Eco-efficiency in Swedish dairy farms : incorporating sustainability into the measure of eco-efficiency

Abstract: This thesis introduces a novel method to incorporate absolute values of GHG emissions into the measure of eco-efficiency. The aim is to assess eco-efficiency in Swedish dairy farms and adjusting the scores towards a threshold of absolute levels of GHG emissions using the proposed method. The Swedish target of net-zero emissions in 2045 is used to specify the threshold of GHG emissions. The secondary objective of this thesis is to test for exogenous variables correlating with eco-efficiency using OLS-regression. The empirical application is on Swedish dairy farms using FADN data from 2016. I use a data envelopment analysis (DEA) to assess eco-efficiency considering the environmental pressures nutrients and contribution to global warming. When assessing eco-efficiency, the result is that Swedish dairy farms are inefficient with a mean-efficiency of 0.33 among conventional farms and 0.56 among organic farms. Adjusting the scores towards absolute levels of GHG emissions increase the farms’ mean-efficiency. Using an environmental target as a threshold when computing efficiency indicates whether the units in the sample reach the environmental target or not. Comparing the unadjusted and adjusted efficiency-scores using a Spearman rank test shows a high similarity between the unadjusted- and adjusted scores. Testing for exogenous variables correlating with efficiency shows that the intensity of farming correlates significantly with both the adjusted and unadjusted eco-efficiency in both conventional- and organic farms. Thus, the policy-recommendation is to increase the intensity of farming in order to increase eco-efficiency. The correlation between eco-efficiency and some other exogenous variables tested for, such as the number of livestock-units and subsidies, differs between the unadjusted and adjusted efficiency scores. This result indicates that the effect of policies aiming to increase eco-efficiency can differ whether considering absolute environmental damages in the measure or not. Differences in scores and correlates with eco-efficiency before and after the adjustment points to the contribution of this thesis.