Economics of marriage payments

I grew up reading about the evils of the dowry system.  And decades of concerted government policies and NGO activism later, dowry still seems to still haunt many a marriage in Bangladesh.  In fact, according to one study, prevalence of dowry has been consistently increading in rural Bangladesh — whereas merely a tenth of families paid a dowry before the 1970s, the proportion rose to three-fifths in the 1990s to over three-quarters in 2003.

Is there an economics literature on dowry, or marriage payments more generally?  Siwan Anderson, a Canadian economist, has done much work on this field.  Drawing heavily on her work, over the fold is my understanding of the literature.

There are two kinds of marriage payments — dowry and brideprice, with significant differences among their prevalences in time and across societies.  In general, brideprice seems to be more prevalent in relatively pre-agridicultural and nomadic societies.  In many ancient cultures, brideprice seemed to have given way to dowry as the respective civilisations grew more complex and stratified.  Brideprice has also been associated with societies where agriculture relies on light tools, which can be used by women.  In contrast, dowry is more common where agriculture is plough-based, which is hard for women to use.  Finally, brideprices are more common in polygynous societies, whereas dowry is correlated with monogamy.

These observations can be explained by a Gary Becker style framework.  Consider an agricultural or nomadic society with homogenous men and women, where women have direct input into production, and hence economic values of their own.  If these societies have polygyny, then there will be competition among men for brides.  Brideprice will occur as a ‘market clearing mechanism’.  Suppose a technological shock (such as plough-based agriculture) leads to specialisation in the production process such that (still homogenous) women are in general consigned to home while men become stratified according to their wealth.  In such a society, brides compete for more desirable grooms, and dowry replaces bridprice.

Brideprice appears to be relatively constant (after accounting for inflation) over time or across social status or wealth.  In contrast, the amount of dowry varies substantially over time and across social groups.  This observation has led to the theory of dowry as a signalling device of one’s status.  Basically, the idea is that in societies where status hierarchies (based on wealth, power, hereditary status, whatever) is important, ‘high level’ individuals typically don’t marry someone from the ‘lower level’ willingly, and dowry becomes one way to maintain/attain social status by attracting a husband of at least equal (if not higher) status for one’s daughter.

There are periods in history, from the Romans to post-Enlightenment Europe to present day South Asia, where dowries have experienced significant inflation — that is, the amounts have grown much faster than the economywide prices.  According to one theory, dowry inflation is a consequence of a modernisation process where  individuals of similar inherited status (in the traditional sector) start having differentiated income levels (in the modern sector).

Suppose there are two individuals from the same social group (castes in South Asia, for example).  Traditionally, a bride would be indifferent between the two.  But suppose the vagaries of capitalist development means now they have different income levels.  What happens now?  Low status brides with wealth can pay a premium to ‘marry up’ irrespective of the groom’s (capitalist) wealth.  Thus, high status but poorer grooms do not see a fall in their dowry.  But the high income high status groom will now ask for a rise in his dowry, and high status brides will oblige.  As a result, the average dowry payment increases.

Finally, dowry seems to cease when preservation of social status through endogamy (marrying within one’s own social group) is no longer important.  That’s what seems to have happened in Europe.  So perhaps it will also happen in South Asia one day.