Men's involvement in decisions about sex, contraception and childrearing strongly influences sexual and contraceptive behavior,¹ significantly strengthens and reduces discord in relationships,² and reinforces a man's responsibility for the children he fathers.³ Few studies, though, have investigated men's perceptions of their roles and responsibilities regarding decisions about sex, contraception and the raising of children. Furthermore, only recently has such research been identified as being important. High levels of nonmarital childbearing, growing concern about the spread of AIDS and other sexually transmitted diseases (STDs) and the concomitant increase in the prophylactic use of condoms has led developers of social policy to include men in efforts to prevent pregnancy and STDs. However, most investigations of men's perceptions about their roles and responsibilities have targeted adolescents and other groups of young, unmarried males. Thus, we have little understanding of how married or older men perceive their roles in these decisions.

Current trends in contraceptive method choice suggest that male-controlled methods are increasingly popular. Indeed, the recent rise in contraceptive use among young, unmarried couples is due almost entirely to an increase in the rate of condom use.⁴ By 1988, about one-third of married couples were using male methods of contraception, including sterilization,⁵ while in a 1991 study, 39% of single men aged 20-39 reported using a condom in the four weeks prior to being interviewed.⁶ Clearly, men have an important role in decision-making regarding contraception and family planning.

Research indicates that there has been an increase in the extent to which family planning is considered a joint responsibility. In a study conducted during the 1970s,⁷ only about one-third of adolescent males thought that men and women should be equally responsible for contraception. However, by the late 1980s, more than two-thirds of young men endorsed this belief.⁸ Moreover, in the later study, substantially more males thought contraception was solely a male responsibility than thought it a female responsibility. While several other studies have shown that adult males tend to view contraception as a shared decision,⁹ only one examined factors that predict such a view: Married men who were older and those who held more egalitarian attitudes were more likely to think that men and women have a shared responsibility for contraception.¹⁰

There is currently renewed interest in the role of the father in family life, and this is especially so for men raising children outside the context of a marital relationship.¹¹ Men's attitudes toward parenting responsibilities have a direct bearing on contraceptive behavior:¹² A man is less likely to take responsibility for effective contraception if he lacks a sense of obligation for the children that may result from his sexual behavior.¹³ While a very high proportion of adolescent males think that men and women have equal responsibility for the children they have together,¹⁴ there is growing evidence that the fathers of infants born to adolescent mothers are likely to be adults.¹⁵

In this study, we use data from the 1991 National Survey of Men (NSM) to examine men's perceptions about their role in a couple's decision-making about sex and contraception, as well as their beliefs about the relative responsibility of men and women for the children that they have together. We explore how a man's individual characteristics may affect his perceptions and beliefs, and identify those groups of men who are likely to feel that they have roles and responsibilities that are greater than, less than or equal to those of women. Examining the effects of couple characteristics on men's perceptions helps us to understand how men's views are shaped by both the nature of the relationship in which such decisions are made (e.g., marital or cohabiting) and the characteristics of their partner in that relationship.

Data and Methods

Sample

The 1991 NSM is a nationally representative household survey of men aged 20-39 living in the coterminous United States. The survey was based on a stratified and clustered area probability sample design.* Black households were oversampled to ensure adequate representation. The sampling frame contained 17,650 housing units, of which 93% were successfully screened for eligibility. A total of 3,321 in-person interviews were completed (70% of the eligible men). The sample was weighted on the basis of population statistics to account for stratification, clustering and disproportionate sampling, as well as for differential nonresponse.

Since our goal is to examine men's perceptions regarding a man's roles within a sexual relationship, the analyses are restricted to the 2,526 respondents who were in a heterosexual relationship at the time they were interviewed; furthermore, only these men were asked to provide detailed information about their partner's characteristics. Thus, the analyses that follow are based on a sample of 958 black and 1,568 white men.

Measurement

One purpose of the NSM was to develop an understanding of factors influencing a man's decisions about sex, contraception (particularly the use of condoms) and fertility. Thus, men were asked a series of questions regarding their perceptions about both men's and women's roles in these decisions. The analyses presented in this article are based on responses to the following five statements: It is generally the man who decides whether or not the couple will have sex; it is generally the woman who decides whether or not the couple will have sex; it is a woman's responsibility to make decisions about using birth control; it is a man's responsibility to make decisions about using birth control; and men have the same responsibilities as women for the children they father.

When presented with these statements, the respondents were handed a card that displayed a five-point scale (with one representing "strongly disagree," three representing "neutral" and five representing "strongly agree") to indicate their level of agreement with the statements. We cross-tabulated responses to the first two statements to create a combined, three-category measure of perceptions about whether decisions about sex were male-oriented, egalitarian, or female-oriented. Men who indicated a higher level of agreement with the male-focused statement than with the female-focused statement were considered to have a male-oriented perception. Men who indicated a higher level of agreement on the female-focused statement were considered to have a female-oriented perception. Those who registered equal levels of agreement on both items were considered to have an egalitarian orientation. Thus, a respondent who disagreed with the statement that it is generally the man who decides when a couple has sex, and also disagreed that it is generally the woman who decides, was considered to have an egalitarian orientation on the contraceptive measure. A similar procedure was used with the third and fourth statements to assess perceptions about contraceptive decision-making.

The item capturing beliefs about responsibility for the children that men and women have together is not based on a combination of two separate questions. Thus, it is not exactly comparable to the combined measures used to examine the other dimensions. Moreover, a very high proportion of men (87%) strongly agreed with the statement that men and women have the same responsibilities for their children. Thus, in the multivariate analyses, this item was collapsed into a dichotomous outcome variable (strongly agree vs. not strongly agree).

Statistical Approach

A multinomial logit regression approach was used to analyze the two combined measures.¹⁶ This procedure provides the likelihood of being in any one category of the dependent variable for respondents in a given category of the predictor variable, relative to respondents in the reference category of that predictor. The regression models included the demographic and socioeconomic characteristics of both the men and their partners that were considered likely to influence a man's perceptions regarding sex and contraceptive decision-making. The age and education of the man and his partner, although shown as discrete categories, were included in the mulitvariate analyses as continuous variables. Because racial homogamy is so extensive in our sample, partner's race (black vs. white) could not be entered into the models that included the man's own race. Variables capturing couple homogamy with respect to ethnicity, education and religion were also tested for inclusion, but as none were found to be statistically significant, they were not included in the final models.

Since it is difficult to interpret the coefficients from multinomial logit models, we used the estimated coefficients to calculate standardized probabilities that men would fall into each of the three cells of either of the composite measures. Thus, for example, to examine the effect of race on perceptions of whether the man or the woman makes the decision to have sex, we show the probabilities that black men and men of other races would fall into each category of the composite measure if they did not differ with respect to the other characteristics in the model. In the analyses, we accomplished this by setting the other characteristics in the model to those of a "standard population."^†

The probabilities reported in this article are therefore not the average probabilities exhibited by a group. Rather, they demonstrate how a factor such as race affects the relative scoring of the two measures when the other factors in the model are statistically controlled. For the nominal variables included in the analyses, standardized probabilities were calculated for each category of the variable. For the two age variables (man's age and partner's age), probabilities were calculated for ages 20, 30 and 40. Education variables for both the man and his partner were calculated for eight, 12, and 16 years of education.

We used a binomial logit regression to analyze the dichotomous statement regarding responsibility for children. Again, to simplify the analysis of the effects of the covariates in the model, we used the estimated coefficients to calculate the probability of strongly agreeing with the statement. These calculated probabilities were also standardized so that the independent effect of each covariate is shown.

Results

The characteristics of the men and their partners are shown in Table 1. Eighty-eight percent of the men in the sample were white and 12% were black; 8% were of Hispanic origin. Thirty percent of the men were 35 and older, and 19% were in the 20-24 year age-group. Two-thirds of the men were married and living with their spouse, while 11% were cohabiting and 23% had a regular partner. Only 11% of the men in the sample had not completed high school, and 22% had completed college. One-third of the men were Catholic, 36% were nonconservative Protestant and 16% were conservative Protestant. Partners were somewhat younger than the men; 21% were 35 and older, and 25% were younger than 25. However, partners were more likely to have had a previous marriage: Twenty-one percent of the partners had been previously married, compared to only 9% of the male respondents.

Only 9% of men registered stronger agreement with the statement that it is generally the man who decides whether the couple will have sex than with the statement that it is generally the woman who decides this. In contrast, 30% reported stronger agreement with the female orientation than with the male orientation. Sixty-one percent of men registered equal levels of agreement with both statements.

Fifteen percent of men registered stronger agreement with the statement that it is a man's responsibility to make decisions about contraception than with the statement that it is a woman's responsibility to make these decisions. This is significantly higher than the 7% who indicated greater agreement with the statement that it is a woman's responsibility, and is consistent with recent research examining the perceptions of male adolescents.¹⁷ Seventy-eight percent of respondents reported an egalitarian orientation on this measure.

Shown below are the weighted percentage distributions of responses to the two composite measures of men's perceptions:

A very high proportion of men (87%) strongly agreed with the statement that men have the same responsibility as women for the children they father, a finding that is also consistent with prior research with adolescents.¹⁸ An additional 8% of men indicated that they somewhat agreed with the statement. In contrast, 5% of men disagreed with the statement or were neutral (not shown).

Decisions About Sex

Table 2 (page 224) presents the standardized probabilities derived from the multinomial logit analysis of the composite measure on decisions about sex. Men's race, age and prior marital history had no significant impact on the relative scoring of the male and female orientations. Hispanic origin, in contrast, had a large impact. Specifically, Hispanic men were substantially more likely than non-Hispanic men to have a male-dominant scoring pattern (.20 vs. .10). They were also less likely than other men to endorse a female-dominant scoring pattern (probabilities of .20 and .36, respectively).

Cohabiting males were more likely than either married men or single men with a regular partner to have a female-dominant scoring pattern (.43 vs. .36 and .35, respectively) and were less than one-half as likely to exhibit a male-dominant scoring pattern. Education was negatively related to the likelihood of scoring the two orientations equally (.61 for men with eight years of education compared with .48 for men with 16 years of education) and was positively related to the likelihood of scoring the male orientation higher: The probability that men with 16 years of education indicated greater agreement with the male orientation was twice that of the probability among those with only eight years of education (.14 vs. .06). Nonetheless, for all levels of education, men with nonegalitarian perceptions were more likely to endorse a female than a male orientation.

Among religious subgroups, Conservative Protestants had the highest probability of scoring both orientations equally (.66) and the lowest probability of having a female-dominant scoring pattern (.24). For other Protestants, this pattern was reversed: These men had a probability of .55 of scoring both orientations equally, compared to a probability of .36 of having a female-dominant response pattern. Catholics were the least likely to have a male-dominant scoring pattern (.08), and those men whose religion was categorized as "other or none" were the most likely to have this pattern (.12).

Whether a man's partner was Hispanic had no significant impact on his scoring patterns. However, partner's previous marriage did influence the pattern of scores. Men with a previously married partner were less likely than men with a never-married partner to score both orientations equally (.49 compared with .55) and were more likely to have a female-dominant scoring pattern (.43 compared with .36). Men with highly educated partners were more likely than those with less educated partners to score the measures equally and less likely to exhibit a male-dominant scoring pattern. The effects of partner's religion were not statistically significant.

Decisions About Contraception

Table 2 also presents results of the analysis of the composite contraceptive responsibility measure. Black men were significantly more likely than white men to have a female-dominant scoring pattern (.16 vs. .06) and were less likely than white men to have a male-dominant scoring pattern (.11 vs. .19). Hispanic origin, in contrast, was associated with an elevated probability of egalitarian scoring and a reduced likelihood of either a male-dominant or female-dominant scoring pattern.

Older age was associated with a less egalitarian scoring pattern: The probability of scoring the two measures equally was .80 at age 20 compared with .67 at age 40. This was due primarily to an increase in the likelihood of female-dominant scoring among older men. A prior marriage was associated with an increased likelihood of having an egalitarian scoring pattern and with a decreased likelihood of a male-dominant scoring pattern. Currently married and cohabiting men were more than twice as likely as unmarried, noncohabiting men to have a female-dominant scoring pattern.

Education was positively related to the likelihood of a male-dominant scoring pattern and was negatively related to the likelihood of a female-dominant scoring pattern. For example, men with 16 years of education were much less likely than men with eight years of education to have a female-dominant scoring pattern (.02 vs. .17). Additionally, men in the category of "other or no religion" had the lowest probability of a male-dominant scoring pattern.

Having a partner of Hispanic origin significantly increased the likelihood of a female-dominant scoring pattern, while having an older partner decreased the likelihood of a female-dominant scoring pattern: Thus, a man with a 40-year-old partner was only about one-fourth as likely as a man with a 20-year-old partner (.03 compared with .12) to display a female-dominant scoring pattern. Partner's age was also positively related to egalitarian scoring. Partner's education, in contrast, was positively associated with a female-dominant scoring pattern among respondents and negatively associated with a male-dominant scoring pattern.

Finally, men with Catholic partners had the highest probability of an egalitarian scoring pattern (.78), and those who did not know their partner's religion had the lowest probability of such a pattern (.62). These men also had the lowest and highest probabilities (.16 and .28, respectively) of exhibiting a male-dominant scoring pattern.

Responsibilities for Children

Table 3 presents the results of the binomial logit analysis of men's beliefs regarding responsibility for the children they father. Hispanic origin was significantly and positively related to the belief that both sexes have an equal responsibility for their children (p<.01). Men with Hispanic partners, however, had a lower probability of strongly agreeing with the statement about equal responsibility than those whose partners were not Hispanic (.73 compared with .87, p<.01).

Men who were previously married were more likely than other men to strongly agree that both sexes have equal responsibility for their children (.94 vs. .87, p<.05). In contrast, men with previously married partners were less likely to have a strong level of agreement (.82 vs. .87, p<.05). No other characteristic of either the man or his partner had a significant impact on this belief.

Discussion

Most men perceive a couple's decision-making regarding sexual behavior and contraception as an egalitarian process. Sixty-one percent of men currently in a heterosexual relationship view decisions about sex as a shared responsibility and 78% view decisions about contraception in this way. Moreover, men are highly likely to perceive that the responsibility for children is a shared effort: Nearly 90% of men strongly endorse such a belief.

Among men who are not egalitarian in their views, decisions about sex are likely to be perceived as a woman's domain, whereas decisions about contraception are likely to be perceived as a man's responsibility. Men with nonegalitarian perceptions are three times as likely to have a female-dominant orientation towards sexual decisions as to express a male-dominant one, but they are twice as likely to register a male-dominant orientation toward contraceptive responsibility as to have a female-dominant view.

Race, while unrelated to the perception of either male dominance or female dominance in the sexual decision-making process, is significantly related to perceptions of relative responsibility for contraception. Black men are more likely than men of other races to view the decision to practice contraception as a woman's responsibility and less likely to view it as a man's responsibility. In comparison, men of Hispanic origin are more likely than non-Hispanics to perceive men as dominant in sexual decision-making and are also more likely than non-Hispanics to indicate that men and women have an equal responsibility regarding contraception. Being black has no significant effect on the level of agreement that both sexes share responsibilities for their children, whereas Hispanic origin is related to stronger agreement in this area. Having a Hispanic partner has no impact on a man's perception of who makes decisions about sex, but it is associated with a perception that women bear a greater responsibility for the decision to use contraceptives and with lower levels of agreement that men and women have the same responsibilities for their children.

Age is unrelated to perceptions of male or female dominance in sexual decision-making. However, older men are more likely than younger men to view women as governing contraceptive decision-making. Men with older partners, in contrast, are less likely than those with younger partners to view women as controlling these decisions. This may reflect a shift by women, as they age, away from the use of oral contraceptives and toward either coitus-dependent methods or male sterilization.¹⁹

A man who has been previously married is more likely than other men to have egalitarian views about the responsibilities of parenthood. However, if a man's partner has been previously married, he is less likely to hold these views. This may reflect perceived differences between men and women in the kinds of experiences they have in dealing with former spouses who are the parents of their children or differences in their expectations about these experiences. Such expectations may be more salient for men who have already experienced a marital dissolution. Previously married men are also more likely to feel that there is joint responsibility in contraceptive decision-making, a relationship that may reflect prior cooperative involvement in such decisions.

Cohabiting men are less likely than their married or noncohabiting peers to view either men or women as primarily responsible for sexual decision-making. This is consistent with research indicating that those in cohabiting relationships have a less traditional sexual ideology, and that cohabiting women initiate sex more often than women in marital relationships.²⁰ Cohabiting men are also most likely to indicate perceived gender equality in the responsibility for contraceptive use. Unmarried, noncohabiting men, in contrast, are more likely than men in coresidential unions to indicate male dominance in contraceptive decision-making, a pattern that may reflect the greater use of condoms for disease prevention among such men.

A man's educational attainment is positively associated with his perceptions of dominance in decisions regarding both sex and contraception. Men whose partners are highly educated, however, are more likely to perceive that decisions about sex are egalitarian, and they are also more likely to perceive that women have greater responsibility in contraceptive decision-making. These findings are consistent with a relative power hypothesis that suggests that the higher the status of the man, the more likely he is to view himself as the dominant decision-maker, while the higher the status of his partner, the more likely he is to adopt a view of her as either an equal or as the dominant decision-maker.²¹

Conservative Protestants are the most likely to perceive men and women as egalitarian and the least likely to adopt a female-oriented view concerning whether a couple will have sex. This is consistent with a conservative view of gender roles and of the family, a view that increasingly accepts sexuality as a positive, mutual aspect of a marital relationship, yet still tends to favor patriarchal authority.²² That men who are affiliated with a Christian denomination are more likely than non-Christians and those with no religious affiliation to adopt a male-dominant orientation toward contraceptive decisions may derive from proscriptions against abortion that lead such men to take greater responsibility over contraceptive decisions, to insure that an unintended pregnancy does not occur.²³

Men who do not know their partner's religious affiliation are very likely to perceive that contraceptive use is a male responsibility. Not knowing the religious affiliation of one's partner may be an indicator of poor communication in the relationship, which also reduces the likelihood of joint decision-making.

Several issues should be kept in mind when interpreting the results presented here. The data are based on perceptions about the behaviors and responsibilities of men and women in general; men's responses therefore reflect ideology more than actual behaviors or the true division of responsibilities in their own relationships. Thus, while about 30% of the men in our sample indicated that the woman generally decides whether or not a couple will have sex, it cannot be assumed that the partners of these men actually exert greater decision-making power regarding sex. Similarly, although a very high proportion of men indicated that men and women have equal responsibility for decisions about contraceptive use, it seems unrealistic to assume that they are all involved equally with their partners in those decisions. Yet, it is likely that such attitudes and perceptions are strongly influenced by an individual's own behavior.

Personal attitudes and perceptions shape sexual and contraceptive decisions. Numerous studies have shown that a woman's partner has a major effect on her sexual, contraceptive and fertility behavior.²⁴ Yet partners may have appreciable differences in their sexual values, and more importantly, one partner's perception of the other's values may be inaccurate.²⁵ Effective contraceptive behavior may depend on joint decision-making to minimize the consequences of such misperceptions.²⁶

The impact of such misperceptions extends beyond their implications for unintended pregnancy to other issues of reproductive health, including the risk of STD and HIV infection. The importance of partner influence underscores the need to include men in interventions to reduce unintended pregnancies and STDs. Yet the prevailing policy and program emphasis on women as the key figures in these decisions often unjustly and unwisely excludes men.

The results reported in this article add to our knowledge about how men perceive their role in decisions about sex and contraception, as well as how they view their parental responsibilities. They also show how men's perceptions and views are shaped by their own characteristics, the characteristics of their partner and the nature of their relationship. Despite the limitations discussed above, the information provided here is useful for understanding the sexual and contraceptive behavior of men, and instrumental for efforts to increase their participation in family planning and reproductive health decisions.