Abstract: | Abstract. We examine a reform to the physician threshold system in Ontario that provides a unique opportunity to assess the effect of fee changes on physician behaviour, free from the biases that potentially affect simple time‐series or cross‐section inference. We find that: (1) the income effects of fee changes are small, but significant; (2) the income effect dominates the substitution effect only for a minority of services with relatively low prices and high volumes; and (3) the cross effects of fee changes tend to be significant only for services with relatively high prices and low volumes. |