The Data Behind Chapter 4:

Leading Through Friction

Source: The American (National) Emotions Survey — Friction, Blankenship & Sass
Sample: 1,188 working U.S. adults, fielded October 2025–January 2026 via a representative panel

About this page: Chapter 4 of the book distills these findings into nine core insights for a general reader. This page reproduces the full underlying data including exact percentages, generational breakdowns, sample sizes, and regression detail, for readers who want to go deeper. All figures below reflect the analytic dataset (N = 1,188) used in the published chapter.

Chapter 4 Findings, In Detail

We already know where we're going.
In the book: Roughly 60% of respondents said emotional intelligence will matter even more in the future of work.
Survey question:“How important will emotional intelligence (EQ) be for human success at work as AI takes on more tasks?” (Top-2 = Very + Extremely important)

Response breakdown: 26.3% of respondents (313 people) said EQ will be Extremely important, and 33.1% said Important, an additional 23.8% said Moderately important. Millennials and Gen Z lead the way at 62–63%; Boomers and the Silent Generation are lowest, but still roughly half.

Survey question: “To what extent do you agree — Showing emotions at work is seen as a sign of weakness” (Top-2 = Somewhat + Strongly agree)

Millennials, the generation most associated with popularizing emotional-intelligence language and “bringing your whole self to work”, endorse the emotions-as-weakness view at a higher rate than any other cohort, including Gen X and the Boomers who are often assumed to be the norm’s originators.

Survey question: “To what extent do you agree — When leaders hide their emotions, it creates more problems than it solves” (Top-2 = Somewhat + Strongly agree)

In regression models controlling for age, gender, leader role and level, industry, and other emotional norms (see Model 1A, Appendix), agreement with this item is a moderate, statistically significant predictor of both emotional drain (standardized β = .135, p < .001) and inauthenticity (β = .186, p < .001).

Leaders vs. Non-Leaders: Emotional Drain

Leaders report meaningfully more emotional drain than the people they lead, not less. This is consistent with the chapter’s “leader-as-thermostat” idea: Leaders don’t just feel pressure to hide emotion, they also personally absorb more of the organization’s emotional weight. 

The generational gap is nearly a full point (1.29) between the oldest and youngest cohorts, though the gap between Millennials and Gen Z specifically is small (0.12).

Supporting comparison — emotional-openness measures where younger cohorts diverge from older ones

External context (not from this survey): A 2024 UKG survey found 83% of Gen Z frontline workers report feeling burned out (higher than any other working group). Eagle Hill Consulting found 66% of Gen Z workers report burnout, vs. 37% of Boomers, and that Gen Z and Millennial workers reach peak burnout around age 25 — roughly 17 years earlier than the survey’s national average of 42.

Top-2 Box across three emotional-openness measures, by generation

Millennials and Gen Z are essentially tied for the lead across all three measures. Millennials edge out Gen Z on two of the three, and the two cohorts trade the top spot on the third by about a point. Both groups run well ahead of Gen X and the Boomers.

Across three systemic-norm indicators combined into a single index, the average American workplace scores 3.04 out of 5 — squarely in the middle, meaning most workplaces send mixed rather than clearly open or clearly restrictive emotional signals when colleagues express emotion, with Millennials and Gen Z together leading a cultural shift.

What actually predicts lower emotional drain (Model 1A, full controls)

The specific, learnable skill of “confidence managing negative emotions” is what buffers drain — not overall EQ as a single composite score.

"Emotions are a weakness at work" as a predictor, controlling for EQ (Model 1A)

This is the single strongest and most robust regression result in the entire dataset. Systemic emotional norms average 3.04 out of 5 across the sample, meaning most workplaces are mixed-to-restrictive rather than clearly emotionally safe.

EQ × Weakness-Norm interaction (moderation model)

The two other tested interactions — EQ × leader-pressure-to-hide, and confidence-managing-negative-emotions × leader-pressure-to-hide — were not statistically significant (p = .302 and p = .432, respectively). The one reliable moderation effect in the data is specifically EQ awareness sharpening the felt cost of a weakness-punishing culture.