6.2  Family formation

People prefer to marry if their benefits of marriage are greater than the benefits remaining single. However, there are costs to marriage—personal freedom is restricted and future options are closed off—that must be considered. Despite this, the vast majority of adults across many societies do marry—marriage therefore must, at least ex ante, confer net benefits.

The theory of marriage associated with Becker (Becker 1973, 1974, 1985, 1991) postulates several benefits from marriage. They include specialisation and exchange, the production of household public goods, economies of scale, risk sharing (insurance).

These benefits may also be realised through other relationships such as co-habitation, sharing accommodation or through the market. However, marriage reduces the transaction costs involved in exchange, since spouses care for one another (Pollak 1985). In addition, the contractual nature of marriage provides a greater ability to monitor and enforce agreements than other, less formal arrangements (Lundberg and Pollak 1996). Formal marriage contracts also protect the specific assets of the marriage, notably the investment in children (Cohen 1987, 2002).

The family is like a firm, in which time and market goods are combined to produce services that confer utility to the members of the family (Becker 1991). These outputs include self-esteem, health, leisure, companionship and children.

Family members maximise production by specialising in duties according to their comparative advantage. When one partner has a comparative advantage in market work relative to home production, the couple can produce more total output through specialisation and exchange. Even with two intrinsically identical individuals, there are gains from household division of labour, through increasing returns to investment in specialised human capital. Typically, it is the husband who has the comparative advantage in the labour market. Even small differences in labour productivity (arising for example from gender discrimination, or when mothers have babies) can reinforce these differences. In the same way, human capital investments after marriage and investments in anticipation of household roles can reinforce the initial differences in productivity and gender specific returns.

A second source of benefits from marriage is the production of household public goods. These are goods where one spouse’s enjoyment of the good does not decrease the amount available for the other spouse to enjoy. They include goods such as a tidy house or pleasant garden, as well as other items such as children.

Marriage also confers benefits from the economies of scale in household production—the “two can live as cheaply as one” phenomenon. Joint consumption also confers benefits—where the consumption by one spouse, of say a meal or television programme, enhances the consumption of the other.

Families also fulfil the function of insurance—risk can be spread across a wider group. In traditional societies food sharing across the kin group afforded some protection against the ever-present chance of any one individual’s failure to kill or locate food. In modern societies the state has adopted many of the functions fulfilled by traditional families. Nevertheless, marriage can still be seen as a pooling of risk in the form of provision of substitute services. If one income earner fails, there is another; if one party becomes ill and cannot undertake domestic work, there is another.

From a law and economics perspective, a fundamental reason to enter into a marriage contract is to allow for optimal investment in assets of peculiar value to the relationship—assets that would be diminished in value if the relationship comes to a premature end. Insuring the investments by contracting is in the interests of both marriage partners (Cohen 1987, 2002).

The most significant marital-specific assets are children. Children are long-term investments: their costs and benefits span a lifetime. Whereas in the past children may have been something of an investment good, (in terms of their contribution to the family labour force) they are now a costly consumption good. Children are particularly valued by their natural parents, and are often seen as a cost to prospective new mates. Although parents still love their children after divorce, the extra value in terms of “being a family unit” is lost.

If both parents value children why is a contract necessary? First, the value of children can vary systematically across individuals, between sexes and over lifetimes. For example, men may find children a burden, or may find them a benefit without requiring their presence, and therefore suffer no loss from not living with them. These differences remove the reliability that self-interest of one’s mate will ensure their performance of the contract.

Furthermore, women as a rule face greater difficulties in mitigating damages by finding a replacement spouse than men do. Women generally gain custody of children and this makes searching challenging for two reasons: the presence of children makes it logistically more difficult for women to search and advertise; and children represent a liability to a potential replacement spouse.

Women also tend to be less highly valued on the remarriage market, as time goes by and men and women age. There are a number of observations that support this assertion: divorced men remarry at a faster rate than divorced women for every age but 14-24; divorced women with children remarry at a slower rate than divorced women without children. Further, higher male mortality rates raise the female/male ratio significantly. This ratio, combined with the fact that women tend to marry men who are older, and the gap increases as they age, mean women have significantly greater competition for potential spouses the second time around.

The most important reason for the marriage contract is to protect the asymmetric investments of men and women in the relationship. The contribution of women to marriage-specific assets is generally more substantial, in terms of procreation and child rearing. Women typically invest more in the early stages of a marriage when they are most fertile; men invest more in the later stages when they are likely to be in a stronger financial position. Women are therefore unwilling to undertake the investment of bearing children without some security. The long term imbalance gives rise to opportunities for a man to act strategically—to perform his obligations under the marriage contract only as long as he is receiving a net positive marginal benefit, and then breach the contract once the marginal benefit falls below his opportunity cost.

The function of the marriage contract is to constrain this sort of opportunistic behaviour.[7] Without such a contract, women would predict the imbalance, marry less frequently and invest less. A contract with penalties for breach makes investment in relationship-specific assets possible by protecting partners” investment, particularly women’s because they are more vulnerable to a breach and suffer more from divorce.

Typically then, it is women who are most anxious to obtain the contractual guarantee of marriage: they invest more in marriage specific assets, they have more to lose from divorce as they cannot mitigate damages as easily, and the temporal asymmetry in investment gives rise to opportunities for breach.

6.2.1  Mate selection

The economic theory of marriage suggests that individuals search for mates within a “marriage market” where there is a supply of and demand for, mates. Individuals have different traits and qualities. Each individual has a “value” (conferred for example by qualities such as looks, status, income, wealth, education, age or personalty) that allows them to attract a mate of equivalent “value”(Becker 1973, 1974).

An efficient marriage market is generally characterised by positive assortative mating, where people select mates in a non-random fashion—they marry people similar to and of equivalent “value” to themselves. Traits such as education, IQ, race, religion, income, family background, and height tend to be correlated. The correlation between spouses by intelligence is especially interesting, since it is as high as that between siblings (Alstrom 1961).

There are advantages to positive assortative mating. Two individuals with similar tastes are more likely to benefit from each other’s company, with lower transaction costs, fewer conflicts and more positive externalities. Furthermore, it maximises output. A superior woman raises the productivity of a superior man and vice versa. Positive qualities of spouses are related multiplicatively, not just additively, so that positive sorting promotes, for instance, children of superior intelligence or beauty (Posner 1998). Also, it reduces the benefits breaking up the marriage. "Well-paired" individuals are those least likely to be made better off by searching for another, higher value mate.

Some traits work best when complements are paired together. In Becker's model, negative assortative mating is most likely to occur around those traits or activities that are easy to substitute between one mate and the other. For example, if one mate likes cooking and the other does not, the traits are complementary.

Negative assortative mating of wage rates can sometimes be important, particularly in terms of maximising the gains from division of labour. By mating a low wage mate with a higher waged mate, the mate with cheaper time can be used more extensively in household production. But in terms of traits other than wage rates positive assortative mating is optimal in maximising aggregate output.

Marriage markets are not perfectly assortative or perfectly efficient. First, the marriage market is a barter market requiring a double coincidence of wants ie, a person needs to find someone he or she wants want who also wants him or her. Secondly, the presence of search costs means that there are trade-offs between the advantages of continued search in terms of the expected benefits from better prospects, and the costs of additional search. This explains why people might end the search and rationally “settle”.

Much search behaviour can be understood as a means to improve information about the qualities of marriage candidates prior to marriage. Quality is difficult to assess, and many traits are proxies for others (eg, education for intelligence, family reputation for character). Marriages frequently occur with erroneous assessments, which are revised early in marriage, accounting for the high frequency of divorce early in marriage (Becker 1991).

The search process involves the complex problem of signalling one’s availability appropriately. In Western cultures marriage helps individuals to signal to each other and to the outside world their desire for a sexually exclusive, permanent union. Modern trends have reduced the credibility of this signal, representing a major loss of information that makes it much more difficult to sort out the committed from the uncommitted (Rowthorn 2002).

Also complex is the challenge of moving from the status of stranger to acquaintance. Historically the process was often well known and highly ritualised. The flexibility of signalling in the modern world, and the abundance of possible channels for approaching candidates, raises the cost and confusion (and often embarrassment) of finding a mate rather than lowering it. A common strategy, to avoid the embarrassment of obvious refusal, is for initial acquaintance to occur in an environment in which mate search is not the principle purpose. The educational institution and workplace are excellent mating grounds in this regard, and an interesting side effect of the move to tighter sexual harassment regulations may be the further complication of mate search (Cohen 2002).[8]

The “inter-temporal lemons model” as a predictor of age at marriage is an example of rational maximisation in a world of imperfect information (Bergstrom 1997). Because information about how a man will perform economically becomes available at a later stage than information about how a woman will perform in the household, men with bad prospects will marry young before the women have a chance to find out; men with good prospects will wait and attract a better wife (Bergstrom and Bagnoli 1993). More successful males will tend to marry later as their careers develop and earnings peak later.

6.3  Family structure

Most economic analysis is concerned with either single-person households or monogamous couples, with or without children.

Alternative family structures are occasionally discussed in the economic literature. Polygyny (men having multiple wives), although rare in the West, is common around the world[9]. Becker extends an economic perspective to polygynous societies, arguing that its existence can improve the lot of married women (Becker 1991). Just as, in a goods market, excess supply tends to lower price, so too, in a monogamous marriage market, an excess of women over men means division of output within marriage favours men and the gains to women of marriage are lower. Under polygyny, Becker predicts higher incomes for women because of increased demand for wives. Even without a sex ratio imbalance, women might prefer only part of the attention of “successful” men to the full attention of “failures”. He also notes that low-status men do worse: “less efficient men are often forced to remain bachelors because they cannot offer women as much as other men can” (Becker 1991).

Unwed, non-cohabitating parents are increasingly prevalent in the West, leading some to describe the rising trend in single parent households as “serial polygyny”. One explanation for this phenomenon is an imbalance in numbers of “marriageable women” and “marriageable men” (men who are employed and not in prison) (Bergstrom 1997, Willis 1999, Wilson 1987). Because more women want to have children than there are marriageable men, some men will benefit from unofficial polygyny—fathering children by several women and marrying none (Willis 1999). Single women gain by having children rather than remaining childless. Having children outside marriage may therefore be a rational choice among the poor and unskilled where men’s wages are not much higher than women’s. Historically non-marital childbearing has occurred disproportionately among the poor (Ermisch 2003).

6.4  Family behaviour

Until recently much empirical work on households assumed that the family acted as if it were maximising a “family utility function” (Bergstrom 1997). However, to understand the family more fully it is necessary to get beneath this assumption and distinguish the incomes and consumption of different family members. Families make decisions about inputs (division of labour, energy, time), investment (in human capital and children) and consumption (how outputs and resources are distributed amongst family members) both contemporaneously and intertemporally (forgoing current consumption for retirement saving).

6.4.1  Sex/gender roles

Increasing returns from specialisation are a powerful force creating a division of labour in the allocation of time and investments. However, economics is agnostic in principle about comparative advantage: increasing returns alone do not imply the traditional sexual division of labour.

However, Becker’s view is that there are intrinsic differences in comparative advantage between men and women, not just in the production of children, but also in contribution to childcare. This determines the direction of subsequent investment and accumulation of human capital. Although there is no necessity that men specialise in market activities and women in household activities, women’s initial advantage in childrearing means that a division of labour based on sex is often the result. Small biological differences result in huge differences in activities as investments are channelled differently (Becker 1991).

6.4.2  Fertility

Decisions about fertility are central to family behaviour. They interact with, and shape, division of labour and differences in specific investments between men and women.

Outside the modern welfare state children can be seen as an investment, providing family labour and care for aged parents. During the nineteenth century fertility and wealth became negatively related among Western urban families. There has been a sharp decline in fertility in developed countries since the late 19^th century.

This reduction in fertility may reflect the growing importance to the economy of well-trained workers, which persuaded parents to substitute fewer, better educated children for traditional large families (Becker 1991). Parents choose to have fewer, high quality children. “A reduction in the number of children born to a couple can increase the representation of their children in the next generation if this enables the couple to invest sufficiently more in the education, training, and “attractiveness” of each child to increase markedly their probability of survival to reproductive ages and the reproduction of each survivor” (Becker 1991: 137).

The cost of children is also significantly affected by the value of the time of married women: as the wage rate for women has risen, the opportunity cost of raising children has also risen. Becker argues that the “contraceptive revolution” was a smaller factor than is often postulated, and was more a response to a decrease in the demand for children, than a cause.

6.4.3  Allocation of resources within families

Individuals within households are not identical, and their differing preferences mean that bargaining is likely to play an important role in household decision-making about such things as expenditure and the household division of labour.

Becker proposed the first model of household collective choice, arguing for efficient allocation and resource pooling[10]. Inspired by game theory, subsequent models have experimented with different assumptions and degrees of complexity, including co-operative bargaining solutions applied to marriages, equilibrium in non-cooperative games and the inclusion of household public goods. For a survey of this literature see Bergstrom (1997) and Zelder (2002).

However, there is evidence to suggest that the distribution of earnings within families determines the distribution of expenditure, directly challenging the resource-pooling hypothesis. Children do better (in terms of mortality and morbidity) when their mothers control a larger fraction of the family resources (Thomas 1990). An increase in a woman’s non-earned income increases her fertility whereas an increase in a man’s non-earned income does not (Bergstrom 1997). A restructuring of the child benefit payments in the UK from tax rebate (generally received by the father) to direct payment (generally collected by the mother) was correlated with a substantial and statistically significant increase in expenditure on children’s and women’s items relative to men’s (Lundberg 1997, Ward-Batts 2000). Control over household resources affects household expenditure and consumption patterns.

The utility of family members is interdependent—the welfare of one individual affects that of the others. Becker suggests that parents have altruistic, or deferential preferences to account for parental investment in children, However, others argue this is not necessary to explain such behaviour and that paternalism is all that is required (Pollak 2002). Posner (1998) argues that love, or altruism, is an emotion that is an efficient control mechanism in families. It facilitates cooperation and is a cheap substitute for formal contracting and monitoring (although does not solve every instance of conflict and exploitation).