Sexual Abstinence

Abstract

Sexual abstinence, which the Medical Subject Headings thesaurus classifies under sexual behavior, is the deliberate restraint from some or all sexual activity for a period of time. Cognitive psychology is less interested in the abstention itself than in the mental machinery that makes it hard to sustain: a standing goal must hold against a recurring, strongly motivating drive. This article treats abstinence as a problem in self-control and decision making. It reviews the self-regulatory-strength account of restraint, the hot–cold empathy gap that makes cold-state resolutions fail, the intention–behavior gap documented in studies of virginity pledges, and what large evaluations of abstinence-based sex education reveal about the limits of resolve as a public-health strategy. A worked example traces why a sincere pledge can change the timing of first intercourse yet leave protective behavior worse.

Keywords: sexual abstinence, self-regulation, hot-cold empathy gap, intention-behavior gap, virginity pledge

Introduction

Sexual abstinence is the voluntary decision to refrain from some or all sexual activity, whether for a defined interval, until a particular milestone such as marriage, or indefinitely. It is a behavior defined by what does not happen, which is precisely what makes it a revealing object of cognitive study. Abstaining is not a single act but the repeated outcome of a standing goal maintained against a powerful and recurring motivational state. Where most behaviors are explained by the forces that produce them, abstinence must be explained by the forces that successfully oppose a drive — and by the regularity with which those opposing forces fail.

The cognitive approach therefore reframes the question. The interesting fact is not that people abstain but that so many who sincerely intend to abstain do not, and that the failure is systematic rather than random. A resolution formed in a dispassionate moment is tested later under conditions — arousal, intimacy, opportunity — that the cold-state planner did not accurately anticipate (Ariely & Loewenstein, 2006). Restraint draws on a self-regulatory capacity that is limited and fatigues with use (Muraven & Baumeister, 2000). And the social meaning of a public commitment can do more to sustain abstinence than private willpower, which is why the same pledge succeeds in one setting and fails in another (Bearman & Brückner, 2001).

The Medical Subject Headings thesaurus indexes Sexual Abstinence beneath Sexual Behavior in its classification of behavior and behavior mechanisms. That placement is a retrieval convenience rather than a theoretical claim: it locates the topic among its neighbors in the biomedical literature but says nothing about the self-regulatory and decision-making machinery the sections below describe. This article uses the classification only to situate abstinence among related behaviors and otherwise treats the psychology on its own terms.

Sexual Abstinence as Self-Regulation

The central cognitive insight is that abstinence is an instance of self-regulation: the capacity to override a prepotent impulse in the service of a longer-range goal. Muraven and Baumeister (2000) proposed that this capacity behaves less like a skill or a stable trait than like a muscle — a limited resource that is temporarily depleted by exertion and that fatigues when drawn upon repeatedly. On this strength model, each act of restraint consumes a common regulatory resource, so that resisting one temptation leaves less capacity available to resist the next. Abstinence is demanding not because the goal is unclear but because the goal must be defended again and again, often after the resource has already been spent on the day's other demands.

This account reframes lapses as something other than a failure of sincerity or character. A person may hold the goal of abstinence with complete conviction and still lapse, because conviction and regulatory capacity are different things: the first sets the target, the second does the work of hitting it, and the second is finite. The practical corollary is that the conditions surrounding a decision — fatigue, stress, prior demands on self-control, alcohol — predict lapses better than the strength of the original resolve. Interventions that reduce the number of in-the-moment battles, by avoiding high-temptation situations rather than enduring them, work with the limitation rather than against it.

The strength model also explains why abstinence is often easiest at the start and hardest to maintain. Early restraint is cheap when motivation is high and the resource is full; sustained restraint is expensive because it must survive the ordinary depletion of daily life. The first demonstration makes this concrete, letting the reader impose a sequence of self-control demands on a modeled regulatory resource and watch how the probability of a lapse rises as the resource is drawn down and recovers with rest.

Decision Making Under Visceral Influence

If self-regulation supplies the capacity for restraint, the visceral states that restraint must overcome explain much of why it fails. Ariely and Loewenstein (2006) studied decision making under sexual arousal directly, asking young men the same questions about their sexual attitudes and intentions in a calm state and again in a state of high arousal. Arousal substantially shifted their answers: judgments about what they would find appealing, what risks they would accept, and what they would do to obtain a sexual outcome all moved in the permissive direction, and the participants in the cold state did not anticipate the size of the shift. The finding is a sharp instance of the hot–cold empathy gap — the difficulty of accurately predicting, from within one affective state, how one will think and act in a different one.

The implication for abstinence is direct. A resolution to abstain is nearly always formed in a cold state, when the drive it must resist is quiescent and therefore easy to underestimate. The resolution is then tested in a hot state, in which the same drive is far stronger than the planner imagined and the deliberative goal correspondingly weaker. The planner and the actor are, in effect, two different decision makers with different preferences, and the cold-state planner systematically overestimates the hot-state actor's compliance. This is not weakness of will in the ordinary sense; it is a predictable failure of affective forecasting, and it recurs however sincere the original intention.

The gap has a practical signature. Because people cannot feel, in the cold state, how the hot state will reorder their choices, advice premised on pure resolve — that a sincere intention to abstain will by itself suffice — is built on a forecasting error. Strategies that acknowledge the gap, by committing in advance to avoid the situations in which the hot state arises, are better matched to how the decision actually unfolds. The second demonstration lets the reader move a single arousal control and watch a stated cold-state intention to abstain diverge from the predicted hot-state behavior as the visceral state intensifies.

The Intention–Behavior Gap

That a sincere intention is a weak predictor of behavior is one of the most robust findings in the study of self-regulation, and sexual abstinence offers an unusually clean natural experiment on it: the virginity pledge, a public, often formal commitment to abstain until marriage. Bearman and Brückner (2001) analyzed a large national sample of adolescents and found that pledging did delay first intercourse substantially — but only under a specific social condition. Pledging worked when pledgers formed an identifiable minority within their school; where almost no one pledged there was no community to sustain the identity, and where pledging was near-universal it conferred no distinctive identity and lost its force. The commitment changed behavior not through private willpower but through the social meaning of belonging to a self-conscious minority.

The same studies exposed the gap's cost. Pledgers who did break the pledge — the large majority, eventually — were less likely than non-pledgers to use contraception at first intercourse (Bearman & Brückner, 2001), and a longitudinal follow-up found that pledgers and non-pledgers contracted sexually transmitted infections at statistically indistinguishable rates, in part because pledgers were less likely to be tested or diagnosed (Brückner & Bearman, 2005). Rosenbaum (2009) tightened the inference by matching pledgers to non-pledgers who were already similar on pre-pledge religiosity and attitudes; five years on, the two groups did not differ in any measure of sexual behavior, yet the pledgers were less likely to use birth control and condoms — and the majority of pledgers denied ever having pledged. A later re-analysis reached the same conclusion, that pledging failed to reduce sexual risk and was associated with less protective behavior among those who became sexually active (Paik et al., 2016).

The pattern is the intention–behavior gap in its most consequential form. The pledge reliably registered an intention and, under the right social conditions, shifted the timing of first intercourse; it did not reliably change whether intercourse eventually occurred, and it worsened the safety of the intercourse that did. The third demonstration models the social mechanism Bearman and Brückner identified, letting the reader vary the proportion of a community that pledges and watch the delaying effect rise and then collapse — an inverted-U that peaks when pledgers are a distinctive minority.

Abstinence Education and Its Evaluation

The cognitive limits of resolve bear directly on a long-running policy question: whether programs that promote abstinence as the expected standard for school-age young people achieve their stated aims. The evidence from controlled evaluation is consistent. A federally commissioned randomized study of four abstinence-only programs found no impact on the rate of sexual abstinence; youth in the programs were no more likely to abstain, and had similar numbers of partners and similar ages at first intercourse, as controls (Trenholm et al., 2008). Reviews reaching across many programs found that comprehensive sex education, which includes but does not restrict itself to abstinence, more reliably delayed initiation and increased protective behavior, whereas abstinence-only instruction did neither (Kirby, 2008).

Syntheses of U.S. policy drew the broader conclusion. Santelli and colleagues (2006) argued that abstinence-only-until-marriage programs were ethically and scientifically problematic because they withheld accurate information about contraception and disease prevention, and an updated review a decade later found the intervening evidence had only strengthened the case: such programs showed little evidence of efficacy in delaying intercourse and raised concerns about their effect on young people's health (Santelli et al., 2017). A quasi-experimental analysis exploiting state-level policy variation found that state-mandated abstinence-based education did not improve, and in some specifications worsened, teen health outcomes (Carr & Packham, 2017). A review of reviews of school-based programs concluded that the interventions with the best evidence for improving sexual health were comprehensive rather than abstinence-restricted (Denford et al., 2017).

The cognitive reading of this literature is not that abstinence is ineffective as a personal choice — a person who does not have sex does not acquire a sexually transmitted infection from it — but that programs built on resolve alone collide with the self-regulatory facts. They ask adolescents, whose regulatory capacity and affective forecasting are still developing (Fortenberry, 2013), to sustain a standing goal against a strong drive, while withholding the protective knowledge that limits harm when the goal fails — which, given the intention–behavior gap, it often does. The surrounding social and motivational context, including religiosity, shapes adolescent sexual decision making more than instruction in resolve does (Rostosky et al., 2004), and the state of U.S. sex education has been described as inconsistent and frequently not evidence-based (Hall et al., 2016).

Worked Example

Consider a cohort of adolescents who take a sincere virginity pledge, and follow the two quantities the cognitive account says will diverge: the timing of first intercourse and the safety of it. The pledge is a genuine intention, so the question is what it predicts.

Timing responds to the pledge — conditionally. Treat the delaying effect as a function of how distinctive the pledge is in a given school. Where only a handful of students pledge, no supporting community forms and the effect is near zero. Where pledging is a self-conscious minority identity — a middling prevalence — the effect is largest, because the pledge confers a membership worth maintaining. Where nearly everyone pledges, the commitment is ordinary rather than distinctive and the effect collapses again (Bearman & Brückner, 2001). The relationship is an inverted-U in prevalence, not a straight line, which is why a single program can appear to work in one school and fail in the next with no change in the pledge itself.

Safety moves the wrong way. Now follow the majority of pledgers who eventually do have intercourse. Because the pledge framed sex as something not to be planned for, the intention that shifted timing does nothing to prepare for the event when it arrives; the pledger reaches first intercourse less likely to use contraception than a comparable non-pledger (Bearman & Brückner, 2001), and over the following years no less likely to acquire a sexually transmitted infection — only less likely to be tested for one (Brückner & Bearman, 2005). Table 1 lays the two quantities side by side against a matched non-pledger.

The example makes the section's point exact. A pledge is a measured, sincere intention, and under the right social conditions it genuinely shifts one outcome — when sex first happens. But an intention is not a behavior, and the long-run behavior it was meant to prevent is largely unchanged while the behavior it never addressed, protection, is made worse. The gap between the resolution and the conduct is not a measure of how insincere the pledgers were; it is a measure of how little a cold-state intention governs a hot-state act.

Discussion

The through-line of the cognitive psychology of abstinence is that restraint is governed by the architecture of self-control and decision making, not by the strength of a moral conviction. A standing goal to abstain must be defended repeatedly against a recurring drive, using a regulatory resource that is limited and fatigues (Muraven & Baumeister, 2000); the defense is mounted under visceral conditions that the goal's author, reasoning in a cold state, systematically failed to anticipate (Ariely & Loewenstein, 2006); and the resulting behavior tracks the intention only loosely, and only when the social context supplies a meaning that private willpower cannot (Bearman & Brückner, 2001). Each piece compounds the others, which is why sincere resolutions to abstain fail so regularly and so predictably.

A broader motivational frame helps explain the terrain on which these mechanisms operate. Baumeister, Catanese, and Vohs (2001) reviewed a large literature and concluded that, on average, men show a stronger and more constant sexual motivation than women across many converging measures — a difference that shapes how demanding abstinence is and for whom. The sexual economics framework extends the point by modeling heterosexual activity as an exchange in which abstinence functions partly as a form of restricted supply (Baumeister & Vohs, 2004); whatever one makes of the model, it underscores that abstinence is embedded in social and motivational systems, not decided in a vacuum by individual resolve.

This integration carries practical weight. Because the limiting factor is regulatory capacity and affective forecasting rather than conviction, the interventions that work are those that reduce the number and intensity of in-the-moment battles and that supply protective knowledge for when the goal fails — which is the comprehensive, not the abstinence-only, approach (Kirby, 2008; Santelli et al., 2017). The person who lapses is not simply weak-willed; the person is a cold-state planner whose hot-state self made a different decision, in a context that did or did not lend the resolution social support. Treating abstinence as a self-regulation problem rather than a test of character is what makes its failures intelligible and its successful conditions identifiable.

Current Directions

Current work has largely settled the program-evaluation question and moved to its implications. The most recent comprehensive review of U.S. abstinence-only-until-marriage policy concluded that the accumulated evidence shows these programs to be ineffective at delaying sexual initiation and ethically problematic for withholding health information, and framed comprehensive sex education as the evidence-based alternative (Santelli et al., 2017). Quasi-experimental designs have strengthened the causal claim: by exploiting variation in state mandates rather than relying on self-selected program participation, Carr and Packham (2017) found that abstinence-based mandates did not improve, and by some measures worsened, adolescent health outcomes — addressing the selection problem that had clouded earlier observational work.

A second strand synthesizes across the whole intervention literature to ask which program features actually move sexual-health outcomes. A review of reviews found the strongest evidence for multi-component, comprehensive school-based programs, and little for abstinence-restricted instruction (Denford et al., 2017), while a survey of the state of U.S. sex education documented how inconsistent and frequently non-evidence-based current practice remains (Hall et al., 2016). The open questions are now less about whether resolve-based programs work and more about how to build the self-regulatory and contextual supports that the cognitive account identifies as the real levers — and how adolescent sexual decision making, shaped by still-developing regulatory capacity (Fortenberry, 2013), can be supported rather than simply exhorted.

A third strand revisits the theoretical foundation itself. The strength model's signature prediction — that a prior act of self-control measurably depletes the capacity for the next — has proved difficult to reproduce: a large multi-laboratory preregistered replication found the ego-depletion effect to be small and not reliably distinguishable from zero (Hagger et al., 2016), prompting an active debate over whether depletion reflects a literal resource, a shift in motivation and attention, or an artifact of particular procedures. The cognitive account of abstinence does not rest on the strongest version of the resource metaphor. Its load-bearing claims — that restraint must be exercised repeatedly, that it is harder under fatigue, stress, and competing demands, and that the situation predicts lapses better than the trait — hold whether the underlying mechanism is a depletable store or a reallocation of motivation, so the contested status of ego depletion sharpens the account rather than undermining it.

Common Misconceptions

“If someone really intends to abstain, they will.”

Intention is a weak predictor of sexual behavior. A resolution formed in a calm state is tested under arousal, which systematically shifts choices in ways the calm-state planner does not foresee — the hot–cold empathy gap, not a failure of sincerity (Ariely & Loewenstein, 2006).

“Virginity pledges reliably keep teenagers from having sex.”

Pledges delay first intercourse only when pledgers form a distinctive minority, and most pledgers eventually have premarital sex. When they do, they are less likely to use contraception than comparable non-pledgers (Bearman & Brückner, 2001; Rosenbaum, 2009).

“Abstinence-only education lowers teen sexual risk.”

Controlled evaluations find abstinence-only programs do not delay sex or reduce partners relative to controls, while comprehensive programs more reliably delay initiation and increase protective behavior (Trenholm et al., 2008; Kirby, 2008).

“Self-control is a fixed trait a person either has or lacks.”

Self-regulation behaves like a limited resource that fatigues with use, so the same person restrains successfully when the resource is full and lapses when it is depleted — the situation predicts lapses better than the trait (Muraven & Baumeister, 2000).

Glossary

Abstinence-only education.

Sex education that promotes refraining from sex as the expected standard and typically withholds instruction on contraception; controlled evaluations find it does not reliably delay intercourse.

Affective forecasting.

Predicting one's own future emotional states and the choices they will drive; a persistent source of error when the predicting state differs from the predicted one.

Comprehensive sex education.

Instruction that covers abstinence alongside contraception, disease prevention, and decision making; the approach with the strongest evidence for improving sexual-health outcomes.

Ego depletion.

The temporary reduction in self-regulatory capacity following an earlier act of self-control, as predicted by the strength model of self-regulation.

Hot–cold empathy gap.

The difficulty of predicting, from within one affective or visceral state, how one will think and behave in a different one; abstinence resolutions formed cold are tested hot.

Intention–behavior gap.

The general finding that a stated intention predicts actual behavior only weakly; especially large for sexual conduct, where the drive opposing the intention is strong and state-dependent.

Prepotent impulse.

A dominant, readily triggered response tendency that self-regulation must override for a longer-range goal such as abstinence to prevail.

Self-regulation.

The capacity to monitor and alter one's own responses to bring them into line with a goal or standard; the core process on which sustained abstinence depends.

Sexual behavior.

The broad class of activity, indexed in MeSH, within which abstinence is classified as the deliberate restraint from some or all sexual activity.

Sexual economics.

A framework modeling heterosexual activity as a social exchange in which restraint can function as restricted supply; a motivational context for abstinence rather than a theory of self-control.

Strength model.

The account on which self-control draws on a single limited resource that depletes with use and recovers with rest, likening regulatory capacity to a muscle.

Variable context effect.

The finding that the same commitment changes behavior differently depending on its social setting, as a virginity pledge delays intercourse only at intermediate pledge prevalence.

Virginity pledge.

A public, often formal commitment to abstain from sex until marriage; a natural experiment on the intention–behavior gap whose effect depends on social context.

Visceral state.

A strong, transient motivational condition such as sexual arousal, hunger, or craving that narrows attention onto its object and reshapes decision making in the moment.

Key Researchers

Dan Ariely

is a professor of psychology and behavioral economics at Duke University, whose experiments on decision making under sexual arousal demonstrated the hot–cold empathy gap that makes cold-state resolutions to abstain unreliable.

Faculty page - Wikipedia

Roy F. Baumeister

is a social psychologist at the University of Queensland whose strength model of self-regulation, and reviews of sexual motivation and sexual economics, supply much of the theoretical frame for treating abstinence as a self-control problem.

ORCID - Wikipedia - Google Scholar

Peter S. Bearman

is a professor of sociology at Columbia University whose analyses of the national Add Health study showed that virginity pledges delay first intercourse only when pledgers form a distinctive minority, isolating the social mechanism behind the commitment.

ORCID - Wikipedia - Google Scholar

Hannah Brückner

is a professor of social research and public policy at New York University Abu Dhabi, whose longitudinal work with Peter Bearman documented that virginity pledgers acquired sexually transmitted infections at rates similar to non-pledgers while being less likely to be tested.

Faculty page - Wikipedia

George Loewenstein

is a professor of economics and psychology at Carnegie Mellon University whose work on visceral influences and the hot–cold empathy gap, including the arousal experiments with Dan Ariely, reframed in-the-moment decision making.

Wikipedia - Google Scholar

Janet E. Rosenbaum

is a public-health researcher at SUNY Downstate Health Sciences University whose matched-comparison re-analysis showed that virginity pledgers and similar non-pledgers did not differ in sexual behavior five years on, yet pledgers used less protection.

ORCID - Faculty page - Google Scholar

John S. Santelli

is a pediatrician and professor of population and family health at Columbia University whose reviews of U.S. abstinence-only policy established the scientific and ethical case against programs that withhold contraceptive information.

ORCID - Faculty page - Google Scholar

Kathleen D. Vohs

is a professor of marketing and psychology at the University of Minnesota whose research on self-regulation and sexual economics clarified how restraint draws on a limited regulatory resource and functions within social exchange.

ORCID - Wikipedia - Google Scholar

Frequently Asked Questions

What is sexual abstinence in psychology?

Sexual abstinence is the deliberate decision to refrain from some or all sexual activity for a period of time. Cognitive psychology studies it as a self-regulation problem: a standing goal that must be maintained against a recurring, strongly motivating drive, so the interesting question is not the abstention itself but why sustaining it is so difficult.

Why do people who intend to abstain often fail?

Because intention is a weak predictor of sexual behavior. A resolution formed in a calm state is tested later under arousal, which shifts choices in a permissive direction that the calm-state planner systematically underestimates — the hot–cold empathy gap (Ariely & Loewenstein, 2006). The failure is a predictable forecasting error rather than a lack of sincerity.

What is the hot–cold empathy gap?

It is the difficulty of predicting, from within one affective state, how one will think and act in a different one. A person reasoning in an unaroused state cannot accurately feel how arousal will reorder their preferences, so cold-state intentions to abstain overestimate hot-state compliance (Ariely & Loewenstein, 2006).

Do virginity pledges work?

Only partly and conditionally. Pledges delay first intercourse when pledgers form a distinctive minority within a community, but most pledgers eventually have premarital sex, and when they do they are less likely to use contraception than comparable non-pledgers (Bearman & Brückner, 2001; Rosenbaum, 2009).

Is abstinence-only education effective?

Controlled evaluations find that abstinence-only programs do not reliably delay sexual initiation or reduce the number of partners relative to controls, whereas comprehensive programs more reliably delay initiation and increase protective behavior (Trenholm et al., 2008; Kirby, 2008).

Is self-control a fixed trait or something that can run out?

The strength model holds that self-regulation draws on a limited resource that depletes with use and recovers with rest, so the same person restrains successfully when the resource is full and lapses when it is depleted (Muraven & Baumeister, 2000). Situational demands on self-control predict lapses better than a fixed trait.

Does abstinence reduce the risk of sexually transmitted infection?

Not having sex does avoid sexually transmitted infection from that activity, but a pledge or program premised on abstinence does not reliably reduce infection at the population level, because most people eventually become sexually active and those who pledged are often less prepared to protect themselves (Brückner & Bearman, 2005).

How does sexual abstinence relate to cognition?

Abstinence is itself a cognitive achievement: it depends on self-regulation, affective forecasting, and decision making under visceral influence rather than on moral conviction alone. Its failures map onto well-documented limits of these processes, which is why cognitive psychology treats it as a case study in self-control (Muraven & Baumeister, 2000).

References