The fight between comprehensive sex education and abstinence-only-until-marriage education is one of the most extensively evaluated policy questions in American public health, and one of the longest to remain politically contested despite the evidence. Federal funding for abstinence-only programs began with the Adolescent Family Life Act in 1981, expanded sharply under Title V Section 510 in 1996 and Community-Based Abstinence Education in 2000, and persisted in modified forms after evidence accumulated that the programs did not work. Over twenty years, federal expenditure on abstinence-only exceeded two billion dollars. Across the same period, dozens of rigorous evaluations, including the congressionally mandated Mathematica study released in 2007, found no effect of abstinence-only programs on initiation of sexual activity, age at first intercourse, contraceptive use, pregnancy rates, or STI rates.
The empirical question is not in serious dispute among researchers. The persistence of the policy reflects political and moral commitments that do not depend on the evidence. The Law 5 problem—revise—is acute. A program that does not work but flatters a particular moral vision is more durable than the same program would be if it were evaluated by its outcomes. The collective's institutions for revising failed plans were not designed for plans whose primary function was symbolic.
John Santelli's series of papers documented the empirical case methodically. Abstinence-only programs frequently contained medical inaccuracies—exaggerated condom failure rates, distorted information about HIV transmission, gendered claims about sexuality that lacked scientific support. Programs based on virginity pledges, the strongest abstinence-only intervention, produced delayed initiation by an average of eighteen months in some studies but were associated with reduced contraceptive use when intercourse did occur, producing equal or higher pregnancy and STI rates over the long run. Hannah Brückner and Peter Bearman's longitudinal analysis showed that adolescents who took pledges and broke them had STI rates indistinguishable from non-pledgers, while having had fewer opportunities to learn protective behaviors.
The romantic lens cuts both ways here. Defenders of abstinence-only frame the policy as protecting the meaning of sex, preserving its connection to commitment, and resisting a culture they see as having stripped sex of consequence. The argument is not stupid. Sex inside a stable commitment is, on most outcomes the literature can measure, better—lower STI exposure, more communication, more reported satisfaction. The mistake is conflating the empirical claim about committed relationships with a policy intervention that demonstrably fails to produce more committed relationships. Telling fourteen-year-olds to wait until marriage does not produce more marriages or later first sex; it produces marginally delayed sex followed by less protected sex.
Comprehensive sex education's defenders have their own characteristic failures. They sometimes treat the curriculum as morally neutral when it is not—any curriculum communicates a model of what relationships are and what they are for. They sometimes oversell the evidence, claiming effects on outcomes that the studies do not measure or only weakly support. They sometimes dismiss the underlying concerns of religious and conservative parents as merely irrational rather than engaging with the genuine question of how a community transmits values about love, commitment, and sexuality to its children.
The Law 2 dimension—think clearly—requires distinguishing several claims that often get conflated. (1) Comprehensive sex education produces better outcomes than abstinence-only on every measured dimension. This is well-supported. (2) Comprehensive sex education produces dramatically better outcomes than no formal sex education at all. This is moderately supported; the effects are real but not enormous. (3) The content of sex education is the main driver of adolescent sexual behavior. This is not supported; family, peer culture, media, and economic factors swamp curriculum effects. The honest position is that comprehensive sex education is the best available curricular option, that its effects are real but bounded, and that policy debates that treat the curriculum as the central determinant of adolescent outcomes are overestimating its reach.
The Law 4 dimension is that the collective has accumulated enough evidence to act decisively, and has not. Federal abstinence-only funding declined in the 2010s under Obama-era reforms, partially reversed in 2017 under different administration priorities, and continues at lower levels through Title V. State-level adoption of comprehensive sex education has grown but remains uneven. Roughly half of U.S. middle and high school students receive comprehensive content; the other half receive abstinence-focused or fragmentary instruction. The Law 5 failure is the inability to consolidate around what the evidence supports.
The romantic lens places a final question. The deeper argument for comprehensive sex education is not that it reduces pregnancy rates—though it does—but that it treats adolescents as moral agents who are capable of making decisions about their own bodies and relationships if given information and skills. The deeper argument for abstinence-only is that it treats adolescents as people whose long-run flourishing is better served by a commitment they have not yet been asked to reason their way into. Both arguments contain something. The comprehensive position is closer to right because it does not require the curriculum to lie about contraception, condom efficacy, or the existence of LGBT students. A curriculum that lies to students about facts they can verify themselves on the internet has lost the authority to teach them about anything else, including the things the curriculum's authors most want them to learn.