Cross-sectional Statistical Analyses of Regional Climate Effects: Ricardian Analysis and Extensions
Final Report Abstract
This project dealt with possible ways to improve the well established Ricardian approach for the assessment of climate change related economic effects on agriculture. First, different ways to control for soil quality were examined. In contrast to Ricardian analyses done so far usually relying upon dummy variables for certain soil types, for the first time also the unique German soil index (“Bodenzahl”, i.e. a climate independent soil productivity index developed in the first half of the past century) was used. Based on data from the 1999 German farm census and accounting for spatial autocorrelation it was shown that in this case different methods to control for physical soil characteristics did not influence the results of the Ricardian analysis. The soil index was finally considered the most precise indicator to account for soil quality when examining effects of climate change on land rents. Thus, it was included among the explanatory variables in the following Ricardian analyses. Second, the influence of very likely errors in variables on the estimated marginal rental prices for climatic attributes was examined. It turned out that the bias from treating interpolated climate variables as exogenous may be quite important. Treating interpolated climate normals as endogenous available cost-effective instruments were used to correct for such errors. Again, spatial autocorrelation was accounted for to correct for other cross-sectional data inefficiencies. Third, the German farm sector’s adaptation to climate change was studied using a very large data set (again based on the 1999 agricultural census) and aggregating dominant farm types at community level. Analyzing the climate-dependent incidence of six farm types (defined by official statistics) as well as the climate-induced economic impacts on farm rents the first socalled structural Ricardian analysis for a whole European country was performed. One important finding consistent with common agronomic knowledge was that permanent-crop farms are more likely to dominate with increasing average temperature. In contrast, higher precipitation levels lead to increasing probabilities of forage or mixed farms being the dominant farm types. The different estimation results produced in this project were used for simulation exercises entering spatially processed climate projections from the regional climate model REMO into the regression equations. By and large, for a moderate climate warming scenario and choosing the period 2011-2040 these simulations yielded overall net benefits to be expected by the German farm sector due to the projected temperature and precipitation changes. Further, the simulation based on the structural Ricardian analysis resulted in an increase in the number of communities where cash-crop, fattening, permanent-crop, horticultural, or mixed farms would dominate, while the number of communities with dominant forage farming would decrease. Technical problems not foreseen at the beginning arose from the new possibility of remote data analysis offered by the German Statistical offices in order to allow for comprehensive econometric analyses despite rather strict data privacy requirements. Running statistical analyses without having direct access to the data sets to be used and waiting for the approval of every output file turned out to be quite time consuming and finally led to a delay of the original time schedule. Also follow-up costs linked to the likely need of future additional statistical analyses because of the ongoing publication process and in order to fulfill the data storage requirements have to be borne by the Institute of Farm Management.
Publications
-
(2012): Explaining the Climate-Dependent Distribution of Crops in Space - the Example of Corn and Corn-Cob-Mix in Baden-Württemberg. In: Balmann, A. et al. (Hrsg.): Unternehmerische Landwirtschaft zwischen Marktanforderungen und gesellschaftlichen Erwartungen. Schriften der Gesellschaft für Wirtschafts- und Sozialwissenschaften des Landbaus e.V., Band 47, Münster, S. 91-102
Lippert, C., Chatzopoulos, T., Schmidtner, E. und J. Aurbacher
-
(2012): Ricardische Analyse zur Produktivität der deutschen Landwirtschaft unter Berücksichtigung verschiedener Bodenqualitätsmaße. Tagungsband der 22. Jahrestagung der Österreichischen Gesellschaft für Agrarökonomie (ÖGA), 20.-21. September 2012, Universität für Bodenkultur Wien, S. 75-76
Schmidtner, E., Dabbert, S. und C. Lippert
-
(2013): Cross-Sectional Statistical Analysis of Regional Climate Effects: Ricardian Analysis and Extensions. In: Bahrs, E. et al. (Hrsg.): Herausforderungen des globalen Wandels für Agrarentwicklung und Welternährung. Schriften der Gesellschaft für Wirtschafts- und Sozialwissenschaften des Landbaus e.V., Band 48, Münster, S. 487-488
Chatzopoulos, T., Schmidtner, E. und C. Lippert
