Weather Shocks, Climate Change and Business Cycles

1. Weather Shocks, Climate Change and Business Cycles Ewen Gallica & Gauthier Vermandelb a CREM UMR CNRS 6211, Université de Rennes 1 & Actinfo Chair b Leda, Université-Paris Dauphine, PSL Universités & France Stratégie Source: Squarzoni, P. (2012). Saison brune. Delcourt. Environmental Economics: A Focus on Natural Resources Orléans, 5-6 April 2018

2. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Motivation 1980 2000 20 40 60 (a) Number of weather disasters per year 1980 2000 0 10 20 current USD (billions) (b) Value of damages per year 1980 2000 0 200 400 600 current USD (millions) (c) Average damage per disaster & year Source: EM-DAT; IMF, World Economic Outlook database. Figure 1: Global frequency and impact of weather shocks (droughts and heat waves) between 1970 and 2016. E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 2/27

3. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Weather Shocks Policy Challenges • With climate change, weather shocks are a growing source of business cycle ﬂuctuations • Variable climate conditions force policy-makers to design new policies aimed at stabilizing the economy following a weather shock • Weather shocks are harmful for agriculture-based economies E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 3/27

6. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Literature Related to the Weather So far, the literature focuses on: • Long run effects of climate, using: Integrated Assessment Models (IAMs) pioneered by Nordhaus (1991) that measure the social costs of carbon emissions on a very long per- spective (no fluctuations). • Short run effects of the weather, using: VAR models that highlight the important role of weather shocks (Buckle et al., 2007; Kamber et al., 2013) Panel models that show that higher temperatures have a detrimen- tal effect on economic growth (Dell et al., 2012; Burke et al., 2015; Acevedo et al., 2017) Yet, there is no macro model which analyzes both short run and long run effects of the weather into an unified setup E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 4/27

10. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References This Paper • Assesses the transmission mechanism of a weather shock both at a sectoral and international levels • Measures short run implications of weather shocks for macroeconomic ﬂuctuations • Measures long run implications of weather shocks for macroeconomic volatility and welfare • Uses a model estimated with full-information methods on New Zealand data E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 5/27

14. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References This paper (cont.) To do so: 1. We build a drought indicator 2. We characterize a weather shock through the lens of a VAR model (not presented today) 3. We design a theoretical model that incorporates the response of an economy to a weather shock and estimate it 4. We analyze the propagation of a weather shock and its implications in terms of business cycle statistics (short-run) 5. We measure the implications of climate change on aggregate ﬂuc- tuations by varying the variance of weather shocks (long-run) E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 6/27

20. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Why New Zealand? • Country subject to weather shocks: 2013 – North Island + West of the South Island drought ($1.3 billion, ≈ .5% of GDP) 2008 – national drought ($2.8 billion, ≈ 1.5% of GDP) Source: Dave Young, on Flickr. • Availability of both weather and agriculture data • Small enough country ⇒ fairly homogeneous weather E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 7/27

21. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Take-away results • Our model suggests that weather shocks play an important role in explaining macroeconomic ﬂuctuations • Propagation of a weather shock: A weather shock (drought event) acts as a standard negative supply shock that hits the agricultural sector and spreads to the entire economy • Welfare costs: Today: welfare cost of weather shocks represents 0.40% of households consumption Tomorrow: this cost would reach up 0.59% of household consumption because of climate change E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 8/27

24. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Roadmap 1 Introduction 2 The Model 3 Estimation 4 Business Cycles Analysis 5 Long Run 6 Conclusion E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 9/27

25. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References The Model in Brief • DSGE model in an RBC framework Households Foreign Households Non-agricultural Sector Agricultural sector Weather (droughts) cons. ct cons. c∗ t hours ht land costs xt investment iA t bonds b∗ t Figure 2: The theoretical model. • Small open economy (home vs world) • Two sectors: weather-dependent agricultural sector (original feature) standard non-agricultural sector E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 10/27

28. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Key mechanisms: 1. Farmers face exogenous weather (no rain dance al- lowed) E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 11/27

29. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Key mechanisms: 1. Farmers face exogenous weather (no rain dance al- lowed) 2. They can oﬀset bad weather conditions by pur- chasing goods in the non- agricultural sector E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 11/27

30. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Key mechanisms: 1. Farmers face exogenous weather (no rain dance al- lowed) 2. They can oﬀset bad weather conditions by pur- chasing goods in the non- agricultural sector 3. Leading to spillover effects between the two sectors E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 11/27

31. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References The Weather • We introduce a weather variable in the model: εW t > 1 • It is a drought index based on soil moisture data • It follows an univariate stochastic exogenous process: log(εW t ) = ρW log(εW t−1) + σW ηW t , ηW t ∼ N (0, 1) (1) E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 12/27

34. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Agricultural Sector and the Weather • Each farmer i ∈ [n, 1] has a land endowment it, whose time-varying productivity (or efficiency) writes: it = (1 − δ ) Ω εW t it−1 + xit, (2) where: δ ∈ (0, 1) is the rate of decay of land efficiency xit are intermediate goods to maintain land efficiency (crops, water, fertilizers...) Ω εW t is a weather damage function: Ω εW t = εW t −θ , (3) where θ is the elasticity of land productivity with respect to the weather variations E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 13/27

35. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Agricultural Sector and the Weather (cont.) • For each farm i ∈ [n, 1], the production function is: yA it = εZ t ω it−1 kA it−1 α κAhA it 1−α 1−ω , (4) where εZ t is an economy-wide technology shock it−1 is the land endowment kA it−1 is physical capital hA it is labor demand ω ∈ [0, 1] is the elasticity of output to land α ∈ [0, 1] is the share of physical capital κA > 0 is a technology parameter E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 14/27

36. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Data The model is estimated using Bayesian techniques which combine likelihood estimation with prior information via dynare. • We estimate 6 sequences of shocks • We estimate 21 structural parameters • Parameters related to weather conditions are estimated agnostically with diﬀuse prior. E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 15/27

40. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Data (cont.) 2000 2010 −2 0 2 Production 2000 2010 −10 0 10 Agriculture 2000 2010 −10 0 10 Investment 2000 2010 0 5 Hours 2000 2010 −2 0 2 Foreign Output 2000 2010 −2 −1 0 1 Weather Figure 3: Observable variables used to estimate the DSGE model E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 16/27

41. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Esimation (cont.) 0.2 0.4 0.6 0.8 1 0 2 4 ρZ productivity 0.2 0.4 0.6 0.8 1 0 5 10 ρG spending 0.2 0.4 0.6 0.8 1 0 2 4 6 ρA preferences −0.20 0.20.40.60.8 0 1 2 3 4 ρI investment 0.2 0.4 0.6 0.8 1 0 2 4 6 ρ∗ foreign 0 0.2 0.4 0.6 0.8 0 1 2 3 4 ρW weather 0 1 2 3 4 0 0.5 1 σH labor disutility 0.4 0.6 0.8 1 0 2 4 6 8 b consumption habits 0 2 4 6 8 0 0.2 0.4 0.6 ι labor sectoral cost 0 2 4 0 0.2 0.4 0.6 0.8 ψ land cost convexity 0 2 4 6 8 0 0.2 0.4 0.6 κ investment cost 0 2 4 6 8 10 0 0.2 0.4 µ sectoral subst. 0 1 2 3 4 5 0 1 2 3 4 µN non-farm subst. 0 2 4 6 8 0 0.2 0.4 0.6 µA farm subst. −10 0 10 20 0 2 4 6 8 ·10−2 θ weather elasticity Figure 4: Prior (green) and posterior (blue) distributions of structural parameters for New Zealand (excluding shocks). E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 17/27

42. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Propagation of a Weather Shock 5 10 15 20 −0.3 −0.2 −0.1 0 total output Yt 5 10 15 20 −0.15 −0.1 −0.05 0 consumption Ct 5 10 15 20 −0.4 −0.2 0 investment It 5 10 15 20 −6 −4 −2 0 2 ·10−2 hours Ht 5 10 15 20 −2 0 2 4 6 ·10−2 non-agriculture Y N t 5 10 15 20 −2 −1 0 agriculture Y A t 5 10 15 20 −0.15 −0.1 −0.05 0 relative price pN t 5 10 15 20 0 0.2 0.4 0.6 0.8 relative price pA t 5 10 15 20 0 5 10 land cost xt 5 10 15 20 −10 −5 0 land productivity t 5 10 15 20 0 0.05 0.1 0.15 0.2 exchange rate rert 5 10 15 20 −0.1 −0.05 0 trade balance tbt DSGE mean DSGE 90% VAR mean Figure 5: System response to an estimated weather shock ηW t measured in percentage deviations from the steady state. E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 18/27

43. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Contribution of Weather Shocks on Aggregate Fluctuations Q1 Q2 Q4 Q10 Q40 Q∞ 0% 20% 40% 60% 80% 100% Production Yt Supply shocks (ηZ t ) Demand shocks (ηG t + ηA t + ηI t ) Foreign shocks (η∗ t ) Weather shocks (ηW t ) Q1 Q2 Q4 Q10 Q40 Q∞ 0% 20% 40% 60% 80% 100% Agricultural production Y A t Figure 6: Forecast error variance decomposition at the posterior mean for diﬀerent time horizons (one, two, four, ten, forty and unconditional). E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 19/27

44. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Climate Change and Business Cycles “ A change in the state of the climate that can be identiﬁed (e.g., by using statistical tests) by changes in the mean and/or the variability of its properties, and that persists for an extended period, typically decades or longer IPCC (2014) ”• Climate is supposed to be stationary in our framework: our set-up is irrelevant for analyzing changes in mean climate values • However, it allows for changes in the variance of climate E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 20/27

47. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Climate Change and Business Cycles (cont.) Less extreme cold weather Less cold weather More hot weather More extreme hot weather Figure 7: Increase in mean of weather shocks More extreme cold weather More cold weather More hot weather More extreme hot weather Figure 8: Increase in variance of weather shocks E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 21/27

48. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Climate Change and Business Cycles (cont.) • How much is the variance of the weather expected to vary? • We consider four emission scenarios: 0 10 20 30 1900 1950 2000 2050 2100 (a) 50 55 60 65 2025 2050 2075 2100 (b) 95 100 105 110 115 120 2025 2050 2075 2100 (c) RCP 2.5 RCP 4.5 RCP 6.0 RCP 8.5 Notes: (a) represents historical CO2 emissions as well as their projections up to 2100 under each scenario. The estimation of the standard errors of projected precipitations σW t for each representative concentration pathway is represented in panel (b). Their linear trend from 2013 to 2100 is depicted in panel (c). E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 22/27

49. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Climate Change and Business Cycles (cont.) • These scenarios are translated in our model through an increase in the standard deviation of drought shocks: RCP 2.6: −4.10% RCP 4.5: +6.82% RCP 6.0: +9.30% RCP 8.5: +23.25% E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 23/27

50. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References 1994-2016 2100 (projections) Benchmark RCP 2.6 RCP 4.5 RCP 6.0 RCP 8.5 sd(ηW t ) Weather shock 100 95.90 106.82 109.30 123.25 sd(Yt) GDP 100 99.82 100.15 100.23 100.72 sd(Y A t ) Agriculture 100 96.89 102.54 103.86 111.53 sd(Ct) Consumption 100 99.94 100.05 100.07 100.22 sd(It) Investment 100 99.98 100.01 100.02 100.07 sd(Ht) Hours 100 99.99 100.00 100.01 100.03 sd(Rt) Real interest rate 100 100.00 100.00 100.00 100.00 sd(rert) Exchange rate 100 99.86 100.12 100.18 100.57 sd(pA t ) Agricultural price 100 99.80 100.17 100.26 100.80 sd(cat) Current account 100 99.90 100.09 100.13 100.40 E(Wt) Welfare -158.02 -158.00 -158.04 -158.06 -158.13 λ (%) Welfare cost 0.4023 0.3562 0.4417 0.4623 0.5873 Table 1: Changes in Standard-Errors of Simulated Observables Under Climate Change Scenarios. E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 24/27

51. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Recap • Our model suggests that weather shocks play an important role in explaining macroeconomic ﬂuctuations • Propagation of a weather shock: A weather shock (drought event) acts as a standard negative supply shock that hits the agricultural sector and spreads to the entire economy • Welfare costs: Today: welfare cost of weather shocks represents 0.40% of households consumption Tomorrow: this cost would reach up 0.59% of household consumption because of climate change E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 25/27

54. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Looking Forward • Fat tailed normal distribution of weather shocks to capture higher order eﬀects. • Analyzing food insecurity adapting this setup for developing economies. • Optimal design of policy aiming at stabilizing weather shocks (pigouvian tax, climate insurance, etc.). E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 26/27

55. Introduction The Model Estimation Business Cycles Analysis Long Run Conclusion References Bibliography I Acevedo, S., Mrkaic, M., Novta, N., Poplawski-Ribeiro, M., Pugacheva, E., and Topalova, P. (2017). The Effects of Weather Shocks on Economic Activity: How Can Low-Income Countries Cope? World Economic Outlook, Chapter 3. International Monetary Fund. Buckle, R. A., Kim, K., Kirkham, H., McLellan, N., and Sharma, J. (2007). A structural VAR business cycle model for a volatile small open economy. Economic Modelling, 24(6):990–1017. doi:10.1016/j.econmod.2007.04.003. Burke, M., Hsiang, S. M., and Miguel, E. (2015). Global non-linear effect of temperature on economic production. Nature, 527(7577):235–239. doi:0.1038/nature15725. Dell, M., Jones, B. F., and Olken, B. A. (2012). Temperature shocks and economic growth: Evidence from the last half century. American Economic Journal: Macroeconomics, 4(3):66–95. doi:10.1257/mac.4.3.66. Kamber, G., McDonald, C., Price, G., et al. (2013). Drying out: Investigating the economic effects of drought in New Zealand. Reserve Bank of New Zealand Analytical Note Series No. AN2013/02, Reserve Bank of New Zealand, Wellington, New Zealand. Narasimhan, B. and Srinivasan, R. (2005). Development and evaluation of soil moisture deficit index (SMDI) and evapotranspiration deficit index (ETDI) for agricultural drought monitoring. Agricultural and Forest Meteorology, 133(1):69–88. doi:10.1016/j.agrformet.2005.07.012. Nordhaus, W. D. (1991). To slow or not to slow: the economics of the greenhouse effect. The economic journal, 101(407):920–937. doi:10.2307/2233864. E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 27/27

56. Data Estimation Business Cycles Analysis Measuring the Weather • Challenge: computing at a macro level a weather index strongly correlated with real agricultural production. • Solution: create soil moisture deficit index: 1. Collect soil water deficit data fril weather stations (on a monthly basis). 2. Compute percentage deviation from monthly median as in Narasimhan and Srinivasan (2005): Dt,m = SW Dt,m−Med(SW Dm) Med(SW Dm) and include persistence of deficits, SMDIt,m = 0.5 × SMDIt,m−1 + Dt,m 50 that captures the evatransporation. 3. Aggregation: compute an average per region, get national index by a weighting each region by its relative size in national agricultural production. E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 28/27

57. Data Estimation Business Cycles Analysis Measuring the Weather (cont.) -45 -40 -35 170 175 Auckland Bay of Plenty Canterbury Gisborne Hawke’s Bay Manawatu-Wanganui Marlborough Northland Otago Southland Taranaki Tasman/Nelson Waikato Wellington West Coast Q1 Q2 Q3 Q4 25 50 75 100 mm Fig a: Weather stations in New Zealand Fig b: Seasonality of Weather Historical average 2012Q1 2013Q1 (0,10] (10,30] (30,50] (50,70] (70,90] (90,110] (110,130] (130,150] Fig c: Historical vs Normal times (2012:1) vs Drought (2013:1) E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 29/27

58. Data Estimation Business Cycles Analysis Measuring the Weather (cont.) weather wt Output gap Agricultural Output gap -2 -1 0 1 2 -0.03 -0.02 -0.01 0.00 0.01 0.02 -0.050 -0.025 0.000 0.025 0.050 1990 1995 2000 2005 2010 2015 1990 1995 2000 2005 2010 2015 1990 1995 2000 2005 2010 2015 Figure 9: Weather and output gaps (Shaded Area denotes drought episodes) Go back to the model E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 30/27

59. Data Estimation Business Cycles Analysis Calibration Variable Interpretation Value β Discount factor 0.9883 δK Capital depreciation rate 0.025 α Share of capital in output 0.33 g Share of spending in GDP 0.22 ϕ Share of good in consumption basket 0.15 ¯HN = ¯HA Hours worked 1/3 σC Risk aversion 1.5 ¯ Land per capita 0.40 ω Share of land in agricultural output 0.15 δ Rate of decay of land eﬃciency 0.10 αN Openness of non-agricultural market 0.25 αA Openness of agricultural market 0.45 χB International portfolio cost 0.007 σ∗ C Foreign risk aversion 1.5 b∗ Foreign consumption habits 0.7 Table 2: Calibrated parameters. E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 31/27

60. Data Estimation Business Cycles Analysis Calibration (cont.) Variable Interpretation Model Data ¯C/ ¯Y consumption-to-GDP 0.56 0.57 ¯I/ ¯Y investment to GDP 0.22 0.22 400 × (¯r − 1) real interest rate 4.74 4.75 (1 − ϕ) αN + ϕαA goods market openness 0.28 0.29 n ¯Y A/ ¯Y farming production-to-GDP 0.08 0.07 Table 3: Steady state ratios (empirical ratios are computed using data between 1990 to 2017). E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 32/27

61. Data Estimation Business Cycles Analysis Estimation Prior distributions Posterior distribution Shape Mean Std. Mean [5%:95%] σH Labor disutility B 2 0.75 1.87 [1.32:2.40] b Consumption habits B 0.7 0.10 0.82 [0.74:0.90] ι Labor sectoral cost G 2 1 2.32 [1.36:3.31] κ Investment cost N 4 1.50 1.83 [0.77:2.91] µ Subst. by type of goods G 1.5 0.8 4.93 [3.53:6.26] µN Subst. home/foreign G 1.5 0.8 1.91 [0.86:2.94] µA Subst. home/foreign G 1.5 0.8 0.41 [0.26:0.56] φ Land expenditures cost G 1 0.60 0.76 [0.02:1.51] θ Land-weather elasticity U 0 10 8.62 [2.3:15.78] Marginal log-likelihood -1012.83 Table 4: Prior and posterior distributions of structural parameters. E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 33/27

62. Data Estimation Business Cycles Analysis Do Weather Shocks Matter? No Weather-Driven Weather-Driven Business Cycles Business Cycles M (θ = 0) M (θ = 0) Prior probability 1/2 1/2 Laplace approximation -1016.853 -1012.835 Posterior odds ratio 1.000000 55.626 Posterior model probability 0.018 0.982 E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 34/27

63. Data Estimation Business Cycles Analysis Historical Decomposition 1994 1996 1999 2002 2005 2007 2010 2013 2016 −4 % −2 % 0 % 2 % 4 % GDP log(Yt/ ¯Y ) 1994 1996 1999 2002 2005 2007 2010 2013 2016 −10 % 0 % 10 % Agricultural production log(Y A t / ¯Y A ) Supply shocks (ηZ t ) Demand shocks (ηG t + ηI t + ηA t ) Foreign shocks (η∗ t ) Weather shocks (ηW t ) Residual Variable Path Figure 10: Historical decomposition of aggregate output and agricultural production. E. Gallic & G. Vermandel Weather Shocks, Climate Change and Business Cycles - 35/27

Weather Shocks, Climate Change and Business Cycles

Recommended

Recommended

More Related Content

Similar to Weather Shocks, Climate Change and Business Cycles

Similar to Weather Shocks, Climate Change and Business Cycles (20)

Recently uploaded

Recently uploaded (20)

Weather Shocks, Climate Change and Business Cycles