Remote vs. In-Office Decision Load: What the Numbers Show

The debate over remote and in-office work tends to center on productivity, collaboration, and culture. What it rarely examines is decision load — specifically, whether working location changes how many unresolved decisions a person carries at any given time.

An earlier version of this note said a Decision Load Index dataset of 901 participants showed a pattern here. That was not true. 901 was roughly the number of people who had signed up for our assessment; almost none completed it, and we hold no dataset comparing remote and in-office workers. What follows is a hypothesis: if there is a gap, it may lie less in decision difficulty than in how reliably decisions get resolved.

The Ambient Resolution Mechanism

In-office environments contain ambient resolution mechanisms that most workers do not consciously notice. A question about project status gets answered during the walk to the coffee machine. A decision about meeting timing resolves when two people happen to be near each other. Uncertainty about a colleague’s reaction is resolved by their visible behavior.

These are not formalized decision processes. They are environmental features that close open loops without requiring a deliberate decision act. The open loop opens — someone asks a question, sends a request, or leaves an ambiguous comment — and the office environment provides resolution pathways that operate in parallel to the formal work stream.

Remote work removes these pathways. Every open loop that would have been resolved by ambient contact must now be resolved by a deliberate action: a Slack message, a calendar invite, a follow-up email. The decision is not harder. But it stays open longer, and while it is open, it occupies working memory.

What We Expected the DLI Subscales to Show

The DLI breaks decision load into five dimensions. Our hypothesis was that a remote/in-office gap would concentrate in one: unresolved inputs, with smaller and less consistent differences in the other four (context switching, prioritization, ambiguity, and emotional weight). None of this was measured.

These are expectations, not results. No participant data supports any row of this table.

The Hybrid Bimodal Pattern

A second hypothesis concerns hybrid workers: their load may be bimodal rather than average. On office days they may face higher context-switching load — concentrated meeting schedules, higher social coordination demands. On home days they may face higher unresolved-input load — async communication cycles extending decision resolution from minutes to hours.

The result is that hybrid workers do not sit between the remote and in-office populations on total DLI score. They experience both patterns on alternating days, with a narrow mid-week window where neither concentration is at peak. This has not been observed in any dataset we hold.

The Mechanism: Async Communication and Open-Loop Duration

Synchronous communication closes open loops quickly. A question asked in person or on a video call has a response latency measured in seconds. An async message — Slack, email, a comment in a document — has a response latency measured in hours. During that window, the open loop remains in working memory.

This is not a claim that async communication is worse than synchronous communication. Async communication has known advantages: it allows deeper work, reduces interruption, and accommodates time zones. The observation is narrower: async communication extends the duration for which decisions stay unresolved, and our hypothesis is that open-loop duration matters more than open-loop count.

In other words, ten decisions that each resolve within five minutes may generate less cognitive load than two decisions that remain unresolved for four hours each.

What This Does Not Mean

This field note is not an argument for returning to the office. The in-office population shows its own load patterns — higher context switching, more social coordination overhead, and meeting density that remote workers do not experience. The finding is about mechanism, not recommendation.

It is also not a claim that remote workers perform worse. Decision load measures cognitive overhead, not output quality. High-load states are uncomfortable and may degrade certain types of performance over sustained periods, and we have no data on output quality.

An earlier version of this note reported a 22% unresolved-input gap. That figure had no dataset behind it and should not be cited.

Corrections, 28 September 2026: An earlier version of this note presented a 22% remote/in-office gap, a dimension-by-dimension comparison and a hybrid-worker pattern as findings from a 901-participant Decision Load Index dataset. No such dataset exists: 901 was roughly our sign-up count, almost no one completed the assessment, and the cited DOI is a conceptual overview with no participant data. We have removed the figures and relabelled the rest as hypotheses.

This field note sets out hypotheses. It is not based on participant data. This content is educational and does not constitute professional advice. Individual results vary.