Coitus Interruptus

Abstract

Coitus interruptus, the withdrawal method, is a contraceptive practice in which the penis is removed from the vagina before ejaculation. Its effectiveness depends not on a device but on an act of self-regulation executed at the moment of peak sexual arousal, which makes it a natural case study in the psychology of self-control. Research on visceral influences shows that intense drive states narrow attention to the present and discount future consequences, widening the gap between a calm intention to withdraw and the behavior that follows. This article examines withdrawal through four converging literatures: visceral influences on decision-making, the hot/cool-system account of self-regulation, the intention–behavior gap, and the epidemiology of contraceptive failure, which records a large divergence between perfect-use and typical-use pregnancy rates.

Keywords: withdrawal method, self-regulation, visceral influences, intention-behavior gap, contraceptive failure

Coitus interruptus is among the oldest recorded methods of birth control, and it remains in wide use worldwide, often alongside or in place of modern contraception. Its defining feature, for a psychologist, is that it relocates the point of possible failure from a manufacturing tolerance to a lapse of behavioral control: there is no device to fail, only a person who must interrupt an act at its most physiologically compelling moment. This places the method squarely within the study of self-regulation, and in particular within the account of how intense drive states reshape judgment. Loewenstein (1996) named these drive states visceral factors and argued that at high intensity they crowd out deliberate goals, bias attention toward immediate gratification, and produce behavior the same person would not have chosen in a cooler state. Withdrawal asks a reasoned contraceptive intention to survive exactly such a moment, and the empirical record of how often it does not is unusually precise.

What Coitus Interruptus Is

Coitus interruptus is defined by a single behavioral requirement: during vaginal intercourse the penis is withdrawn and ejaculation occurs away from the partner's genitals. No barrier, hormone, or device is involved; the method consists entirely of the timing and execution of one action. For that reason it has no shelf life, no cost, and no side effect profile, and it is always available, which partly explains its persistence across cultures and its frequent use as a backup or stopgap method (Jones et al., 2009). The scientific indexing vocabulary files it under contraception and under the broader heading of coitus, but its substance for cognitive psychology is not reproductive physiology. It is the problem of enacting a plan whose execution coincides with the moment that most undermines planned behavior.

Two physiological facts set the difficulty of that execution. First, the window for the action is brief and its cue, the sensation preceding ejaculation, is the same sensation that most strongly pulls toward continuing. Second, pre-ejaculatory fluid may carry viable sperm in some men, so even flawless timing does not guarantee the absence of risk. The first fact is psychological and is the focus of this article; the second sets a floor on the method's effectiveness that no amount of self-control can lower. Together they explain why withdrawal is simultaneously the method most dependent on the user and one whose best-case performance is still bounded.

Self-Regulation Under Arousal

The core psychological question withdrawal poses is why a firmly held intention to act fails precisely when it must be executed. The most direct evidence comes from an experiment designed to measure the effect of arousal on judgment. Ariely and Loewenstein (2006) had young men rate their willingness to engage in a range of sexual and risk-taking behaviors both in a calm state and while sexually aroused, and found that arousal produced large, systematic increases in reported willingness to take risks, including risks the same men rejected when calm. The study operationalized what Loewenstein's earlier theory had proposed: a visceral state does not merely add a competing motive, it reshapes the entire decision, inflating the appeal of immediate action and shrinking the weight placed on consequences that are certain but deferred. A further consequence is the hot-cold empathy gap: a person in a cool state systematically underestimates how a future aroused state will alter their own preferences, so a contraceptive plan laid in calm deliberation is built on a forecast that the aroused self will predictably violate.

This maps onto a general architecture of self-regulation. Metcalfe and Mischel (1999) proposed that behavior is governed by two interacting systems: a cool system that is reflective, slow, and strategic, and a hot system that is emotional, fast, and reflexive, driven by appetitive cues. Self-control succeeds when the cool system governs and fails when the hot system, activated by a strong stimulus, takes over. Withdrawal is a textbook case of hot-system provocation: the cue that should trigger the planned action is also the most powerful activator of the system that opposes it. The same research tradition grew out of the study of delay of gratification, in which Mischel et al. (1989) showed that the ability to defer a smaller immediate reward for a larger later one depends on strategies that keep attention off the hot, appetitive features of the tempting stimulus. A contraceptive plan that relies on withdrawal is a delay-of-gratification problem in which the tempting stimulus cannot be cognitively distanced, because it is the act itself.

The demonstration above models the hot-system effect directly. A person's baseline willingness to follow a cautious plan is set in a cool state; raising the arousal level applies a multiplicative pull toward the immediate option, so the same plan that is easily honored at low arousal becomes progressively harder to execute as the hot system is engaged. The pattern reproduces the central finding of Ariely and Loewenstein (2006): the shift is not a small perturbation but a reordering of what the decision-maker, in the moment, prefers.

Figure 1

How a Hot State Overrides a Cool Contraceptive Plan

The Efficacy Gap

Contraceptive effectiveness is reported in two numbers that make the psychology of withdrawal quantitatively visible. Perfect-use failure is the proportion of couples who experience a pregnancy in a year while using the method exactly as intended on every occasion; typical-use failure is the proportion among all couples who report the method, including occasions of incorrect or inconsistent use. For most methods the two differ, but for withdrawal the divergence is wide, because the method's correct use is precisely the behavior that arousal undermines. Trussell (2011), compiling the standard United States estimates, placed withdrawal's perfect-use failure near 4% and its typical-use failure near 18–22% in a year. Sundaram et al. (2017), drawing on the National Survey of Family Growth, produced broadly comparable typical-use estimates and confirmed that the gap is not a statistical artifact but a stable feature of how the method is actually used.

The size of that gap carries a specific message. A device-based method's perfect-use and typical-use rates differ mainly through forgetting or misuse that is incidental to the moment of intercourse; withdrawal's gap is generated at the moment of intercourse itself, by the failure of in-the-moment control. This is why Jones et al. (2009) argued that withdrawal should be taken seriously rather than dismissed: for a couple who would otherwise use nothing, a method with a typical-use failure around 20% is a substantial improvement over the roughly 85% annual pregnancy rate of using no method, even though it is clearly inferior to consistent modern contraception. The method's value, in other words, is bounded both below by its dependence on behavior and above by its real but partial efficacy.

The demonstration above compounds a chosen annual failure rate over several years using the standard relationship, letting the reader see how a per-year rate becomes a cumulative probability. Moving between the perfect-use and typical-use rates shows how the behavioral gap, small in a single year, grows into a large difference in the probability of at least one pregnancy over the span a couple might rely on the method.

The Intention-Behavior Gap

Withdrawal's efficacy gap is a specific instance of a general finding in the psychology of behavior: intentions predict actions only weakly. Webb and Sheeran (2006) synthesized experimental studies in which intentions were actually changed and asked how much behavior changed as a result; they found that a medium-to-large change in intention produced only a small-to-medium change in behavior, establishing that intention, though causal, is far from sufficient. Sheeran and Webb (2016) framed the residual as the intention-behavior gap and catalogued its sources, prominent among them the hijacking of planned behavior by competing situational and visceral forces at the point of action. Contraceptive use is one of their recurring examples precisely because the behavior must be enacted in a state that opposes it.

The gap has an important asymmetry for withdrawal. The population that relies on the method is not one that lacks the intention to avoid pregnancy; by definition its members intend to use a method on every occasion. Their typical-use failures are therefore not failures of intention but failures of its translation into behavior under load. This is why interventions that only strengthen the motivation to avoid pregnancy tend to disappoint, and why approaches that target the moment of action have the stronger rationale. The best-evidenced of these is the implementation intention, an if-then plan that links a specified situational cue to a goal-directed response in advance; a meta-analysis by Gollwitzer and Sheeran (2006) found that forming such plans has a medium-to-large effect on goal attainment, precisely by delegating the action to a pre-committed cue rather than to in-the-moment deliberation. Cue planning and a shift to a method that removes the decision from the hot moment entirely share the same rationale.

The demonstration above represents the gap as a transfer function from intention strength to the probability of enacting the behavior, with a situational-load control that flattens the transfer. At low load a strong intention translates almost fully into action; raising the load, the analogue of a strong visceral state, progressively decouples behavior from intention, so that even a maximal intention no longer guarantees the act.

Worked Example

Consider the cumulative-risk relationship used in the second demonstration. If a method's probability of failure in a single year is p, and the years of use are independent, then the probability of avoiding pregnancy for n consecutive years is (1 − p)^n, and the probability of at least one pregnancy over that span is P = 1 − (1 − p)^n.

Take withdrawal at its typical-use rate of p = 0.20. Over a single year the risk is simply 0.20. Over two years, P = 1 − (1 − 0.20)^2 = 1 − 0.80^2 = 1 − 0.64 = 0.36. Over five years, P = 1 − 0.80^5 = 1 − 0.3277 = 0.672, so a couple relying on typical-use withdrawal for five years has about a two-in-three chance of at least one pregnancy. Now take the perfect-use rate of p = 0.04 over the same five years: P = 1 − 0.96^5 = 1 − 0.8154 = 0.185, under one in five. The five-year gap between perfect and typical use, 0.672 against 0.185, is far larger than the one-year gap of 0.20 against 0.04, because the behavioral shortfall compounds. The arithmetic makes the psychological point exact: the cost of the in-the-moment control failure is not paid once but accrues with every year of reliance.

Discussion

The convergent reading of these literatures is that coitus interruptus is less a contraceptive technology than a recurring test of self-regulation, and that its measured failure rate is a field estimate of how often that test is failed. The visceral-influence and hot/cool-system accounts explain the mechanism of failure; the intention-behavior gap locates it within a general theory of why planned behavior miscarries; and the epidemiological record quantifies its consequence. Each framework points at the same pressure point, the moment of peak arousal at which a cool-system plan must be executed by a hot-system-dominated agent.

This has a practical corollary that the psychology makes clear and the efficacy data confirm. Because the shortfall is located at the moment of action rather than in the strength of the underlying intention, the most effective way to improve outcomes for someone who wishes to avoid pregnancy is not to exhort greater resolve but to move the decisive decision out of the hot moment. Methods that are set in advance and require no in-the-moment act, from the daily routine of an oral contraceptive to the complete decision-independence of a long-acting method, succeed precisely by not asking the hot system for cooperation. Withdrawal, by contrast, asks for that cooperation at the worst possible instant, which is why it occupies a stable middle ground: markedly better than nothing, reliably worse than methods that do not depend on in-the-moment control, and unlikely to close the gap through motivation alone.

Cognitive Implications

Coitus interruptus is indexed as a contraceptive behavior, not as a cognitive construct, yet its entire reliability profile is a cognitive phenomenon. The method is an instrument for reading, in a real population and with hard outcome data, the degree to which deliberate control survives a strong appetitive state. The perfect-use rate measures what the cognitive system can do when it governs; the typical-use rate measures what it does when a competing drive state is engaged; and their difference is a naturalistic estimate of the strength of the visceral override that laboratory studies such as Ariely and Loewenstein (2006) induce under controlled conditions. In this sense the contraceptive literature and the self-regulation literature are measuring the same thing by different means.

The topic also sharpens a point about the architecture of control. Delay-of-gratification research established that successful self-regulation depends heavily on strategies that keep attention off the hot features of a tempting stimulus, cognitively distancing the reward (Mischel et al., 1989). Withdrawal is a case in which that distancing is structurally unavailable, because the tempting stimulus and the required action are the same event, with no possibility of looking away. It is therefore a boundary condition for the hot/cool model: a situation engineered, by its nature rather than by an experimenter, to maximize hot-system dominance over the very behavior the cool system has planned. The hot/cool account is one member of the broader family of dual-process models, which hold that a fast, automatic process and a slow, deliberate process jointly determine behavior; withdrawal is valuable to that framework precisely because it isolates the moment at which the automatic process wins.

Current Directions

The experimental study of arousal and risk has broadened from its original single-sample design toward questions of generality and mechanism. Skakoon-Sparling et al. (2016) extended the arousal-and-risk paradigm to both men and women and found that induced sexual arousal increased sexual risk-taking across sexes, addressing a limitation of the earliest work and reinforcing that the visceral effect on sexual decision-making is not specific to one group. The convergence of these results with the field data strengthens the claim that the typical-use failure of withdrawal reflects a robust feature of decision-making under arousal rather than an idiosyncrasy of any particular population.

A second current runs through contraceptive epidemiology. Sundaram et al. (2017) refined typical-use failure estimates using nationally representative survey data, part of a broader methodological effort to measure real-world effectiveness with less reliance on clinical-trial populations whose adherence is atypically high. Because the perfect-to-typical gap is where the psychology lives, improvements in measuring typical use are, in effect, improvements in measuring the population-level cost of in-the-moment control failure, and they keep the quantitative anchor for the self-regulation account current.

Common Misconceptions

Withdrawal is essentially useless as contraception.

Its typical-use failure of around 20% is far better than the roughly 85% annual pregnancy rate of using no method, which is why it is treated as a real if imperfect option rather than as equivalent to nothing (Jones et al., 2009). The dismissal persists because the method's failures are salient and its successes invisible.

The gap between perfect and typical use is just carelessness.

The gap is generated at the moment of intercourse by the override of a reasoned plan by a strong visceral state, a systematic effect demonstrated experimentally, not a matter of indifference (Ariely & Loewenstein, 2006). Framing it as carelessness misreads a predictable feature of decision-making under arousal as a character failing.

A strong enough intention to avoid pregnancy is sufficient to use withdrawal reliably.

Changing intention produces only a modest change in behavior, because intention is regularly overridden by situational and visceral forces at the point of action (Webb & Sheeran, 2006). The belief survives because intention feels decisive from the inside, even though it predicts behavior only weakly.

Glossary

Contraceptive failure rate.

The proportion of couples using a method who experience an unintended pregnancy within a year, reported separately for perfect and typical use.

Cool system.

The reflective, slow, strategic mode of control that forms and holds deliberate plans, including a plan to use a contraceptive method correctly.

Delay of gratification.

The capacity to forgo a smaller immediate reward for a larger later one, shown to depend on strategies that keep attention off the tempting stimulus.

Dual-process model.

An account of cognition in which a fast, automatic process and a slow, deliberate process jointly determine behavior; the hot/cool system is one such account for self-control.

Hot system.

The emotional, fast, reflexive mode of control activated by appetitive cues, which can override the cool system at the moment of action.

Hot-cold empathy gap.

The systematic failure of a person in a calm state to predict how a future visceral state will change their own preferences and behavior.

Implementation intention.

An if-then plan that links a specified situational cue to a goal-directed response in advance, shown to improve goal attainment by delegating action to the cue rather than to in-the-moment deliberation.

Intention-behavior gap.

The weak translation of intentions into actions; a changed intention produces only a smaller change in behavior.

Pearl Index.

A common measure of contraceptive failure expressing pregnancies per 100 woman-years of method use.

Perfect use.

Use of a contraceptive method exactly as intended on every occasion; its failure rate reflects the method's inherent limits rather than user behavior.

Pre-ejaculatory fluid.

Secretion released before ejaculation that may in some men contain viable sperm, setting a floor on withdrawal's effectiveness independent of timing.

Self-regulation.

The processes by which a person governs their own thoughts, emotions, and actions in the service of a goal, against competing impulses.

Sexual arousal.

A visceral state shown experimentally to increase willingness to take sexual and other risks relative to a calm baseline.

Typical use.

Use of a method as it actually occurs across a population, including inconsistent and incorrect use; its failure rate captures real-world performance.

Visceral influences.

Drive states such as hunger, pain, and sexual arousal that at high intensity bias judgment toward immediate gratification and crowd out deliberate goals.

Withdrawal method.

A synonym for coitus interruptus: removal of the penis before ejaculation so that semen is deposited away from the partner's genitals.

Key Researchers

Dan Ariely

. James B. Duke Professor of Psychology and Behavioral Economics at Duke University; demonstrated experimentally that sexual arousal sharply shifts decision-making toward risk-taking, quantifying the hot-cold gap that undermines plans to withdraw. Faculty Page - Wikipedia - Google Scholar

George Loewenstein

(b. 1955). Herbert A. Simon University Professor of Economics and Psychology at Carnegie Mellon University; formulated the theory of visceral influences on behavior, explaining how drive states crowd out deliberate goals at the moment of action. ORCID - Wikipedia - Faculty Page

Walter Mischel

(1930-2018). Robert Johnston Niven Professor of Humane Letters in Psychology at Columbia University; pioneered the study of delay of gratification and self-regulation, and with Metcalfe formalized the hot/cool-system account of willpower. Wikipedia - Wikidata

Paschal Sheeran

. Stephen J. Walsh Distinguished Professor in the Department of Psychology and Neuroscience at the University of North Carolina at Chapel Hill; quantified the intention-behavior gap and the conditions under which good intentions fail to control action. ORCID - Faculty Page - Google Scholar

James Trussell

(1949-2018). Charles and Marie Robertson Professor of Public and International Affairs, Emeritus, at Princeton University; produced the standard United States estimates of contraceptive failure rates, including the large perfect-use to typical-use gap for withdrawal. Faculty CV - Wikidata

Frequently Asked Questions

How effective is the withdrawal method?

Its perfect-use failure is near 4% a year, while its typical-use failure is near 18 to 22%, a wide gap that reflects how often the required in-the-moment control is not executed (Trussell, 2011).

Why does withdrawal fail so much more in typical use than in perfect use?

The correct use of withdrawal is a deliberate act that must be performed at peak sexual arousal, and intense visceral states systematically shift decisions toward immediate reward and away from planned consequences (Ariely & Loewenstein, 2006).

Is withdrawal better than using no contraception at all?

Yes. A typical-use failure around 20% is substantially lower than the roughly 85% annual pregnancy rate of using no method, which is why it is regarded as a real if limited option (Jones et al., 2009).

What is the hot/cool-system explanation of why withdrawal is hard?

A cool, reflective system forms the plan to withdraw, but an appetitive cue activates a hot, impulsive system that can override that plan at the decisive moment (Metcalfe & Mischel, 1999).

Does simply intending to use withdrawal correctly make it work?

Intention predicts behavior only weakly: a large change in intention yields only a small-to-medium change in action, because competing forces intervene at the point of behavior (Webb & Sheeran, 2006).

Does the effect of arousal on risk-taking apply to women as well as men?

Yes. A later study extending the paradigm found that induced sexual arousal increased sexual risk-taking in both men and women, indicating the effect is not specific to one sex (Skakoon-Sparling et al., 2016).

How does the risk of pregnancy from withdrawal change over several years?

Because annual risk compounds, a 20% typical-use rate accumulates to about a two-in-three chance of pregnancy over five years, far larger than any single year's risk (Sundaram et al., 2017).

What does the psychology imply about improving contraceptive outcomes?

Since the shortfall lies in executing a plan rather than in holding it, the stronger approach is to move the decision out of the hot moment toward methods that need no in-the-moment act (Sheeran & Webb, 2016).

References