With the aim of determining the welfare implications of price change in consumption data, we introduce a revealed preference relation over prices. We show that an absence of cycles in this preference relation characterizes a model of demand where consumers trade-off the utility of consumption against the disutility of expenditure. This model is appropriate whenever a consumer’s demand over a strict subset of all available goods is being analyzed. For the random utility extension of the model, we devise nonparametric statistical procedures for testing and welfare comparisons. The latter requires the development of novel tests of linear hypotheses for partially identified parameters. In doing so, we provide new algorithms for the calculation and statistical inference in nonparametric counterfactual analysis for a general partially identified model. Our applications on national household expenditure data provide support for the model and yield informative bounds concerning welfare rankings across different prices.