Estimates of unmet need for family planning derived from fertility and family planning surveys have now become standard statistics. The measure has become popular in policy and program circles for its usefulness as an indicator of the market for services, and for assessing future demand and the potential for fertility reduction.

All existing estimates of unmet need are cross-sectional, however, and thus do not take into account changes over time. Although repeat surveys of different samples of women permit a view of aggregate changes in the proportions of women who are in need of contraception,1 we still cannot determine what happens to women in need. Do they become pregnant or do they become contraceptive users? What fraction remain in need? Do nonusers with an unmet need for spacing births adopt contraception when they reach their desired number of children or do they shift into the category of having an unmet need for limiting births? Of special interest is the subsequent behavior of women in need who said they did not intend to use a method, for one reason or another. We are concerned with what prevents women from using a method, and which of these impediments are the least intractable.

Data and Methods

Transitions over time in contraceptive needs and fertility intentions can only be investigated with a longitudinal research design; the 1995 Morocco Demographic and Health Survey (DHS)2 incorporated a longitudinal panel that permits such an analysis. A subsample of approximately half of the women who were initially interviewed in the 1992 Morocco DHS were reinterviewed in 1995 (3,324 women who were aged 15-46 in 1992), in addition to 1,429 new respondents aged 15-17 and other new women in the selected households. No attempt was made to locate women from households from the original sample that had moved during the period between the surveys.

The analysis reported in this article is based on a subset of 1,682 women of the total of 3,324 who were interviewed twice. We arrived at this final sample after removing cases suspected to be incorrect matches (because of gross inconsistencies in their reported ages or numbers of children) and excluding women who were not married to the same partner at both times. Compared with women who could not be successfully located and reinterviewed, those who could were younger, were better educated, were more likely to reside in cities and had fewer children.

Several modifications of the standard measure of unmet need have been recently proposed, and are incorporated here; the rationales for these changes are described elsewhere.3 Briefly, in defining unmet need, the new algorithm excludes the intentions of pregnant or amenorrheic women who stated that they either wanted more children or had used a method in the past and intended to resume use. (Other analyses of the Morocco data have shown the reported planning status of the last pregnancy among pregnant or amenorrheic women to be very unreliable.4) The reproductive intentions of all women are substituted instead. In an analysis of data from 19 countries, use of this revised definition typically reduced the proportion of women with an unmet need for contraception from an average of 25% to 21%.5

Unmet Need and Subsequent Births

According to the new measure, the proportion of Moroccan women in need of contraception declined substantially between 1992* and 1995, from 21% to 12%, which represents a decline of roughly 43% over the period.† Conversely, the proportion of women who were using a method increased over the three-year period from 46% to 56%.

A very high proportion (84%) of women who were classified in 1992 as being in need of methods to space births gave birth in the subsequent three years or were currently pregnant at the time of the interview (see Figure 1). This high proportion was not unexpected, however, because all of these women said in 1992 that they intended to have another child and were only hoping to delay a pregnancy. Even most of the women who in 1992 were using a method to space that pregnancy would have interrupted use in the period between the surveys to become pregnant.

Notably, among the women who were classified in 1992 as having an unmet need for contraception to limit births, 58% had given birth by 1995 or were currently pregnant. These were essentially fecund women who wanted no more children but were not practicing contraception (a total of 15% of the reinterview sample in 1992). By definition, most of those births would have been unwanted, although changes of mind certainly occur. However, only one-third of these births were actually reported, in retrospect, as unwanted. (The tendency to redefine a birth as wanted, which occurs quite frequently and increases with time since the birth, is one reason why the measure of unmet need was modified.) Even among women who were using a method in 1992 and wanted no more children, 25% gave birth over the three-year period, but less than half (46%) of these women said the birth had been unwanted at the time of conception.

Transitions in Planning Status

Our analysis focuses on the transition of women through the different categories of contraceptive need and use. The basic cross-classification (Table 1) shows the changes in three categories of women between 1992 and 1995: those with an unmet need for contraception, current users and "all others" (i.e., those who were not using a method because they were trying to become pregnant, were pregnant or amenorrheic and either wanted more children or had used contraceptives and intended to resume use, or were infecund). This analysis concerns only the classification at the time of each survey and ignores women's experiences during the three-year interim period; it is thus analogous to examining two snapshots of the same individual's reproductive status, one in 1992 and one in 1995.

Most of the women classified in 1992 with an unmet need for contraception had moved into the current user category or the "other" nonuse category in 1995. However, 29% remained in the unmet need category. Most of the women who were using a method in 1992 remained in that category three years later (77%); very few users stopped practicing contraception and moved into the unmet need category (5%). Just over half (51%) of women in the "other" nonuser category were still there three years later (mostly infecund women and those who still wanted more children soon), while 37% had moved into the user category.

Table 2 shows a more detailed picture of these transitions. Here, we disaggregate both the women in need of contraception and users of contraceptives into spacing and limiting categories. Almost all contraceptive users in 1992 remained users in 1995. The most serious concern involves women who in 1992 were classified as having an unmet need (either for spacing or for limiting) and who were classified as having an unmet need for limiting births in 1995. Although these women accounted for only 5% of all married women, they will contribute most of the unwanted fertility: Two-thirds of these women had had at least one birth in the three-year interval between surveys, compared with nearly half of women overall. Moreover, 20% of such women reported 2-3 births over the period, compared with 7% of all women in the sample. Two-thirds of women who were in need at both surveys had never used any method of contraception.

What distinguishes these women from others who subsequently became users? Those who shifted from being in need to using a contraceptive method tend to be younger, to live in cities, to have some formal education, to have more exposure to mass media, and to have used a method in the past. When we examined all of these variables simultaneously in a regression analysis to determine the likelihood that contraceptive need would be fulfilled, only past use and education independently predicted the transition to current use.

Intentions for Future Use

An important finding from our earlier work on unmet need was that a high proportion of women in need did not intend to practice contraception in the future.6 In Morocco in 1992, two out of five women in need were in this category. Among these women, 38% changed their minds and began contraceptive use between 1992 and 1995, while 62% did not.7

In 1992, all women who had no intention of using a method were asked why this was so. Most responded that they felt ambivalent about the timing of their next birth. The next most common responses were that they lacked information about methods or had health concerns, that they or their husband opposed contraception, that they feared side effects and that they had other reasons for not practicing contraception.

We exploited the longitudinal design of the Morocco DHS to examine the extent to which these reasons were "overcome" in the ensuing three years. Figure 2 shows the percentage of women who in 1992 did not intend to practice contraception, but who used a method between 1992 and 1995, classified by their stated reason for not intending to do so. Clearly, opposition to contraception (mainly for religious reasons and partner's objections) was the most difficult reason to overcome, and a fear of side effects was the weakest impediment to eventual use.

These conclusions are based on small numbers of women, however. Moreover, the results apply to all women with unmet need, whereas the need for limiting births is clearly the more important subcategory. However, if we were to focus on intentions about future use among women in need for limiting only, the number of observations would be reduced further. Thus, to avoid even smaller sample sizes, we relaxed the definition of unmet need and examined intentions about future use among the 42% in the sample who, in 1992, wanted no more children, were not practicing contraception and did not intend to use a method in the future.

Figure 3 presents the proportions of these women who used a method between 1992 and 1995, by the reason they originally offered for not intending to do so. Since women who wanted no more children tend to be older, on average, a higher proportion are at a relatively low risk of pregnancy--that is, they have difficulty conceiving or have sex infrequently--and fewer are ambivalent about their intentions. Thus, their profile of reasons for nonuse differs from that of all nonusers (Figure 2). Although the numbers are still relatively small, the same conclusions hold: Women who were opposed to use for religious reasons (or whose husbands opposed use) in 1992 were the least likely to have become users by 1995, and those who were concerned about side effects were the most likely to have done so.

Discussion and Conclusions

The availability of data from a national longitudinal study made it possible to track Moroccan women who were classified in 1992 as having an unmet need for family planning over the following three years. A high proportion--58%--of those who were classified as having an unmet need for limiting births in 1992 had reported one or more births between 1992 and 1995 or were currently pregnant at the 1995 interview. Moreover, 25% of women who were using a contraceptive method in 1992 (mostly the pill) and who wanted no more children reported a birth in the ensuing three years. Even granting changes of mind, these are strong indications of unwanted fertility.

Nearly 29% of women in need of contraception in 1992 were still classified as being in need three years later; the fact that many of these women in persistent need wanted to limit births is a serious concern. Although the group still in need for limiting births constitutes only 5% of the total reinterview sample, they will contribute most of the unwanted fertility. Almost all (97%) of these women were without formal education, and two-thirds had never practiced contraception.

Our analysis revealed opposition to use for religious reasons or because of husbands' objections to be the most intractable reason for nonuse. In contrast, fear of side effects and lack of information were the most easily surmounted obstacles, as women citing these reasons were the most likely to be persuaded to adopt family planning.