Couples' Reports of Their Contraceptive Use: Do Husbands in Africa Overstate the Case?
When monogamous spouses in Kenya and Zimbabwe are asked if they are currently using a contraceptive method, the majority give the same response; when their answers differ, however, men are more likely than their wives to say that the couple use a method. The difference is particularly striking in Kenya: Only the husband reports current method use in 29% of couples, whereas only the wife reports use in 6% of couples. In Zimbabwe, these proportions are 11% and 7%, respectively. According to an analysis of Demographic and Health Survey (DHS) data, key factors in the consistency of spouses' reporting include whether they approve of and whether they discuss family planning.1
To explore patterns of contraceptive reporting, the analyst examined data on 1,055 matched married couples from the 1993 Kenya DHS who said that they were monogamous (i.e., not in a polygynous union) and 592 such couples from the 1994 Zimbabwe DHS. He conducted bivariate analyses and then two sets of multinomial logit analyses: one to assess the factors influencing the risk that only one partner (rather than both) would report that the couple used a method, and one to examine the factors affecting the risk that both partners or only the wife (rather than the husband) would report using a contraceptive.
Two-thirds of couples in Kenya gave consistent responses about their contraceptive use: In 38%, both partners said that the couple were not currently using a method, while in 28%, both said that they were. When spouses disagreed, the husband was more likely than the wife to report using a method (29% vs. 6% overall).
The overwhelming majority of Kenyan men whose wives said they had been sterilized corroborated that report (86%), but the proportion was somewhat lower among those whose wives reported using the pill (72%), IUD (65%) or injectable (61%). Only about half of men whose wives said they used periodic abstinence (52%) or no method (57%) gave the same account.* While some men reported using a different method than their wives reported, others said that the couple used no method at all--6% of those whose wives said they were sterilized; 13-16% of those whose wives said they used the pill, IUD or injectable; and 36% of those whose wives reported using periodic abstinence.
By contrast, women were most likely to agree with their husbands' account of the couple's contraceptive use if their husbands reported using no method or the IUD (87% each), sterilization (77%) or the pill (71%); agreement was moderate for those whose husbands said that the couple were protected by an injectable (64%) and lowest for women whose husbands said they used periodic abstinence (10%) or condoms (6%). Roughly three-quarters of women whose husbands reported relying on condoms (73%) or periodic abstinence (79%) said they used no method; the proportion was much smaller for the injectable (32%), pill (21%) and sterilization (16%), and was negligible for the IUD (2%).
Results of bivariate analyses suggested that partners' background characteristics, attitudes toward family planning, spontaneous knowledge of methods and discussions about contraceptive use affected patterns of spouses' reporting. Findings from the multivariate analyses illustrate the independent effects of these factors.
The first multivariate analysis showed that compared with the risk that both partners would report contraceptive use, the risk that only the husband would report using a method was doubled if the husband knew about periodic abstinence (risk ratio, 1.9). This risk was reduced, however, if either the wife alone or both partners reported that the couple discuss family planning (0.1-0.2), if both partners approved of family planning (0.6) or if the wife knew about rhythm, understood the ovulatory cycle or had at least a secondary education (0.4-0.6). It also was reduced among urban couples (0.5), and it fell by 4% for each year that the husband aged. The odds that the wife alone would report using a method were substantially elevated if another adult was present at her survey interview (3.0). They were greatly reduced if only the husband or if both partners reported having talked about family planning (0.1 for each), and they dropped as the wife's number of living children increased (0.8).
More striking results emerged when the researcher explored factors that influenced the risk that both partners or only the woman reported contraceptive use, rather than the man. Several factors--the couple's joint approval of family planning; the wife's knowledge of periodic abstinence and of the ovulatory cycle, and her having at least a secondary education; and urban residence--roughly doubled the likelihood of both spouses' saying they used a method (1.6-2.4). The greatest increase, however, occurred among couples in which either the wife alone or both partners said they had discussed family planning (8.2 and 5.6, respectively). Furthermore, couples in which the husband knew of periodic abstinence were half as likely to have both spouses report contraceptive use as to have only the man report use (0.5). The likelihood that wives, but not husbands, would report method use was elevated if the wife reported discussing family planning with her husband (3.3) and if she knew of periodic abstinence (4.3); it was reduced if only the husband reported discussing family planning (0.2).
In four-fifths of Zimbabwean couples, both spouses gave the same answer when asked about their contraceptive use: Thirty-three percent were consistent in reporting that no method was being used, and 47% that they were using a contraceptive. In 11% of couples, the man alone said that the couple were using a method, and in 7%, the woman alone gave this response.
Virtually all men whose wives said they were using the pill (91%) corroborated that report; agreement also was high for nonuse of any method (74%). However, only 23% of Zimbabwean men whose wives said they were protected by withdrawal gave the same response. Half (53%) of men whose wives said that the couple used withdrawal reported that they used no method, compared with only 6% of those whose wives said they took the pill. Zimbabwean women were highly likely to corroborate their husbands' reports of nonuse or pill use (83% for each), but 13% of those married to men who reported pill use said they used no method.
A number of factors had independent effects on the likelihood that a man would report contraceptive use but his wife would not. The risk of this disparity was dramatically reduced if both partners approved of family planning or if the wife had at least a secondary education (risk ratio, 0.2 for each); it declined by 24% for every additional living child the wife reported. For each year that husbands aged, however, the likelihood that only the male partner reported contraceptive use increased by 7%. A couple's joint approval of family planning reduced the odds that only the wife would report method use (0.2), but two other factors substantially elevated this likelihood: the presence of another adult when the wife was interviewed (6.0) and the wife's knowledge of withdrawal (9.6).
The risk that both spouses (rather than the husband alone) would report using a method was sharply elevated if both approved of family planning (4.5) and if the wife had a secondary education or more (5.9); the wife's number of living children had a more moderate but still positive effect (1.2). With each year that husbands aged, the odds that both partners would say they used a contraceptive fell by 10%. One factor significantly affected the likelihood that only wives would report contraceptive use: the wife's knowledge of withdrawal (4.2).
Summing up his findings, the analyst notes that women's reports of contraceptive use receive greater corroboration from their husbands than men's reports receive from their wives. Furthermore, most of the significant multivariate findings are associated with increased odds that only wives would report contraceptive use or decreased odds that only husbands would do so. As a result, the analyst concludes that women may be making "a conscious effort...to use contraception even if their husbands are not supportive of it," and wives' report of current use "may more correctly reflect the couple's true contraceptive use status."--D. Hollander
1. Ezeh AC, Differences in the Reporting of Contraceptive Use Among Marital Partners, African Population and Health Research Center Working Papers, Nairobi, Kenya: Population Council, 2000, No. 15.
*For both countries, method-specific data are discussed only for methods reported by at least 25 respondents.