Non-parametric productivity analysis with undesirable outputs: An application to the Canadian pulp and paper industry

This article extends the Chavas-Cox approach to non-parametric analysis by incorporating undesirable outputs to provide a more complete representation of the production technology. Inner and outer non-parametric technology bounds are constructed. The methods are illustrated with application to time series data for the Canadian pulp and paper industry. Conventional measures that ignore changes in pollutant outputs underestimate true productivity growth. Further, there is a large gap between estimates generated with reference to inner and outer bounds to the technology, suggesting that researchers need to be aware of the limitations of results derived from analyses relying only on DEA methods.