New Indexes for Measuring Resource Use Efficiency

Abstract

The concept of Resource Use Efficiency (RUE) is about understanding how effectively are various resource inputs (e.g. energy, water, fertilisers,…) converted into outputs in production systems. There are number of indicators for RUE currently in use, but all of them have significant shortcomings, as they are mostly based on simple physical accounting based on the ratio of resources used as inputs to resources embodied in the output. In this paper, we address these shortcomings by designing a set of new indicators for RUE. Indicators that we propose are firmly grounded in economic theory and are based on long established indicators for evaluating efficiency of productive units. Some of the proposed indexes also take explicitly into account the production of undesirable outputs and their release into the environment. In that sense, the index is an environmentally adjusted efficiency index. In addition, indexes are dynamic in nature, and allow to track RUE of a given unit of observation over time, as well as to consider possible changes in technology over time. To explore the dynamics of RUE over time, we construct Luenberger indexes of RUE that allow for decomposing overall RUE into changes in efficiency relative to the technology frontier at each point in time, and changes in technology as movements of the frontier through time. The indexes are constructed from directional distance functions estimated by combining frontier technologies, and input, output, and undesirable output quantities for two time periods.

The empirical analysis consists of two separate applications. In the first application we estimate the directional distance functions using Data Envelopment Analysis (DEA) based on data for the Upper Mississippi River Basin (UMRB), which is taking a large portion of the US Mid-West, is an area with major use of nitrogen fertilisers, and consequently contributes significantly to nitrogen (N) runoff that pollutes the river basin and the Gulf of Mexico. Data on N loss come from a biopsychical study, whereas data on input use, output production, and N use come from the US Agricultural Census for years 2002, 2007, and 2012. Our findings indicate that nitrogen use efficiency (NUE) across the UMRB is very varied, and it also varies over time. These findings can be used to identify productive activities and specific areas with very low NUE, and to potentially target them with specific policy.

The second empirical application, which is still in progress, involves network DEA estimation, but this time looking at individual stages of production, in this case, stages in wheat production. We are again concerned with nitrogen fertilizer, and we use a biophysical model to simulate nitrogen uptake, nitrogen fluxes, and nitrogen stock on a daily basis during crop development. We then group the daily observations in seven production stages based on crop development. We link these stages in a network DEA model, and we estimate directional distance functions using specified constraints for each stage. This allows us to identify stages in production, and the corresponding management practices where NUE is highest/lowest, based on which we can consequently provide management advice.

About the presenter's visit

Tiho Ancev will be visiting the School of Economics on 13.11.19.  While here he will be using room 520A Colin Clark Building. If you would like to meet with him or have lunch or dinner with him please contact Fu Ouyang who will be his host while at The University of Queensland. Fu Ouyang can be contacted on f.ouyang@uq.edu.au.