US Worker Engagement Hits 29% as Remote Staff Fall Fastest

Twenty-nine per cent of American employees were engaged at work in the third quarter of 2026, the lowest share Gallup has recorded since 2011. The figure, published on 29 October 2026, is down two percentage points from the previous quarter. Alongside it, the share Gallup classes as actively disengaged rose from 17 to 18 per cent, and 53 per cent were neither engaged nor actively disengaged. The readings come from self-administered web surveys of 21,648 adults conducted between 5 and 19 August 2026, with a margin of sampling error of plus or minus 0.9 percentage points.

Gallup has asked the same questions since 2000, when 26 per cent of US employees were engaged. The series sat between 26 and 30 per cent through 2013, peaked at 36 per cent in 2020, fell to 31 per cent in 2024 and stayed there through the first half of 2026 before this quarter's drop.

Gallup puts a headcount on each point: roughly 1.6 million US employees. On that arithmetic the quarterly fall is about 3.2 million fewer engaged workers, and the decline since the 2020 peak is around 11 million.

Where the two points came from

The fall is not spread evenly. Engagement dropped five points among employees who work exclusively remotely and two points among hybrid employees. Among on-site staff, and among people whose jobs cannot be performed remotely at all, it moved by about one point. Gallup states that the declines among remote-capable, hybrid and exclusively remote employees exceeded the margin of error while the changes in the other two groups did not.

Two underlying measures moved with it. The share who strongly agreed they knew what was expected of them at work had been 49 per cent across the first two quarters of 2026; in the third it fell three points to 46 per cent, matching the 2024 annual low and sitting ten points below the series high of 61 per cent in 2015. The share who strongly agreed that their supervisor or someone at work cares about them as a person fell from 40 per cent to 37 per cent, against a peak of 47 per cent in March 2020. Gallup had already flagged both items as its largest five-year decliners in its analysis of 2025 trends.

The AI premium is thinning where people are furthest away

In July, Gallup reported that engagement ran six points higher at organisations that had adopted AI, fifteen points higher where the employer had communicated a clear AI plan, and eighteen points higher where managers actively supported AI use. Where frequent use, a clear plan and manager support were all present, engagement reached 53 per cent. That was the encouraging version of the story.

The third-quarter data complicates it. In the second quarter, exclusively remote employees at AI-integrated organisations were five points more engaged than their counterparts elsewhere, and hybrid employees were eight points ahead. By the third quarter both gaps had narrowed to two points. Among on-site workers the advantage held at roughly four to five points. The benefit, in other words, is surviving where people share a building and eroding where they do not.

Jim Harter, Gallup's chief scientist for workplace and the author of the release, writes that since 2020 hybrid and fully remote employees have seen larger declines in clarity of expectations than on-site employees in remote-capable jobs, and adds:

That pattern sits alongside what employers have been disclosing about their own workforces. Anthropic's employment data, which Parallax Nexus reported on 16 October, showed the junior share of its technical hires falling from 27 per cent to 9 per cent since early 2025. The composition of work is changing at the same time as the instructions for doing it are getting harder to read, and the people least likely to overhear a clarification are the ones logging in from home.

The judgment worth stating plainly is that this is a management-communication reading more than a technology reading. Gallup's two moving items are both about being told things: what the job is, and whether anyone has noticed you doing it. Both are cheaper to fix than an AI deployment and neither is fixed by one.

The strongest case against reading too much into it is that engagement is a self-reported attitude, not a measure of output, and a two-point quarterly move in a series that has sat in a nine-point band for a quarter of a century is not a structural break. Gallup's own note that on-site changes fell inside sampling error limits the claim further: this is a specific deterioration among remote and hybrid staff, not a general collapse. What the data cannot say is whether AI adoption caused the divergence or merely arrived during it, because Gallup is measuring correlation across work arrangements rather than running an experiment.