The Low-Cost-Hypothesis - Theoretical foundations and empirical

The Low-Cost-Hypothesis (LCH) postulates that the effect of attitudes on behavior varies with the costs at stake in the situation. The effect is deemed to be higher in low-cost-situations, compared to high-cost-situations. We argue that a closer look at the literature reveals two distinct versions of the LCH, which have not been disentangled so far. We reconstruct both versions - the "simple" and the "specific" version of the LCH - and discuss their theoretical foundations and empirical implications. The "simple" version can be derived from a simple expected utility model. In this model, attitudes and (tangible) behavioral costs exert independent effects on the net utility of the action alternatives. A conditional effect of the attitudes that depends on the behavioral costs only exists with respect to the probability of choosing an action alternative. However, this dependence of marginal effects on the actor's initial level of utility or probability holds for any independent variable or utility argument. A second, "specific" version of the LCH postulates a variable-specific interaction effect between attitudes and costs. We point to the problems of previous approaches to deriving such a specific hypothesis and identify dual-process-theories as an alternative theoretical foundation that allows understanding the limited scope conditions of the specific LCH. The article carries important conclusions for empirical applications and tests of the LCH in diverse fields of sociological research and, more generally, for the decision-theoretic analysis of social action.