The supply of resources in an economy is limited. However, the demand for resources is generally unlimited. This is particularly the case with healthcare; if the technology and expertise to treat people exist, then there will be a demand for such resources. If the technology and expertise do not exist, then there will be a demand for their provision. Scarcity of resources leads to competing alternatives. In the healthcare arena, policy-makers and clinicians are confronted with having to decide how healthcare resources should be allocated. What should guide decision-makers in this task of resource allocation? Obviously we would expect treatments and interventions to be looked at favourably if they are known to produce good outcomes. However, it is also necessary to know what the costs of achieving these outcomes are. This is particularly relevant given the high and increasing cost of healthcare. But economics should not be concerned with simply cutting costs. Costs are in effect proxies for units of production. More units of production could well result in improved outcomes. Therefore, the more expensive option might be the preferred one. If an evaluation does not include a cost component, then inefficient services may go undetected and resources will be used inappropriately. High-quality outcome and cost data are therefore required to advise policy-makers and clinicians as to the best use of their limited resources. This chapter examines the extent to which costs have been calculated in an appropriate way in mental healthcare evaluations.

Cost measurement has been undervalued in many studies. Sometimes costs have been omitted. In other cases, not all relevant costs have been collected. Costs are frequently measured incorrectly or not interpreted correctly. When costs have been measured, it has often been in an ad hoc fashion. Hardly any costing instruments have been published and described. One exception in the UK is the Client Service Receipt Inventory (Beecham, 1995; Beecham & Knapp, 2001). This is unfortunate, as costs are one of the few measures that have regularly exhibited variations among treatment interventions.

We need to be sure that the cost data being collected are appropriate.