| Abstract: | The piecemeal application of second-best arguments to practical policy-making is highly dangerous. This paper attempts to demonstrate this by taking the recent Australian debate over ‘tariff compensation’ as an example. The attempt to ‘compensate’ for tariff-induced distortions in output prices by means of input subsidies is shown to lead to informational problems so serious that the pursuit of such a policy in practice seems unwise. This point is argued by means of extensive numerical examples. These suggest that the potential welfare gains from tariff compensation, optimally applied, are small relative to the potential costs of well-intentioned but erroneous interventions. |
