SI Appendix A: Do declines in cultivated areas for one crop precipitate increases in

cultivated areas for other crops?

The significance of our findings hinges on the degree to which declines in cultivated

areas in one crop lead either to land saving or to an increase in land used for another crop. As

farmers, for example, decrease the number of fields in which they cultivate soy, do they increase

the amount of land that they devote to barley? If so, then gains in productivity may reduce

acreage in a particular crop, but acreage in other crops will climb. Under these circumstances he

intensification – cultivated area decline dynamic would produce little land saving and

investigations of the causes for intensification – cultivated area changes based on some but not

all crops, like this study, would not be indicative of patterns in total crop area. For this reason, it

is important to try and ascertain, to the extent possible given data limitations, the degree to which

farmers shift between crops in large scale, asymmetrical ways. If they do make these shifts

frequently from crops in our analysis to crops outside our analysis, it would undermine the

validity of our work.

Three types of evidence allow us to estimate the extent of large scale, asymmetrical crop

shifting. First, using FAO agro-ecological zone data for crops (1), we can calculate the degree of

overlap between zones that are suitable for the ten crops in our analysis. A correlation matrix

from these analyses reveals three agro-ecological clusters among the ten crops: wheat – potatoes,

corn-soy-cotton, and sugarcane-coffee-bananas-cocoa. For example, we might expect to see

farmers shift from growing cotton to growing soy because the extent of overlap in suitable zones

is very high (.97), but we would not expect to see a farmer shift from wheat to rice because the

extent of the overlap in suitable zones is quite low (.15). Seventy-four percent of the pairs of

1
crops in our analysis (coffee and cocoa excluded because of lack of data) showed less than 70%

overlap. Because this analysis includes even minimally suitable sites for crops, it probably

overestimates the possibilities for crop shifting. Farmers use much more fine grained criteria in

their decision making about crops for particular fields than we have used here. In sum this

analysis demonstrates that farmers’ possibilities for crop shifting are substantially limited by

agro-ecological conditions.

Second, the large scale, asymmetrical shifts from one crop to another crop could,

depending on the crops involved, introduce error into our analyses. These crop to crop shifts in

response to changing yields should, if they occur, leave markers in the aggregate data on crops.

In particular we would expect to see inverse correlations between the two crops. To be sure,

farmers will shift back and forth between crops, but it is only large scale, asymmetrical shifts

from one crop to another that would threaten the validity of our analyses. To ascertain the

frequency of these inverse associations, we cross-correlated changes in cultivated areas for all

ten crops plus pasture for the 1990 to 2005 period (Table A1). Of the 55 correlations in the

table, only one, between pasture and sugarcane, exhibits the inverse association we would expect

to see if declines in the cultivated areas of one crop are precipitating increases in the cultivated

area of another crop.

Of course the declines in the cultivated areas of one crop could produce a less

conspicuous pattern of expansion in other crops in which the increases in cultivated areas spread

across several other crops. If this pattern prevails, we would expect that the decline in the

cultivated area for one crop would not extend to the cultivated areas of other crops in the

country. There should, in fact, be a disjunction between the trends in the one crop and the

national trends in cultivated areas if asymmetrical crop shifting occurs frequently between that

2
crop and other crops. Table A2 correlates the changes in cultivated areas for eleven crops (the

ten crops plus pasture) with changes in the total agricultural land (pastures plus cropland) in

nations. Seven of the eleven crops show significant (.10 > p.) positive associations between

trends in the cultivated areas for the specific crop under study and trends in cultivated areas for

all crops in a nation. We would not expect to see these positive associations between trends for

individual crops and the aggregated cultivated areas for nations if the declines in cultivated areas

of one crop were accompanied by increases in the cultivated areas of other crops. The prevalent

pattern is that the changes in the cultivated areas for one crop seem to get repeated for other

crops in a nation. Given that the cultivators are subject to a common set of political economic

conditions, this finding should not seem surprising.

The preponderance of the evidence from these three analyses suggest that farmer

decisions to curtail the cultivation of some crops did not, at least for the 1990-2005 period,

induce a corresponding increase in the cultivation of other crops outside our analysis on the same

lands. In this context it seems more likely that declines in cultivated land could have led to land

sparing. By extension it seems worthwhile to pursue analyses of agricultural intensification and

changes in cultivated areas.

References

1. Food and Agriculture Organization, Global Agro-Ecological Zones 2000, FAO. Rome. At:

http://www.fao.org/ag/AGL/agll/ gaez/index.htm.

3
 Table A1: Associations between Changes in Cultivated Areas for Crops across Nations, 1990-
2005
wheat pasture cotto sugar rice potato coffee corn soy banana cocoa
pasture -.026
.822
cotton .044 .002
.740 .984
sugar- .025 -.248* .021
cane .868 .024 .867
rice -.018 -.019 .033 .049
.886 .856 .781 .671
potato -.012 -.021 -.019 .217* .210*
.902 .830 .876 .069 .052
coffee .042 .374* .055 .121 -.111 -.070
.824 .002 .719 .353 .397 .621
corn .056 -.049 .035 .044 .047 -.022 .221*
.605 .614 .758 .693 .644 .815 .077
soy -.006 .194 .261* .173 .264* -.034 -.100 .056
.962 .118 .048 .214 .031 .771 .533 .625
banana -.060 .014 .561* .009 .159 .001 -.012 .197* .115
.684 .895 .000 .941 .172 .991 .924 .071 .430
cocoa .394 -.055 .271 .043 .247 .106 -.102 .426* -.020 .032
.106 .700 .140 .784 .102 .540 .486 .003 .918 .828
Notes: the top number is the Pearson correlation coefficient; the bottom number is the p value.
*p<.10.

Table A2: Changes in Areas in Specific Crops and in all Agriculture Land, 1990-2005, across Nations

Chnge Chnge Chnge Chnge Chnge Chnge Chnge Chnge Chnge Chnge Chnge
in in in in in in in in in in in
Past. Wheat Cotton Sugarca Rice Potato Coffee Corn Soy Banana Cocoa

Chnge
in all .262* -.035 .242* .181* .207* .231* .335* -.037 .193* -.024 .065
Ag. .002 .732 .028 .096 .037 .009 .006 .678 .087 .823 .649
Lands

Notes: the top number is the Pearson correlation coefficient; the bottom number is the p
value. *p<.10.

4
Appendix B: Countries by Region:

Sub‐Saharan Africa: Angola, Benin, Burundi, Cameroon, Cape Verde, Congo ‐ Dem Rep., Congo ‐ Rep.,

Cote D'Ivoire, Djibouti, Ethiopia, Gabon, Gambia, Ghana, Guinea‐Bissau, Guinea, Kenya, Liberia,

Madagascar, Malawi, Mali, Mauritania, Mauritius, Mozambique, Niger, Nigeria, Rwanda, Sao Tome and

Principe, Senegal, Seychelles, South Africa, Sudan, Tanzania, Togo, Uganda, Zambia, Zimbabwe.

Near East and North Africa: Algeria, Egypt, Iran, Israel, Jordan, Lebanon, Libya, Morocco, Saudi Arabia,

Syria, Tunisia, Turkey. UAE, Yemen

Europe: Albania, Austria, Belarus, Belgium, Luxembourg, Bosnia and Herzegovina, Bulgaria, Croatia,

Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland,

Italy, Latvia, Lithuania, Macedonia, Malta, Moldova, Netherlands, Norway, Poland, Portugal, Romania,

Serbia and Montenegro, Slovakia, Slovenia, Spain, Sweden, Switzerland, United Kingdom

East and South Asia: Bangladesh, Brunei‐Darussalam, Cambodia, India, Indonesia

China, Comoros, Japan, Korea ‐ Dem. Rep., Korea‐Rep., Laos, Malaysia, Maldives, Myanmar, Nepal,

Pakistan, Philippines, Sri Lanka, Thailand, Timor‐Leste, Viet Nam

Central Asia: Armenia, Azerbaijan, Georgia, Kazakhstan, Kyrgyzstan, Mongolia, Russian Federation,

Turkmenistan, Uzbekistan

Oceania: Australia, Fiji, French Polynesia, Kiribati, New Zealand, Samoa, Solomon Islands, Vanuatu,

Anglo‐America: Bermuda, Canada, United States

Middle America: Antigua and Barbuda, Bahamas, Barbados, Belize, Costa Rica, Cuba, Dominica,

Dominican Republic, El Salvador, Grenada, Guatemala, Haiti, Honduras, Jamaica, Mexico, Nicaragua,

Panama, Saint Lucia, Trinidad & Tobago,

5
South America: Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador, Guyana, Paraguay, Peru,

Suriname, Uruguay, Venezuela

6
 Appendix C: the Land Sparing Effects of Intensification: Determinants by Crop (Source: FAOSTAT)
All Corn Rice Soy Wheat Potatoes Bananas Cocoa Coffee Sugarcane Cotton

Cons.Res. .09** -.4*** .10* .21**

Program (.03) (.10) (.05) (.06)

Production, -66** .18*

1990 (18) (.08)

Irrigated .02 -.07 -.05

Land, 1990 (.02) (.07) (.04)

Cereal Imp. .02* .17*** .86* .13***

.01 (.03) (.41) (.02)

Prop. Land 43** 289* .73 -292** 1553*** 2360***

Cultivated (14) (130) (.54) (85) (196) (578)

GDP, 1990 .12* -.23** -.33*

(.06) (.07) (.14)

Prop. Econ. 42** -1.9*** -.70*

Active Ag (.12) (.15) (.29)

SubSaharan .15*
Africa (.06)

South -.58*** .27** .99*** -.16* .78***

America (.12) (.13) (.27) (.09) (.17)

Cntl. Asia -.36***

(.08)

Adj r2 .020 .172 .099 .468 .525 .693 .066 .480 .368 .044 .203
N of Cases 927 130 102 81 101 127 85 45 63 78 80
Appendix D: Countries with Conservation Set Aside Programs during the 1990s: A list
with information sources.

Countries

Austria Italy

Belgium Luxembourg

Canada Netherlands

China Portugal

Denmark Spain

Finland Sweden

France Switzerland

Germany United Kingdom

Greece United States

Ireland

Sources: For Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy,

Luxembourg, Netherlands, Portugal, Spain, Sweden, United Kingdom (Common Agricultural

Policy at a glance – at: http://www.epha.org/a/495); for Canada, in some provinces (Agriculture

Canada – at http://www4.agr.gc.ca/AAFC-AAC/display-

afficher.do?id=1204137480722&lang=eng#e); for China (Xu et al., 2006, World Dev. 34:130-

48); for Switzerland (Muller, Grether, 2001, Long run effects of the common agricultural policy

for Switzerland – at: http://www.unige.ch/ses/metri/mueller/cahier0201.pdf); for United States

(Osborn, 1997, Wheat and the Conservation Reserve Program, USDA, Washington, DC).