basic trends affected all subgroups of the married population, although important disparities remained.

Previous studies, particularly the Princeton Fertility Study,20 showed a pattern of nonuse or indifferent use of contraception in the early years of marriage until the desired number of children was reached or exceeded, followed by more regular and systematic use. The 1970 study suggests that this pattern is now shifting toward initiation of contraception earlier in marriage and use of more effective methods from the start of married life. Both processes have important implications for the ability of couples to control their fertility successfully.

Table 10 presents the 1970 data on the proportion of couples who initiated contraception before the first pregnancy and who had never used contraception. The question is retrospective and covers, for some respondents, experience dating back to the 1940's, before the advent of the Pill and the IUD. Presumably, couples who initiate contraception before the first pregnancy are better able to space the arrival of their children deliberately and control their total number.

Among all couples, there were still some differences by color, poverty status, and marriage duration in the proportion who had never used any contraception: More poverty-status and nonwhite couples had not yet done so. Interestingly, although couples married less than five years have had substantially less exposure to risk than couples married longer, in each subgroup a smaller proportion of these younger married couples had never used contraception than of older couples.

The most striking difference is the proportion initiating contraception before the first pregnancy between couples married less than five years and those married longer. In each poverty-status and color group, a considerably larger proportion of younger than of older married couples initiated contraception before their first pregnancy. Above the 150 percent-of-poverty cutoff, the proportion of both white and nonwhite couples married between 1966 and 1970 who initiated contraception this early in marriage was two-thirds greater than among couples married before 1966. Below the poverty cutoff, the increase among wives married since 1966 was even greater: Contraception was initiated before the first pregnancy by three times as many couples married in the last five years as by couples married before 1966. Among all younger married couples, fewer povertystatus and nonwhite couples began contraception this early in marriage. But more poverty-status couples married between 1966 and 1970 initiated contraception before the first pregnancy than couples above the poverty cutoff married before 1966 (Figure 2). Lowincome and nonwhite couples appeared to lag behind the overall pattern by about five years.

tion of the childbearing population, the contraceptive profile tends to converge toward the most effective medical methods.

In 1970, 17 percent of all couples with wives under 45 had a sterilization operation either for contraceptive reasons or to remedy pathological conditions. The highest rates of sterilization were found among white non-Catholics (19 percent) and blacks (18 percent). The incidence of contraceptive sterilization in 1970 was 11 percent. Table 11 shows the percent of couples, by age of the wife, in which either husband or wife had undergone a medical sterilization or a contraceptive operation. More than one-fifth of couples with wives aged 30 to 34 had been sterilized and this proportion increased to one-third of couples with wives aged 40 to 44. Contraceptive operations were almost evenly divided between vasectomy and tubal ligation: About five

Despite these important changes, nearly two-thirds of poverty-status couples and half of couples above poverty status married less than five years in 1970 did not use contraception until after the birth of the first child. Other research has shown that, from 1964 to 1966, one-third of all first births in the United States were conceived premaritally-15 percent resulting in out-of-wedlock births and 18 percent occurring less than eight months after marriage.21 Initiation of contraception before the first pregnancy thus depends, for some couples, on access to effective contraception before marriage. The implications of these findings for family planning programing will be explored below.

During the 1966-70 period, earlier use of contraception in marriage was accompanied by a major shift toward use of more effective methods. Use of the Pill and the IUD increased significantly. These two methods are the most effective currently available; for analytical purposes, they can be grouped with the next most effective techniques-the diaphragm and condom-in a classification of "most effective and moderately effective" methods. By 1970, three-quarters of all contraceptors (including two-thirds of Catholic contraceptors) were using these methods. The corollary of this shift is a smaller proportion of contracepting couples who relied on less effective methods. A major factor in this change was the adoption of the most effective medical methods by younger married couples in all subgroups. As couples married more than five years who began family planning in pre-Pill days become an increasingly smaller propor

Table 11.-Proportion of Couples with Contraceptive or Medical Sterilizations, by Age of Wife, 1970

Total

Below 19 years 20-24 years 25-29 years 30-34 years

35-39 years

40-44 years

Source: 1970 National Fertility Study.

percent of all couples had undergone each operation, suggesting that vasectomy is being utilized to a greater degree. Examination of the base data reveals, as would be expected, that the incidence of contraceptive sterilization was heavily concentrated among older couples: While four percent of couples with wives aged 20 to 29 had undergone contraceptive sterilizations, 17 percent of couples with wives above 30 had done so.22 (This compares to a rate of 10 to 11 percent of white couples with wives older than 30 in 1965).23

A cross-sectional view of the fertility control practices and exposure to conception of the entire married

population in 1970 is presented in Table 12. Some of the traditional differences in practice remained: A smaller proportion of poverty-status couples than of couples above the poverty cutoff utilized the most

effective or moderately effective methods, and a larger proportion used the least effective methods or none at all. But the main finding is the picture of convergence on a very similar pattern among all sectors of the married

population. The proportions pregnant or seeking pregnancy in each subgroup are quite similar. The proportions sterilized vary slightly within a narrow range, with more poverty-status couples (and particularly nonwhite poverty-status couples) having undergone sterilization. The most striking similarity is in the proportion of couples married less than five years who had used the most effective or moderately effective methods: Nearly half of these couples did so, with no large differences associated with poverty-status or color.

Table 12 begins to focus on the directions United States family planning programs would need to take to assist couples in avoiding unwanted pregnancy. Some couples would not be expected to be using contraception at a given point in time-for example, those who are pregnant or seeking pregnancy or those who are already sterile. But almost all other couples (including those whose natural sterility is not definitively established) will require effective contraception to avoid number and timing failures. Unmarried persons have special family planning needs; current users of the most effective methods will require continuing supervision and reinforcement; and about half of the younger married couples will need assistance in initiating effective contraception before the birth of their first child. In addition, there are two other groups of married couples who need help from a family planning program-the nonusers, who comprise 15 percent of all couples, and users of less. effective methods, who make up 14 percent. (If Catholic rhythm users are excluded from this calculation, the proportions using less effective methods would be 13. percent of white non-Catholics, 10 percent of Catholics, and 12 percent of nonwhites.)

who

The proportion of couples who use no methods or the least effective methods is summarized in Table 13. Two measures are shown-one including all rhythm users among those practicing the least effective methods, and the other excluding Catholic rhythm users on the assumption that they use rhythm for compelling religious reasons. Both are minimum estimates since they do not include currently pregnant women previously used less effective or no methods and who will require improved contraceptive services after delivery; nor do they include users of condoms and diaphragms who experience failure rates two to four times greater than users of the Pill and IUD. Below 150 percent-of-poverty, about one-third of married couples fall into this minimum category of need, compared to somewhat more than one-fourth of couples above poverty status. In both socioeconomic groups, couples married less than five years appear to be better protected against unwanted pregnancy than those married longer.

We would expect the shift toward increased use of more effective methods of fertility control since 1965 to result in fewer unwanted births. The 1970 study, like its predecessors, asked respondents to characterize each child as wanted or unwanted at the time of conception. Wives characterized 15 percent of all births occurring between 1966 and 1970 as unwanted; the proportion reported as unwanted increased sharply by birth order, as would be expected, climbing to one-third of fourth births, nearly half of fifth, and nearly two-thirds of sixth and higher order births. (See Table 14.)

A greater proportion of births to couples below the 150 percent-of-poverty level were reported as unwanted than of those to couples above poverty-status-27 percent compared to 12 percent. Almost all of the overall white-nonwhite difference is attributable to the very high proportion (40 percent) of births reported as unwanted by nonwhite poverty-status wives.

The definitions of unwanted births employed in the 1970 study are somewhat different from those used in 1965; thus, the present data are not entirely comparable with the 1965 medium estimates that 19 percent of all births occurring between 1960 and 1965 were unwanted, including 32 percent of births to poor and