Surrogate modelling of a detailed farm-level model using deep learning

Technological change co-determines agri-environmental performance and farm structural transformation. Meaningful impact assessment of related policies can be derived from farm-level models that are rich in technology details and environmental indicators, integrated with agent-based models capturing dynamic farm interaction. However, such integration faces considerable challenges affecting model development, debugging and computational demands in application. Surrogate modelling using deep learning techniques can facilitate such integration for simulations with broad regional coverage. We develop surrogates of the farm model FarmDyn using different architectures of neural networks. Our specifically designed evaluation metrics allow practitioners to assess trade-offs among model fit, inference time and data requirements. All tested neural networks achieve a high fit but differ substantially in inference time. The Multilayer Perceptron shows almost top performance in all criteria but saves strongly on inference time compared to a Bi-directional Long Short Term Memory.     

Does access to improved grain storage technology increase farmers' welfare? Experimental evidence from maize farming in Ethiopia

Seasonal price variability for cereals is two to three times higher in Africa than on the international reference market. Seasonality is even more pronounced when access to appropriate storage and opportunities for price arbitrage are limited. As smallholder farmers typically sell their production after harvest, when prices are low, this leads to lower incomes as well as higher food insecurity during the lean season, when prices are high. One solution to reduce seasonal stress is the use of improved storage technologies. Using data from a randomised controlled trial, in a major maize-growing region of Western Ethiopia, we study the impact of hermetic bags, a technology that protects stored grain against insect pests, so that the grain can be stored longer. Despite considerable price seasonality—maize prices in the lean season are 36% higher than after harvesting—we find no evidence that hermetic bags improve welfare, except that access to these bags allowed for a marginally longer storage period of maize intended for sale by 2 weeks. But this did not translate into measurable welfare gains as we found no changes in any of our welfare outcome indicators. This ‘near-null’ effect is due to the fact that maize storage losses in our study region are relatively lower than previous studies suggested—around 10% of the quantity stored—likely because of the widespread use of an alternative to protect maize during storage, for example a cheap but highly toxic fumigant. These findings are important for policies that seek to promote improved storage technologies in these settings.     

Agriculture and climate change in global scenarios: why don't the models agree

Agriculture is unique among economic sectors in the nature of impacts from climate change. The production activity that transforms inputs into agricultural outputs involves direct use of weather inputs (temperature, solar radiation available to the plant, and precipitation). Previous studies of the impacts of climate change on agriculture have reported substantial differences in outcomes such as prices, production, and trade arising from differences in model inputs and model specification. This article presents climate change results and underlying determinants from a model comparison exercise with 10 of the leading global economic models that include significant representation of agriculture. By harmonizing key drivers that include climate change effects, differences in model outcomes were reduced. The particular choice of climate change drivers for this comparison activity results in large and negative productivity effects. All models respond with higher prices. Producer behavior differs by model with some emphasizing area response and others yield response. Demand response is least important. The differences reflect both differences in model specification and perspectives on the future. The results from this study highlight the need to more fully compare the deep model parameters, to generate a call for a combination of econometric and validation studies to narrow the degree of uncertainty and variability in these parameters and to move to Monte Carlo type simulations to better map the contours of economic uncertainty.     

Viagra(ization) or Technoromanticism

Viagra can rather obviously be termed "techno-romantic"--literally it is a technology to make sex possible. But technoromanticism is also a philosophical term referring to the belief that advances in technology serve "truth" "unity" and "wholeness" according to Coyne (1999). Surprisingly many pharmaceutical companies seem to have developed "Viagra" or at least substances that were very similar, but to have ignored and/or rejected commercializing them. This article assumes that the primary activity of research and development in pharmaceutical companies is to produce texts that interpret chemicals as "meaningful." Assumptions of what constitutes (medical) meaning within prevailing technoromanticism have seen R and D as a form of "production" that demands rejection of "consumption." This technoromanticism succeeded in rebuffing the sexual "techno-romanticism" of Viagra in the company studied, but Viagra's enormous commercial success has reversed priorities. R and D's meaning system has proven to be no match for Viagra-ization--the article examines the resulting crisis in meaning systems.     

Simplified Risk Analysis in Agricultural Extension

Attention is given to the major methods of assessing risk in order to make choices. The literature frequently reports studies on mean-variance (E-V) analysis and stochastic dominance rules. Both of these methods are compared, in an agricultural application, with the approach outlined by Hanoch and Levy. The empirical results suggests that Hanoch and Levy's criterion appear to reduce the size of the risk-efficient set, and that the sets of recommendations from the methods under comparison were consistent. The two recommendations generated by Hanoch and Levy's rule were found to be reasonable for this particular decision process.     

Gender equality still on the European agenda—but for how long?

While the momentum has certainly not yet disappeared from the gender equality agenda within Europe, the impact of that agenda remains patchy. Progress within individual member states has been significant but not always steady or cumulative. At the European level more attention has been paid to the gender pay gap but policy initiatives remain weak. The loss of the equal opportunities pillar in the new employment guidelines puts this momentum at risk but the new phase still includes commitments to an integrated strategy of gender mainstreaming and equal opportunities.     

The ups and downs of European gender equality policy

In 2003, equal opportunities policy in the European Union suffered both ups and downs. New opportunities came in the guise of the hotly contested new directive on gender equality outside the field of employment, in the invitation to present the first of an annual report on equality between women and men to the Spring Council, in the consolidation of gender mainstreaming within the second round of the National Action Plans on social inclusion and in the new commitments to ‘substantial reductions by 2010’ in gender gaps in employment, unemployment and pay that were included in the new employment guidelines in 2003. These new guidelines presented, however, a major challenge to gender equality as the new phase of the European Employment Strategy dispensed with the four pillars, and therefore the equal opportunities pillar. Instead gender equality became just one of 10 new guidelines. In December the launch of the Employment Taskforce report appeared to push employment policy back more to a ‘full employment with flexibility’ approach and away from concerns with job quality. The focus was therefore more on the integration of women into employment rather than on closing the equality gap.     

Shared Sectoral Growth Versus the Dual Economy Model: Evidence from Côte d'Ivoire,Ghana, and Zimbabwe

Abstract: Though recently challenged, the dual economy view has affected development economics and development policy, either explicitly or implicitly, for more than half a century. According to this view, agriculture merely serves to build the industrial sector — in particular, agriculture has no role as an engine of growth in the long term. Examining agriculture, industry and service sector growth in Côte d'Ivoire, Ghana, and Zimbabwe over more than three decades, we find little empirical support for this view. On the contrary, our analyses find a large degree of interdependence in long‐run sectoral growth, implying that the sectors ‘grow together’ or, similarly, that there are externalities or spillovers between sectors. Policy implications are also discussed; these include directing more attention towards the interdependencies in sectoral growth broadly defined. In particular, our findings have implications for the design of education and health programs, as well. This improved understanding of inter‐sectoral dynamics at all levels may facilitate policy implementation aimed at increasing economic growth — and thereby ultimately improving peoples' livelihoods — in Africa and elsewhere.     

Opening services markets within Europe: Modelling foreign establishments in a CGE framework

Foreign direct investment (FDI) in services is often more important to supply foreign markets than cross-border trade. A complete analysis of services liberalization therefore requires the modelling of FDI. This paper presents the treatment of FDI in our CGE model WorldScan based on the ideas of Petri [Petri, P.A., 1997. Foreign direct investment in a computable general equilibrium framework. Paper Prepared for the Conference, Making APEC work: Economic Challenges and Policy Alternatives, March 13–14. Keio University, Tokyo] and Markusen [Markusen, J.R., 2002. Multinational Firms and The Theory of International Trade. MIT Press] that firms which establish affiliates abroad also transfer firm-specific knowledge. Consequently, capital owned by suppliers from home and foreign countries are not perfect substitutes. We apply this model to the proposals of the European Commission to open up services markets. Even when FDI in services could increase by 20% to 35%, the overall economic impact is limited. Our assessment suggests that GDP in the EU25 could increase up to 0.4%. These effects could be up to 0.8% higher if foreign capital also increases the overall productivity of the services sector.