August 2008 - Volume 2, Issue 4

Factors Affecting Age at Marriage in Two Major Divisions of Bangladesh: A Multiple Classification Analysis (MCA)

Md. Kamruzzaman1, Dr. Md. Nurul Islam1, Md. Mosiur Rahman2 & Md. Mahfuzar Rahman2 Institutions 1.Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh 2.Department of Population Science and Human Resource Development, University of Rajshahi, Rajshahi-6205, Bangladesh Md. Mosiur Rahman Department of Population Science and Human Resource Development University of Rajshahi, Rajshahi-6205, Bangladesh E-mail: swaponru_2000@yahoo.com

ABSTRACT The present study describes the factors affecting age at marriage in Bangladesh by examining the situation prevailing in two major divisions using data from the 2000 Bangladesh Demographic and Health survey (BDHS, 2000). The findings reveal that among all the variables considered, women's education is one of the most important correlates, which strongly affect for explaining the variability in age at marriage for these two major divisions. Although the average level of education is very low, education still has a strong positive relationship with age at first marriage. The husband's education has come out to be a strong determinant of age at marriage. Childhood residence is also quite important; affecting age at marriage through education and work participation, though the direct effect is not as high as it was for education Key Words: Age at marriage, Multiple Classification Analysis (MCA) and BDHS

INTRODUCTION

Bangladesh, situated in South Asia, is a unitary, independent and sovereign Republic known as the People's Republic of Bangladesh and is the seventh most populous country in the world (PRB, 2007). Ethnically, Bangladesh is homogeneous, having only one major ethnic group (98.8%) known as Bangalees. There are some ethnically different tribal populations in the hilly regions of the country constituting 1.2%. Islam is the predominant religion with 88.30%, Hinduism 10.5%, Buddhism 0.6%, Christianity 0.3% and others 0.3% (BBS, 2003).

In Bangladesh, marriage means the prescribed legal union between a man and woman, establishing them in new social roles as husband and wife. Pre-marital cohabitation does not exist and it is looked upon as a social evil. Marriage is a desirable event and a universal phenomenon in Bangladesh. Early marriage for women in Bangladesh is widespread and most of them become married before reaching the age of 20 years. In rural Bangladesh, there is a common belief that a girl aged 20 years is too old to marry. Constraints on marriage that didn't exist before are, however, now making an appreciable change - shortage of land, unemployment, etc. Also marriage among the highly educated elite is becoming less desirable as they are capable of supporting themselves free from the contractualism of arranged marriages. As early as possible a girl must go to her husband's household. For parents a post puberty unmarried daughter is considered a danger because of the secret desire to indulge in illicit coitus that might be in their daughter's mind, which may result in social scandal and for which the parents may be socially condemned and have to face rancour and social boycott. If a daughter's marriage is delayed, the parents feel guilty and begin to think of her as a burden. Parents of girls who can arrange early marriage for their daughters feel very proud. 30% of women age 10-49 have never married, while 65% are currently married. At age 10-14, the proportion never married is 93% and by age 25-29, marriage is nearly universal for females only 4% have never married. The proportion divorced, separated or deserted is small in Bangladesh and widowhood is quite limited until older ages. 12% of women age 40-44 and 16% of those 45-49 are widowed. The proportion divorced or deserted is relatively even across most age groups (to 4 percent).

Islam and Abedin (1996) show that fertility takes place in the country over time and is contributed much by the change in marital fertility, marriage pattern and increasing use of contraception. Shaikh (1997) considered the age at marriage and other aspects of nuptiality in Bangladesh in the context of opportunities to accelerate national development. It suggests that concerted efforts to increase the female age at marriage could produce a number of beneficial effects ranging from reduction in the incidence of divorce and widowhood to a lowering of fertility. Islam and Ahmed (1998) investigate marriage patterns and their determinants in Bangladesh and believed that, whatever is the impact of the differentials that could be discerned from the analysis, a lower age at marriage (a) among rural women, (b) among those who are Muslims and (c) among those without pre-marital exposure to work is associated with a low level of education. Bittles and Hussain (2000) show that the consanguineous marriage is widely favored in a large majority of the world's Islamic populations. Fowers et al. (2001) shows that the common tendency to describe one's marriage in unrealistic terms has been conceptualized as a positive illusion about marriage. Kiecolt-Glaser et al. (2003) shows that Neuroendocrine function, assessed in 90 couples during their first year of marriage (Time 1), was related to marital dissolution and satisfaction 10 years later. Compared to those who remained married, epinephrine levels of divorced couples were 34% higher during a Time 1 conflict discussion, 22% higher throughout the day, and both epinephrine and nor epinephrine were 16% higher at night.

Marriage policies can be directed to a variety of objectives, such as the social condition of unmarried women, the welfare of the unmarried elderly population, the effect of migration on the marriage market and the social and legal aspects of illegitimate children. The implications deal primarily with demographic aspects and more specifically, with the fertility implication of marriage in countries where the current level of fertility is not deemed satisfactory. This is why, the main aim and objective of this article is to identify the direct and indirect effects of socio-economic and demographic factors on marriage.

DATA AND MATERIALS

The data of the present study are taken from Bangladesh Demographic and Health Survey (BDHS) of 1999-2000. Survey collected data on marital status classifying the status as Single (never married), Married, Divorce and Widow Categories addressing female population. The 1999-2000 BDHS is a nationally representative survey of 10,544 ever-married women aged 10-49 and 2,817 currently married men aged 15-49. The marriage patterns are analyzed by means of a statistical measure of arithmetic mean. A multivariate technique-Multiple Classification Analysis (MCA) has been used for the analysis of the determinants of age at marriage (Yates, 1934 and Anderson Bancraft, 1952). In 1963, the computerized MCA program was prepared by a group of researchers at the survey Research Center of the University of Michigan. Since then, the MCA program has been widely used in social science research.

RESULTS AND DISCUSSION

The variables that are considered to look at the extent of effect on the marriage pattern of the women are education of wife, childhood residence, religion of wife, work status of woman before first marriage, husband's education and husband's occupation. In this case, age at first marriage is taken to be the dependent variable and the stated socio-economic variables as explanatory variables. Table 1 shows that the mean age at first marriage both unadjusted and adjusted by different socio-economic characteristics with the values of h2 and b2 for Dhaka and Rajshahi divisions of Bangladesh.

For Dhaka division, the proportion of variance in age at marriage explained by female education is the highest (unadjusted value of h2 =0.492 and adjusted value of b2=0.426) among all other variables while the proportion of variance explained by work status of women is the lowest (h2 =0.019 and b2=0.006). For Rajshahi division, the proportion of variance in age at marriage explained by female education is the highest (unadjusted value of h2 =0.418 and adjusted value of b2=0.372) among all other variables while the proportion of variance explained by work status of women is the lowest (h2 =0.002 and b2=0.017).

The proportion of variance in age at marriage explained by childhood residence is h2 =0.204 and b2=0.052 for Dhaka division and h2 =0.156 and b2=0.064 for Rajshahi division. Among the included variables, it is observed that the effect of childhood residential differential has been found to be the third strongest differential for explaining the variation on age at first marriage for both the divisions.

The proportion of variance explained by pre-marital work status (unadjusted) is h2 =0.019 and b2=0.006 for Dhaka division and h2 =0.002 and b2=0.017 for Rajshahi division. For instance, Dhaka division has a density population, so the land people ratio is very small. Rajshahi division is sparsely populated, but has a relatively higher labor demand than the other areas (Ahmed, 1982).

Religion of women has also shown a significant contribution on age at marriage. The strength in explaining the variability by this variable is (unadjusted) h2 =0.160 and b2=0.152 for Dhaka division and h2 =0.023 and b2=0.026 for Rajshahi division. Husband's education has a significant effect on the variation of age at marriage of women for these two divisions. The proportion of variance explained by husband's education (unadjusted) is h2 =0.361 and b2=0.073 for Dhaka division and h2 =0.295 and b2=0.082 for Rajshahi division.

Husband's occupation has also shown a positive association with age at first marriage and has a significant effect on the variation of age at first marriage for Dhaka and Rajshahi divisions. The strength of explaining variability (unadjusted) is h2 =0.174 and b2=0.007 for Dhaka division and h2 =0.120 and b2=0.026 for Rajshahi division.

Tables 2 and Table 3 corresponding to the divisions Dhaka and Rajshahi produce the results of zero order correlation coefficient of age at marriage with various socio-economic variables. Using the values of the coefficient b2 with the values of Tables 1, it is possible to posit a causal model of socio-economic structural influences on age at first marriage.

** Correlation is significant at the 0.01 level.
* Correlation is significant at the 0.05 level.

** Correlation is significant at the 0.01 level.
* Correlation is significant at the 0.05 level.

CONCLUSION AND POLICY IMPLICATIONS

Among the selected socio-economic variables as childhood residence of women, education of both wife and husband, work status of women before first marriage, religion of women and husband's occupation suggest that all the variables considered, women's education is, by far, one of the most important correlates which effects is the strongest for explaining the variability in age at marriage for these two divisions. Although the average level of education is very low, education still has a strong positive relationship with age at first marriage. The husband's education has come out to be a strong determinant of age at marriage. Childhood residence is also quite important; affecting age at marriage through education and work participation, though the direct effect is not as high as it was for education. In this regard, the following policy recommendations are to be considered:

• The early marriage of women (before age 18) should be discouraged by creating its effect on reproductive health and fertility. To do this, the law relating to it should be uniformly enforced all over the country.
  
• The government can deliberately manipulate economic incentives and disincentives to achieve the national goals in education and fertility by involving civil societies and various NGO's.
  

REFERENCES

Ahmed, A. (1982). "Marriage and Divorce in Bangladesh". Unpublished Ph. D. Dissertation Submitted to the Department of Sociology, University of Chicago.

Bangladesh Demographic and Health Survey (BDHS) (1999-2000). "National Institute of Population Research and Training (NIPORT)". Ministry of Health and Family Welfare, Dhaka, Bangladesh.

Fowers, B. J. (2001). "Positive illusions about marriage among married and single individuals". J Fam Psychol, 25(1): 195-209.

Islam, N. and S. Abedin (1996). "Some Observations on Marriage, Contraception and Fertility in Bangladesh". Genus-vol.-II n. 1-2.

Islam, M. N. and A.U. Ahmed (1998). "Age at first marriage and its differentials in Bangladesh". Asia Pacific Journal, 13(2): 73-92.

Kiecolt-Glaser et al. (2003). "Love, marriage, and divorce: newlyweds' strees hormones foreshadow relationship changes". Consult Clin Psychol, 71(1): 176-88.

Shaikh, K. (1997). "Recent changes in marriage patterns in rural Bangladesh". Asia Pacific Population Journal, 12(3): 31-48.

Population Reference Bureau (PRB) (2007). World Population Data Sheet, Washington DC.

Yates, F. (1934). The Analysis of Variance with Unequal Numbers in the Different Classes, Journal of American Statistical Association, Vol. 29, pp.51-66.