D. Byerlee et R. Murgai, Sense and sustainability revisited: the limits of total factor productivity measures of sustainable agricultural systems, AGR ECON, 26(3), 2001, pp. 227-236
Many economists have advocated and applied total social factor productivity
(TSFP) (i.e., total factor productivity estimated with both market and non
-market inputs and externalities, and with all factors valued at social pri
ces) as a single all-embracing measure of agricultural sustainability. This
paper reviews the conceptual and practical issues in measuring TSFP and sh
ows that no one measure alone will be theoretically or empirically robust a
s an indicator of sustainability. TSFP is a conceptually flawed measure sin
ce inclusion of non-mark-et inputs and outputs and social price-based valua
tion, in most cases, violates the theoretical basis underlying those estima
tes. Trends in TSFP also have limited value in diagnosing the nature of sus
tainability problems, unless changes in productivity are related to underly
ing changes in technology, human and physical infrastructure, and indicator
s of resource quality.
More attention needs to be given to defining key indicators of agro-ecosyst
em health and relating these measures to trends in productivity. This analy
sis must be sufficiently disaggregated and for a long enough time period to
allow for spatial and temporal variability inherent in agricultural produc
tion. Secondary data at the district level on both conventional inputs and
outputs and resource quality have recently allowed more quantitative estima
tes of sustainability and its causes. With limited data, yield growth decom
position analysis can often be used to provide valuable insights into susta
inability problems. Meanwhile, there is a need to invest in long-term exper
imental and panel surveys of farmers and their fields for key production sy
stems in order to provide long-term data that will allow full productivity
accounting, using more formal statistical procedures. Regardless of the app
roach selected, the findings of this paper strongly suggest a need for econ
omists, agronomists and soil scientists to collaborate in integrating appro
aches in order to provide more robust and informative measures of sustainab
ility. (C) 2001 Elsevier Science B.V. All rights reserved.