(c,c+Q) would not rule out purchases by uninformed consumers under the commission regime. However, since the price is higher under the commission regime than under the fee regime, it is always the case that expected consumer surplus of uninformed consumers is greater under the fee regime. The welfare comparisons below will not therefore be qualitatively affected. 14 The last step follows because (p-c)F = kF = r, and equation (5) holds at the equilibrium. 15 See Gravelle (1991) where it is also demonstrated that it is possible for the commission equilibrium to be worse than the allocation produced by a cartel of firms. Introduction Potential purchasers are likely to be initially ignorant about the complex characteristics of life insurance products. They can be provided with information and advice by brokers who are remunerated by commissions paid by insurers. It has been suggested in the United Kingdom (see Office of Fair Trading, 1990), the United States (see Ferling, 1991, and Otis, 1991), and Australia (see McIlwaine, 1990) that it would be better if consumers directly paid a fee for advice about financial products rather than having the information providers receive a commission from the insurer if a sale is made. The suggestions are prompted by the argument that a commission system gives greater incentives to provide biased advice to unsophisticated potential customers. However, a change to a fee-for-advice system would also affect consumers by increasing the price of advice and reducing the price of the product. This article models the determination of the price of advice, the price of the product and the quality of advice under the two remunerations systems. The model is used to compare the welfare properties of the two systems and, particularly, to test the intuitively plausible assertion that a fee-for-advice system is better than the commission system because it leads to a lower product price and better advice. The maintained assumption of regulators and commentators is that brokers take advantage of consumers' ignorance and lack of sophistication. This article shows that the weak market position of consumers will be exploited under both the commission and fee-for-advice systems. Neither remuneration system will achieve even a second-best efficient allocation. It is also demonstrated that the fee system is not necessarily superior to the commission system once the number of brokers and purchases under the two regimes are taken into account. The next section describes the basic framework of the model that is used in the following two sections to examine the equilibrium with a commission system and with a fee-for-advice system. Next, the efficiency of the two systems is compared.(1) The model is then extended to allow for heterogeneous consumers who differ in the value they place on information. The penultimate section investigates the implications of the remuneration regime for the quality of advice. The last section summarizes the argument and contains some conclusions. Basic Framework The model of the life insurance market is deliberately stylized with many important features of the market ignored or assumed away. The justification for the strong assumptions and simple specification is that the aim of the article is to investigate the plausible view that a fee-for-advice regime is obviously better than a commission regime. If it can be shown that the welfare comparison of the two remuneration regimes is ambiguous even in such a simple framework then there is no need for a more elaborate and less tractable model. A simple framework also has the advantage of starkly illustrating the influence of different parameters and assumptions and points the way for further investigations by indicating the significant features of the specification. Life Insurers Life insurers are risk neutral, competitive, and produce identical policies and break even in equilibrium. …