In this paper we are concerned with studying the technical efficiency of Tax Monitoring Agencies using DEA estimation procedure. Recently, the discussion of technical efficiency has recovered its interest basically because of the need to rationalize the expenditures of the public sector. However, DEA procedure only allows the point estimation of the technical efficiency. That is, the distribution of the technical efficiency estimator is unknow and therefore it is not possible to compare how different are two units in terms of its efficiency. The novelty of this paper is the application of two bootstrap procedures to approximate the distribution of the technical efficiency estimator. We find that, despite of the fact that the point estimates of the technical efficiency are different, they are not significantly different.

Key words: efficiency, DEA, bootstrap.
JEL classification: C14, D24, H59.