Advice about the strategic pause is usually delivered as though silence were a lever with one setting. Hold your tongue, let it hang, watch the other person fill the gap. It is the kind of tip that sounds like insider knowledge and is almost never accompanied by a number.
Two pieces of research put numbers on it. They measured silences of almost exactly the same length, in two different rooms, and found opposite results. Read side by side they are far more useful than either would be alone, because between them they say what a pause actually does and where it stops working.
Three seconds at a negotiating table
Jared Curhan and colleagues at MIT and four other business schools recorded simulated negotiations and ran the audio through speech detection software working in 25-millisecond frames, so that every silence in every conversation could be found and timed. They took three seconds as the threshold for an extended pause, following an older convention in the field.
The Artful Age
A weekly letter on aging well, family across generations, and the creative life after the kids leave home.
Across 54 negotiating pairs, the frequency of pauses between three and 17.5 seconds correlated with value creation, meaning the sum of both sides’ points rather than either side’s share, at .32. Pauses shorter than three seconds showed no significant association. The strongest associations sat between roughly three and a half and nine and a half seconds.
The obvious objection is that value creation takes time and longer negotiations have more room for pauses. The researchers controlled for the length of the negotiation and the association held, while length by itself predicted nothing.
The more interesting check concerns which came first. Offers that put several issues on the table at once, which is the behavior that grows the pie rather than splitting it, were 1.73 times more likely to appear in the five speaking turns after a pause than at any other point in the negotiation. In the five turns before a pause, they were no more likely than usual. Whatever was happening, the silence came first and the creative offer came after.
Then there is the result that undoes the folklore. The pauses had no relationship with how the winnings were divided. Not a small one. None, across the observational study and the experiments, including a follow-up analysis checking whether silence made outcomes more lopsided in either direction. Neither party rated the process or the relationship worse for having been on the receiving end of it. Whatever a strategic pause is doing, it is not squeezing the other person.
Three further experiments point at what it is doing instead. Instructing one party to pause raised joint value against a control, and instructing both was no better than instructing one. The effect ran through what the researchers call a deliberative mindset, and pausing was associated with a drop, measured before and after, in the assumption that the other side’s gain must be your loss. In the last study, telling negotiators to use silence beat telling them to problem-solve.
Four seconds among friends
Now the other room. Namkje Koudenburg, Tom Postmes and Ernestine Gordijn at Groningen edited a four-second silence into a six-minute video of three students talking, placed immediately after the person participants had been asked to identify with voiced an opinion. A pilot study had tested 2.5, 4 and 6 seconds and settled on four as the length people would not consciously register.
It worked. Asked afterwards how long the gap had been, participants in the edited condition gave the same answer as everyone else, around three and a half seconds either way. They had not noticed.
They were affected anyway. Compared with both the uninterrupted version and a no-video baseline, the four-second group reported more feelings of rejection, more negative emotion, less positive emotion, less belonging, lower self-esteem, less sense that their view had been validated, and less confidence that the group agreed with each other. A scenario study with 102 participants found the same pattern.
Four seconds. In a conversation the participants were only watching.
Why the same silence does two opposite things
The difference is not the length. It is what the pause follows and what the room expects.
In a negotiation the silence lands after an offer, in a setting where taking a moment is the expected activity, and the person doing the pausing is the one it benefits. The gain was almost entirely internal: it changed what the pauser thought, not what the counterpart did.
In an ordinary conversation the silence lands after somebody has said what they think, and the Groningen researchers suggest we treat conversational smoothness as a rough proxy for agreement, without ever deciding to. When the flow breaks, the speaker updates, and what they update toward is the sense that the room did not endorse them. Notably, the group that saw no video at all reported the same contentment as the group that saw the smooth one, which suggests fluency is what we assume unless something interrupts it.
One variable turns up in both bodies of work. When the negotiators’ relative status was made salient, pauses by the higher-status party created value and pauses by the lower-status party did not, and the lower-status parties rated the process worse when they tried it. The negotiation researchers took that idea from the Groningen group’s own later work, so this is two literatures in contact rather than two independent discoveries, but it points the same way: the same three seconds means one thing from a person whose deliberation is assumed to be worth waiting for, and another thing from someone whose is not.
The honest limits
All four negotiation studies were run in a lab or a class exercise, and the authors say field research is needed. The first one was observational, so it cannot establish cause, and the researchers could not tell whether a given silence was deliberate or someone checking their notes. Everything was done in the United States with mostly native English speakers, and the authors state plainly that silence is a nonverbal cue whose meaning may not survive translation.
The Groningen studies used a written scenario and an observed video, not a real conversation. The authors argue this makes their test conservative and that real life would hit harder, which is plausible but not demonstrated.
And there is no established upper limit. The longest pause in the entire negotiation data set was 25.5 seconds, so nothing beyond that was tested. The negotiation authors do note, citing earlier work by Harinck and De Dreu, which is reported here secondhand, that breaks of three to five minutes made negotiators more competitive rather than less, apparently because they spent the time ruminating. Somewhere between three seconds and three minutes, the benefit turns around.
What is left
Pause after a proposal. Be more careful pausing after a person has just told you what they believe, where the silence you experience as thinking is being read across the table as a verdict. And hold the advantage loosely, because what the evidence actually found was not a lever on other people at all. It was three seconds in which somebody stopped assuming the other side’s gain had to be their loss.