Options to reduce the environmental effects of livestock production – Comparison of two economic models

Abstract

Global livestock production accounts for about 80% of global land use, is one of the main drivers of biodiversity loss, and is responsible for about 18% of global greenhouse gas emissions. These impacts are likely to become more pressing as a consequence of rising demands for meat, eggs and dairy products. Theoretically, these impacts could be reduced by making the global food system more efficient or by dietary changes, as recent studies suggest. However, multiple feedbacks exist in the agricultural system, which may reduce the effectiveness of any promising change. Estimation of these effects is highly uncertain and depends on the tools applied. In this study, we used two different economic models (IMPACT and LEITAP), coupled to the integrated assessment model IMAGE, to examine different options to reduce the environmental impact of agriculture: dietary changes (less meat and dairy), increased production efficiency, and reduced food waste. In a detailed model comparison, we assessed the model results on consumption, agricultural production, commodity prices, land-use change and greenhouse gas emissions, and identified feedbacks in the global agricultural system. In both models, all options resulted in a reduction in agricultural land use and greenhouse gas emissions, as well as in agricultural commodity prices. The model results show that for most options less than the theoretical environmental gains would actually be achieved, due to price feedbacks leading to increased consumption and less intensive production. On the other hand, larger than expected effects could occur as a result of reduced European consumption.However, large differences were found between the IMPACT and LEITAP model calculations. We attribute these differences to model design and parameterisation, discuss implications and sketch ways forward to improve studies of future dynamics in the global agricultural system. The most pertinent discrepancies between the model results were related to the models’ implementation of international trade, determining to what extent current trade patterns are retained, the assumptions on technological change, which have major implications for future price developments, and the treatment of agricultural expansion, which strongly affects how agricultural land use reacts under certain policy options. To our knowledge, this is the first attempt to compare different economic models in a consistent scenario study, and the results indicate the need for model improvements and show that data harmonisation and more extensive model comparisons are needed.