Thirty years after contraceptives were legalized in France, effective methods have become easily available and are widely used. Although contraceptive advertising is still prohibited, there are no barriers to speaking freely about the subject. Therefore, the question is: Are French contraceptive behaviors relatively stable, or are they still changing?
In this article, we describe current contraceptive practices in France, using results of a 1994 demographic survey, and focus on the impact, if any, of changes in family formation and dissolution. For example, increasing numbers of French men and women are delaying entry into cohabitational relationships, and this social change may be affecting contraceptive use patterns. We also analyze the main trends in contraceptive use during the last three decades, using results of comparable surveys conducted in 1978 and 1988.
During the last 10 years, the growing AIDS epidemic has given the French a new perception of sexually related risks. In the second half of this article, we review contraceptive behavior that may have changed in response to the fear of AIDS, focusing on the contraceptive behavior of single men and women. We also analyze recent trends in condom use that may have been influenced by increasing fear of sexually transmitted diseases (STDs), as well as by contraceptive needs.
We distinguish the percentages of men and women who currently use condoms (at the time of the survey), who have ever used condoms, who used them during the year before the survey and who used them during first sexual intercourse. When compared through successive cohorts, these indicators allow us to analyze the recent changes in condom use. Such changes can be difficult to measure because condom use is often limited to short intervals.
The Institut National d'Etudes Démographiques (INED) and the Institut National de la Statistique et des Etudes Economiques (INSEE) conducted the Fertility and Family Survey (FFS) in March 1994, with financial support from the Ministry of Social Affairs and the Caisse Nationale des Allocations Familiales, a French family policy agency. The 1994 FFS was the fourth in a series of general surveys on fertility and contraception in France. It followed a 1971 survey of married women1 and a 1978 survey by the INED, which was the first survey to include the contraceptive practices and attitudes of all women irrespective of their marital status. The 1978 survey, which was part of the World Fertility Survey, interviewed a random sample of 2,941 women aged 20-44.2
In 1988, INED conducted a third survey. The Fertility and Family Planning Survey (FFPS) used a random sample of 3,183 women aged 18-49, focusing on current contraceptive use and contraceptive and pregnancy histories. This survey described oral contraceptive and IUD use, as well as the decline of withdrawal as a family planning method between 1967 and 1987.3
A total of 5,900 households—selected from among 75,000 households interviewed in the Enquête Emploi, an annual survey on employment conducted by INSEE—participated in the 1994 FFS.* The survey selected respondents after considering the number of adults aged 20-49 living in each household and the structure of the household.
The probability of being sampled in the FFS varied—from approximately 92% for a household with a child living with only one or neither parent and for an unmarried couple with children to about 21% for a married couple. The survey selected one adult as a respondent from each household.4 Overall, 2,944 women and 1,941 men aged 20-49 were interviewed.
There was an 8% nonresponse rate for households in the employment survey—5% because no member of the household could be contacted and 3% because of refusal. The probability of FFS nonparticipation among the 5,900 selected household members was 16% for women and 19% for men. Nonresponse rates were similar for married persons and unmarried cohabitants (10% for women, 16% for men). Rates were higher, however, among unmarried, noncohabiting persons (23% for both women and men).
We computed a first set of weights from the inverse inclusion probabilities in the employment survey and the FFS, including a poststratification by sex, age and size of locality. To eliminate the bias from differences in nonresponse rates, we constructed final weights based on a second poststratification by sex, age and conjugal status (in a married union, an unmarried union or no union). Prestratification and poststratification caused the variance of the estimates to be lower than in a comparable random sample.
The FFS found that about two out of three French women aged 20-49 used reversible contraceptives in 1994 (Table 1). The pill was the most commonly used reversible method; about 36% of women aged 20-49 were using this method during the month of the survey. Reliance on oral contraceptives declined with age, from a high of 59% among women aged 20-24 to 14% among women aged 45-49. The IUD followed the pill in prevalence, with 16% of women aged 20-49 using the method. IUDs were mostly used by women aged 35-44: About 26% of women in this age-group reported using IUDs.
Other reversible contraceptive methods were less commonly used. Around 5% of French women used condoms, 4% practiced periodic abstinence and only 2% relied on withdrawal, which had been the most popular reversible method in France until the 1960s.5 Out of 100 women using reversible methods, 56 were taking the pill, 25 used the IUD, and 19 reported using some other method.
Seven percent of women aged 20-49 had been sterilized, with some 4% reporting that they were sterilized for contraceptive purposes and 3% for other reasons. Male sterilizations were extremely rare: Much fewer than 1% of women said that their partner had been sterilized for contraceptive purposes. French laws on male or female sterilization—even when voluntary—are unclear except in cases where it is medically necessary.† In practice, however, the boundaries between medical and contraceptive purposes are unclear, and no physician has ever been sued or prosecuted by the government for performing sterilizations. Nevertheless, sterilization before age 40 remains rare, and it is not commonly regarded as a contraceptive method in France. When sterilization for contraceptive purposes is included, then the overall percentage of women aged 20-49 practicing contraception reaches 69%.
Among the 31% of women who reported that they were not using contraceptives, 3% were noncontraceptively sterilized, 4% were infertile, 4% were pregnant, 4% wanted to become pregnant and 11% reported "having currently no (or few) sexual relationships."‡ The remaining 5% were considered to be at risk for an unintended pregnancy, with 3% saying that they want no more children.
In 1994, about 83% of women aged 20-24 reported having ever used the pill, although their contraceptive practice was not limited to this method. Seventy percent had ever used some other method—mainly condoms (58%). Other contraceptive methods that had also been used by young women included withdrawal (32%) and periodic abstinence (12%). But these two methods appear to have been used only on a short-term basis: The proportions of women aged 20-24 who reported using these methods alone or combined with another method during the month the survey was conducted were 12% for condoms, 1% for withdrawal and 2% for periodic abstinence.
In summary, one of the main features of French contraceptive practices found in the 1994 FFS is the high rate of use of reversible medical methods. Oral contraceptives or IUDs are used by more than one-half of women aged 20-49. In contrast, use of sterilization is limited, especially among women younger than 40. The typical contraceptive sequence of French women begins with 10 years of pill use, followed by use of the IUD. This pattern was reported by one-half of women during the mid-1980s, according to the 1988 FFPS. During that time, one out of five women reported being sterilized, usually between the ages of 40 and 50.6
Single Men and Women
In France, as in many other developed nations, the percentage of single men and women has been rising. During the 1970s, the increasing number of unmarried cohabiting couples compensated for a declining marriage rate in the country. Since 1980, men and women have been delaying couple formation, and an increasing number of divorces has led to a greater percentage of people living outside a couple arrangement.7
Because both men and women were included in the FFS survey, it was possible to compare contraceptive behaviors reported by each gender (Table 2). Among people living as a couple—either married or unmarried—the results were similar for men and women: When we restricted the analysis to couples where both members were aged 20-49, we found that the percentages reporting the use of any method never differed by more than one percentage point. We may conclude that for people living as a couple, contraception was seen as a joint behavior, whether the contraceptive actually used was considered a "female" method or a "male" method.
Because there was no question in the survey about sexual intercourse during the preceding month, it is difficult to directly compare single men's and single women's contraceptive behaviors. Nevertheless, men and women not living as a couple reported different contraceptive behaviors. Table 2 indicates that oral contraceptive use was reported by the highest proportion of single men, followed by condoms; for single women, oral contraceptives were again the most frequently used contraceptive method, but far fewer women reported condom use. One obvious explanation for these differences was that even when they are not currently sexually active, women who had relied on oral contraceptives or the IUD may still take the pill or have an IUD in place, and so continue to identify themselves as users of those methods. Furthermore, men may declare themselves to be condom users, even if they are not currently having sexual intercourse, if they have used that method in the past or if they expect to use it in the future.
Dual Method Use
Contraceptive practices were not always limited to a single method. For example, 5% of all women and 7% of all men aged 20-49 in the FFS reported using two methods simultaneously. The hierarchy (described in Table 1) used to define a "main method" favored the most effective methods. This leads to an underestimate of the use of nonmedical methods, which are, in fact, used by 15% of women and 19% of men aged 20-49 (compared with 12% of women and 15% of men using a nonmedical method as their "main" method).
Women who used medical contraceptives seldom used another method. Only 6% of pill users and 3% of IUD users said they were also using another method. In both groups, the additional method was usually the condom; 5% of pill users and 2% of IUD users also used condoms in the month preceding the survey.
Women who used nonmedical methods frequently combined use with another method. Second-method use was reported by 42% percent of condom users (those using condoms either as a main or a complementary method), 37% of withdrawal users, 43% of periodic abstinence users and more than 50% of vaginal methods users. The most common method associations were between oral contraceptives and condoms (used in conjunction by 2% of all women and 3% of all men aged 20-49), withdrawal and periodic abstinence (by about 1% each) and condoms and periodic abstinence (also by about 1% each).
Trends in Contraceptive Use
Examining the results of the 1978, 1988 and 1994 surveys along with some contraceptive sales data allows us to describe trends in French contraceptive behavior during the last 30 years. The use of contraceptives became legal in December 1967. Between 1968 and 1988, pill and IUD use increased sizably among women aged 20-44, while the percentage of sterilized women decreased (Table 3).
The percentage of women using the pill also increased between 1988 and 1994, while the percentage of IUD users decreased during that period. Finally, while reliance on most nonmedical methods (such as withdrawal and periodic abstinence) declined between 1978 and 1994, the percentage of women aged 20-44 using condoms increased between 1988 and 1994, most likely because the method offers protection against HIV.
Pill use increased rapidly until 1975, when growth in pill use slowed. Sales statistics show that pill use has plateaued8 in France since 1990, which contradicts the 1994 survey results that show a continuing increase. (This gap can be explained by the use of progestin-only pills before or around menopause, however, as those are not included in the sales statistics.) By 1988, pill use appeared to have crested, with 88% of women aged 25-29 having ever used the method (not shown).
IUD insertions have almost halted in France, partly because the age range of possible users has narrowed. (This decline is unlike that in the United States, where IUD use has become very uncommon because of legal liability concerns.) Medical standards in France forbid physicians from prescribing IUDs to childless women, because of the risks of infection and subsequent sterility. (This risk is not higher for childless women, but the consequences are considered more serious for this group.) Because French women have been delaying childbearing, the mean age at which they become "eligible" for IUDs has therefore been increasing.
The number of sterilizations has been declining in France since 1978. Between 1978 and 1988, this decrease was due mainly to a decline in the number of sterilizations for medical purposes only. Between 1988 and 1994, sterilizations for contraceptive purposes also fell, especially among women younger than 40. Vasectomies remain extremely rare in France, and women are sterilized only after the age of 35 or 40—and only if they already have two or more children—because French doctors are reluctant to perform "early" sterilizations.9
The motives for prescribing oral contraceptives in France have become more diverse. Doctors usually first prescribe them to young women in order to regulate their hormonal cycles (regardless of sexual activity). Pills are also prescribed to older women in order to prevent troubling menopausal and premenopausal symptoms. Moreover, because minipills or third-generation pills supposedly have fewer side effects, there is less resistance to prescribing oral contraceptives for more than 10 years of use. Consequently, oral contraceptive use has grown steadily through all age-groups of French women (Figure 1); IUD use, on the other hand, has remained negligible among younger women, and was prescribed less frequently in 1994 than in 1988 among women younger than 35 (Figure 1).
During the 1970s, use of medical methods (oral contraceptives and IUDs) was higher among socioeconomic groups more familiar with medical prevention, such as more educated women, women living in urban areas and white-collar employees. By 1988, however, contraceptive behavior had become more homogeneous. In particular, the percentage of pill users did not vary by education, profession or place of residence. (IUD use was still more frequent among more educated women.10)
In 1994, the percentages of women using IUDs and pills were almost identical within various socioeconomic groups, despite an increase in oral contraceptive prices. Because we do not have any information about the type of pills used by survey respondents, we can only presume that the use of nonreimbursed oral contraceptives did not introduce new differentials in contraceptive behavior patterns.
Condom Use Patterns
Another major recent change in French contraceptive practices has involved condom use. In 1994, condoms were used by 8% of women aged 20-44, up from 5% in 1988 (Table 3). In both 1988 and 1978, condom use as a primary contraceptive method was about the same in all age-groups, at around 5%. Condoms were mostly used as a short-term method during this period, while women took a break from the pill or before they adopted a medical contraceptive method.11 In France, the first AIDS prevention campaigns were launched in 1988. The recent increase in condom popularity has involved mostly people younger than 30. In 1994, 53% of women aged 25-29 and 58% of women aged 20-24 had ever used condoms (Table 4). (These proportions represent an increase from 38% and 23%, respectively, in 1988.) The percentage of women older than 30 who had ever used condoms remained stable, at around 42%.
The 1994 FFS revealed that 83% of men aged 20-24 had ever used condoms, compared with about 57% between the ages of 30 and 39. According to a survey on sexual behavior in France conducted in late 1991, 40% of women aged 20-24 reported ever having used condoms;12 this percentage falls between the level of 23% reported in 198813 and that of 58% reported in 1994. The percentage of men and women aged 30 and older who reported ever using condoms in 1994 was nearly constant with age, at around 40% for women and 50-60% for men. The recent increase in condom use occurred mainly among men and women born after 1965, and seems to have been as large between 1992 and 1993 as between 1988 and 1991.14
The 1991 survey also revealed that condoms were more likely to be used during the early stages of new relationships, and couples then turned toward other methods as their relationship became more stable. In 1994, condoms were still mainly considered a temporary rather than a long-term method: Sixty-one percent of men aged 20-49 had ever used condoms, but only 32% had used them during the 12 months preceding the survey, and just 13% had used them during the preceding month. For women, the corresponding figures were 46%, 19% and 7%, respectively (Table 4).
Not surprisingly, condom use during the month the survey was conducted occurred most frequently among single men and women: Twenty-six percent of single men and 10% of single women used condoms in the survey month, compared with only 7% of cohabiting men and 6% of cohabiting women (Table 4).
Among couples, the percentages were similar for men and women, which suggests that the practices were accurately reported. Among respondents who lived alone, condom use was reported more frequently among men than among women. Many surveys on sexual behavior have shown that men tend to report more sexual partners and more frequent acts of sexual intercourse than women. (A small part of these discrepancies may result from homosexual relationships or sexual relations with prostitutes.15) It has been assumed that men tend to overreport sexual activity, and that women may tend to underreport sexual activity by selecting their more "significant" relationships. We may also assume that similar overreporting by men occurs with condom use, and that single men probably tend to overestimate their use of the method.16 On the other hand, single women may underestimate their sexual activity and condom use.
As condoms apparently are often employed for only a short period of time, their reported use during the past year may be a better indicator than current use. In 1994, nearly two out of three noncohabiting men (61%) and one out of three noncohabiting women (33%) used condoms during the last 12 months, according to the FFS. Condom use was reported more frequently among young single men and women. The percentage of single men who used condoms during the year before the survey increased as age declined—from 34% among single men aged 45-49 to 74% of single men aged 20-24; the corresponding figures for single women were 13% and 45%, respectively. Single men and women more frequently have sexual intercourse with new partners,17 and younger single people also may adopt condom use more easily than their older counterparts.
Among married or unmarried cohabiting men and women aged 20-49, only 18% of men and 14% of women said they had used condoms during the last 12 months. When we separated those who had been cohabiting for more than one year from more recently formed couples, the latter reported more frequent condom use: Seventeen percent of men and 13% of women who were cohabiting for one year or longer before the survey reported using condoms during the last 12 months, compared with 40% of men and 32% of women who were cohabiting for less than one year (not shown). However, condoms probably are replaced by another contraceptive method (usually the pill) when couples settle down together: Among men and women who were cohabiting for less than one year, only 9% and 7% were using condoms at the time of the survey, indicating that many respondents who had used condoms in the past year might already have switched to another method.
In 1988, condoms were mostly used by respondents who were older than 30 who already had one child and who wanted to have another. This level of use was similar among different social groups. When we controlled for demographic variables, we found no significant contrasts in the percentages of women using condoms for contraception by profession, place of residence or importance of religion. The only significant variation was found between the most educated women and women without any secondary education: Nearly 7% of the most educated women and 3% of the least educated had used condoms in the last month—an odds ratio of 2.2 after standardization for demographic variables.18
In 1994, condom use was at the same level among less-educated women as in 1988 (3%), but had increased to 10% among the most educated women, producing a standardized odds ratio of 5.0.§ Moreover, differences by profession and place of residence were significant. Condoms were used more by white-collar workers and residents of cities with populations over 100,000. When we examined condom use in the year preceding the survey (Table 5), we found that it varied among men, from 24% in small cities to 43% in the Paris metropolitan area. Condom use among women in these areas ranged from 16% to 23%, respectively.
Differences in condom use were also pronounced among men with various levels of education. Twenty-three percent of men with less than a secondary school education reported using condoms, compared with 49% of men who completed their secondary schooling and 42% of men with university degrees. Women reported even greater educational differences in condom use, ranging from 11% among women who completed less than a secondary level of education to 30% among those with university degrees. These differences remained after standardization by demographic variables, with the odds increasing along with education in both genders (p<.01). In 1994 as in 1988, the importance of religion had no impact on condom use or on contraceptive behaviors.
Contraception and STD Prevention
Condoms are not only used for contraceptive purposes, but also for protection against STDs and HIV. People who used condoms during the month before the survey said they did so mainly for contraceptive motives, and women equated condoms with contraception more often than men. When asked for their reasons for condom use,** 58% of women who used condoms in the previous month referred only to the risk of pregnancy, 19% referred only to STDs and AIDS and 23% mentioned both reasons (Table 6). Men most frequently referred to both reasons (42%), followed by contraceptive motives (30%) and then by protection against STDs and AIDS (24%).
Motives for recent condom use were described differently by users who were cohabiting—either married or unmarried—than by those who lived alone (Table 6). Seventy-four percent of cohabiting men and 90% of cohabiting women who used condoms in the month before the survey reported that pregnancy prevention was their main motive; only 6% in each group said they used condoms in the last month to prevent infection with HIV or other STDs. Among those living alone, 33% of men and 35% of women had used condoms for STD prevention in the month before the survey, while 58% of men and 49% of women who lived alone used the method for both pregnancy and STD prevention.
Women mentioned contraceptive motives for condom use more often than men, for two reasons. First, single women used the method less frequently than single men: Sixty percent of women who used condoms were cohabiting, compared with only 39% of men. Furthermore, women in cohabiting relationships were more likely to give contraceptive motives for condom use than were men. Second, women who were cohabiting were less likely than were cohabiting men to refer to STD prevention as their main motive (9% vs. 19%); women who lived alone also mentioned this reason less often than did single men (84% vs. 92%).
Overall, in 1994, contraceptive motives appear to have driven the majority of condom use in France (including among married and single men), despite the fact that pregnancy prevention was not included in any of the educational campaigns about condoms.
Condom Use at First Intercourse
In 1970, the majority of people who had first sexual intercourse did not use a contraceptive method (Figure 2). During the 1970s and the 1980s, the percentage of women who did not use birth control at first intercourse began to decline, from 60% in 1970 to 30% in 1987. Since 1987, this decline has accelerated, and the proportion reached 16% in 1993. In the 1970s and the 1980s, the pill became the contraceptive choice of young people and was adopted before first intercourse by nearly one-half of women.19
After 1988, condom use at first intercourse increased substantially, from 8% in 1987 to 45% in 1993, as AIDS educational campaigns focused on preventing sexual transmission of the disease. A combination of pill and condom use was uncommon at first intercourse (5%), and the increase in the percentage of women who used condoms at first intercourse was associated with stagnation in the trend in pill use at first intercourse.
By 1994, condoms were the most frequently used contraceptive method at first intercourse. A 1994 adolescent sexual behavior survey revealed that condom use at first intercourse might be even higher than we estimate here. In that survey, 74% of young women and 79% of young men aged 15-18 reported that they had used condoms at first intercourse.20
Men reported more often than women that their first intercourse was unprotected (Figure 2), and their partners' level of pill use was lower than that reported by women at first intercourse. (Men probably underestimated pill use, however, since oral contraceptives are a female-controlled method and not directly linked to sexual intercourse.) During the 1970s and the 1980s particularly, many men may not have known about their partners' contraceptive practices, since birth control may not always have been discussed.
Men's condom use at first intercourse began to increase sharply in 1985. Starting in 1990, condom use at first intercourse rose at a slower pace, while the percentage of men who reported that their first sexual partner used oral contraceptives began to increase. Men's survey answers probably reflected new behavior patterns, in which partners discussed condoms and other contraceptive methods before first intercourse. They then chose between using oral contraceptives or condoms, which gave men a larger role in their partners' contraceptive choices than in the past.
French contraceptive behavior appears to be unique: Female sterilization is rare, especially among women younger than 40, and male sterilization is virtually nonexistent. Considered together, oral contraceptives and IUDs are more commonly used in France than in any other country.21 The pill is by far the most frequently used contraceptive method. It has been estimated that 36% of women aged 15-49 were using oral contraceptives, which amounts to more than five million women. By the 1980s, all social groups were reporting similar levels of pill use,22 and the IUD appears to have become the method of choice for women 35 and older after they stop using the pill.
The recent availability of new pills, such as minipills and third-generation pills, has led to an increase in pill use by older women, while IUD use has remained constant among women older than 35 but decreased among younger women. In France, medical contraceptives are reimbursed by the social security system, but there have been lags between the appearance of new pills and the government's agreement to reimburse their costs.23
During the last 10 years, the number of nonreimbursable oral contraceptives has been increasing, following the introduction of "third-generation" combined pills. In 1991, about 53% of prescribed oral contraceptives were not reimbursed by the social security system, up from 10% in the mid-1980s. In early 1992, the government agreed to reimburse the cost of some of these pills, and the percentage of nonreimbursable contraceptive pills now stands at around 42%.
However, an increasing number of French doctors have been prescribing pills to treat menopausal and premenopausal symptoms, and these uses are covered by the social security system. Since such use also has contraceptive features, it may have been reported as contraception in our survey (although they are not defined as contraceptives by the Ministry of Health and are not included in contraceptive sales statistics).
In 1993, more than 17 million of these pills had been sold to 1.4 million French women for menopausal and premenopausal symptoms. Under the plausible assumption that one out of three of those users were younger than 50 and reported their pill use as contraception in the 1994 survey, the gap can be filled between the sales data and survey answers to estimate about 5.2 million pill users in 1994. When we added these women into conditions of reimbursement, the percentage of nonreimbursed pills fell to 33%.
Use of withdrawal, periodic abstinence and sterilization have all declined in France. Condom use increased since 1988, in contrast, especially among men and women younger than 30, and particularly at first sexual intercourse. Recent public educational campaigns about the role of condoms in preventing the transmission of HIV have increased the method's popularity among persons who frequently consult medical professionals and are familiar with AIDS prevention behaviors.
However, condom users reported that they relied on the method more for pregnancy prevention than for AIDS prevention. The condom also was employed most often by people who did not live as couples. The method apparently is frequently abandoned for the pill (or less frequently, for the IUD) when relationships become stable. The recent upswing in condom use, however, did not slow down the increase in pill use, with the possible exception of very young women at first intercourse.
The increasing use of contraceptives by French men and women contrasts with the relatively constant number of induced abortions in the country (approximately 30 abortions per 100 live births), indicating that the use-effectiveness of contraceptives may not be as high as expected.