WPS4162

                                   An Analysis of Crop Choice:

               Adapting to Climate Change in Latin American Farms1



                                                 Niggol Seo
                        University of Aberdeen Business School, UK
                                                     and
                                           Robert Mendelsohn
       School of Forestry and Environmental Studies, Yale University, USA



                                                   Abstract

This paper explores how Latin American farmers adapt to climate by changing crops. We
develop a multinomial choice model of farmer's choice of crops. Estimating the model
across over 2000 farmers in seven countries, we find that both temperature and precipitation
affects the crops that Latin American farmers choose. Farmers choose fruits and vegetables
in warmer locations and wheat and potatoes in cooler locations. Farms in wetter locations
are more likely to grow rice, fruits, and squash and in dryer locations maize and potatoes.
Global warming will cause Latin American farmers to switch away from wheat and potatoes
towards fruits and vegetables. Predictions of the impact of climate change must reflect not
only changes in yields or net revenues per crop but also crop switching.



World Bank Policy Research Working Paper 4162, March 2007


The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the

exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if

the presentations are less than fully polished. The papers carry the names of the authors and should be cited

accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the

authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries

they represent. Policy Research Working Papers are available online at http://econ.worldbank.org.




1 We thank Emilio Ruz, Flavio Avila, Jorge Lozanoff, Luis José María Irias, Magda Aparecida de Lima, César
Teherán, Jorge González, Flavio Játiva, Alfredo Albin, Liubka Valentina Trujillo, Luisa Caraballo Silva, and
Ariel Dinar for their contributions to this effort. This project was funded by the Research Committee of the
World Bank under the study 'Climate Change and Rural Development' that was tasked managed by Ariel Dinar.


                                                       1

1. Introduction


This paper uses cross-sectional evidence to explore how farmers adapt to exogenous

environmental factors such as climate and soils. By comparing choices of farmers who face

different conditions, the model uncovers how farmers adapt. In this paper, we apply this

technique to study how climate affects the crop choice of Latin American farmers. We

quantify which crops farmers are likely to choose and how dependent this choice is on

climate. Understanding adaptation is an important goal in itself to assist planning by policy

makers and private individuals (Smit and Pilifosova 2001). However, understanding

adaptation is also important if one is interested in quantifying the impacts of climate change.

Forecasts of the impact of climate on agriculture cannot rely solely on how climate affects a

specific crop.    The forecasts must also capture crop switching.           Unfortunately, data

limitations make it difficult to study the crop specific impacts of climate change in this paper.

However, independent data on the effect of climate on yields of specific crops could be

combined with the crop switching results of this study to obtain an overall measure of

damages.


        Climate impact studies have consistently predicted extensive impacts to the

agricultural sector from climate change across the globe (Pearce et al. 1996; Tol 2002). A

large set of these studies have focused on the reduction of yields of specific crops in warmer

temperatures (Reilly et al. 1996; McCarthy et al. 2001). Because these studies assume that

farmers make no changes in crops, these studies predict large yield losses from climate

change and therefore large losses in net revenue. Studies that do allow crops to change

(Adams et al. 1999; Mendelsohn et al 1994) predict that farmers will move away from crops

with low yields and substitute new crops that will perform better in the new climate. Studeis

that allow adaptation predict smaller damages. However, empirical analyses of just how

much farmers are likely to switch crops in response to climate are rare in low latitude



                                                2

countries. The only exception is a new study of farmers in Africa (Kurukulasuriya and

Mendelsohn 2006). This paper follows the approach taken in the African paper but explores

the choices of farmers in Latin America.


        The theoretical choice model is developed in the next section. Section 3 discusses

how data were collected from over 2000 farmers in seven countries across Latin America.

Section 4 discusses the estimation procedure and the empirical results. Three climate change

scenarios from Atmospheric Oceanic General Circulation Models (AOGCM's) are then

examined in Section 5. The paper concludes with a summary of results and policy

implications.


2. Theory


In this paper, farmers are assumed to maximize their profits. Farmers choose the desired

species to yield the highest net profit. Hence, the probability that a crop is chosen depends

on the profitability of that crop. We assume that farmer i's profit in choosing crop j (j=1, 2,...,

J) is




ij =Vj(Ki,Si)+ (Ki,Si)                                                                (1)
                     j




where K is a vector of exogenous characteristics of the farm and S is a vector of

characteristics of the farmer. For example, K could include climate, soils, and price variables

and S could include the age of the farmer and family size. The profit function is composed of

two components: the observable component V and an error term, . The error term is

unknown to the researcher, but may be known to the farmer. The farmer will choose the crop

that gives him the highest profit. When farmers select multiple crops, the crop choice is

defined as the single crop with the greatest net revenue. Alternatively, we could have


                                               3

examined all combinations of crops that farmers select.          However, the number of

combinations is large and becomes difficult to model. Given the assumption that only the

most important crop matters, we look at all available choices. The farmer must pick one and

only one of the available crops.

        Defining Z = (K,S), the farmer will choose crop j over all other crops k if:




 (Zi) >k (Zi)fork  j.[orif k(Zi)- (Zi) <Vj(Zi)-Vk(Zi)for k  j]
   *         *                                                                      (2)
 j                                            j




More succinctly, farmer i's problem is:




  argmax[       1 (Zi),2 (Zi),...,J (Zi)]
                   *       *         *                                              (3)
        j




The probability Pji for the jth crop to be chosen is then




Pji = Pr[k (Zi ) -  (Zi ) < Vj -Vk ]  k  j where Vj = Vj(Zi )                       (4)
                    j




Assuming  is independently Gumbel distributed and Vk = Zki k +k ,




Pji =     eZji j
          J   eZkik                                                                 (5)
          k=1




                                              4

which gives the probability that farmer i will choose crop j among J species (McFadden 1973,

Train 2001).


         The parameters can be estimated by the Maximum Likelihood Method, using an

iterative nonlinear optimization technique such as the Newton-Raphson Method. These

estimates are CAN (Consistent and Asymptotically Normal) under standard regularity

conditions. (McFadden 1999)


3. Data


The data this study relies upon came from a World Bank project to study climate change

impacts on agriculture in Latin America. The project collected economic surveys at the

farm level from the following seven countries: Argentina, Brazil, Chile, Colombia, Ecuador,

Uruguay, and Venezuela. The countries were selected to represent the wide range of climate

throughout South America and include representatives from both the Southern Cone and

Andean regions. Districts within each country were selected to provide as much within

country climate variation as possible. The original survey interviewed over 2000 farmers of

which 949 farmers selected one of the crops that we are modeling. Some farmers focused

strictly on livestock, some farmers picked other crops, and some farmers did not reveal which

crops they grew.


         Climate data came from two sources: US Defense Department satellites and weather

station observations. We relied on satellite temperature observations and interpolated

precipitation observations from ground stations (see Mendelsohn et al 2006 for a detailed

explanation). Soil data were obtained from the FAO digital soil map of the world CD ROM.

The soil data were extrapolated to the district level using GIS (Geographical Information

System). The data set reports 26 dominant soil types.




                                               5

4. Empirical results


In this analysis, we focus on the seven most important crops in the region: fruits and

vegetables (31%), maize (24%), wheat (15%), squash (11%), rice (8%), potatoes (7%), and

soybeans (4%). Altogether these seven crops generated about 85 % of the total revenue

from crops.


        In Table 1, we estimate the probability each species is selected using a multinomial

choice model. The choice of fruits and vegetables has been left out of the regression as the

base case. The probability of choosing each crop was assumed to be a function of summer

and winter temperature and summer and winter precipitation. Other explanatory variables

included three soil variables, farmer age, farmer education, household size, prices, and a

dummy variable for a computer. The model is significant according to three tests of global

significance. Most of the individual coefficients are significant. P-values show how

significant each variable is. The positive coefficients imply that the probability of choosing

each crop increases as the corresponding variable increases.


        The coefficient on education is positive and significant for every crop in Tab1e 1.

This effectively implies that lower educated farmers tend to grow fruits and vegetables, the

omitted choice. Fruits and vegetables tend to be grown in more tropical climates. The

association with lower education may simply reflect the fact that farmers in tropical zones are

less educated.    Maize and squash are more likely to be chosen if a farm has lithosols and

luvisols. Potatoes and soybeans are more likely to be chosen if a farm has planasols.

Farms with computers are more likely to choose potatoes, rice and squash. It is not clear

whether this equipment actually enhances the profitability of these crops or whether the

computer is a proxy for a missing variable. Larger farm families are less likely to choose

maize and soybeans. These crops are easily mechanized and so may be selected by farmers

with smaller families. Older farmers are more likely to choose wheat.



                                              6

        Only two of the own prices are significant: maize and wheat. Both coefficients are

positive as expected. Farmers are more likely to choose these crops when their prices are

higher. The remainder of the price effects are cross price terms. When wheat prices are

higher, farmers are more likely to pick potato and soybean. When maize prices are higher,

they are more likely to pick rice but less likely to pick squash. When squash prices are

higher, they are more likely to pick maize, soybean, and wheat. Higher tomato prices are

associated with maize and wheat. These positive cross price terms imply a complementarity

between the two crops in question.


        Maize and soybeans do not have any significant climate coefficients but all the other

crop choices do.     There are many varieties of maize and soybeans so that they can

effectively grow in many climate zones in Latin America. The two crops are in this sense

"generalists". In contrast, the other crops are specialized to grow under certain temperature

or precipitation ranges. Rice for example is temperature sensitive. Potatoes and squash are

precipitation sensitive. Wheat is both temperature and precipitation sensitive. Fruits and

vegetables generally prefer warmer temperatures.


        Figure 1 reveals that the choice of crop varieties in Latin America is temperature

sensitive. The graph describes the relationship between the probability of choosing a crop

and annual mean temperature measured in °C. Note that the mean annual temperature in

Latin America is 18°C. The probability of choosing wheat and potatoes is high in the farms

at the cooler place but much lower in the farms at the warmer place. By contrast, the

probability of choosing fruits and vegetables is low in cool farms but much higher in warmer

farms.    The rest of the crops have hill-shaped relationships with temperature.         The

probability of being selected at first increases as one moves from cool to warm farms and

then it decreases as one moves to even warmer farms. With maize, the peak probability of

being chosen is about 13°C. With rice and soybeans, the peak is closer to 16°C. With squash,



                                               7

the peak is closer to 20°C.


         Figure 2 displays the estimated relationship between the probability of choosing a

crop and annual precipitation measured in mm/mo. The mean annual precipitation in Latin

America is 118 mm/month.         The probability of choosing maize and potatoes declines

precipitously as one moves from dry to wet farms. By contrast, moving from dry to wet

farms increases the probability of selecting fruits, rice, and squash. Wheat and soybeans

exhibit a hill-shaped pattern. They are less likely to be picked in very dry farms, more

likely to be picked in moderately dry farms, and then less likely to be picked in wet farms.

The peak condition for wheat and soybeans is around 70 mm/mo which is well below the

average precipitation in Latin America.


5. Climate scenarios


In this section, we simulate the consequences of climate change using the parameter estimates

in the previous section.     We examine a set of climate change scenarios predicted by

AOGCMs. The climate scenarios reflect the A1 SRES scenarios from the following three

models: the Canadian Climate Center (CCC) scenario (Boer et al. 2000), Centre for Climate

System Research (CCSR) scenario (Emori et al. 1999), and the Parallel Climate Model

(PCM) scenario (Washington et al. 2000). We use country level climate change scenarios in

2020, 2060, and 2100 from each climate scenario. The change in temperature predicted by

each climate model is added to the baseline temperature in each district. The percentage

change in precipitation is multiplied by the baseline precipitation in each district. This gave

us a new climate for every district in Latin America for each scenario.


         Table 2 summarizes the climate scenarios of the three models for the years 2020,

2060, and 2100. The models predict a broad set of scenarios consistent with the range of

outcomes in the most recent IPCC (Intergovernmental Panel on Climate Change) report

(Houghton et al. 2001). In 2100, PCM predicts a 2°C temperature increase in Latin America


                                              8

whereas CCC predicts a 5°C increase. Rainfall predictions are noisier: PCM predicts rainfall

to increase by 8% by 2100 whereas CCC predicts rainfall to decrease by 8%. Examining the

path of climate change over time reveals that temperatures are predicted to increase steadily

until 2100 for all three models but precipitation will vary across time.


        We assume that the cross sectional evidence used in the estimation is appropriate to

predict future changes in long run equilibriums. The parameters from the estimated choice

model in Table 1 are used to simulate the impacts of climate change on the probabilities of

choosing a particular crop for each climate scenario in Table 2.


        Table 3 describes the results. The dryer and hotter CCC and CCSR scenarios predict

that farmers would choose fruits and vegetables more often and maize, potatoes, squash, and

wheat less often by 2020. There is no noticeable effect on rice. With the milder and wetter

PCM scenario, farmers will pick potatoes, rice, and wheat more often in addition to fruits and

vegetables. In all three climate scenarios, the magnitude of the crop changes grow over time

as the climate scenario becomes more severe. For example, the crop switching in 2060 and

2100 is more common.        The more severe is the scenario, the more crop switching is

predicted.


6. Conclusion


This paper uses a multinomial choice model to capture the choice of crops made by farmers.

The model is estimated across almost 1000 farmers in Latin America. We observe that the

choice of species varies with climate. Farms that are cooler are more likely to choose

potatoes and wheat, average temperature farms tend to choose maize, soybeans and rice, and

farms in warm locations choose fruits and vegetables and squash. Farms in dry locations

tend to choose maize and potatoes, farms in moderately dry conditions tend to pick soybeans

and wheat, farms in wet conditions choose fruits and vegetables, squash, and rice. These

cross sectional results suggest that farmers have adjusted crop choice to fit their local climate


                                                9

conditions.


         Although crop switching has not generally been captured by the climate change

impact literature, crop switching is quite consistent with broad observations of where species

are currently located.     Maize is grown from Argentina to Venezuela.             Potatoes are

concentrated in the mountains of Chile and Colombia. Rice is the crop of choice in Ecuador.

Soybeans and squash are concentrated in Uruguay, northern Argentina, and southern Brazil.

Wheat is chosen in cooler parts of Chile. Fruits are the primary choice of hot Brazilian

farms.


         The crop choice model is quite consistent with the response functions from Africa

(Kurukulasuriya and Mendelsohn 2006).          This study also found that maize was grown

across many temperature zones, that wheat favored cool dry regions, and that fruits and

vegetables tended to be chosen in warmer, wetter places.


         We simulate climate change impacts for the three AOGCM scenarios based on the

parameter estimates from the choice model. The dryer and hotter CCC and CCSR scenarios

predict that farmers would choose fruits and vegetables more often and maize, potatoes,

squash, and wheat less often by 2020. There is no noticeable effect on rice. With the milder

and wetter PCM scenario, farmers will pick potatoes, rice, and wheat more often in addition

to fruits and vegetables. These differential effects on crops are magnified over time.


         In interpreting these results, there are several caveats that should be kept in mind.

First, this analysis does not include price effects. Large changes in crop prices may alter the

results. Second, the analysis does not take into account carbon fertilization. If it affects all

crops identically, it may not matter. However, evidence suggests that some crops may

benefit more from carbon fertilization than others. Third, we assume that adaptations can

take place as needed. For example, farmers can switch from one crop to another as

temperature increases and rainfall decreases. However, this may not be the case if the


                                                10

adjustment requires a heavy capital investment.      Fourth, we assume that in forecasting

climate change impacts, the only thing that changes in the future is climate. Many things,

however, will change over the century such as population, technologies, institutional

conditions, and reliance on animal power. Future studies should address these issues and

provide ever more accurate measures of climate change impacts.


        Unfortunately, it was not possible to estimate incomes per crop as a function of

climate because there were not enough observations. However, results from other studies

that predict the yields of crops in different climates could be combined with the crop

switching results in this paper in order to predict the overall economic impacts of climate

change on Latin American farmers.




                                             11

REFERENCES

Adams, Richard, McCarl, Bruce, Segerson, Kathy, Rosenzweig, Cynthia, Bryant, Kelley,

Dixon, Bruce, Conner, Richard, Evenson, Robert, Ojima, Dennis. "The economic effects of
       climate change on US agriculture." in Mendelsohn, Robert and Neumann, James, eds.
       The

Impact of Climate Change on the United States Economy, Cambridge, UK: Cambridge
       University Press, 1999.

Boer, G., G. Flato, and D. Ramsden (2000), "A transient climate change simulation
  withgreenhouse gas and aerosol forcing: projected climate for the 21st century", Climate
  Dynamics 16, 427-450.

Dubin, Jeffrey A., and McFadden, Daniel L. "An Econometric Analysis of Residential
  Electric Appliance Holdings and Consumption", Econometrica, 1984, Vol.52, No.2 , pp.
  345-362.

Emori, S. T. Nozawa, A. Abe-Ouchi, A. Namaguti, and M. Kimoto (1999), "Coupled ocean-
  atmospheric model experiments of future climate change with an explicit representation of
  sulfate aerosol scattering", J. Meteorological Society Japan 77, 1299-1307.

Houghton, John, Yihui, Ding, Griggs, Dave, Noguer, Maria, Van der Linden, Paul,
  Dai,Xiaosu, Maskell, Kathy and Johnson, Cathy, (eds.) Climate Change 2001: The
  Scientific Basis, Intergovernmental Panel on Climate Change: Cambridge University Press,
  2001.

Kurukulasuriya, P. and R. Mendelsohn. 2006. "Crop Selection: Adapting to Climate Change
  in Africa" World Bank Working Paper (forthcoming).

McCarthy, James, Canziani, Osvaldo F., Leary, Neil A., Dokken, David J. and White, Casey,
  (eds.) Climate Change 2001: Impacts, Adaptation, and Vulnerability, Cambridge:
  Cambridge University Press, Intergovernmental Panel on Climate Change, 2001.

McFadden, Daniel L. "Conditional Logit Analysis of Qualitative Choice Behavior", in P.
  Zarembka (ed.), Frontiers in Econometrics, Academic Press, 1973.

McFadden, Daniel L. "Chapter 1. Discrete Response Models", University of California at
  Berkeley, Lecture Note, 1999.

Mendelsohn, R., A. Basist, A. Dinar, F. Kogan, P. Kurukulasuriya and C. Williams. 2006.
  "Climate Analysis with Satellites Versus Weather Station Data" Climatic Change
  (forthcoming).

Pearce, D. et al. 1996. "The Social Costs of Climate Change: Greenhouse Damage and
  Benefits of Control" in Climate Change 1995: Economic and Social Dimensions of


                                             12

  ClimateChange, J. Bruce, H. Lee, E. Haites (eds.) Cambridge Univ. Press, Cambridge,
  UK, pp.179-224.

Smit, Barry and Olga Pilifosova. 2001. "Adaptation to Climate Change in the Context of
  Sustainable Development and Equity." In J.J. McCarthy, O.F. Canzianni, N.A. Leary, D.J.
  Dokken, and K.S. White, eds., Climate Change 2001: Impacts, Adaptation, and
  vulnerability - Contribution of Working Group II to the Third Assessment Report of the
  Intergovernmental Panel on Climate Change. Cambridge, U.K.: Cambridge University
  Press.

Tol, R.S.J. (2002), `New Estimates of the Damage Costs of Climate Change, Part I: Benchmark
  Estimates', Environmental and Resource Economics, 21 (1), 47-73.

Train, K., 2003, Discrete Choice Methods with Simulation, Cambridge, U.K.: Cambridge
  University Press.

Washington, W., et al. (2000), "Parallel Climate Model (PCM): Control and Transient

Scenarios". Climate Dynamics, 16: 755-774.




                                             13

Table 1: Multinomial logit crop selection model


                                     Maize                   Potato


Variable                     Est.       2    P-value  Est.     2      P-value


Intercept                     4.444     1.35   0.246 -23.338    2.10     0.147


Temperature summer            0.025     0.00   0.944   0.130    0.03     0.857


Temperature summer sq        -0.001     0.02   0.890  -0.029    1.45     0.228


Precipitation summer         -0.004     0.42   0.519   0.234  112.40<.0001


Precipitation summer sq       0.000     0.00   0.991  -0.002 1008.65<.0001


Temperature winter           -0.122     0.50   0.480   1.844  15.75<.0001


Temperature winter sq         0.003     0.25   0.620  -0.073  16.78<.0001


Precipitation winter          0.005     0.73   0.393  -0.058    4.50     0.034


Precipitation winter sq       0.000     0.16   0.691   0.000    8.34     0.004


Soil_Lithosols                0.013     2.39   0.122   0.074  17.49<.0001


Soil_Luvisols                -0.021     3.11   0.078   0.039    4.01     0.045


Soil_Planasols                0.007     0.84   0.361   0.024    4.98     0.026


Computer_dummy                0.269     1.37   0.241   0.761    2.49     0.115


Age_head                      0.009     0.56   0.456   0.038    2.22     0.137


Log household size           -0.909     8.40   0.004  -1.215    5.11     0.024


Log education                 0.106     0.21   0.645   1.004    4.32     0.038


maize_pr                      0.460     11.95  0.001   0.688    2.31     0.129




                                             14

wheat_pr      18.724 21.92<.0001  -44.745  5.86  0.016


squash_pr    -26.401 17.74<.0001   79.127  2.02  0.156


mango_pr      -2.015  1.80   0.179 -7.649  0.43  0.513


tomato_pr      4.290  4.92   0.027 -2.876  0.43  0.511




                          15

Table 1: continued


                                 Rice                Soybean


Variable                 Est.     2    P-value Est.    2     P-value


Intercept                -11.823   0.49  0.486 -6.536   0.52    0.471


Temperature summer         4.046   4.29  0.038  0.528   0.37    0.543


Temperature summer sq     -0.151   4.79  0.029 -0.010   0.18    0.670


Precipitation summer       0.045   5.63  0.018  0.002   0.09    0.762


Precipitation summer sq    0.000   6.95  0.008  0.000   0.16    0.691


Temperature winter        -1.380   3.85  0.050  0.088   0.09    0.760


Temperature winter sq      0.078   6.89  0.009 -0.013   1.92    0.165


Precipitation winter       0.097   2.56  0.110  0.052   3.06    0.081


Precipitation winter sq    0.000   1.72  0.190 -0.001   4.16    0.041


Soil_Lithosols             0.000   0.00  0.996  0.006   0.32    0.573


Soil_Luvisols              0.031   1.74  0.187 -0.015   0.96    0.328


Soil_Planasols            -0.052   0.00  0.996  0.026 10.67     0.001


Computer_dummy             0.656   2.19  0.139 -0.108   0.19    0.667


Age_head                   0.004   0.03  0.867  0.017   0.95    0.330


Log household size        -0.937   1.95  0.163 -0.998   5.39    0.020


Log education             -0.030   0.00  0.953  1.085   7.21    0.007


maize_pr                   1.563 15.83<.0001    0.099   0.08    0.771




                                       16

wheat_pr      27.751   2.87  0.090 16.910   5.96  0.015


squash_pr    -113.900  2.09  0.148 -20.114  6.78  0.009


mango_pr      -11.597  2.88  0.090  0.391   0.01  0.911


tomato_pr     10.581   2.43  0.119  -2.765  0.50  0.478




                           17

Table 1: continued


                               Squash                Wheat


Variable                 Est.    2    P-value Est.   2     P-value




Intercept                 7.774   0.66  0.418  5.292  1.07    0.301


Temperature summer       -1.255   1.86  0.173 -0.091  0.04    0.846


Temperature summer sq     0.036   2.30  0.130  0.006  0.22    0.642


Precipitation summer      0.044   7.19  0.007  0.009  0.60    0.438


Precipitation summer sq   0.000   4.85  0.028  0.000  5.66    0.017


Temperature winter       -0.149   0.23  0.630 -0.561  3.96    0.047


Temperature winter sq    -0.002   0.06  0.809  0.004  0.13    0.714


Precipitation winter      0.014   1.18  0.278 -0.005  0.14    0.708


Precipitation winter sq   0.000   0.44  0.506  0.000  4.77    0.029


Soil_Lithosols            0.011   1.40  0.237 -0.015  1.44    0.231


Soil_Luvisols            -0.152   0.44  0.506 -0.015  1.00    0.316


Soil_Planasols           -0.005   0.20  0.652  0.031 14.36    0.000


Computer_dummy           -0.066   0.09  0.759  0.057  0.05    0.828


Age_head                  0.005   0.10  0.752  0.035  4.12    0.042


Log household size       -0.031   0.01  0.940 -0.841  3.78    0.052


Log education             0.629   4.83  0.028  1.512 15.03    0.000




                                      18

maize_pr                     -2.459     3.76   0.053   0.104    0.15      0.703


wheat_pr                      4.026     0.53   0.468 33.562    23.18<.0001


squash_pr                    -7.197     1.13   0.288 -27.264    9.51      0.002


mango_pr                     -1.387     4.92   0.027  -2.009    0.64      0.425


tomato_pr                     0.451     0.02   0.882 -19.580    6.29      0.012




Fruits and vegetables are the omitted choice. Likelihood ratio test: P<0.0001, Lagrange

multiplier test: P<0.0001, Wald test: P<0.0001.




                                             19

Table 2: Marginal Effect of Climate Change on Crop Choice in Latin America


          Maize       Potato     Rice       Soybean   Squash      Wheat    Fruits


Baseline      19.5%        6.8%      4.8%       7.9%       8.0%     14.4%    38.6%


Temp          -0.2%        0.5%      0.4%       0.2%       0.7%     -2.3%      0.8%


Prec          -0.3%        0.2%      0.1%       0.0%       0.1%     -0.1%     -0.2%




                                           20

Table 3: Latin American Average AOGCM Climate Scenarios


                        Current      2020          2060         2100




Temperature (°C )


         CCC             18.1     19.5 (+1.4)   20.8 (+2.7)   23.2 (+5.1)


        CCSR             18.1     19.4 (+1.3)   20.4 (+2.2)   21.3 (+3.2)


         PCM             18.1     18.7 (+0.6)   19.5 (+1.3)   20.1 (+2.0)


Rainfall (mm/month)


         CCC              119     116 (-2.6%)   107 (-9.5%)   109 (-7.7%)


        CCSR              119     120 (+1.5%)   119 (0.0%)    114 (-3.8%)


         PCM              119     128 (+8.2%)  133 (+11.9%   129 (+8.4%)




                                         21

Table 4: Effect of Climate Change Scenario on Crop Choice in Latin America


           Maize      Potato     Rice       Soybean   Squash       Wheat   Fruits


Baseline       19.5%       6.8%      4.8%       7.9%       8.0%     14.4%    38.6%


                                         2020


CCC            -0.7%       -1.4%     1.3%       -0.5%      1.9%      -0.7%     0.2%


CCSR           -1.5%       -1.8%     1.7%       -0.4%      2.2%      -0.2%     0.0%


PCM             1.9%       4.9%      0.0%       -1.2%      -2.2%     -4.9%     1.4%


                                         2060


CCC            -1.2%       -0.8%     0.3%       -1.1%      4.3%      -2.2%     0.7%


CCSR           -0.3%       -2.0%     0.3%       0.3%       3.2%      -3.1%     1.6%


PCM             2.2%       3.8%      0.9%       -1.1%      -1.7%     -6.3%     2.1%


                                         2100


CCC            -3.3%       2.2%      1.1%       -3.3%      9.7%      -5.0%    -1.3%


CCSR           -0.7%       -2.5%     0.1%       -0.6%      5.5%      -3.0%     1.1%


PCM             3.1%       2.7%     -0.1%       -1.2%      -1.3%     -6.5%     3.2%




                                            22

Figure 1: The effect of annual temperature on the probability of choosing a crop.

PROB
1.0



0.9



0.8



0.7



0.6



0.5



0.4



0.3



0.2



0.1



0.0

    0                            10                           20                     30

                                     ANNUAL MEANTEMPERATURE

              PLOT         mai ze        potato         ri ce           soybean
                           squash        wheat          Frui ts




                                             23

Figure 2: The effect of annual precipitation on the probability of choosing a crop


PROB
 1.0



 0.9



 0.8



 0.7



 0.6



 0.5



 0.4



 0.3



 0.2



 0.1



 0.0

     0                            100                             200                 300

                                      ANNUAL MEAN PRECIPITATION

              PLOT         maize           potato           rice         soybean
                           squash          wheat            Fruits




                                                 24