1.1.1 Sample Characteristics

The first NSFG was done in 1973 and sampled only female persons aged 15-44 years who had ever been married. The 1982 edition of the NSFG expanded the sample to all females regardless of marital status, and the 2002 edition of the NSFG was the first to sample males in addition to females. More recently, the 2015-2017 edition of the NSFG expanded the age range for sampling from 15-44 years to 15-49 years.

In the NSFG, female respondents and male respondents are given different questionnaires, and NSFG results are provided in separate data tables for females and males with survey design information specific to the data table. Therefore, separate estimates for females and males are given throughout this report. Estimates for both females and males are only combined when no attempt is made to quantify sampling error.

The 2017-2019 edition of the NSFG has a sample size of 6,141 females and 5,206 males. The overall response rate is 63.4%, with response rates of 65.2% for females and 61.4% for males. Results are weighted to approximate the population at July 2018, the midpoint of sampling. The sampled population is persons aged 15-49 living in households in the United States and includes neither those without a current residential address nor those living in group housing, such as military bases, correctional facilities, etc. However, the population does include college students living in dormitories by way of their parents’ households.

1.1.2 Sexes of Sexual Partners

Use of contraceptive methods for the purpose of contraception per se is only relevant for sexual intercourse with the opposite sex, though, as seen in Section 2.5, contraceptive methods are sometimes used for other purposes than contraception. Therefore, the NSFG and this report focus extensively on sexual intercourse between individuals of opposite sex. All mentions of sexual activity in this report should be interpreted to refer to sexual intercourse with someone of the opposite sex unless stated otherwise.

1.1.3 Disclosure Risk Reduction

Before the NSFG public use data files are released, actions are taken in order to reduce the risk that a specific respondent can be personally identified based on the information to be contained therein. This disclosure risk reduction can be as severe as suppressing certain variables from the public use data files entirely, though many variables are merely modified to be less specific. For instance, while respondents are asked about the month and year that pregnancies begin and end in the NSFG, only the year that pregnancies begin and end are available in the public use data files.

This report is based entirely on the NSFG public use data files. Whenever analysis is constrained in some way by the disclosure risk reduction, this is noted in the report.

1.1.4 Documentation

More information about the NSFG is available in its official User’s Guide and appendices (National Center for Health Statistics 2020c), the Summary of Design and Data Collection (National Center for Health Statistics 2020e), and its more detailed design documentation (National Center for Health Statistics 2020a, 2020b, 2020d).

2.2.1 Reason for Nonuse

Some persons intentionally refrain from using contraception for various reasons. The most common reason for intentional nonuse of contraception is a desire to conceive a pregnancy, followed by the female partner currently being pregnant. (Table 2.4)

The “otherwise did not use contraception during last intercourse” category in Table 2.4 estimates those who are at increased risk for unintended pregnancy due contraceptive nonuse.

“Postpartum” is categorized separately in Table 2.4 for females, but not for males because this information is not available in the public use data files for males due to disclosure risk reduction. “Otherwise did not use contraception during last intercourse” for males therefore includes some persons who last had intercourse with a postpartum female. However, as can be gleaned from the estimates for females, this is a small number of persons.

The “physically infertile not due to sterilizing procedure” category is explored in Section 2.2.4.

2.2.2 Contraception Use Rate

Combining the estimates from Table 2.3 and Table 2.4, an overall contraceptive use rate can be calculated with the number of persons using contraception in the numerator and the number persons at risk for unintended pregnancy in the denominator. Those “at risk for unintended pregnancy” refers to persons who had sexual intercourse with someone of the opposite sex in past 3 months and are not themselves nor had intercourse with a partner who is intending to become pregnant, currently pregnant, postpartum, or physically infertile.

Generally, the contraceptive use rate is high, though this still leaves several million persons at increased risk for unintended pregnancy due to contraceptive nonuse. (Table 2.5)

The reported contraceptive use rate is higher for females than for males. A chi-squared test of independence with second-order Rao-Scott adjustment is statistically significant \((p = 0.0048)\). Therefore, there is sufficient evidence to conclude that a person’s sex and reported contraception use during last intercourse are dependent, and the difference in contraception use rates between females and males in Table 2.5 is not attributable to sampling error.

2.2.3 Sexual Inactivity

Among persons who have not had intercourse with someone of the opposite sex in the past 3 months, a sizable minority have never had intercourse with someone of the opposite sex. (Table 2.7)

Figure 2.1: Persons who have never had sexual intercourse with the opposite sex by age group.

Most persons who have never had sexual intercourse with someone of the opposite sex are in younger age groups, and totals of persons who have never had sexual intercourse decrease with increasing age. (Figure 2.1)

2.2.3.1 Reason for Never Having Intercourse

Figure 2.2: Females by reason for not having had sexual intercourse with the opposite sex, among those who have not. Dots represent point estimates and brackets represent 95% confidence intervals.

Figure 2.3: Males by reason for not having had sexual intercourse with the opposite sex, among those who have not. Dots represent point estimates and brackets represent 95% confidence intervals.

Many persons who have not had intercourse with someone of the opposite sex report “haven’t found the right person yet” as a reason, though this appears to be more common for males than for females. “Against religion or morals” is another commonly cited reason. It is notable that “other” is also a popular reason chosen for not having had intercourse with someone of the opposite sex. (Figures 2.2 and 2.3) This is indirectly explored in the subsequent Section 2.2.3.2.

2.2.4 Reason for Physical Infertility Not Due to Sterilizing Procedure

Figure 2.4: Females by reason reported for their inability to become pregnant, among those who report being physically infertile not due to a sterilizing procedure.

Figure 2.5: Females by reason reported for their partners’ inability to father a child, among those who report having a partner who is physically infertile not due to a sterilizing procedure.

The reason for physical infertility not due to a sterilizing procedure is part of the questionnaire for female respondents of the NSFG, but not the questionnaire for male respondents. As Figures 2.4 and 2.5 illustrate, because these conditions are relatively rare, the effective sample sizes of the relevant subdomains from a two-year sequence of the NSFG are small, and the sampling error is quite large relative to the estimates.

2.2.5 Number of Methods

While the vast majority of both females and males used only one method of contraception during their last intercourse, a substantial minority of persons used multiple methods. (Table 2.10)

2.2.6 Method(s) Used

Figure 2.6: Females categorized by whether or not each contraceptive method was used during last intercourse with a male in the past 3 months, with persons using multiple methods counted multiple times.

Figure 2.7: Males categorized by whether or not each contraceptive method was used during last intercourse with a female in the past 3 months, with persons using multiple methods counted multiple times.

Overall, there are similar distributions in contraceptive method(s) used during last intercourse reported by females and males. (Figures 2.6 and 2.7) For instance, condoms and pills are the most commonly reported methods by males, and they are the second and third most common methods reported by females. Indeed, female sterilizing operations, condoms, and pills are in the top three most common methods reported both by females and males. However, the most salient difference between females and males is that approximately 10.5M females report a female sterilizing operation being used during their last intercourse, whereas only approximately 6.7M males report a female sterilizing operation being used.

Other differences include a greater number of males reporting use of condom, pill, contraceptive patch, and “other” methods than females, and a greater number of females reporting use of intrauterine device (IUD), calendar-based, and temperature or cervical mucus test methods. Possible reasons for these and other differences are discussed in Section 3.1.

Estimates of the number of female and male persons using withdrawal, vasectomy, implant, injectable, and ring methods are similar between females and males. Methods such as suppository, female condom, diaphragm, film, foam, jelly, and cream are rarely reported used by either females or males. Withdrawal is the fourth most commonly reported method both by females and males.

2.2.7 Method Used Exclusively

Figure 2.8: Females categorized by which contraceptive method was used during last intercourse with a male in the past 3 months, among those who used exactly one method during last intercourse.

Figure 2.9: Males categorized by which contraceptive method was used during last intercourse with a female in the past 3 months, among those who used exactly one method during last intercourse.

Restricting analysis to persons who used only one method of contraception during their last intercourse, there is a similar overall pattern of contraceptive method use as seen in Section 2.2.6. (Figures 2.8 and 2.9) However, in this subdomain, the number of females and number of males reporting pill use is roughly the same. Another salient difference in this subdomain is that there are relatively fewer persons who report using withdrawal, indicating a substantial number of withdrawal users do not rely on withdrawal as their only method of contraception, which is investigated further in Section 2.2.8.

2.2.8 Proportion of Method Users Who Also Used Other Method(s)

Figure 2.10: Proportion using another method of contraception among females using each method of contraception during last intercourse with a male in the past 3 months.

Figure 2.11: Proportion using another method of contraception among males using each method of contraception during last intercourse with a female in the past 3 months.

A direct way to measure the extent to which specific contraceptive methods are used exclusively or in combination with other methods is to estimate the proportion of users of each method that used it with another method. (Figures 2.10 and 2.11.) The number of persons using safe period by temperature or cervical mucus test method, contraceptive patch method, emergency contraception, and “other” method are relatively few, constituting small effective sample sizes with large sampling errors. Indeed, estimates for contraceptive patch are contradictory between female and male respondents.

Ignoring those, calendar-based methods have the greatest proportion of users who also used another method during their last intercourse, particularly among females. As alluded to in Section 2.2.7, about half of withdrawal users used another method during their last intercourse. Thus, withdrawal use is split between those using withdrawal alone as a method of contraception and those using withdrawal in addition to another method.

While a simple majority of users of pill and condom methods used them exclusively during their last intercourse, a sizable minority of pill and condom users also used another method. On the other hand, female sterilizing operation, IUD, and perhaps contraceptive ring are the methods most often used exclusively.

The remaining subsections of Section 2.2 explore what other contraceptive methods are used in conjunction with the methods commonly used with others, i.e., calendar, withdrawal, condom, and pill methods.

2.2.9 Other Method(s) Used with Calendar Method

When used in conjunction with other methods, calendar-based methods are most commonly used with withdrawal or to a lesser extent condoms. (Table 2.11)

2.2.10 Other Method(s) Used with Withdrawal, Condom, and Pill

Figure 2.12: Females by other contraceptive method(s) used during last intercourse, among those who used withdrawal during last intercourse with a male in the past 3 months.

Figure 2.13: Males by other contraceptive method(s) used during last intercourse, among those who used withdrawal during last intercourse with a female in the past 3 months.

Figure 2.14: Females by other contraceptive method(s) used during last intercourse, among those who used condom during last intercourse with a male in the past 3 months.

Figure 2.15: Males by other contraceptive method(s) used during last intercourse, among those who used condom during last intercourse with a female in the past 3 months.

Figure 2.16: Females by other contraceptive method(s) used during last intercourse, among those who used pill during last intercourse with a male in the past 3 months.

Figure 2.17: Males by other contraceptive method(s) used during last intercourse, among those who used pill during last intercourse with a female in the past 3 months.

When used in conjunction with other methods, withdrawal, condom, and pill methods are used with a variety of other methods, though the most commonly used other methods with any one of these three are the other two methods. (Figures 2.12, 2.13, 2.14, 2.15, 2.16 and 2.17) However, this is to be expected, since these three methods are more common generally. (Section 2.2.6) One exception to this general trend is that calendar-based methods are unusually common among methods used in conjunction with withdrawal as reported by females. (Figure 2.12)

2.3.1 Number of Methods

A majority of persons sexually active in the past 12 months used at least one method of contraception at least once in the past 12 months, and a sizable minority used more than one method of contraception in the past 12 months. (Table 2.13)

The estimates of those who have not used any method of contraception in the past 12 months despite being sexually active in Table 2.13 should not be interpreted as an estimate of those at increased risk for unintended pregnancy due to not using contraception, since the estimates include sexual partners trying to conceive, partners with a pregnant or postpartum female, and partners with one or more physically infertile persons.⁶

2.3.2 Method(s) Used

Figure 2.18: Females categorized by whether or not each contraceptive method was used at all during the past 12 months, with persons using multiple methods counted multiple times.

Figure 2.19: Males categorized by whether or not each contraceptive method was used at all during the past 12 months, with persons using multiple methods counted multiple times.

The distribution of method(s) used by females in the past 12 months in Table 2.18 parallels the distribution of method(s) used during last intercourse in Figure 2.6, with the counts of persons using each method increased. The major exception is the number of females using a sterilizing operation, which is barely greater in Figure 2.18, causing it to be the third most rather than most prevalent method of contraception among females when examining the past 12 months.

Similarly, the distribution of method(s) used in the past 12 months in Table 2.19 parallels the distribution of method(s) used during last intercourse in Figure 2.7, but with counts of those using female sterilizing operations not increasing as much as other methods, causing it to decrease from the third most to the fourth most prevalent method among males when examining the past 12 months. A salient feature of Figure 2.19 is the extent to which the number of males who have used condoms at least once in the past 12 months is greater than other methods.

2.3.3 Method Used Exclusively

Figure 2.20: Females categorized by which contraceptive method was used during past 12 months, among those who used exactly one method in past 12 months.

Figure 2.21: Males categorized by which contraceptive method was used during past 12 months, among those who used exactly one method in past 12 months.

When restricting analysis to those who reported using only one method of contraception, the distributions of method used during the past 12 months (Figures 2.20 and 2.21) parallels the distributions of method used during last intercourse (Figures 2.8 and 2.9) very closely, with female sterilizing operation being the most commonly method reported by females, and condoms then female sterilizing operation being the most commonly methods reported by males.

2.4.1 Any Method During Past 12 Months

Among those who have used a method of contraception in the past 12 months, the majority of persons report using a method of contraception every time, though a substantial minority report more inconsistency in their contraceptive use. Some respondents to the NSFG report their frequency of using any method of contraception to be “none of the time.” This is contradictory with their answers to other questions that indicate that they used at least one method of contraception in the past 12 months. These self-contradictory answers represent several million persons in the population. (Table 2.17)

In the NSFG, different series of questions about consistency of contraceptive use are asked of males who are married or cohabitating than of those who are not, whereas the same series of questions pertaining to consistency are asked of all females. This is why the subdomains for the estimates in Table 2.17 differ slightly for females and for males. The estimates for females exclude all persons who have active sterilizing operations. The estimates for males exclude such persons and also exclude persons who are married or cohabitating with a female partner who has an active sterilizing operation.

In cases of males who have answers for both series of questions because they have had intercourse both with a wife or cohabitating partner and with a more recent female partner in the past 12 months, the answers pertaining to contraceptive use with the more recent partner are used for the estimates of Table 2.17.

Two methods of contraception, condoms and pills, are both popular and require user interventions either during each intercourse or daily, respectively. They are explored further in Sections 2.4.2 and 2.4.3.

2.4.2 Condom Use During Past 12 Months

Consistency of condom use is slightly worse than consistency of the use of any method of contraception described in Section 2.4.1, though a simple majority of those relying on condoms as their exclusive method of contraception report using condoms every time. (Table 2.18)

2.4.3 Pill Use During Past 4 Weeks

A majority of pill users report never missing a pill, though there is substantial amount of inconsistency of pill use among a minority of pill users. (Table 2.19)

2.5.1 Condom

The size of the subdomain of condom users who are surveyed regarding their motives for condom use by the NSFG is approximately 16.4M females and 20.0M males. (Table 2.20) However, it should be noted this subdomain specifically contains users of condoms for vaginal intercourse with the opposite sex and does not address non-vaginal or same-sex condom use.

The vast majority of persons who use condoms during vaginal intercourse are motivated by contraception at least in part, though these persons are split approximately equally between those who are exclusively motivated by contraception and those motivated by both contraception and prevention of sexually transmitted diseases. (Table 2.21)

2.5.2 Pill

The size of the subdomain of pill users who are surveyed about their motives for pill use by the NSFG is approximately 11.1M females. (Table 2.22)

Figure 2.22: Females by reason(s) mentioned for using contraceptive pills with persons mentioning multiple reasons counted multiple times, among those who used pills either this or the previous month.

The most commonly mentioned reason for use of contraceptive pills is indeed birth control. (Figure 2.22) However, a minority of pill users mention birth control exclusively as a reason for their use. About half of all pill users are motivated both by birth control and other health reasons, and a substantial minority of pill users are motivated exclusively by reasons other than contraception. (Table 2.23)

2.5.3 Hormonal IUD

The size of the subdomain of hormonal IUD users who are surveyed about their motives for hormonal IUD use by the NSFG is approximately 5.3M females. (Table 2.24)

Figure 2.23: Females by reason(s) mentioned for using hormonal IUD with persons mentioning multiple reasons counted multiple times, among those who used an hormonal IUD either this or the previous month.

The most commonly mentioned reason for use of hormonal IUD is birth control. (Figure 2.23) For approximately half of hormonal IUD users, contraception is the sole motivation for use of an hormonal IUD, but approximately a third of hormonal IUD users are motivated by other health reasons in addition to birth control, and a minority of hormonal IUD users are motivated exclusively by reasons other than contraception. (Table 2.25)

2.6.1 Among Those Who Have Ever Had Intercourse

Figure 2.25: Females categorized by whether or not each contraceptive method was ever used, with persons using multiple methods counted multiple times, among those who have ever had intercourse with someone of the opposite sex.

There are approximately 64.0M females who have ever had intercourse with someone of the opposite sex. (Table 2.27) Of these, approximately 99.1% have used a method of contraception at least once in their lives. (Table 2.28) The major difference between the totals estimated in Figure 2.25 and Figure 2.24 is that the estimates in 2.25 are slightly lower for some methods, such as birth control pills, owing to persons who have never had sexual intercourse with someone of the opposite sex using these methods for reasons other than contraception.

2.7.1 Condom

Figure 2.28: Females categorized by reason for dissatisfaction with condoms, with persons mentioning multiple reasons counted multiple times, among those who have ever discontinued using condoms due to dissatisfaction.

Figure 2.29: Females categorized by reason described for dissatisfaction with condoms, with persons mentioning multiple reasons counted multiple times, among those who have ever discontinued using condoms due to dissatisfaction for reasons “too difficult to use,” “side effects,” or “other.”

When condoms are discontinued due to dissatisfaction, the most common reasons are due to decrease in sexual pleasure or dislike from male partners, though a nontrivial number of persons report experiencing side effects. (Figure 2.28) The most common side effects are allergy or discomfort. (Figure 2.29)

2.7.2 Pill

Figure 2.30: Females categorized by reason for dissatisfaction with contraceptive pills, with persons mentioning multiple reasons counted multiple times, among those who have ever discontinued using pills due to dissatisfaction.

Figure 2.31: Females categorized by reason described for dissatisfaction with condoms, with persons mentioning multiple reasons counted multiple times, among those who have ever discontinued using condoms due to dissatisfaction for reasons “too difficult to use,” “side effects,” or “other.”

Side effects are by far the most commonly reported reason for discontinuation of contraceptive pills. (Figure 2.30) Side effects are varied, but mood, weight gain, or nausea issues are commonly reported. (Figure 2.31) It is also commonly reported that pills are too difficult to use or that persons cannot remember to take pills regularly.

2.7.3 IUD

Figure 2.32: Females categorized by reason for dissatisfaction with IUDs, with persons mentioning multiple reasons counted multiple times, among those who have ever discontinued using an IUD due to dissatisfaction.

Figure 2.33: Females categorized by reason described for dissatisfaction with IUDs, with persons mentioning multiple reasons counted multiple times, among those who have ever discontinued using a IUD due to dissatisfaction for reasons “too difficult to use,” “side effects,” or “other.”

Like with contraceptive pills, side effects are by far the most commonly reported reason for discontinuation of IUDs. (Figure 2.32) For IUDs, commonly reported side effects involve abdominal pain or bleeding. (Figure 2.33)

2.8.1 Intent to Conceive

Among those in this subdomain, a substantial minority do not have any intent to become pregnant and so are at risk for unintended pregnancy. (Table 2.29)

2.8.2 Reason(s) for Nonuse When Not Intending to Conceive

Figure 2.34: Females categorized by reason(s) for not using contraception, with persons reporting multiple reasons counted multiple times, among those at increased risk for unintended pregnancy due to contraceptive nonuse.

A nontrivial number of up to half a million persons are falsely included in this subdomain, because some respondents to the NSFG spontaneously respond that they actually are using a method of contraception when asked why they are not. (Figure 2.34) This is evidence of an opposite type of measurement error to that found in Section 2.4.1.

Because there are a relatively few number of persons estimated in Figure 2.34, the effective sample size is small, leading to large sampling errors. Nonetheless, a person’s belief she cannot get pregnant is either the most common or one of the most common reasons for not using contraception. This itself is another form of measurement error because respondents are elsewhere asked whether they are physically unable to conceive, and if so, are not asked this series of questions.

2.8.3 Reason(s) Respondent Believes She Cannot Become Pregnant

Figure 2.35: Females categorized by reason(s) they believe they cannot become pregnant, with persons using reporting multiple reasons counted multiple times, among those who select “you do not think you can get pregnant” as a reason for contraceptive nonuse.

Among the approximately million females who select “you do not think you can get pregnant” as a reason for not using contraception, there is no clearly most common reason for the belief that becoming pregnant is impossible. (Figure 2.35) Some answers, such as having prior unprotected sexual intercourse without a pregnancy, are likely spurious reasons to believe pregnancy is impossible, while others, such as medical diagnosis of a reproductive condition, should have led to respondents reporting their physical infertility earlier in the NSFG questioning.

3.1.1 Ignorance of Partners’ Contraceptive Use

Few males report ignorance of their partners’ contraceptive use in the NSFG. (Section 2.2.2.1) However, direct questions about this are only asked of some male respondents to the NSFG, and thus some respondents may not think to give such an answer, even if such an answer would be the most accurate description. Further, some males may not know that they do not know that their partners are using contraception. This is particularly the case for methods of contraception – such as tubal ligation, IUDs, injectables, implants, etc. – that are largely not visible to male sexual partners. The effect of this would be for males to under-report contraception use, which might explain some of the variation in both the lower contraception use rate reported by males and the lower prevalence of female sterilizing operations reported by males.

3.1.2 Younger Female Partners of Males and Older Male Partners of Females

As part of disclosure risk reduction efforts, the 2017-2019 edition of the NSFG (as well as the 2015-2017 edition) does not release age of most recent sexual partners in its public use data files. However, prior research based on previous editions of the NSFG has found that male sexual partners of females are more likely to be older, and female sexual partners of males are more likely to be younger. (Darroch, Landry, and Oslak 1999) The effect of the age asymmetry of sexual partners is that, while the age ranges of female respondents and of male respondents surveyed by the NSFG are both 15-49 years, the average age of the partners of male respondents is likely younger than the average age of the female respondents.

As seen in Section 2.9, the probability that a female has a sterilizing operation apparently increases with increasing age. Thus, the effectively older sample of females in the female respondents sample are more likely to have had sterilizing operations than the effectively younger partners of respondents in the male respondents sample. This might explain some of the difference in reported prevalence of female sterilizing operations between females and males.

3.1.3 Oversampling of Those with Multiple Recent Partners

The last sexual partner of the opposite sex within the past 3 months for one person might also be the last sexual partner of the opposite sex within the past 3 months for another person. Thus, the behavior of persons who have many recent sexual partners are effectively over-sampled in any analysis examining behavior during most recent sexual intercourse.

If the contraceptive use of those who have multiple recent sexual partners differs from the more general population and between females and males, this might explain some of the reported differences in contraceptive use between females and males. For instance, if males, but not females, who have multiple recent partners are more likely to use contraception than the general population, this would cause the contraception use rate reported by females to be greater for females than that reported by males. Similarly, if females who have multiple recent partners are less likely to have had sterilizing operations, this would cause the prevalence of female sterilizing operations reported by males to be lesser than that reported by females.

6.11.1 Among Those Who Have Ever Had Intercourse