Do trends in Bangladeshi rice yields support Conway’s hypotheses about the consequences of modern agroecosystems?

Abstract

Gordon Conway postulated that in comparison to traditional agroecosystems, crop yields from modern agroecosystems are: (1) higher; (2) less sustainable; and (3) more variable, particularly in the early adoption stages. The existing literature predates more recent developments and does not employ long-term time series data to identify time phases using any econometric tests for structural breaks. This paper fills this gap by investigating the applicability of Conway’s hypotheses to Bangladeshi rice yields over sixty years examining trends, their sustainability, and their variability. We find mixed support for Conway’s hypotheses. Rice yields rose substantially following the Green Revolution. Overall, higher yields have been sustained. However, they have now virtually plateaued, and various factors (including environmental and ecological) have increased the risk of these yields becoming unsustainable. Until recently, absolute variations in Bangladesh’s rice yields were found to be higher than before the Green Revolution. This supports Conway’s third hypothesis. However, the relative variability of Bangladesh’s rice yields has consistently fallen since the Green Revolution. We highlight the trio of general factors determining rice yields and underscored the need for holistic analysis. Non-economists often overlook economic factors while economists often do not pay adequate attention to eco-environmental factors.