Cost of Children
COST OF CHILDREN
At least since 1800 there has been a growing awareness that children are a cost factor for the household. To raise children, parents, or the household at large, must undertake a variety of direct expenditures. Children also require a time investment by their parents, some of which might otherwise be used in gainful economic activity. The latter costs, measured not as money spent on children but in lost earnings, are termed the parents' opportunity costs of children.
Examining parents' decisions about having children in sheer economic terms is often undertaken only with reluctance. The rewards of having children tend to be primarily emotional, associated with a mixture of altruistic and self-serving impulses: to nurture, to continue the bloodline, to balance family life, to satisfy curiosity about one's offspring, and to create relationships based on love. At the same time, and most evident in traditional agricultural societies, children yield economic rewards (1) by supplying labor resources to the family enterprise, often from an early age; and (2) when grown, by providing a measure of physical and economic security to parents, especially in the latter's old age, for which no equivalent institutional means may be readily available. If only implicitly, parents are likely to balance the costs of raising children against expectations of children's material and immaterial utility. An example of the cost of child-rearing becoming the object of a cold economic calculus can be seen in the decisions made by slave-owners as described by the economic historians Robert Fogel and Stanley Engerman. For slave-owners, the cost of raising slaves was in economic terms analogous to cattle-breeding.
In the nineteenth century the cost of children became a prominent consideration in policies addressing poverty. In order to ensure welfare equality among poor families irrespective of their number of children, families with many children would have to be supported by a cost-of-children subsidy. Later on, assessment of child costs also became pertinent in establishing the levels of family allowances and tax rates.
The first and classical study bearing on the cost of children is by the statistician Ernst Engel, who in 1895 published a study for the Prussian government. His name is still known through Engel's Law, which states that the food share in the household budget falls with rising income. There was considerable interest in the cost of children through most of the twentieth century, but some waning of interest in the later years. The main reason seems to be that there is no generally accepted scientific solution to the problem of what the costs of children are and that it seems impossible that one will be found.
Socializing the Costs of Children
Nevertheless most modern societies accept the idea that families and especially poor families should be partly subsidised in order to mitigate the welfare differences that may be caused by differences in the number of children. There are two questions. (1) What subsidies would be needed to neutralize the differences in family size? (2) If those amounts can be identified, should the state's subsidy be set at the level needed for complete neutralization, or at a lower or even a higher level? (A subsidy higher than needed for neutralization might be adopted as a means to raise the birth rate.) The second question is clearly a political issue, which is not the subject of this article. The first question is the relevant one.
It was noted that there is no generally accepted scientific solution to the problem. The practical solution chosen is just to let politicians and/or civil servants make decisions on the basis of intuition and compromises among interest groups such as the parents of large families. This is the route taken in most countries and explains the diversity in child-friendly regulations among countries.
A distinction may be drawn between household subsidies delivered in kind and those provided as income supplements. In-kind subsidies are in the form of state provision of below-cost education, health facilities, childcare, and even school meals. The objective is not only welfare equalization. For instance, education and the removal of illiteracy not only benefits the child who gets the education but also has a large positive external effect for others. But even if those policies are in force, the household itself still incurs substantial child-related expenses for food, clothing, shelter, insurance, etc. These can be offset by family allowances or family-size-dependent tax deductions. Benchmark estimates are needed in order to establish welfare-neutral compensations. Whether the actual subsidies provided are under-, over-, or just compensating is a matter of political choice.
Methods of Estimating the Cost of Children
A naive, but still popular, approach is to determine expenditures for children by bookkeeping and budget surveys. However, this approach raises two problems.
(1) What expenditures are necessary and thus may be called costs and what expenditures should not be seen as necessary costs? This depends very much on the cultural environment (establishing expenditure norms) and on the level of wealth of the household (wealthier families spend more on their offspring than poorer families [can] do). If the normative cost of children that is to be established is to be used for taxation and social security purposes, it is felt by many that it is unacceptable to assign to richer families a higher cost of children and as a consequence a higher per-child tax deduction. This leads to the ethical presumption that the cost of children is to be set equal to the cost incurred by a normative household, which is typically chosen at the minimum income level.
(2) The second problem in the bookkeeping solution is that of joint costs. There are many items in household expenditures that cannot be assigned exclusively as benefiting one particular member. Examples are expenditures on housing, health, insurance, television, etc. Decisions on how to assign such expenditures between parents and children are partly arbitrary. Recognition of economies-of-scale effects (cheaper by the dozen) introduces further complexities. The best-known household expenditure scale is probably the so-called Oxford scale, which is adopted more or less as the official scale by the OECD and the EU. It sets the first adult at 1.0, other adults at 0.7, and children (below 16 years) at 0.5. This scale, although not based on firm research results, is frequently used in official statistics to compare household welfare.
More sophisticated methods of estimating costs of children are the so-called 'adult good' method and the 'food-share' method. The adult good method (developed by E. Rothbarth and Angus Deaton and John Muellbauer) is based on the idea that one may identify a specific part of expenditures in the household budget, e.g. cigarettes, alcohol, etc., as adult expenditures. Let the monthly household expenditures be $2000. Assume that the adult's expenditures amount to $800 before the birth of the first child and $400 after the child has been born. Then the cost of the first child would be $800 to $400. This method also involves arbitrary elements. What are 'adult goods' and is the consumption of adult goods representative for 'adult welfare'? It may be assumed that if the birth of a child is the result of a voluntary choice by the parents, parental well being is increased by having that child–that is, increased by more than the loss in adult welfare caused by the resulting reduction in adult expenditures.
The 'food-share' method, which is influential in social security policy in the United States, is based on seminal research by economist Mollie Orshansky. She assumed that the share of food in family expenditures is an index of the family's welfare. For example, suppose food makes up 33 percent of total family expenditures before the first child's birth in a household with a $2000 per month income, and increases to 40 percent after the child is born. Assume further that the one-child household would again spend 33 percent on food if its income increased to $2500. Then the cost of the child is considered to be $500. Notice that the food share in accord with Engel's Law falls with increasing income. The method assumes that food-share represents household's welfare, which is dubious, and that the cost of children varies with income level.
A third approach to estimating the cost of children is subjective. Individuals are asked how satisfied they are with their income. It is found that respondents are less satisfied with the same income, the more children they have to support from it. This method was proposed by Bernard van Praag and Arie Kapteyn, who worked at the time at Leyden University in the Netherlands and is known in the literature as the Leyden method. On the basis of these satisfaction surveys it can be estimated how much has to be given to a household to keep it at the same financial satisfaction level after the birth of an additional child.
Although this method is less arbitrary than the methods previously described, it too has problems. First, the cost of children thus defined again increases with income. Second, it is unclear whether financial satisfaction is a good proxy for the metaphysical concept of household well-being. It is quite probable that non-financial satisfaction with life increases with the birth of a child. Third, the cost of the child will differ between one- and two-earner families. However, it may be argued that the additional money spent on external childcare will be roughly equivalent to the income forgone by the mother, if she would stay at home to care for her own children.
Although there is no generally accepted definition of the cost of children or accepted method of assessment, the various methods have some common features. Generally each shows an economies-of-scale pattern. For the objective methods, costs per additional child seem to diminish with rising family size at an exponential power of about 0.50. For subjective methods, the corresponding profile is consistently flatter, at about 0.30. Subjective estimates, however, implicitly take existing family allowance systems and the price of education into account. So the estimated costs of children are in a sense complementary to the family allowance system. In countries with liberal family allowance systems, and education and health care costs covered by the state, the estimated subjective costs will be smaller than in countries where parents receive fewer or no such benefits. For instance, significant differences in estimated child costs have been found between the Netherlands and Germany on one hand and the United Kingdom and Russia on the other.
In most countries even at the minimum income family income levels there is not a complete state compensation for the costs of children. For higher income levels, this holds a fortiori. If the financial consequences are among the determining factors in a couple's decisions concerning numbers of children, then most countries have a system that is discouraging childbearing. A few countries–for example, France and Canada–have a much more liberal system with the explicit purpose of increasing numbers of births. However, the effects are modest.
A final question that should be asked is whether children not only impose costs but also offer benefits in terms of an addition to general well-being. It might be argued that parents decide on having a child by comparing the expected benefits and costs to their well-being. For the first child the benefit will be evaluated as being larger than the cost. As the marginal benefit falls with each new child, at some specific family size the additional cost will be assessed to be larger than the additional benefit. Then the optimum family size is reached. If this theory were true the cost of a child is still more difficult to define. An attempt in this direction has been made by van Praag and Erik Plug, who found that the optimum number of children increases with income. Hence children are a luxury good.
Almost all states utilize some cost-of-child concept for minimum income families but the operationalization is mainly based on political decisions, sometimes enriched by the ideas of experts and social workers. If most of the population supports that practice, it is clearly acceptable, although not founded on widely accepted economic science. For an extensive survey, refer to the article by van Praag and Marcel Warnaars in the Handbook of Population and Family Economics.
Deaton, Angus S., and John Muellbauer. 1986. "On Measuring Child Costs: With Applications to Poor Countries." Journal of Political Economy 94: 720–744.
Engel, Ernst. 1895. "Die Lebenskosten Belgischer Arbeiterfamilien früher und jetzt–ermittelt aus Familienhaushaltrechnungen." Bulletin de l'Institut International de Statistique 9:1–149.
Fogel, Robert W., and Stanley L. Engerman. 1974. Time on the Cross: The Economics of American Negro Slavery. Boston: Little Brown.
Orshansky, Mollie. 1965. "Counting the Poor: Another Look at the Poverty Profile." Social Security Bulletin 28: 3–29.
Rothbarth, E. 1943. "Note on a Method of Determining Equivalent Income for Families of Different Composition." In War-time Patterns of Saving and Spending, ed. Charles Madge. Cambridge, MA: Cambridge University Press.
van Praag, Bernard M. S., and Arie Kapteyn. 1973. "Further Evidence on the Individual Welfare Function of Income: An Empirical Investigation in the Netherlands." European Economic Review 4: 33–62.
van Praag, Bernard M. S, and Erik J. S. Plug. 1999. "The Cost and Benefits of Children." In Fighting Poverty: Caring for Children, Parents, the Elderly and Health, ed. Stein Ringen and Philip R. deJong. Aldershot: Ashgate.
van Praag, Bernard M. S., and Marcel F. Warnaar. 1997. "The Cost of Children and the Use of Demographic Variables in Consumer Demand." In Handbook of Population and Family Economics, Vol. 1A, ed. Mark R. Rosenzweig and Oded Stark. Amsterdam: Elsevier.
Bernard M.S. van Praag