+Bioline International Official Site (site up-dated regularly)

African Population Studies/Etude de la Population Africaine, Vol. 11, October/octobre 1996

EXPLORING INTER-SPOUSE RESPONSE CONSISTENCY IN A DEMOGRAPHIC SURVEY OF THE WESTERN AREA OF SIERRA LEONE

Several studies have drawn attention to the importance of consistency in couples' responses to question related to the attitudes and behaviours of their spouses. Even among educated couples in developing societies, there are considerable response inconsistencies. This implies minimal discussion between spouses about apparently sensitive issues. Where spouses hardly discuss their fertility and sexuality, there is strong likelihood of the occurrence of differences in their attitudes and subsequent behaviours. National efforts aimed at addressing family planning, women's empowerment and children's welfare may be frustrated where husbands maintain a culture of superiority over their wives. The results of this study reveal a promising situation. But inter-spouse discussion is not universal; and, though the level of response consistency is encouraging, there is a substantial response inconsistency. There is a need to improve the environment for uninhibited discussions between spouses if the negative implications of response inconsistency are to be successfully dealt with.

Plusieurs études ont attiré l'attention sur l'importance de la cohérence des réponses données par les couples aux questions relatives aux attitudes et comportements de leurs conjoints. Même chez les couples instruits dans les pays en développement, il existe un nombre considérable d'incohérences dans les réponses qui sont données. Ceci implique qu'une discussion minimale est nécessaire entre les conjoints sur des questions apparemment sensibles. Là où les conjoints discutent à peine de leur fécondité et sexualité, il est fort probable que des différences apparaissent dans leurs attitudes et leurs comportements futurs. Les efforts nationaux destinés à aborder la planification familiale, la responsabilisation des femmes et le bien-être des enfants peuvent être frustrés là où les maris maintiennent une culture de supériorité à l'égard de leurs femmes. Les résultats de cette étude révèlent une situation prometteuse. Mais la discussion entre les conjoints n'est pas universellement pratiquée; et, quoique le niveau de la cohérence des réponses soit encourageant, il y a une large part d'incohérence dans les réponses. Il y a donc lieu d'améliorer ce milieu de manière à le rendre propice à des discussions à coeur ouvert entre les conjoints si les conséquences négatives des incohérences dans les réponses doivent être abordées avec succès.

Several studies have shown that husbands and wives may have difficulty in correctly perceiving their spouses' desired family size (Williams and Thomas, 1985). In developing countries, and particularly in Africa, the frequency of inter-spouse discussions was very limited prior to the 1980s; and this is primarily due to the subservient status of the wife vis-a-vis the husband (Molnos, 1968; Getaweh, 1978; Ware, 1981). Though the socio-economic status of women has improved, statements by a spouse about the partner's attitude and behaviour should be treated with caution. One approach toward attaining reliable response about a spouse is to interview the couple. This has the advantage of obtaining information on the level of agreement about their attitudes and behaviours as well as improving the quality of information reported (Coombs and Fernandez, 1978; Coombs and Chang, 1981; Koenig et al, 1984; Beaujot, 1988). It also helps answer the question about which gender of the population provides more reliable responses. It has been suggested that responses from men are more reliable than those from women where fertility related issues are concerned (Yaukey et al, 1965; Streatfield and Lucas, 1985; Ahmed et al, 1987). In some developing countries the indications are that where preferences exist, the husband's opinion was usually adopted (Mitchell, 1972).

The importance of maximizing response reliability is particularly relevant to policies which deal with the empowerment of women as well as enhancement of children's quality of life. Substantial inconsistency in a couple's responses reflect poor interspouse discussions; and this implies husbands' intolerance of their wives' opinions. This has the effect of reducing progress toward the attainment of reasonable family sizes in developing societies.

The Western Area is the most urbanized district in Sierra Leone, with literacy rate exceeding 70.0%. It consists of Greater Freetown (the city) and the Western Rural Area. The city has 87.3% of the Western Area's population; and it is the administrative, economic and political centre of the country. Though Sierra Leone had a fairly good economy when it gained independence in 1961, it was economically bankrupt by 1981 (World Bank, 1985). Stresses related to the rapid social and economic and political decline in the country are currently manifest in a civil war within the country.

This paper was prepared from data collected in a demographic sample survey on family size preferences among ever married men (and women) in the Western Area of Sierra Leone in 1986-87. In this survey (which was conducted by the first author) 539 couples were enumerated. Since the target population was primarily male, the sample size of men was 3,006. The sample of wives was selected systematically so that one in every seventh spouse of men who were enumerated was also interviewed. The questionnaire was designed to cover men and women.

A common problem related to the protection of subjective efficacy is the presence of third parties during the interview. Though Anderson and Silver (1987) observed no significant contamination of Soviet emigrants' responses by the presence of others during interviews, our respondents were isolated during interviews. Moreover, the spouses were interviewed simultaneously in separate parts of the house or compound. In cases where there was only one enumerator, some men asked to be interviewed first so that they could get away from the house when the wife was being interviewed. This helped to reassure the wife of the freedom to express her views. Where spouses in polygynous unions are interviewed, the selection of one wife could be difficult. But though 64.7% of the total sample population in the Western Area were muslims, there were only 15.0% cases of polygynous unions. To avoid complications, it was decided that where polygynous couples fell within the sample of couples selected, they should be replaced by the next household with a monogamous couple. Only two such replacements occurred.

Out of 539 couples interviewed, the analysis covers 528 couples where the husbands were heads of households. This is due to difficulties encountered when attempting to match the spouses of men who were not heads of households. The tables include frequencies for men only because the frequency of agreement and disagreement sum up to the same value. For ease of reference, positive and negative signs were assigned to the frequencies corresponding to disagreement (ie inconsistency). A positive sign (+) indicates that the disagreement came more from the husband than the wife, and a negative sign (-) indicates otherwise. For purposes of convenience, inter-spouse responses are considered here to be inconsistent if the disagreement ratio is 20.0% or more.

Subjects related to family size preferences require statements on actual numbers. It is therefore not expected that responses will be identical on singular numbers. Even where the effects of similar backgrounds are considered if assortative mating occurs, the appearance of total agreement on each family size would be more or less by chance than a reality. It was therefore deemed necessary to group family size preferences into four categories (ie 0-2, 3-4, 5-6 and 7+ children).

Table 1 shows that generally, there was some consistency in responses of couples on the desired family size (with 83.1% agreement).¹ About 59.0% of the couples desired 3-4 children while 24.0% wanted 5-6 children. Husbands were more in favour of large families (5-6 children) whereas their wives preferred moderate family sizes (3-4 children). The proportion of relative disagreement was fairly low, with 6.5% more husbands disagreeing on the desire for 3-4 children and the same percentage more wives disagreeing with their husbands over 5-6 children². This pattern of disagreement is plausible because it has been found that African men tend to want more children than their female counterparts (Belcher et al, 1978). The percentage of couples (53.0) who agreed on the ideal family size was considerably lower than that observed for the desired family size.

It should be noted that the standard deviation for the desired family size of men and women in the Western Area (3.01) was lower than that for the ideal family size (11.00). This indicates that the ideal family size is a good reflection of the subjective normative preference, which is unaffected by immediate changes in environmental conditions. Under normal circumstances, people can freely choose their family sizes. But under conditions of rapid economic decline, the tendency is to have preferences clustering around numbers which are perceived by the population to be cost effective. This explains partly the standard deviation being smaller for the desired than the ideal family size.

About one percent of the couples had similar ideal family sizes of 0-2 children, while 18.7%, 21.1% and 12.2% expressed similar ideal family sizes of 3-4, 5-6 and 7 or more children. Relatively more husbands disagreed with their wives on the ideal family sizes of 0-2 and 7+ children. Since the ideal is more representative of lifetime underlying preferences than the desired family size, the percentage agreement as well as the pattern of disagreement are revealing. For the majority of couples, the preference for 5-6 children is subject to economic and social conditions being favourable. As anticipated, relatively less husbands wanted less than five children (preferring five or more).

With respect to the decision maker on family welfare issues, there was marginal inconsistency in responses from spouses (see Table 2). The level of inter-spouse agreement was 78.0%. There was 84.3% agreement that the husband was the sole decision maker on family welfare issues in the household and 68.0% agreement that decisions were made jointly (ie by husband and wife). It was never agreed that the wife was the sole decision maker. This is a manifestation of the subservient position of women who, traditionally, are not recognized by Sierra Leonean men and the society in general as capable of playing dominant roles in family issues (Smart, 1983). Men disagreed with their spouses 16.7% more times on the issue that the husband alone makes family decisions. The corresponding percentage of disagreement from women over the spouse's response that family decisions were made jointly is 14.5.

With respect to fertility decisions, there was substantial inconsistency in the responses of husbands and wives (with 64.7% agreement). There was 39.2% more disagreement from the husband that fertility decisions were made jointly, while the wife disagreed 58.5% more times that the husband alone made such decisions.

As Table 3 indicates, there was high consistency (94.9% agreement) between spouses' responses on current use of effective contraceptives. The percentages of husband-wife agreement on inter-spouse discussions were similarly high (93.5% in each case). There was 81.9% agreement between husbands and wives that the husband actually helped in domestic (house work, etc.) activities. It would seem that the problem surrounding the disagreement is one of the respondent addressing personal issues. Since perceptions of what constitutes real assistance is highly subjective and may be determined by the relationship between couples at the time of enumeration, the dissonance is partly explained. On whether wives accompanied their husbands to social functions, there was response inconsistency (with 77.8% agreement). There was 30.4% more disagreement from the wife to the husband's affirmative response, and same percentage more from the husband to wife's response that she did not accompany him to social functions.

Table 4 reveals high consistency (92.0% agreement) between husband and wife's response on whether she ever had an induced abortion. This topic is not independent of inter-spouse discussions on sexual matters. It is noteworthy that the wife's disagreement with the husband over his statement that she ever had an abortion constitutes 66.7% of all agreement. Moreover, three (or 0.6%) of the men reported that their wives had had no abortion; but the wives disagreed. There was 55.3% agreement between spouses over the subject of the value in having many children; and this constitutes considerable inconsistency. About 20.4% more women disagreed with men that there was some benefit in having many children. This is not surprising when we recall that men expressed desires for large families much more than women did.

There was a high response inconsistency (69.3% agreement) on who made the decision about the choices of spouse (see Table 5). Indeed, 27.3% more wives disagreed with husbands that the choice of spouse was made by both of them. Meanwhile, 20.3% and 19.1% more husbands disagreed with wives that the decision was made jointly with parents or by parents alone.

Our findings indicate that in the Western Area of Sierra Leone there is considerable consistency in statements made by spouses about the attitudes and behaviours of their partners. In the light of improvements in women's responses to effective contraception to the point that fertility has declined in Kenya, Tanzania, Zimbabwe and Botswana (Blanc and Rutstein, 1994; Mturi and Hinde, 1994; Thomas and Muvandi, 1994), it may be conjectured that men have contributed significantly to these changes through a revolutionary approach to their relationships with their spouses. The indications are that men have moved away from the traditional and are currently experiencing a transition towards the western type of inter-spouse relationship (Bertrand et al, 1989; Mbizvo and Adamchak, 1991). Particularly due to labour migration to South Africa, men in Botswana have been observed to be more ready to discard traditional practices than women (Schapera and Comaroff, 1991). In southern Africa, cultural changes towards the weakening of male arrogance in his relationship with the opposite sex appears to be a response to education as well as economic growth. But in the rest of sub-Saharan Africa, the transition is due to rapid economic decline as it is to education. The transition from traditional to modern values among Sierra Leoneans owes much to their long exposure to especially formal education, given that the university was established in Greater Freetown over two centuries ago. But it took the shock of experiencing the effect of rapid decline in real income in the 1980s for an erosion of the male arrogance towards women to be manifest in families in the Western Area. It is logical to assume that a strengthening of wives' status, through the empowerment of women would accelerate the process of fertility decline in the country. And this may be deemed to be the position in other African countries with similar socioeconomic conditions.

Notwithstanding the positive aspects of inter-spouse response consistencies, the implications of the inconsistencies should not be overlooked. Indeed, recognition of these inconsistencies help in the understanding of what goes on in families which are at a state of demographic transition. The observation that disagreement between spouses occurred because husbands prefer larger families than their wives weakens the reliability of the husband alone response about the couple's ideal family size. The pattern of family size preference by men and women in Sierra Leone varies from what has been observed elsewhere. In Asia, Malaysian men prefer smaller families than the women (Coombs and Fernandez, 1978). Even in Benin City, Nigeria, Okojie (1993) found considerably lower preferences for family sizes of five or more children among men than women. These observations are plausible because though women bear most of the health risks associated with childbearing, their fertility preferences are partly explained by their perceptions of their spouses' preferences. Moreover, in Malaysia, contraceptive awareness and use are fairly high compared to the situation in West African countries. With respect to inter-spouse disagreement over fertility decisions, the position in the Western Area lends support to speculations that men cannot claim dominance over the consequences of their spouses' sexual behaviour. Indeed, this has been acknowledged about the Akans in Ghana (Bleek, 1987). From a study of men and women in Botswana, Kokwane (1993) observed that among couples in marital and consensual unions, joint fertility decisions are common; and the indications are that women have considerable power over family size decisions. In effect, in transitional societies, it cannot be established that husbands have significant power over their wives' fertility decisions.

The fact that more husbands disagreed with their wives over current non-use of effective contraceptives may be partly explained by the incidence of their ignorance over the wife's use of contraceptives. From the field, it was discovered that several wives have reportedly obtained and used contraceptives from mobile and permanent clinics (even in remote villages) without their husbands' knowledge. This demonstrates that wives can control their fertility effectively in developing societies, regardless of their spouses' wishes. It is noteworthy that the disagreement may also be due to tendentious misreporting by the wife. Having considered herself socially liberated, it may be deemed embarrassing for her to report that no discussion on sexual issues actually occurs with her spouse. The indications are that some wives are not aware of their husbands' uses of effective contraceptives. As Harrell-Bond (1975) observed (and our survey showed), currently married Sierra Leonean males frequently indulge in extra-marital relationships. This practice, which is tantamount to informal polygyny, is expected to delay the occurrence of fertility decline. However, our study revealed no significant effect of informal polygyny on men's fertility in the Western Area.

Wives' disagreement with their husbands over statements about a past occurrence of induced abortion may be explained by the possibility that frequent inter-spouse discussions may generate the climate for disharmony more than where such discussions are rare (Coombs and Fernandez, 1978). Recently, West Africans have experienced considerable improvement in the domestic, economic and political roles of women (Onuwejeogwu, 1975; Oppong, 1984; Adepoju and Oppong, 1994).

Though findings from the survey revealed that couples in the Western Area frequently discuss family and sexual topics, it is unlikely that a woman would share a history of abortion with her husband if the act occurred during her relationship with someone else. Such confessions could threaten her chances of being married to her current spouse; and it may lead to marital instability. The disagreement may also be explained by the wife's reluctance to disclose information about her abortion history to a stranger (ie the enumerator). Other researchers have found that women tend to be shy over questions related to sex related behaviour (Yaukey et al, 1965).

Should actual inconsistency be determined by the level of disagreement alone without taking individual perception of events into consideration? For instance, with respect to whether or not husbands attend social functions regularly with their wives, and whether husbands actually assist in house cleaning, cooking, childcare, etc., disagreements were expected due to subjective interpretation of the husband's relationship with his family. Our view is that such inconsistencies do not represent errors because it does not appear that tendentious misreporting would take precedence over subjective efficacy.

Some of the problems are associated with the complex mechanisms of the family and fertility decision processes in West African societies. Interventions by parents, other relatives and friends are crucial to the final decision that both spouses make. The right to individual decisions is however interfered with by factors considered normative in the society. This is especially so where parents of one (or both) of the couple hold strong traditional values. It is not uncommon to find cases in the Western Area where spouses were matched by parental manipulations. It may be difficult for one of the spouses to realize this before it is manifest in stresses within the marriage. Inconsistencies observed about spouse selection could have implications for marital destabilization; and a possible effect of this is the raising of fertility beyond the desired family size.

The results of this study reflects the attitude of a population in the demographic transition stage. Particularly due to the rapid economic decline in the country, Sierra Leonean men prefer small families now (Campbell, 1993). Meanwhile, their women have tried to regulate their fertility before the 1980s, even in the face of strong opposition from their spouses. But they have not fully received the blessing of men to implement fertility control on their own. The attitude of over 50.0% of the sample men suggests that, though men currently appreciate the need for fertility regulation in the family, they would prefer to be the decision maker on this issue. The presence of inconsistency in inter-spouse reports about discussions on sexual matters indicates men's poor acceptance of women's rights over their sexuality.

It is probable that because of men's dominance in social and economic issues, Sierra Leonean men have been living in the illusion that feminism is synonymous with weakness. However, since the late 1970s, several factors have contributed to assist Sierra Leonean women in their quest for the power to exercise their rights to independent decisions over matters concerning their reproductive capacity. Education and socioeconomic mobility have contributed partly. But the major influence is a combination of economics and politics (Smart, 1983). The quest to survive within an oppressive economy led women to shift from the formal to the informal economic sector. For many, this meant travelling to Liberia, Ivory Coast, Guinea, the Gambia and Senegal to purchase goods for sale in Sierra Leone. More enterprising women enhanced their support systems and profits through assistance from politicians. As women became more powerful economically and socially, men's reliance on them for household assistance increased.

Another strong contributor to women's expression of individuality is the devastating effect of the civil war which began in 1991. Since women and children are the major victims of wars, this has heightened the desires and skills of women toward surviving in a previously male dominated society. Notwithstanding the negative impact of a civil war, the process of survival should add to improve the self-esteem of women as they successfully elude death and, possibly, prosper. The common struggle by the sexes to survive a poor economy and civil war should place men and women at equal positions. These should contribute to a better environment in future families for uninhibited inter-spouse discussions. In effect, we predict that in the future the level of inter-spouse consistency on family and fertility issues will increase considerably as men concede to women's empowerment. The overall effect will be significant fertility decline.

Though the Western Area's population is in a transitional stage of social development, there is consistency in the responses of couples about their spouses' attitudes and behaviours. Still, the areas where inconsistency abound reveal the need for men to accept the rights of women to express their individual positions on issues related to their sexuality and fertility. The men do intend to reduce their family size significantly, given the current poor state of the national economy. Since their fertility intentions are consistent with those of women, eventual fertility decline will occur through couples sharing their opinions on equal basis. This requires that women are exposed to greater opportunities for educational and occupational mobility. This should begin at the childhood stages. Girls should not be exposed to prostitution through serving as street sellers for their parents. The school curriculum should be revised to include sexual education and boys (as well as adult males) should be made to account for any irresponsible act of sexual relationship with teenage girls. Meanwhile, girls should be encouraged to improve their self-esteem; and one factor in this is the assurance of employment and promotion on equal basis with their male counterparts.

Normally, it would be informative to examine inter-spouse agreements on the desired and ideal family sizes while controlling for the effects of age and the actual family size. The relationship between desired/ideal family size and actual family size is quite strong (Campbell, 1993; Pritchett, 1994). However, the chi square corresponding to the results of each analysis indicated no significant difference between agreement and age (as well as the actual family size). We strongly suspect that the results were biased by our controlling for the characteristics of men alone. A meaningful analysis should consider the effects of differences in the ages and fertility preferences for each couple. The effect of standardizing the data for such analysis on the eventual result may remove realism from the results. However, this aspect of the study could be considered in future research.

Table 1: Percentage distribution of husbands, by agreement with wife on family size preferences (Western area)

Number of children	Desired family size				Ideal family size
	A	D	Total	N	A	D	Total	N
0-2 3-4 5-6 7+	0.0 87.8 73.7 0.0	0.0 12.2(+) 26.3(-) 0.0	0.0 100.0 100.0 0.0	0 353 175 0	21.4 65.8 52.6 46.4	78.6(+) 34.2(+) 47.4(-) 53.6(-)	100.0 100.0 100.0 100.0	28 149 211 138
All N	439	89		528	279	247		526

Table 2: Percentage distribution of husbands, by agreement with wife about family and fertility decision (Western area)

Subject of decision	WHO MAKES DECISION								Total N*
	Husband & Wife			Wife			Husband
	A	D	N	A	D	N	A	D	A	D
Family Fertility	68.0 76.4	32.0(-) 23.6(+)	132 263	0.0 40.0	100.0(-) 60.0(+)	2 10	15.7(+) 46.5(-)	332 254	412 341	116 186

Table 3: Percentage distribution of husbands, by agreement with wife about whether or not behaviour was implemented by husband and/or wife (Western area)

Behaviour of couple

Yes

Total N*

Current use of contraceptive

Discuss family welfare matters

Discuss sexual matters

Assist wife with cooking, etc.

Attend social functions with wife

89.8

96.0

95.6

81.9

74.0

10.2(-)

4.0(-)

4.4(+)

18.1(-)

26.0(-)

100.0

147

494

295

409

265

96.8

42.4

90.9

73.9

81.7

3.2(+)

57.6(+)

9.1(-)

26.1(+)

18.3(+)

100.0

380

232

119

262

500

488

494

423

410

105

117

Table 4: Percentage distribution of husbands, by agreement wibout whether wife has had induced abortion and whether there is some benefit in having many children (Western area)

Topic

Yes

Total N*

Abortion

Have many children

93.9

55.2

6.1(-)

44.8(-)

100.0

252

94.6

61.9

5.4(+)

38.1(+)

100.0

465

218

486

274

196

Table 5: Percentage distribution of husbands, by decision maker on choice of spouse (Western area)

Spouse only

Spouse and parent

Parent only

Total

79.9

20.1(-)

384

42.0

58.0(+)

39.7

60.3(+)

366

162

¹ The desired and ideal family sizes of the sample population were obtained from a multiple question approach with application of analysis of variance and partial correlations (see Campbell [1993] for explanation).

² Percentages referring to the relative dominance of husbands or wives where disagreements are concerned cannot be read from the tables. Such percentages were computed from the separate data on husbands' and wives' responses.