Despite the fact that divorce is discussed in almost any sociology or anthropology book, no unified explanations for family instability are available. Levinson and Malone (1980: 69) state that "anthropologists have offered little in the way of trustworthy, universal explanations for divorce." The differences in cultures, traditions, and practices among societies make it almost impossible to offer general explanations for family instability. The widely held belief is that when a significant amount of money or property is exchanged, the marriage is more stable. This belief is supported by Minturn et al. (1969: 308) findings that absence of bride price, dowry, and gift exchange at marriage make divorce easier. The purpose of this paper is to investigate how the presence or absence of a marriage transaction and how marriage payments such as bride wealth and dowry affect the frequency of divorce.


Definition of Marriage

I began my study by looking for a definition of marriage. For I believe how people in these societies view marriage may give me an explanation on why divorce is practiced. Stephens (1963: 5) defined marriage as follows ". . .is a socially legitimate union, begun with a public announcement and undertaken with some idea of permanence; it is assumed with a more or less explicit marriage contract, which spells out reciprocal rights and obligations between spouses, and between the spouses and their future children." Similar to Stephens’ definition, many experts defined marriage as some sort of a permanent relationship between two people. Then, why do societies permit divorce? It can be argued that divorce serves as a way for individuals to get out of abusive and/or unhappy marriages.


Different Types of Marriage Transactions

After looking at the definition of marriage, I went on with my study by looking at the different types of marriage-transaction practices. Among them, bride price or bride wealth is one of the most frequently used methods. In other words, the most common type of marriage payment is that the husband or his family pays the bride’s family. The least common method of payments is probably dowry, which is paid to the bride from her family. Other types of practices in these societies include gift exchange, bride service, token bride price, sister or female relative exchange. Levinson and Malone (1980: 73) also point out that one other way of obtaining a spouse is known as bride theft, which is defined as ". . . the forcible abduction of a woman for the purpose of marriage, without the knowledge or consent of her parents or guardians." There are, of course, societies that do not practice any type of marriage payment.




After looking at the definition of marriage and the different types of marriage transactions, I began my research with some cross-cultural literature. According to Minturn et al. (1969: 308), a bride price often serves as a mechanism to discourage divorce. Family pressure on the woman to keep the marriage relationship intact is tremendous. For her family is reluctant to give up what they have received for her. Furthermore, her father often relies on the installments paid on her to meet payments due on her mother. In short, the literature suggests that divorce should be infrequent in marriages that involve a significant marriage payment from the groom to the bride’s family.



My hypothesis is that the presence of a marriage transaction, be it a bride price, bride service, a token bride price, gift exchange, female exchange, or a dowry, reduces the frequency of divorce. In other words, divorce is moderate or rare in societies which have marriage-transaction practices. On the other hand, where marriage-transaction practices are absent divorce is universal or common. The variables that I used to test this hypothesis include "revised marriage transactions" (v-1240) and "frequency of divorce" (v-744). The correlation between these two variables is .252 and the p value is .020 as shown in Table 1. This suggests that there is a high association between "revised marriage transaction" and "frequency of divorce."



However, as I looked more carefully at the definitions of the variables, the table is really not testing my hypothesis—a re-coding process is required. I went ahead and re-coded variable 1240 by grouping all transactions together. Thus, the new variable 1240 has two groups: transaction is present and transaction is absent. As shown in Table 1a, the correlation between the "revised marriage transactions recoded" and "frequency of divorce" is -.156 and the p value is .154. Hence, the test does not support my hypothesis.



Test of Hypothesis

I continued to test my hypothesis using SPSS to run some crosstabulations. Contradictory to my hypothesis, as seen in Table 2 (I rearranged the order and re-coded v-1240 so the table makes more



sense), divorce is common or universal in 27 out of 53 cases where a marriage transaction is involved (adding "universal or almost universal" and "common, frequent, not uncommon"). In other words, even when a marriage is created through some sort of marriage transactions, divorce is common in over 50% of the cases.


Furthermore, divorce is moderate or rare only in 21 out of 53 cases where a marriage transaction is involved. That is, the frequency of divorce is rare or moderate only at around 40%. On the other hand, divorce is moderate or rare at a higher percentage (46.9%) than divorce is universal or common (37.5%) when marriage transaction is absent. This test, all in all, does not support my hypothesis. This is expected since the correlation in Table 1a suggests that my hypothesis should not be supported by the data. Note that "divorce is frequent in 1st year of marriage and before children" is left out of this analysis since it cannot be clearly classified as the "common" or "rare" category.



As suggested by much literature, including Stephens: 1963 and Minturn et al: 1969, a significant marriage payment discourages divorce. I went on to modify my hypothesis slightly to see if the data will support it. My modified hypothesis is as follows, the presence of a bride price reduces the frequency of divorce. In other words, divorce is less common when a marriage involves a bride price. The intention of this analysis is to test if a marriage payment discourages divorce when it requires a repayment at the time a marriage terminates. Gifts, token bride price, female exchange, bride service, and dowry are left out of this analysis since these types of marriage modes can not generally be returned if a marriage is dissolved.


Variables 744 and 1240 are still used in testing this hypothesis. However, some data have been removed from Table 2 to derive Table 3 to minimize the confusion of having to look at unrelated data.


As seen in Table 3, the data still does not support my modified hypothesis. Divorce is still quite common or universal (56.5%) rather than rare or moderate (39.1%), even if the marriage transaction requires a repayment at time of divorce. Interestingly, divorce is moderate or rare at a higher percentage (46.9%) than divorce is universal or common (37.5%) when marriage transaction is absent. Again, "divorce is frequent in 1st year of marriage and before children" is left out because it cannot be clearly classified as the "common" or "uncommon" category.


This test, in short, supports neither my modified hypothesis nor the widely held assumption that when a sizable amount of money or property is exchanged at a marriage there will be great pressure brought by the spouses’ families to keep the couple together and will therefore reduce the frequency of divorce. This is an interesting finding, though not what I expected. As a result, I went on to replicate the crosstabulations with variable 200 to see in which region the bride-price practice is concentrated or it is at least practiced. As seen in Table 4, the replication table shows that the highest concentration of bride-price practice is in Africa, 26.1%. Hence, I went on to search for some explanations for the result of my test.



According to Cooper’s study on Nigeria (in Africa), the amount of a bride wealth ranges from 20 to 20,000 cowries. The consequence of the relatively low bride wealth is that women in general, and certainly in the Maradi region, enjoy a high degree of "mobility in and out of marriage," since the payment is not so high that it can not be repaid to the groom’s family. This study explains for one case why the test fails to support my hypothesis.


Another explanation offered by Tambiah is that the bride price often does not have to be repaid. Tambiah did his study on bride wealth and dowry in Sub-Saharan Africa. As claimed in his study, the repayment of the bride wealth in Africa is kept in relation to the children born of the union. In other words, the calculation of a repayment in the case of a divorce is as follows: a maximum in an infertile marriage and certain sums subtracted for each child. Hence, practically nothing being repaid if the woman has borne numerous children.


These two studies, conducted by Cooper and Tambiah, explain for two cases why my hypothesis A is not supported by the data. For the idea behind the modified hypothesis is that a bride price discourages divorce because it requires a repayment if a woman breaks out of the marriage. However, since the payment does not always have to be repaid or it is relatively small and it can easily be repaid, bride wealth, therefore, can not be claimed to stabilize marriages.



As I was searching for explanations for the result of my test, something ran across my mind, what about dowry or indirect dowry? Similar to a bride wealth, a dowry or an indirect dowry involves some kind of payment. Unlike a bride price, however, dowry is a payment to the bride from her family, not from the groom or his family. An indirect dowry is a payment made to the bride’s family from the groom, but the bride’s family then gives the payment to the bride.


According to Cooper, the payment of the dowry announces the wealth and value of the bride. I then propose my next hypothesis to be as follows: the presence of a dowry or an indirect dowry reduces the frequency of divorce. In other words, if a bride receives a payment from her family, the payment serves to stabilize the marriage. This hypothesis is supported by the data. As shown in Table 5, divorce is rare in almost 63% of the cases when a dowry or an indirect dowry is present. On the other hand, divorce is universal or common only in about 37% of the cases.


The support of the data on my hypothesis can be explained as follows: the presence of a dowry not only announces to the public the high worth of the bride, but also tells the groom’s family of her important status. Because a new bride spent much of her time "in the company" of other women in her new household, the dowry acts as a mechanism to prevent her in-laws to demand of her labor or treat her as a subordinate. This no doubt helps stabilize the marriage. For instance, the groom does not often have to deal with the conflicts or stand in between the bride and her in-laws. Hence, reducing fights or arguments between the two people created by the in-laws and allowing them to spend more time together to better understand each other.



I am convinced that my modified hypothesis is not strongly supported by the data due to the primary reason that it is too general. A look at each individual case is essential to intelligently evaluate if a bride wealth helps stabilize a marriage. For instance, if a bride price is so small that the bride or her family has no difficulties in returning the payment, bride price does not reduce the frequency of divorce. Furthermore, if there are children from the union, a repayment is often not required at the time of divorce. Therefore, bride price can not be claimed to reduce the frequency of divorce. On the other hand, the presence of a dowry or an indirect dowry reduces the frequency of divorce because it represents the independent status of the bride. Hence, helping her cope with the in-laws and giving the couple more time to spend together to allow more communication. As a result, a dowry or an indirect dowry is quite effective in reducing the frequency of divorce.




Cooper, Barbara M. Women’s Worth and Wedding Gift Exchange in Maradi, Niger, 1907-89. Journal of African History V36, N1 (Jan, 1995): 121-141.

Levinson, David and Martin J. Malone. 1980. Toward Explaining Human Culture: A Critical Review of the Findings of Worldwide Cross-cultural Research.

Minturn, Leigh, Martin Grosse and Santoah Haider. 1969. Cultural Patterning of Sexual Beliefs and Behavior. Ethnology 8:301-318.

Stephens, William N. 1963. The Family in Cross-cultural Perspective. New York: Holt, Rinehart and Winston.

Tambiah, Stanley J. Bride Wealth and Dowry Revisited. Current Anthropology V30, N4 (Aug-Oct, 1989) 413-435.