional signals they spent decades learning to read at the head of the table simply are not transmitted anymore.

My participants described attention, empathy, and cultural awareness as the human factors that AI cannot replace. The meta-analysis agrees with them, and adds a warning they did not always voice: in a distributive phase, empathy without assertiveness costs money.

It also confirms a design choice in the Business Negotiation Framework that came out of the research, which I introduced in “Introduction to Disruptive Negotiations”. In the framework, EI complements preparation built on data and a structured process. It does not carry the deal on its own. This paper is the quantitative case for putting it there.

What the Paper Cannot Tell Us

A few things worth knowing before anyone quotes r = 0.21 on a slide.

Most of the strong relational findings come from studies in which the same person self-reported both their EI and the outcome, so common method bias cannot be ruled out; the authors say as much. The ability-based stream rests on four studies. The primary research is dominated by student simulations, mostly buyer-seller dyads in a classroom, which is a long way from a contract that runs for several years. And, most relevant to me, nobody in this dataset tested the communication medium as a moderator. Face-to-face, video, and email were never separated. Cultural norms and power dynamics could not be tested either, because fewer than five primary studies report them.

Those gaps are, more or less, the space my dissertation works in: European B2B, multiparty, virtual and hybrid, cross-cultural, with real power asymmetries. Qualitative research cannot produce an effect size.

My Takeaways

If you negotiate large, complex deals remotely, the evidence points to three things.

Invest in emotional intelligence because it reliably buys you trust and a working relationship. Do not expect it to buy you a bigger number on the term sheet.

Expect its effect to weaken as the call grows and the channel thins, and compensate on purpose through process and visible reliability. Every summary email, every kept promise, every session with the camera on is a brick in the trust architecture. In a virtual multiparty deal, those bricks carry more weight than your ability to sense a mood.

And in competitive phases, keep empathy on a leash. The studies where high-EI negotiators lost money are the ones where they felt too much and claimed too little.

Thanks again, Tim H.

@misc{herglotz2026dn09,
  author       = {Herglotz, Tim},
  title        = {{Feel Less, Claim More: What 105 Studies Say About Emotional Intelligence in Negotiation}},
  year         = {2026},
  month        = {9},
  number       = {9},
  howpublished = {Tim Herglotz, entry № 09},
  url          = {https://timherglotz.com/feel-less-claim-more-emotional-intelligence-in-negotiation/}
}