What Our 2020 Data Got Right (and Wrong) About Crisis Resilience

In the middle of 2020, a few months into the pandemic, we looked at our own diagnostic data and found something that surprised us. Resilience hadn’t fallen. On average, across a sample of over 1,600 professionals and executives, it had gone up. Six years on, that finding is still worth revisiting, not as current data, but as a genuinely useful case study in how resilience actually behaves during the early phase of a major disruption, and why reading that phase correctly matters.

What the 2020 Data Actually Showed

We compared Resilience Diagnostic results from March to June 2019 (1,599 participants) against the same window in 2020 (1,659 participants), drawn from a similar pool of professionals and executives across comparable organisations. The average resilience ratio rose from 1.51 to 1.61, a roughly 7% increase during the acute early phase of a global crisis, not the decline anyone would have predicted going in.

The pattern wasn’t uniform. Male participants improved more clearly, from an average of 1.57 to 1.70, with the proportion at risk actually falling slightly. Female participants showed no change in the average ratio, but the proportion at risk rose by 15%, with increases specifically in anhedonia, self-doubt, and depressive symptoms. This tracks with separately well-documented research from the same period on the disproportionate caregiving and domestic burden many women carried during pandemic-era lockdowns and school closures, a genuine and important asymmetry inside an otherwise positive headline number.

At a category level, mental strength, physical energy, stress mastery, and related risk categories were slightly better during the crisis window. Emotional engagement, disengagement, and depressive symptoms were slightly worse. Specific strength factors that improved included purpose, integrity, fulfilment, compassion, situational agility, optimism, and sleep quality. Factors that declined included trust, flow, decisiveness, emotional agility, and empathy.

Why This Isn’t as Strange as It Sounds

Psychologists Leonard Zunin and Diane Myers developed a framework, widely used in disaster psychology, describing the emotional phases most communities move through during a prolonged crisis: warning, impact, heroic, honeymoon, disillusionment, and reconstruction. The heroic phase is marked by altruism and mobilisation, first responders, healthcare workers, and communities pulling together with real intensity. The honeymoon phase that often follows brings a genuine, if short-lived, sense of optimism and community bonding, along with a belief that things will return to normal reasonably quickly.

March to June 2020 sits squarely inside that window for most of the world. Our data wasn’t capturing a crisis that had resolved well. It was capturing a population still inside the heroic and honeymoon phases, genuinely mobilised, purposeful, and pulling together, before the far longer disillusionment phase that followed. That later phase, well documented separately across 2020 to 2022, brought the more severe and prolonged mental health decline most people actually associate with the pandemic: rising burnout, anxiety, and depression as the crisis stretched on far longer than the initial mobilisation could sustain.

Seen through that lens, the 2020 finding isn’t really a story about resilience being immune to crisis. It’s a story about which phase of a crisis a snapshot happens to catch, and how easy it is to mistake a heroic-phase mobilisation for a resolved, stable new normal.

The Transferable Lesson

This matters well beyond one pandemic. Any organisation moving through a major disruption, a restructuring, a market downturn, a leadership change, a merger, will likely see something similar in its early resilience or engagement data: a genuine, real mobilisation effect in the first weeks or months, purpose and cohesion measurably up, as people pull together around the immediate challenge. Reading that early signal as evidence the disruption has been successfully absorbed is a real risk. The honest interpretation is closer to: the team is currently in its heroic or honeymoon phase, and the real test comes later, when the mobilisation naturally fades and the underlying, unresolved strain has had time to surface.

Two practical implications follow. First, treat early positive signals during any crisis with appropriate caution rather than as confirmation the hard part is over. Second, plan deliberately for the later phase, the point where disillusionment risk peaks, rather than being caught off guard by a decline that the Phases of Disaster framework would have predicted from the outset.