No method of contraception or disease prevention is effective when practiced incorrectly or inconsistently. A 1988 National Survey of Family Growth found abstinence to have a contraceptive failure rate of 26% when not practiced consistently. So, in abstinence, as in condom use, consistency is key.
Mighty odd-sounding, as OpinionJournal reported: Is it really abstinence when you aren't consistently abstaining? Or is their claim that people who decide to be abstinent nonetheless in practice end up backsliding — and often enough that they get pregnant in 26% of the first year in which they chose to be abstinent? (That's how contraceptive failure rates are generally measured.) If that's so, then that would suggest that abstinence decisions, at least unless reinforced by some belief system that will deepen the commitment to the decision, are remarkably ineffective.
But here's what's really going on: (1) a confusing term being used by public health scholars, which (2) likely led to confusion on the part of the person writing the Web page, which (3) translates into false claims being passed along to the public. If you look at abstracts of the 1988 Survey, you find that 26% is the failure rate for "periodic abstinence", which means "rhythm and natural family planning."
That's right: 26% is the failure rate for the rhythm method, not for deliberate decisions to abstain. Public health scholars apparently refer to the rhythm method and similar practices as "periodic abstinence," which is literally accurate, but potentially confusing to nonexperts, since it's close to a term ("abstinence") that means something quite different in lay discussion. The Web page author seemed to have been confused, interpreting "periodic abstinence" simply as "inconsistently practiced abstinence," and thus labeling it simply as "abstinence." And readers will therefore be getting false information: "Abstinence" in lay discussion generally refers to a deliberate decision not to have sex at all — rather than just to a decision to have sex only on one's presumably less fertile days — so people will read the claim as pointing to the dangers of abstinence, rather than the dangers of the rhythm method.
So three tips: (1) If you're a scholar, especially in a discipline where laypeople or nonacademic public policy analysts might consume your output, try to avoid adopting terms that may confuse outsiders. (I realize that sometimes there may have to be trade-offs between clarity to outsiders and clarity or convenience to fellow scholars, hence my use of "try.") Lawyers and judges, by the way, are frequent offenders here, creating legal terms that sound like one thing in English (e.g., "actual malice" in libel) but mean something quite different in legalese. This is especially dangerous in law, where the lay public is especially likely to be interested in the information, though also especially hard to root out, since many of these terms evolve over decades, and are hard to clarify because of the weight of precedent and tradition.
(2) If you're consuming academic works, be on the lookout for terms that don't mean quite what they seem to mean at first, and for qualifiers that are sometimes overlooked but are tremendously important. There are false friends in translation from technical English into lay English just as there are in translation between languages.
(3) Don't believe everything you read.