Tax compliance studies usually focus on the effect of enforce-ment spending on tax evasion. Reliable estimates are difficult to obtain because evasion data are often suspect. This note shows how tax revenues can be used instead of evasion data to estimate the impact of changes in enforcement spending. Applying our method to Chilean data, we find that $1 (USD) of additional enforcement spending increases VAT revenues by $31. Moreover, current levels of spending could increase by 40% and still be within sample values. Hence, a 10% increase in spending could reduce evasion from its current rate of 23% to 20%.

In this paper we show that fixed‐term contracts, which are commonly used to franchise highways, do not allocate demand risk optimally. We characterize the optimal risk‐sharing contract and show that it can be implemented with a fairly straightforward mechanism—a least‐present‐value‐of‐revenue auction. Instead of bidding on tolls (or franchise lengths), as in the case of fixed‐term franchises, in an LPVR auction the bidding variable is the present value of toll revenues. The lowest bid wins and the franchise ends when that amount has been collected. We also show that the welfare gains that can be attained by replacing fixed‐term auctions with LPVR auctions are substantial.