Demand-withdraw: the research on one partner pushing and the other pulling away

By EQTiger Editorial. Review status: needs-expert-review. Qualified review is pending; the review date records an editorial check. This is general education for adults, not therapy or couples counseling, and EQTiger does not diagnose anyone, including your partner.

What researchers mean by demand-withdraw

In the research literature, "demand" covers nagging, criticizing, pressing for a discussion, and pushing for change. "Withdraw" covers going silent, changing the subject, giving minimal answers, refusing to discuss, and physically leaving. The pattern is the combination: the more one person demands, the more the other withdraws, and the more the other withdraws, the more the first demands. Researchers measure it in two main ways, by asking partners to rate how often typical sequences happen at home, and by coding couples' behavior while they discuss a real disagreement in the lab.

The phrasing matters because it shifts attention from two individuals to a loop. A partner who "nags" and a partner who "shuts down" are each reacting to the other in real time.

How strongly is it linked to unhappiness?

The best single answer comes from a 2014 meta-analysis by Paul Schrodt, Paul Witt and Jenna Shimkowski, which combined 74 studies with a total of 14,255 participants (Schrodt, Witt, and Shimkowski, 2014). Across those studies the pattern had a moderate association with worse outcomes (an average correlation of about .36). The links were strongest for relationship outcomes such as satisfaction (about .42) and somewhat stronger in distressed samples (about .41).

A correlation of that size is meaningful in social science, but it is not destiny. Plenty of couples with some demand-withdraw are satisfied, and plenty of unhappy couples have other problems. Most of the pooled studies were also correlational (Schrodt, Witt, and Shimkowski, 2014), which leaves the direction open: the pattern may wear a relationship down, unhappiness may make the pattern more likely, or both.

Is it a gender thing?

Early studies of married heterosexual couples found wife-demand, husband-withdraw sequences more often than the reverse, and this became a popular stereotype. But the more careful work points to something else.

Andrew Christensen and Christopher Heavey observed 31 married couples in two discussions each: one about a change the wife wanted and one about a change the husband wanted (Christensen and Heavey, 1990). Wife-demand, husband-withdraw was more common overall, but each spouse was more likely to demand when it was their own requested change and to withdraw when it was their partner's.

Brian Baucom, Pamela McFarland and Andrew Christensen then studied 75 couples, including 20 lesbian, 15 gay male, 20 cohabiting heterosexual and 20 married heterosexual couples (Baucom, McFarland, and Christensen, 2010). Same-sex and different-sex couples showed the pattern in very similar ways. Partners demanded more on their own topic, and more demand-withdraw went with lower satisfaction for every type of couple. The subgroups were small, but the result points the same way as the earlier study.

The meta-analysis also found that the two directions (wife-demand and husband-demand) were associated with poor outcomes to a similar degree, around .38 and .39 (Schrodt, Witt, and Shimkowski, 2014).

The practical reading: the person who wants something to change is in the demand seat. In many heterosexual relationships that is more often the woman, possibly because women more often want something in the relationship to change, but the roles flip when the topic flips.

The same couple, two topics

Here is what the role-flip finding looks like in an ordinary week. Priya wants Sam to stop committing them both to weekend plans with his college friends. When she raises it on Tuesday, she is the one pressing ("we need to talk about this, it keeps happening") and Sam gives short answers and starts loading the dishwasher. On Thursday, Sam brings up something he wants changed: Priya checking work email in bed. Now Sam is the one following up ("can we actually decide something?") and Priya says she is too tired and turns off the light.

Neither of them is "the avoider" or "the nag". Each is guarding the thing they do not want to give up and pushing for the thing they want. Seeing that is often the first time a couple can talk about the loop without arguing over whose fault it is.

Why withdrawing is not simply "not caring"

Withdrawal often has a physical side. In a study of 73 married couples followed over four years, Gottman and Levenson found that couples whose interactions were poorly regulated showed more stubbornness, defensiveness and withdrawal, along with lower satisfaction and greater risk of separation (Gottman and Levenson, 1992). The Gottman Institute's practitioner writing describes "flooding", a state of high physiological arousal in which listening becomes hard, and recommends about twenty minutes of something calming before resuming (The Gottman Institute, 2013). Those specific numbers are presented as guidance rather than results from a cited trial.

Is demanding the real problem?

It would be easy to conclude that the partner raising the issue should just stop. The research does not support that. In a year-long study of 61 couples, direct attempts to get a partner to change, including blunt negative ones, felt unsuccessful at the time but predicted more actual change over the following year, while softer indirect approaches felt better and produced little change (Overall and colleagues, 2009).

So the goal is not silence from the demander and endurance from the withdrawer. It is a version of the conversation that keeps the request clear while lowering the threat.

What the research cannot tell you

• Which partner is "right". The pattern describes a process, not a verdict on whose request is reasonable.
• Whether your relationship will last. Moderate average links do not predict a single couple's future.
• Whether something more serious is happening. Some withdrawal is fear, and some "demanding" is control. The studies above describe ordinary conflict between partners who are safe with each other. If one person is afraid, the issue is safety, not communication style.

What to try

1. Notice which seat you are in on this topic, and remember you may be in the other seat on a different topic.
2. If you are demanding: make one clear request, then ask for a time to discuss it rather than pressing now. See how to bring up a recurring problem.
3. If you are withdrawing: replace disappearing with a named pause and a return time. See my partner shuts down during arguments.
4. If the loop has gone on for months, couples therapy is a reasonable next step: a meta-analysis of 58 studies found large average gains in satisfaction and better communication (Roddy and colleagues, 2020).

Key studies

Start with the free EQTiger assessment (adults 18+), or browse more relationship help.