Farm Risk Management Policies under Climate Change


Published on

Presentation delivered at the conference: Climate Smart Agriculture: Global Science Conference University of California – Davis, March 20th, 2013.

© FAO:

Published in: Education, Technology
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Farm Risk Management Policies under Climate Change

  1. 1. Andrea Cattaneo (FAO) & Jesús Antón (OECD), Shingo Kimura (OECD), and Jussi Lankoski (OECD) Session on risk management Climate Smart Agriculture: Global Science Conference University of California – Davis, March 20th, 2013 Farm Risk Management Policies under Climate Change
  2. 2. Overview 2 1. Issues and Research strategy 2. Uncertainty in downscaled climate outcomes 3. Uncertainty in farmer response 4. How do policies perform under different combination of these uncertainties? 5. Robust policy choice in the contexts of Australia, Canada, and Spain
  3. 3. The issues From CC concerns... • More extreme weather events • Farmers not aware, or without tools • Production disrupted in some areas • Government’s role on management of new risk ... to adaptation policy issues • CC modifies the farming risk environment, but how? • High uncertainty about the impact and adaptation response (ambiguity) • Policies often can hinder the needed adaptation to climate change • Different needs/impacts by farm type • What is policy objective? Market failure, stabilization, low incomes? 3
  4. 4. Tools & Research Strategy 4 MICROECONOMIC MODEL under uncertainty: Expected utility of returns Montecarlo simulations on P & Q Farm Household panel data: Australia, Canada & Spain Empirical Literature on CC CC Risk Scenario Policy Calibration: Y, A, & I Insurance & ex post Production decisions Risk Management decisions Impactonvariability Costofpolicy Optimal Policy Under risk • Uncertainty in CC scenarios and farmer expectations of climate • Robust policy needs to incorporate these uncertainties, not just risk
  5. 5. Uncertainty in CC impacts on Yield Risk • Increased temperatures, CO2 fertilization, change in rainfall, “new” pest and diseases, climatic extremes • Impacts: GCM + Econometric/Agronomic • Australia: winter time decrease rainfall • Canada (Sask.): increase and changed precipitation • Spain: decrease rainfall 5 Australia1 Canada2 Spain3 Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation Wheat -7.2 10.3 -3.0 -2.0 -1.8 110.5 Barley -20.0 0.0 -10.0 -17.0 7.3 89.3 Oilseeds -19.9 -6.1 -13.0 2.0 Sources: 1. Luo et al. (2010), Van Gool and Vernon (2006), 2. Zhang et al. (2011), and 3. Guereña et al. (2001). • But uncertainty persists on changes in mean and variance of yields caused by climate change at disaggregated scales
  6. 6. Uncertainty about Behavioural response • Adaptation • Structural / anticipatory: new technologies/type of farming • Reactive /autonomous: timing, diversification • Misalignment - What if farmers do not have sufficient information to update their perception of risk environment? 6 Change in diversification index in response to marginal climate change (percentage change) Australia Canada Spain Low risk farm 17.6 -3.6 19.8 Medium risk farm 16.3 -2.7 n.a. High risk farm 13.7 3.1 22.4
  7. 7. CC Scenarios & Behavioural Response 7 Description Climate Scenarios Baseline (No climate change) Marginal climate change Extreme events BehaviouralSub-scenarios Business-as-usual Expresses how policy instruments would function without climate change Baseline Diversification (No adaptation) Based on expected impact on yields assuming farmers can only adapt by diversifying among existing varietals Marginal Extreme Structural adaptation Expected impact on yields based on the literature, assuming farmers can switch to crop varietals that reduce impact of climate change Marginal with adaptation Extreme with adaptation Misalignment Farmers make production decisions based on their historical experience and therefore do not take into account the increase in systemic risk (no adaptation) Marginal with mis- alignment Extreme with mis- alignment
  8. 8. Policy choices analyzed (1) Individual yield insurance - Insurance triggers by an individual yield loss - Premium differentiated by farm - High administrative cost (2) Area-yield insurance - Insurance triggers by a systemic yield loss - Uniform premium for all farms - Relatively low administrative cost (3) Weather index insurance - Insurance triggers by a systemic rain fall - Uniform premium for all farms - Low administrative cost (4) Ex-post payment - Payments triggers when all crop yields decline systematically - No premium payments by farmers
  9. 9. Marginal C.C.& No misalignment Australia: • Strong specialization (crowding out) • Best performance: Area ins. and ex post Canada: • Low effectiveness of policies • Area Ins. Performs well, Index improved with CC (more correlation) Spain: • Large increase in demand for insurance • Non-irrigated farms more affected by CC, but no improvement on cost- effectiveness General: Implications of CC: – does not dramatically modify results and effectiveness – costs increase (e.g. Spain) Best policy differ by farm type Individual Ins. is well demanded and reduces risk, but it is expensive Ex post payments are cheaper, but more effective (less crowding out) for low income objectives OECD Trade and Agriculture Directorate 9
  10. 10. Why robustness of policy? • Budgetary costs out of control (e.g. Canada) Baseline Marginal Climate change Extreme events NoStruct. Adapt. Structural Adaptation Misalignm ent NoStruct. Adapt. Structural Adaptation Misalignm ent No policy 0 0 0 0 0 0 0 Individual yield 68 179 198 227 185 236 399 Area yield 82 80 90 630 87 134 1070 Weather index 36 32 31 95 41 49 88 Ex-post payment 56 41 42 199 35 48 308 Percentage of triggering 3.9 4.9 4.0 14.0 3.8 3.4 17.0 Budgetary cost when triggered 867 840 945 1419 917 1374 1925 10
  11. 11. Why robustness of policy? • A robust policy should avoid – Budgetary costs out of control (e.g. Canada) – Policies increasing risk (e.g. Australia) • How to define robustness? – Bayesian Criterion: • Best performance “on average” – Satisficing Criterion: • Best policy or within 35% of the best – MaxiMin Criterion: • Best performance in its worst outcome. 11
  12. 12. Robust Policy Choice Depends on Objective Country case Bayesian optimum Satisficing MaxiMin Australia Variability Low incomes gain Area yield Ex-post payment Area yield - Area yield Ex-post payment Canada Variability Low incomes gain Weather index Weather index Weather index - Weather index Ex-post payment Spain Variability Low incomes gain Area yield Weather index Area yield Weather index Area yield Weather index 12
  13. 13. Conclusions on Robustness • Extreme events and misalignment significantly change the decision environment – Misalignment imply high cost and low adaptation – Information policies can be useful • Reducing income variability focuses on “normal” risk: – Crowding out of adaptation is more likely – Area yield and weather insurance tend to be cheaper than individual risk and effective enough • Reduce incidence of low income more justified: – Ex post payments are effective – Individual yield with deductible targeted, but costly 13
  14. 14. Caveats and further insights • Stylized model does not capture intricacies of policy design, transaction costs & governance • However, some issues identified here will be even more relevant in developing country context – Uncertainty in CC impacts and farmer response – The role of different farm types (asset availability) – Design of safety nets and interaction with adaptation • Other relevant issues not addressed here – Barriers to change (financial, institutional, technical) 14
  15. 15. For more information • Paper available on OECD website: • Contact : 15