Effects of Economic Factors on Adoption of Robotics and Consequences of Automation for Productivity Growth of Dairy Farms. Factor Markets Working Paper No. 32, December 2012

In the long term, productivity and especially productivity growth are necessary conditions for the survival of a farm. This paper focuses on the technology choice of a dairy farm, i.e. the choice between a conventional and an automatic milking system. Its aim is to reveal the extent to which economic rationality explains investing in new technology. The adoption of robotics is further linked to farm productivity to show how capital-intensive technology has affected the overall productivity of milk production. The empirical analysis applies a probit model and an extended Cobb-Douglas-type production function to a Finnish farm-level dataset for the years 2000–10. The results show that very few economic factors on a dairy farm or in its economic environment can be identified to affect the switch to automatic milking. Existing machinery capital and investment allowances are among the significant factors. The results also indicate that the probability of investing in robotics responds elastically to a change in investment aids: an increase of 1% in aid would generate an increase of 2% in the probability of investing. Despite the presence of non-economic incentives, the switch to robotic milking is proven to promote productivity development on dairy farms. No productivity growth is observed on farms that keep conventional milking systems, whereas farms with robotic milking have a growth rate of 8.1% per year. The mean rate for farms that switch to robotic milking is 7.0% per year. The results show great progress in productivity growth, with the average of the sector at around 2% per year during the past two decades. In conclusion, investments in new technology as well as investment aids to boost investments are needed in low-productivity areas where investments in new technology still have great potential to increase productivity, and thus profitability and competitiveness, in the long run.

Identifying Factor Productivity from Micro-data: The case of EU agriculture. Factor Markets Working Paper No. 34, January 2013

The classical problem of agricultural productivity measurement has regained interest owing to recent price hikes in world food markets. At the same time, there is a new methodological debate on the appropriate identification strategies for addressing endogeneity and collinearity problems in production function estimation. We examine the plausibility of four established and innovative identification strategies for the case of agriculture and test a set of related estimators using farm-level panel datasets from seven EU countries. The newly suggested control function and dynamic panel approaches provide attractive conceptual improvements over the received ‘within’ and duality models. Even so, empirical implementation of the conceptual sophistications built into these estimators does not always live up to expectations. This is particularly true for the dynamic panel estimator, which mostly failed to identify reasonable elasticities for the (quasi-) fixed factors. Less demanding proxy approaches represent an interesting alternative for agricultural applications. In our EU sample, we find very low shadow prices for labour, land and fixed capital across countries. The production elasticity of materials is high, so improving the availability of working capital is the most promising way to increase agricultural productivity.