Contraceptive prevalence measures the extent to which contraception is being used in a population, in particular among women of reproductive age (conventionally, between 15 and 50). Prevalence is the proportion, expressed as a percentage, of women of reproductive age currently using a method of contraception.
This definition, however, presents several ambiguities regarding the base population, the reference period, and what constitutes a method of contraception, lending itself to variations in measurement. As to the base population, ideally prevalence should cover all sexually active partners–men and women–of reproductive age. In practice, however, information on contraceptive use is most often sought only from women, often only married women. (The "married" group usually includes those in consensual unions in countries where such unions are common.) In most countries, the vast majority of women of reproductive age is married or in an informal consensual union, so this restriction does not greatly affect estimates of contraceptive prevalence. However, in countries where large proportions of unmarried women are sexually active, prevalence estimates based solely on currently married women may not reflect the true level of overall contraceptive use. For example, in the developed countries around 1990, contraceptive prevalence among unmarried women ranged from 47 percent in the United States to 75 percent in Belgium.
Another problem in defining the base population concerns the extent to which all women of reproductive age are exposed to the risk of conception at a particular time, given that some women may be infecund or may not be sexually active while others may be seeking to become pregnant. It is for this reason that, in practice, contraceptive prevalence does not attain the theoretical maximum value of 100 percent.
The definition of contraceptive prevalence centers on current use, and the distinction between past and current contraceptive use can be problematic. Most surveys that have asked about the current use of a method of contraception have asked about use "now" or "within the last month"; sometimes other reference periods are specified. Moreover, there is usually no information collected about the regularity with which the method is employed or about the respondent's understanding of correct use. The fuzziness in the timeframe for measuring use and the difficulty of identifying exactly the women who are exposed to the risk of conception during the specified period undermine the status of prevalence as a rate. It can be recorded, rather, as a simple percentage.
What is considered as contraceptive use is also somewhat subjective, given the differing effectiveness of different methods and the varying motives for use. Contraceptive methods are usually grouped into two broad categories, modern and traditional. Modern methods are those that require clinical services or regular supply: they include female and male surgical contraception (sterilization), oral contraceptive pills, intrauterine devices (IUDs), condoms, injectible hormones, vaginal barrier methods (including diaphragms, cervical caps, and spermicidal foams, jellies, creams, and sponges), and, more recently, subdermal contraceptive implants. The traditional methods–also known as non-supply methods to distinguish them from modern supply methods–include the rhythm method, withdrawal (coitus interruptus), abstinence, douching, prolonged breastfeeding, and a variety of folk methods. Nonetheless, the labels "modern" or "traditional" are inexact: for example, both the condom and the rhythm method have a long history of use, yet the condom is considered modern and the rhythm method traditional.
Almost all surveys about contraceptive use have asked about rhythm and withdrawal, but there has been less consistency regarding other traditional methods. A particular difficulty arises with practices whose main motivation may not have been to prevent pregnancy but which may do so in fact–notably, abstinence and breastfeeding. Some surveys have explicitly excluded such practices from the definition of contraception.
In some African countries, there is a tradition of lengthy abstinence from sexual relations following a birth, but surveys often report prolonged abstinence as the method currently used by a substantial proportion of women. The distinction between contraceptive and noncontraceptive motives for this traditional practice is not clear-cut, and many women who practice lengthy postnatal abstinence evidently do not regard it as contraception. Most surveys do not include abstinence, or postnatal abstinence specifically, in the definition of contraception–including surveys conducted in sub-Saharan Africa. When women spontaneously report that they were practicing prolonged abstinence for contraceptive reasons, they may be recorded under the category of "other" methods.
Breastfeeding has fertility-inhibiting effects and in societies that practice prolonged breastfeeding, fertility is depressed. As in the case of abstinence, most surveys have not included breastfeeding in the list of contraceptive methods. In cases where it has been included, the number of women that identify breastfeeding as their contraceptive method is typically a small fraction of the number that are currently breastfeeding.
Depending on the society, folk methods of contraception may include a large number of herbal preparations, manipulation of the uterus, vigorous exercise, adoption of particular postures during or after intercourse, incantations, and the wearing of charms. The effectiveness of these methods has never been scientifically evaluated: some are wholly fanciful, others may be highly unreliable, and still others probably act as abortifacients rather than as contraceptives. Women often do not mention folk methods unless the survey inquires about them specifically, and most surveys do not include probing questions dealing with specific folk methods.
Sources of Information on Contraceptive Prevalence
Surveys are considered the best source of data on contraceptive practice, since they can record the prevalence of all methods, including those that require no supplies or medical services. Most surveys ask respondents broadly similar questions to measure contraceptive use. Women are first asked what methods they know about, and the interviewer then names or describes methods that were not mentioned by the respondent. Respondents are then asked about use of each method that was recognized. This procedure helps make clear to the respondent what methods are to be counted as contraceptive. When methods are not named by the interviewer, the level of use tends to be underreported. In particular, it does not occur to many persons to mention methods such as withdrawal and rhythm, which require no supplies or medical services.
Organized family planning programs keep records on their clients who come for contraceptive supplies or services. These records are another main source of information about contraceptive prevalence. However, data from this source have the serious drawback of excluding use of contraception obtained outside the program, including modern methods supplied through nonprogram sources as well as methods that do not require supplies or medical services. In addition, the process of deriving reasonably accurate prevalence estimates from the information in family planning program records is much less straightforward than the direct questions posed in representative sample surveys.
Prevalence levels range from 4 to 10 percent in pretransitional societies, where fertility is typically high, to 70 to 80 percent in posttransition, low-fertility countries. (As mentioned above, in practice, contraceptive prevalence never attains the maximum value of 100 percent.) In 1997 contraceptive prevalence for the world as a whole was estimated to be 62 percent–that is, 62 percent of currently married women between ages 15 and 50 were using a method of contraception. Regional average levels of prevalence range from 25 percent in Africa to over 65 percent in Asia and Latin America and the Caribbean. The average prevalence for developed countries was 70 percent.
The reported level of contraceptive use in pretransitional societies is very low for both modern methods and traditional methods. For example, contraceptive prevalence in Chad in 1996 was 4 percent (Chad's total fertility rate exceeded 6); in Uganda in 1995 it was 15 percent. The prevalence of modern method use was 1 percent in Chad and 8 percent in Uganda. A large proportion of married contraceptive users in Chad reported the use of traditional methods of contraception: rhythm and withdrawal. It is likely that many women in pretransitional societies use traditional methods that are not captured in the standard surveys.
In the low-fertility countries, the great majority of women not using contraception are pregnant, seeking to become pregnant, infecund, or sexually inactive. Because of the relatively high levels of prevalence already reached in these countries, there is little room for further increase. In developed countries, certain traditional methods–including withdrawal and various forms of the calendar rhythm method–are commonly used: together they account for 26 percent of total contraceptive use in the low-fertility developed countries, compared with just 8 percent in the less developed regions. However, recent trends indicate that the prevalence of modern methods is increasing at the expense of traditional methods. In France, for example, between 1978 and 1994 the use of modern methods increased from 48 to 69 percent, even as the use of all methods decreased by 4 percent. Contraceptive prevalence in the United States in 1995 was estimated to be 76 percent of women who were married or in a union. Female sterilization was the most popular method, with a prevalence of 24 percent, followed by the pill, at 16 percent.
Empirical Relationship between Prevalence and Fertility
There is a strong relationship between contraceptive prevalence and the overall level of childbearing as measured by the total fertility rate. (The total fertility rate indicates the average number of children that would be born per woman according to childbearing rates of the current period.) Cross-national data show that the total fertility rate decreases, on average, by 0.7 children for every 10 percentage-point rise in contraceptive prevalence. This translates into 1 child fewer for every 15 percentage-point increase in contraceptive prevalence. Contraception is the most important of the proximate determinants of cross-national differences in fertility. (Other major proximate determinants of these differences are patterns of marriage and sexual activity outside of marriage, the duration of breastfeeding, and the practice of induced abortion–none of them as strongly associated with fertility as contraceptive use.)
Abma, Joyce, et al. 1997. "Fertility, Family Planning, and Women's Health: New Data from the 1995 National Survey of Family Growth." Vital and Health Statistics 23(19): 61.
United Nations. 2000. Levels and Trends of Contraceptive Use as Assessed in 1998. New York: United Nations.