Food Versus Fuel: Extractive Industries, Insecure Land Tenure, and Gaps in World Food Production

World Development Vol. 51, pp. 62–70, 2013 Ó 2013 Elsevier Ltd. All rights reserved. 0305-750X/$ - see front matter www.elsevier.com/locate/worlddev

http://dx.doi.org/10.1016/j.worlddev.2013.05.015

Food Versus Fuel: Extractive Industries, Insecure Land Tenure, and Gaps in World Food Production THOMAS K. RUDEL * Rutgers University, New Brunswick, NJ, USA Summary. — Corporations now go “to the ends of the earth” to extract natural resources like oil and diamonds from the earth at the same time that farmers, investors, and development experts try to expand the supply of food, sometimes through large land acquisitions in remote regions. These two processes of globalization interact in important ways. Cross-national analyses indicate that oil and mineral dependent nations with neo-patrimonial elites have lower than expected areas under cultivation and yields from cereal crops. Booms in extractive sectors and neo-patrimonial practices in governance have debilitated agricultural enterprises within nations and conceivably throughout the globe. Ó 2013 Elsevier Ltd. All rights reserved. Key words — Dutch disease, agriculture, yield gaps, land tenure, world food supplies, neo-patrimonial ethos

1. INTRODUCTION

act? In particular how do agricultural producers and markets change as revenues from extractive industries transform the larger political economic context in which they work? For two decades economists have argued that farmers leave agriculture when oil, gas, and mineral booms transform the national economies of poor producer states (Corden & Neary, 1982). In a comparative study of eight tropical countries Wunder (2004) demonstrated that most oil producing countries experienced relatively little tropical deforestation during the latter part of the 20th century because the agricultural sectors of their economies shrank. This paper extends Wunder’s analysis, using a modified version of the “Dutch disease” theoretical apparatus, supplemented by neo-patrimonial theories of the state, to explain in part the global agricultural stagnation that has contributed to the recent, worldwide surge in the prices of foodstuffs. A new Geographic Information System (GIS) has simplified this analytic task by making it possible to assess the social and economic causes for variations in agricultural productivity while controlling for variations in the soils, climate, and topography of cultivated areas. The Food and Agricultural Organization of the United Nations (FAO), working with the International Institute for Applied Systems Analysis (IIASA), created the GAEZ (Global Agro-Ecological Zone) GIS. It can identify places where, controlling for the extent of lands unsuitable for agriculture, cultivated areas seem small and places where, controlling for the constraints of soil and climate, large gaps exist between actual and potential crop yields (IIASA, 2011). Understanding the social and economic conditions that have contributed to agricultural underproduction in these places would be an important preliminary step in crafting a strategy for expanding or intensifying agricultural production to meet growing human demands for food. The Dutch disease argument, outlined above and supplemented

Seismic shifts in world food markets and environmental policymaking since 2000 have intensified competition between people who would use lands in rural areas of the developing world differently. 1 After nearly a century of declines in the prices of agricultural products relative to manufactured goods, these trends reversed during 2000–05. Global food prices began to rise faster than the prices for other goods and services. During 2003–10 global food prices increased by almost 50% in real terms at the same time as the volatility of food prices increased significantly (OECD, 2011). These price trends prompted executives in corporations and officials in sovereign wealth funds to begin buying up large tracks of land in poor nations of the developing world in order to produce foodstuffs for consumers in more affluent countries. 2 In this context the projected increases in the world population to approximately 9 billion people in 2050, coupled with recurrent crop failures and increasing numbers of people adopting protein rich diets, has raised alarming questions about how the world’s farmers might meet the additional demands for foodstuffs. Where are the additional lands that might be brought into cultivation? What are the possibilities for further agricultural intensification in different parts of the world? How do societies cope with the now more pressing questions of food sovereignty and food security? Under these circumstances it seems important to inquire about possible interaction effects between global scale efforts to feed human populations and fuel human activities. This question about agriculture occurs in a global context marked by ever more insistent searches by multi-national corporations for new deposits of valuable natural resources like oil, natural gas, and metal bearing ores from mines. 3 Because these corporations, historically headquartered in either Europe or North America, explored areas in their own countries first and only later regions in the developing world, their focus for oil, gas, and mineral exploration has increasingly turned to the same poor countries that, with the world’s largest gaps between actual and potential yields, might be expected to contribute the most to future increases in global food production (Tilman, Balzer, Hill, & Befort, 2011). How, then, do these efforts inter-

* This paper was originally presented at a workshop on the “Global Land Reserve: Where is it? What are the constraints?: a Hectare by Hectare Approach” in Rome, Italy, October, 2011. I would like to thank Eric Lambin, the Academia Belgica, and the Fundacion Franqui for providing the occasion for preparing this paper. Final revision accepted: May 9, 2013. 62

FOOD VERSUS FUEL: EXTRACTIVE INDUSTRIES, INSECURE LAND TENURE, AND GAPS IN WORLD FOOD PRODUCTION 63

by a theory of new states, offers one possible explanation for these variations in agricultural production. The article begins with a theoretical discussion that revises Dutch disease theory and pairs it with neo-patrimonial understandings of public expenditures in order to explain how the growth of oil and mineral sectors has affected agriculture in the developing world. This discussion links extractive activities with government expenditures, labor migration, and rural land tenure. Brief descriptions of the quantitative methods and results follow. A subsequent discussion specifies the link between insecure land tenure and agricultural production. The paper ends with a consideration of what the depletion of oil and mineral deposits might mean for agriculture in countries with depleted deposits. 2. DUTCH DISEASE THEORY, DISAPPEARING MARKETS, AND INSECURE LAND TENURE “Dutch disease” occurs when a boom in one sector of an economy cripples enterprises in other sectors of the economy. To understand how a boom in the extractive sector of an economy could transform other economic sectors, it is important to remember that these markets are embedded in larger, evolving social structures (Granovetter, 1985, 1992). Most of the new extractive sectors and associated booms have occurred in the developing world during the past 50 years, in part because, before World War II, prospectors had not looked for deposits in these areas as thoroughly as they had in the more developed countries. For this reason prospectors have made most of their new discoveries in fledgling states in which a neo-patrimonial ethos of enriching oneself at the public’s expense has prevailed among officials. 4 During the last three decades of the 20th century this political culture gradually eroded the capacity of states to deliver services, like the titling of lands, in rural areas (Van de Walle, 2001). The dynamics of Dutch disease would reinforce the decline in the capacity of states in rural areas. This theory, so named because the first documented episode of it occurred in the Netherlands (Corden & Neary, 1982), describes a dynamic in labor markets that appears shortly after the discovery of valuable oil or mineral deposits touches off an economic boom in that sector of a nation’s economy. Disparities in wage rates grow, with workers in the boom sector earning much higher wages than workers in other sectors. The differences in wage rates, predictably, prompt flows of workers out of the nonboom sectors and into the boom sectors. Because most of the newly discovered deposits of oil and minerals occurred in the developing world where most workers still labored in agricultural sectors, Dutch disease should induce flows of workers out of agriculture and into booming oil or mining sectors. The rapid loss of workers would create problems for farmers. If their operations were labor intensive, the loss of workers should have forced some farmers to abandon agriculture. Currency fluctuations triggered by the boom in the oil-mineral sector would exacerbate the destructive impact of the boom on the agricultural sector. The influx of dollars and euros to purchase oil or other minerals would strengthen the local currency and make it cheaper to import goods from overseas. These goods usually would include inexpensive cereals produced overseas with government subsidies. In urban food markets the imported foodstuffs would take market share away from local farmers. So the price of labor for local farmers rises about the same time that they face more competition

in urban food markets. If farmers produce for overseas markets, the appreciating currency makes their produce more expensive relative to produce from economies with weaker currencies. These intersecting trends would compel some farmers to close down their operations. They would find more lucrative nonfarm opportunities in which to invest (Fuglie, 2008). Despite the logic apparent in the market dynamics outlined above, Dutch disease does not appear to apply to most developing country economies because it assumes full employment labor markets, not the surplus labor market conditions that typically prevail in the developing world (Ross, 2012). In a surplus labor market, the extra demand for labor from the boom sector might have little effect on the price of labor in other sectors because the small number of jobs created by a capital intensive industry like mining or oil extraction would not change the price of labor in agriculture (Pegg, 2009; Ross, 2012). An alternative pathway for a Dutch disease dynamic might occur through government expenditures (Dunning, 2008; Pegg, 2009). Government expenditures typically represent an unusually large proportion of all economic activity in oil and mineral dependent economies (Ross, 2012). Under these circumstances oil and mineral booms alter labor dynamics, but they do so largely through their effect on government expenditures in the urban areas that house government offices. The urban focus of the boom stems from the surge in revenues that the central government receives from royalties or taxes assessed on the extracted resources (Wunder, 2004). 5 While a few oil or mineral enriched governments like Indonesia focused their increased expenditures on rural areas (Timmer, 2007), most oil or mineral dependent governments concentrated their increased expenditures on urban areas. With increased revenues, a continuing neo-patrimonial ethos, and an urban headquarters, government agencies have typically expanded the number of employees, built more office buildings, and indirectly, through their growth in staff, spurred the construction of residential real estate for government workers. Service sector activities expanded in part because service providers did not suffer from the increases in foreign competition that afflicted farmers whose products were traded on international markets. These tendencies led to a boom in the urban real estate market and a surge in migrants from rural areas who took jobs in the construction and service sectors (Wunder, 2004). Agricultural enterprises lost workers, but to a different economic sector than envisaged in the original formulation of Dutch disease theory. Economists have examined this dynamic extensively, usually in an effort to assess or explain the “resource curse,” the often commented upon tendency for populations in nations rich in natural resources to remain poor despite the wealth of natural resources around them (van der Ploeg, 2011). Political scientists have used the Dutch disease formulation to investigate the impact of oil or mineral extraction on the transition to democracy (Dunning, 2008). This body of work leaves largely unanswered important questions about the agro-ecological impacts of Dutch disease. I offer several preliminary answers to these questions below. People create markets and then, when conditions change, they dismantle them (Granovetter, 1985, 1992). These dynamics are very evident when oil booms turn to busts. Ex-farm workers and their children may want to return to agriculture, only to find that the farms, markets, and jobs in agriculture no longer exist. Other circumstances make it difficult to enter or re-enter agriculture. There are few other farmers around, so the opportunities for experiential learning are not there. With

64

WORLD DEVELOPMENT

the declines in numbers of farms, they frequently become scattered physically, so learning by observation is no longer possible. The intergenerational transfer of knowledge that socialized many younger farmers into the occupation did not occur after people left the land, so would be farmers in the younger generation do not know how to farm. When Granges and other farmer organizations disbanded, another way of teaching farming disappeared. Few governments offer extension programs for beginning farmers (Brislen, 2012; Fake, Filan, & Ingram, 2012). Middlemen, who purchase crops in rural areas and make markets, no longer exist, so it is often not clear who would purchase the crops of fledgling farmers. Furthermore, in some instances oil extraction damages nearby croplands through spills, as for example in the delta of the Niger in Nigeria (Okonta & Douglas, 2001). These damages make it impossible for local residents to return to agriculture. Trends in the political arenas also disadvantage small farmers. First, with the boom in the oil and mineral sectors, agriculture becomes a less important source of receipts for the central government. The accompanying change in political priorities reduces government expenditures for agricultural research and development. During 1980–2000, a period of rapidly growing mineral extraction in sub-Saharan Africa, agricultural research and development expenditures in the region declined from 7.8% to 6.3% of global expenditures (Pardey, Beintema, Dehmer, & Wood, 2006). Fiscal pressures stemming from adverse price trends for agricultural products and from the siphoning off of state funds by neo-patrimonial oriented elites further reduced agricultural assistance to smallholders (Van de Walle, 2001). These adverse trends in the agricultural sector reinforced a pre-existing dualism in the land tenure system. Only wealthy landowners near cities had secure land tenure. Everyone else, including all farmers in remote rural areas, had insecure land tenure (Alston, Libecap, & Mueller, 1999). With states having reduced services to rural areas, smallholders rarely, if ever, acquired titles to the land that they worked. Insecure land tenure would make farmers reluctant to expand the cultivated area or intensify agriculture. This dualistic system originated in the colonial era when colonial officials concentrated in coastal cities governed subjects who lived in small villages in the interior. At independence the heads of the new states asserted that all lands belonged to the state. In almost all of these instances the states became passive owners of the lands. State officials exerted no practical control over land or land uses in remote rural areas (USAID, 2007). These dualistic systems of land tenure became, if anything, more prevalent during the 1970s and 1980s when governments adopted neo-liberal land tenuring practices. During the 1960s and 1970s governments had subsidized the acquisition of titles to land by paying the salaries and traveling expenses of the teams of surveyors who measured the farmers’ boundaries. The imposition of neoliberal political agendas in the aftermath of the debt crises in the early 1980s changed these titling procedures. Citing fiscal austerity, governments in all but a small number of countries (El Salvador, Brazil) eliminated new land settlement schemes, shut down agrarian reform programs, and eliminated subsidies for land titling. This last policy change made prospective landowners responsible for the costs, up to $30 per hectare, of the land surveying and titling (Holland et al., 2013). These neo-liberal reforms reduced the incentives of smallholders, already struggling with the loss of labor, to acquire titles to their lands. Only the holders of valuable lands close to cities or large tracts of land would find it to be in their financial interest to pay the costs of obtaining secure titles (Alston et al., 1999).

The land tenure system in Sierra Leone, described below in a recent USAID report (USAID, 2007), typifies the institutional pattern common to many tropical countries in Africa. “Community based tenure prevails in most of the country, save for the coastal regions and the area around Freetown (the capital), where land is largely held in freehold tenure. Land that is not held privately is owned by the state. In customary tenure areas of the country . . . most customary land is controlled and managed by families of the original settlers of a given village . . . These founding families grant land to new families.” Because customary land tenure arrangements rarely have legal force, smallholders who rely upon it become vulnerable to legal challenges over their ownership of land from outside organizations like mining companies or corporate groups engaged in land grabs. A sudden flood of oil or mining revenues into this context would encourage corruption because office holders no longer have to rely on their taxpaying constituents for revenues, so they no longer feel an obligation to deliver government services effectively. In this manner one possible restraint on venal behavior would disappear with the arrival of oil revenues (Petermann, Guzman, & Tilton, 2007; Wunder, 2004). 6 In a polity rife with corruption, improved lands could attract the attention of politically powerful persons with an interest in wresting control of the lands from their original owner, a sequence of events recounted in the popular Indonesian parable of the “Little Man with the Big Stone” (Dove, 1993). People with valuable resources, a large diamond in the parable, attract the unwanted attention of powerful figures who, through the favoritism of public authorities, take the resource away from the “little man.” By implication people without secure land tenure in corrupt polities take a considerable risk when they improve the productivity of their land in expensive, attention attracting ways (Lipton, 2009). For these reasons farmers in societies with a high perceived incidence of corruption may show some reluctance to invest in land improvements that would raise crop yields. The retreat of the neo-liberal state from rural areas worked in concert with Dutch disease induced effects to disadvantage small farmers. This urban oriented social structure did not preclude agricultural pursuits, but it favored part-time, peri-urban farmers who focus on foodstuffs like fruits and vegetables that cannot be readily imported. Some of these farmers circulate between a small farm and an urban occupation, following “roads of many returns” (Waters, 1997). For many of them, farming is part of a diverse mix of livelihoods that households pursue to make economic ends meet (Collins, Morduch, Rutherford, & Ruthven, 2011; Lerner & Eakin, 2011). The part-time character of these agricultural pursuits would blunt the full force of the Dutch disease effects and explain why this type of agriculture persists in agriculturally disadvantaged settings. In this manner Dutch disease and neo-patrimonial state officials, assisted by neo-liberal reforms, should have contributed to the atrophy of agriculture and state capacity in the rural areas of many developing countries (Van de Walle, 2001; Wunder, 2004). Dualistic societies would have reemerged. Modern, sometimes overseas oriented urban societies would co-exist with shrinking farm sectors in rural areas of the same polity. This social structure shows little sign of disappearing. With the intensified global extraction of oil and mineral deposits over the past four decades, more than 100 nations are now classified as “mining or oil dependent” (Haglund, 2011). 7 The large number of these extractive societies underscores the

FOOD VERSUS FUEL: EXTRACTIVE INDUSTRIES, INSECURE LAND TENURE, AND GAPS IN WORLD FOOD PRODUCTION 65

importance of analyzing the national and regional scale interactions between expanding extractive activities and contracting agricultural enterprises. The remainder of this paper draws upon newly available agro-ecological data and economic measures of extractive activity to investigate these links between the food and fuel sectors. The previous discussion suggests two general expectations about the links between extractive industries and food production in countries. 1. Mineral and oil dependent countries will cultivate less land than expected and exhibit larger than expected yield gaps in agricultural production. The lower than expected extent of cultivated land would stem from the increased competition from imported agricultural products and the migration of potential cultivators to cities. The gap between potential yields and actual yields should be especially large in countries with large oil or mining sectors because the urban focus of these regimes would contribute to ineffective agricultural extension services and insecure land tenure in rural areas which, in turn, would depress yields. 2. Countries with high perceived frequencies of corruption, founded on a prevailing neo-patrimonial ethos in government and exacerbated in some instances by oil and mineral production, will cultivate less land than expected and exhibit larger than expected gaps between potential and actual yields. A high incidence of corrupt activities would discourage farmers from cultivating land without titles to it. Similarly, a high incidence of corruption would obstruct the delivery of extension services to farmers which in turn would increase the gap between potential and actual yields. The data and measures used to conduct preliminary tests of these hypotheses are outlined below. 3. DATA AND METHODS (a) Independent variables (i) Control variables Over the past two decades the Food and Agricultural Organization (FAO) of the United Nations, working with the International Institute for Applied Systems Analysis (IIASA), developed a geographical information system of Global Agro-Ecological Zones (GAEZ). It identifies areas that are not suitable for agricultural production and areas that currently are cultivated. The remaining areas represent potential places for expanding agriculture, albeit in some cases at a considerable ecological cost. In addition, by bringing together information on topography, rainfall, temperature, and soil quality, GAEZ maps potential crop yields at varying levels of inputs for different crops and agricultural regions around the world. By comparing the potential yields with actual yields in a place, observers can identify places with “yield gaps,” large differences between the actual yields and potential yields of crops. These places would appear to be excellent candidates for intensified agriculture. The GAEZ provides two control variables for the following analyses. They are: Lands not suitable for agriculture: IIASA (International Institute for Applied Systems Analysis) scientists used measures of topography, soil quality, length of growing season, and precipitation to calculate the extent of lands suitable for agriculture in a country. They entered these data into a GIS and calculated the extent of lands not suitable for agriculture in a country.

Potential yields: IIASA analysts used the same biophysical data to calculate the yield potentials of lands for cereals (maize, rice, and wheat). 8 The yield potentials are then calculated for varying levels of inputs. In the analyses reported below, the yield potentials assume an intermediate level of inputs over a long period of time. For details see Global Agro-Ecological Zoning, version GAEZ 3.0 (IIASA, 2011). The potential yields in GAEZ refer only to rain-fed agriculture. Gross Domestic Product (GDP) per capita in 1990: This variable (logged) provides a third control variable for the analysis. It controls for variations in the human capital available to agricultural producers across societies. A Dutch disease variable: oil or mineral dependency: The proportion of a country’s export earnings that come from fuels (oil or natural gas) and minerals provides a measure to a country’s past exposure to Dutch disease. Only countries with high scores on this variable would have been susceptible to Dutch disease like phenomena. A neo-patrimonial state variable: the corruption perception index: A high perceived incidence of corruption would indicate the prevalence of a neo-patrimonial ethos among state officials. This patrimonial understanding of the state would justify corrupt practices. Oil and mineral producing societies would also encourage corruption because the revenue from the industries frees politicians from having to deliver services to a tax paying public in a competent way. In addition, by concentrating income, the oil revenue makes it easier for politically favored elites to use their financial advantages to ignore laws when they choose to do so. A high incidence of corruption would in turn make landowners with insecure titles reluctant to invest in land improvements that would intensify agricultural production because the newly valuable land would make them vulnerable to usurpation by more favorably placed elites in the society. The Corruption Perception Index (CPI), created out of multiple surveys of businessmen and citizens, measures the perceived incidence of corruption in a country. It is constructed out of responses to questions in multiple surveys about the frequency with which citizens and businessmen have experienced corruption in their daily affairs. A higher score on the index indicates a lower perceived incidence of corruption. The use of this measure raises questions about the relationship between perceived and actual corruption in places. Studies that explore this relationship are hard to find, but one recent, reputable study in Indonesia finds a strong relationship between perceived and actual corruption in a place (Olken, 2009). Because the CPI measures the perceived incidence of corruption in all government—citizen interactions, some portion of them will concern landownership, especially in predominantly agrarian societies. Each time a smallholder seeks title to a tract of land, s/he may have to pay a bribe to an official in order to obtain a title. When conflicts over landownership erupt, the contending parties typically go before a local authority who assesses the relative merits of “pre- titles,” “clouded titles” or poorly documented previous rulings by local authorities. Legal authorities have much discretion in these situations, so corruption sometimes occurs. While it would be preferable to have a measure of corruption for just land titling activities, these data do not exist. Under these circumstances the CPI would seem to provide a valid surrogate measure for tenure insecurity. (b) Dependent variables The dependent variables in the two analyses are respectively cultivated area and yields per hectare for cereal crops. The data for these variables come from FAO. These data are col-

66

WORLD DEVELOPMENT Table 1. Variables, operational definitions, and data sources Variable

Role in multivariate analysis

Operational definition

Source

Area of country

Control

Km2

Corruption Production Index (CPI), 2002

Independent variable

Cultivated area

Dependent variable

GDP per capita, 1990

Control

Oil and mineral dependence, 2000

Independent variable

Yields, actual, 2005

Dependent variable

Yields, potential

Control

Operational definition for neo-patrimonial practice in governance The extent of all currently cultivated lands, including fallows, in country, 2005 Human and financial capital in a country Operational definition for Dutch disease: Proportion of export earnings from fuels and minerals in 2000 Yields for rice, wheat, and maize, averaged across the three crops for each country Maximum attainable, long-term achievable yields for rainfed wheat, rice, or grain maize averaged over all arable land

World Resources Institute at http://earthtrends.wri.org/ Transparency International at http://www.transparency.org/policy_research/ surveys_indices/cpi/2010/in_detail Food and Agricultural Organization of the United Nations (FAO) at http://faostat.fao.org/site/567/default.aaspx

lected by the agricultural ministries of each nation that belongs to the UN. Table 1 specifies these measures further and lists the sources for them. The Supplementary materials for this article, the mmc1 data set, contain the data used in the analysis and the associated variable definitions. Tables 2 and 3 present descriptive statistics derived from the GAEZ: the extent of cultivated areas and unsuitable areas for agriculture (Table 2) and regional variations in the size of the gap between potential and actual yields (Table 3). The multivariate analyses reported in Tables 4 and 5 address the different ways in which agricultural production could expand, either through increases in the extent of cultivated areas (Table 4) or through increased yields from intensification (Table 5). The analyses in Table 4 try to predict the absolute amount of cultivated land in a nation in 2005. Presumably, nations suffering from Dutch disease would cultivate significantly less land than other nations. The analyses in Table 5 try to predict a compos-

Penn World Tables at http://pwt.econ.upenn.edu/php_site/pwt_index.php United Nations Conference on Trade and Development (UNCTAD) at http://unctad.org/en/ Pages/Statistics.aspx

Food and Agricultural Organization of the United Nations (FAO) at http://faostat.fao.org/ site/567/default.aspx#ancor Global Agro-Ecological Zoning (GAEZ) website at: http://www.iiasa.ac.at/Research/ LUC/Products-datasets/index.html?sb=21

ite yield for cereal crops in 2005. Nations with a Dutch disease experience would presumably exhibit larger yield gaps (the difference between actual and potential yields) than other nations. The multivariate analyses reported in Tables 4 and 5 are ordinary least squares regressions. These equations have normally distributed residuals. They do not exhibit high levels of multicollinearity. The highest variance inflation factors for individual variables are between 2 and 3. Highly influential cases, never more than two in number, have been dropped from these equations. The independent variables are either invariant with respect to time or they measure conditions prior to the observations about agricultural areas and yields, so the equations do not suffer from simultaneity bias. Because agricultural underproduction clusters in the developing world, Tables 4 and 5 each contain, in addition to analyses of the global set of nations, analyses that focus on nations in the developing world.

Table 2. Cultivated lands and agriculturally unsuitable lands across the world’s regions. Source: Global Agro-ecological Zoning (GAEZ) database, International Institute for Applied Systems Analysis Region

Sub-Sah. Africa Mid. East–N. Africa Europe South, East Asia Central Asia Oceania North America Middle America South America Means for regions

(1)

(2)

(3)

Mean proportion of lands in a country that are cultivated (# of countries)

Mean proportion of lands in a country that are not considered suitable for cultivation (# of countries)

Sum: columns 1 and 2: where are the idle agricultural lands?

.1613 (47) .1056 (19) .2659 (37) .2568 (22) .0957 (11) .1209 (9) .1413 (3) .2187 (19) .0562 (12) .1814 (179)

.5255 (43) .8583 (19) .4615 (35) .6063 (19) .8232 (11) .8345 (5) .7377 (2) .6829 (12) .5226 (12) .6060 (158)

.6868 .9639 .7274 .8631 .9189 .9554 .8790 .9016 .5746 .7874

FOOD VERSUS FUEL: EXTRACTIVE INDUSTRIES, INSECURE LAND TENURE, AND GAPS IN WORLD FOOD PRODUCTION 67 Table 3. Yields and yield gaps by region across crops. Source: Hengsdijk and Langeveld (2009) Region

Yields (% of potential yields)

Yield gaps (1—Yield)

Number of nations

Std. dev.

.7478 .8460 .3459 .2328 .4419 .2106 .2704 .5635 .4493 .1657 .3388

.2522 .1540 .6541 .7672 .5581 .7894 .7296 .4365 .5507 .8343 .6612

4 6 5 7 5 3 3 5 6 7 5

.20447 .12280 .13957 .15697 .10550 .03629 .01751 .26756 .23365 .05779 .18321

Northeast Asia Western Europe Eastern Europe South Asia Southeast Asia Oceania CIS North America Latin America Humid Africa Mid. East–N. Africa

Table 4. The extent of agricultural lands across nations: Dutch disease effects (1) (2) All nations Global south Control variables Area of country Area not suitable for crops GDP 2000 Dutch disease related variables Dutch disease: oil-mining (% of exports) Corruption perception index, 2002 r2 (adj.) N

.837*** (.066) .003 (.006) .099 (.105)

.777*** (.090) .008 (.013) .057 (.126)

.019*** (.003) .116* (.056) .641 136

.017*** (.003) .138 (.094) .617 94

** *

p < .01. p < .10. p < .001.

***

4. RESULTS Table 2 displays the regional variations in agriculturally suitable lands. Column three in Table 2 sums together areas

of land already cultivated (column one) and lands unsuitable for agriculture (column two). In so doing, it provides an indirect measure of the possibilities for expanding lands under cultivation. In a region like Oceania the sum of these two percentages approaches one, indicating little possibility for an expansion in cultivated areas. The two regions with the most potential for expanding cultivated areas are South America and sub-Saharan Africa. Table 3 reports the median gap between actual yields and potential yields for cereal crops for the different regions of the world. The calculations of potential yields assume an intermediate level of agricultural inputs like fertilizers. Sub-Saharan Africa exhibits the largest gaps between actual and potential yields, with actual yields averaging only 16.5% of potential yields. Clearly, the opportunities for expanding global agricultural production through increases in cropland and growth in yields concentrate in the developing world. Tables 4 and 5 present the results of regressions that explore the causes for the more limited agricultural output in some nations. The unmistakable influence of oil and mining exports in these equations establishes that oil, gas, and mining booms have, in indirect ways, retarded the growth and productivity of agricultural sectors both globally and regionally. The significance of the corruption perception index in the global equation (column 1) in Table 5 indicates that the reliance on oil or mineral exports has an important governance dimension. Countries with relative little perceived corruption (a high score on the corruption perception index) have, ceteris paribus, ac-

Table 5. Yields for cereal crops across nations, 2005: Dutch disease effects (1) (2) (3) (4) Globe (all nations) Global north Global south Sub-Saharan Africa Control variables Yield potential of lands Area not suitable for crops GDP 1990 (logged) Dutch disease variables Fuel mining exports (%) Corruption perception index, 2002 r2 (adj.) N *

p < .10. p < .01. p < .001.

**

***

.281 (.242) .003 (.008) .839*** (.180)

.043 (.077) .001 (.002) .255* (.131)

.091 (.091) .008 (.005) .263*** (.068)

.306 (.186) .082** (.021) .431* (.166)

.013** (.005) .427*** (.095) .611 125

.006* (.003) .045 (.033) .557 37

.005** (.002) .021 (.051) .261 87

.013** (.004) .085 (.143) .462 32

68

WORLD DEVELOPMENT

tual yields that come close to the potential yields for crops in a country. In Table 4, in the extent of the cultivated land equation, a high perceived incidence of corruption, contrary to expectations, associates with larger areas of cultivated land, controlling for other factors. 5. DISCUSSION The results from the regression analyses in Tables 4 and 5 suggest that Dutch disease dynamics have played a significant role in both the slow rate of expansion in cultivated areas and the emergence of yield gaps in world food production over the past two decades. In Table 4, as expected, a high dependence on oil or mineral exports associates with relatively low areas of cultivated land. High levels of perceived corruption do not, in contrast, produce the expected smaller amounts of cultivated land in part because corruption co-varies with poverty and the most impoverished societies have predominantly agrarian economies. Recent work on deforestation rates in twelve sub-Saharan countries with tropical moist forests offers a plausible extension of the observed patterns in cultivated areas (Fisher, 2010). The total volume of foodstuffs produced and consumed locally, like cassava and yams, grew substantially during the 2000–05 period. Because production per hectare did not increase, virtually all of the increases in harvest size came from increases in acreage under cultivation. Not surprisingly, increases in the production of cassava correlated positively with losses of forest in a country. These findings, along with the findings in column 2 of Table 4, suggest empirical support for the idea that peri-urban agriculture, with its focus on local production and consumption, persists even in the inhospitable macro-economic conditions created by oil and mineral booms. A high incidence of corruption appears to have depressed agricultural yields (Table 5) without diminishing the extent of arable land in countries (Table 4). Endemic corruption would reduce the effectiveness of government extension services, thereby contributing to lower actual yields relative to potential yields. It is less likely to affect the extent of cultivated lands. Given the prevalence of customary and informal land tenure in so many places, smallholders surrounded by other farming families without titles to their lands are not likely to regard the prospect of working the land without a title as a decisive deterrent to farming (Rudel, 1995). For smallholders subsistence needs and the prospect of some income from the sale of crops represent compelling reasons to expand the area under cultivation. Rural poverty encourages smallholder agricultural expansion, even when titles to new landholdings remain difficult, if not impossible, to obtain. Land grabbing outside investors take a different position from smallholders regarding land tenure. They insist on it, because it provides them with a defense against “weapons of the weak” strategies (Scott, 1985) in which poor neighbors, aggrieved by the sudden acquisition of nearby lands by outsiders, encroach on the newcomers’ lands. Because payments for environmental services, like REDD+ (Reducing Emissions from Deforestation and Degradation), require that the recipients of the payments have titles to their land (Wunder, 2005), countries with tropical forests who want to participate in REDD+ may begin in the near future to extend secure land tenure to landholders in peripheral regions. While secure land tenure would undoubtedly benefit smallholders in these regions, it would provide even more crucial assistance to large, land grabbing enterprises who want to raise yields through land improvements like irrigation, but also want legal protections for their investments in the land.

The finding in Tables 4 and 5 that Dutch disease discourages agriculture does not define the range of ways in which it happens. While economists developed the theory largely to explain labor market dynamics and sectoral imbalances, Dutch disease afflicted societies exhibit distinct social structures that handicap agricultural production in myriad ways. As noted earlier, they exhibit an economic dualism between economically affluent urban areas and impoverished rural areas. They become highly urbanized, with as much as 85% of the population residing in urban areas. Outside of the oil and mineral sectors, most economic activity focuses on the provision of services rather than the production of either manufacturing or agricultural goods. Consistent with this profile, a “comprador bourgeoisie” (Frank, 1966) develops in centers of commerce. These people specialize in importing a wide range of consumption goods, ranging from basic foodstuffs to luxury goods consumed by the elites. Colonial regimes generated dualistic social structures, and in some senses Dutch disease afflicted societies represent a “neocolonial dualism.” In its original usage dualism featured two sectors, a capitalist sector founded on a plantation economy that supported a colonial elite and a rural subsistence sector containing significant numbers of the subject population (Boeke, 1953; Lewis, 1954). Dutch disease societies also feature radical disparities between rural and urban sectors, but these inequalities have their origins in royalties that go to urban based governments from agreements that they have negotiated, on increasingly favorable terms, with multi-national corporations in the oil or mining industries. While colonialism encouraged dualism through the creation of economic enclaves organized through political stratification between colonial elites and subject peoples, oil and mineral dependent dualism has emerged in large part out of market dynamics and deal making between political elites in government and economic elites in the oil and mining companies. The disparities in levels of living between urban and rural areas, coupled with the Dutch disease dynamics discussed above, have spurred “depeasantization” or “de-agrarianization” and continued migration from rural to urban places (Bryceson, 2002). As cities house progressively larger proportions of the national population, a condition of “over-urbanization” emerges in which cities in Dutch disease afflicted societies, by historical standards, contain relatively few manufacturing jobs given the size of their populations (Timberlake & Kentor, 1983). The dearth of productive enterprises, manufacturing as well as agricultural, reflects the difficulties that a Dutch diseased economy with an appreciating currency poses for producers who have to sell their products in international markets. In contrast, the service sector grows because most service providers do not face foreign competition. To provide the volume of services necessary to stay in business, service providers usually have to locate in densely populated places, so their prevalence in an economy favors the growth of urban areas. Taken together, these patterns contribute to a pronounced urban bias in both economic and political affairs (Jones & Corbridge, 2010; Lipton, 1977). This urban oriented social structure does not preclude agricultural pursuits, but it favors part-time, peri-urban farmers who focus on foodstuffs like fruits and vegetables that cannot be readily imported. Some of these farmers circulate between a small farm and an urban occupation, following “roads of many returns” (Waters, 1997). For many of them, farming is part of a diverse mix of livelihoods that households pursue to make economic ends meet (Collins et al., 2011; Lerner & Eakin, 2011). The part-time character of these agricultural pursuits blunts the full force of the Dutch disease effects and

FOOD VERSUS FUEL: EXTRACTIVE INDUSTRIES, INSECURE LAND TENURE, AND GAPS IN WORLD FOOD PRODUCTION 69

explains why this type of agriculture may persist in less than hospitable macro-economic settings. Could Dutch disease take a globalized form? Analysts usually discuss it as a characteristic of national economies with impacts on governments, currencies, and national labor markets. It could be construed, in modified form, as a transnational phenomenon that would have global scale effects on agricultural production. In other words Dutch disease could be a phenomenon of global proportions, with a global economy that has a large fuel sector and a somewhat shrunken agricultural sector. In this line of analysis a succession of countries experience Dutch disease over time. In each case the labor market dynamics, related currency fluctuations, and associated urban biases reduce the numbers of full-time farmers in the societies, prevent the socialization of new generations of farmers, and weakens the network of organizations that support farmers. Because these Dutch disease effects occur sequentially over time, following discoveries of deposits and surges in commodity prices, the short-term price increasing effects of declines in agricultural production could easily be offset by interacting effects like fluctuations in weather conditions and changes in biofuel policies in the large, grain producing countries. Eventually, the loss of productive agricultural communities should induce a global increase in commodity prices, but the long-term erosion of the productive and institutional capacity in the agricultural sector suggests that the price increases would make the influx of people back into agriculture occur slowly, if at all. In this instance, global society could become, at least in the production of foodstuffs, over-urbanized. A watered down, but globalized version of Dutch disease will have emerged. A test of this thesis and a calculation of the magnitude of these Dutch disease effects would require work

with time series data and a model of the global economy that goes far beyond the analyses that have been attempted here. 6. CONCLUSION: WOULD A DECLINE IN FUEL PRODUCTION SPUR AN AGRICULTURAL REVIVAL? Could a decline in a country’s economic reliance on oil and minerals spur an agricultural revival in these nations? The preceding discussion argues that oil and mineral driven Dutch disease has ramifications for governance, settlement patterns, and occupational structures in societies (Ross, 2012). While oil and mineral dependent economies have shown remarkable persistence (Davis, 1995), an eventual decline in extractive activities, either through the exhaustion of deposits or more speculatively, through a decline in demand for fossil fuels, could have a transformative impact on the agricultural economies of these countries. Agriculture would become a more important livelihood to citizens and a more important source of revenue to governments, so agricultural assistance in various forms might become more available to farmers. The growing importance of agrarian communities might also reinvigorate democracies and, in so doing, create pressures to counteract neo-patrimonial practices in government. The growing reliance on revenues from agriculture production could also make national governments more receptive to land grabbing by tax paying outside funds. It may seem naı¨ve to point out these possibilities, but, given the premise of declining oil and mineral activity, the reinvigoration of agriculture in its various forms does follow logically from the depletion of oil and mineral reserves given the empirical linkages between the fuel and food sectors outlined in this paper.

NOTES 1. The developing world refers to all countries outside of western Europe, North America, Japan, Australia, and New Zealand. It is sometimes referred to as the Global South. 2. The locations and partners in recent land grabbing deals can be found at http://landportal.info/landmatrix. 3. The argument and analysis presented here applies to both mineral and oil dependent economies. 4. The ethos of a people expresses their beliefs of about ideal behaviors. A neo-patrimonial ethos among state officials would sanction the siphoning off of public resources in order to benefit an official’s family. It is sometimes referred to as pre-bendalism. Weber (1968) first articulated the idea of a patrimonial state. 5. The discovery of oil or mineral wealth in rural areas has led to armed conflict in a significant number of countries (Collier & Venables, 2010; Ross, 2004), and in these instances the magnitude of the urbanization associated with the growth of the extractive sector may increase even further as the cities absorb people displaced by the violence.

6. The effects of mining and oil revenues on the incidence of corruption are subject to some controversy. Petermann and colleagues (2007) found that in poorer states surges in both mining and oil revenues generated more corruption. Ross (2012), in a conservative test, finds that increments in oil and mining revenues did not produce increases in corruption. Given that corrupt practices require cultural supports and cultural changes occur in spurts (Swidler, 1986), it is unlikely that Ross’s test with its focus on increments in revenues, would affirm the link between oil–mining revenues and the incidence of corruption. 7. Nations for whom more than 25% of their export earnings come from either fuels or minerals would be considered oil or mineral dependent. 8. These three cereal crops are consumed throughout the world, so variations in their yields have consequences for food security in almost all nations. As a reviewer pointed out, a more precise measure of yield gaps would incorporate yield data on a more diverse array of crops, including crops like yams, manioc, and other tubers that are consumed in great quantities, for example, in sub-Saharan Africa. A yield index could be constructed, with weightings for different crops depending on their importance in a country. While clearly an empirical advance, the construction of this kind of index is beyond the scope of this paper.

REFERENCES Alston, L. J., Libecap, G., & Mueller, B. (1999). Titles, conflict, and land use: The development of property rights and land reform on the Amazon frontier. Ann Arbor, Michigan: University of Michigan Press. Boeke, J. H. (1953). Economics and economic policy of dual societies. New York: Institute of Pacific Relations.

Brislen, L. (2012). Produces food that is good: Engaging beginning farmers of Kentucky on their own terms. Paper presented at the Rural Sociological Society, Chicago, Illinois. Bryceson, D. F. (2002). The scramble in Africa: Reorienting rural livelihoods. World Development, 30, 725–739.

70

WORLD DEVELOPMENT

Collier, P., & Venables, A. (2010). Natural resources and state fragility. EUI Working Paper 2010:36. Robert Schuman Centre for Advanced Studies, European University Institute, Florence, Italy. Collins, D., Morduch, J., Rutherford, S., & Ruthven, O. (2011). Portfolios of the poor: How the world’s poor live on $2 a day. Princeton, NJ: Princeton University Press. Corden, W. M., & Neary, J. P. (1982). Booming sector and deindustrialization in a small open economy. The Economic Journal, 92, 825–848. Davis, G. A. (1995). Learning to love the Dutch disease: Evidence from the mineral economies. World Development, 23, 1765–1779. Dove, M. (1993). A revisionist view of tropical deforestation and development. Environmental Conservation, 20, 17–24. Dunning, T. (2008). Crude democracy: Natural resource wealth and political regimes. Princeton, NJ: Princeton University Press. Fake, C., Filan, T., & Ingram, R. (2012). Sierra Nevada Foothills beginning farmer and rancher training and networking project: Methods, successes, and lessons learned. Paper presented at the Rural Sociological Society, Chicago, Illinois. Fisher, B. (2010). African exception to drivers of deforestation. Nature Geoscience, 3, 375–376. Frank, A. G. (1966). The development of underdevelopment. New York, NY: Monthly Review Press. Fuglie, K. O. (2008). Is a slowdown in agricultural productivity growth contributing to the rise in commodity prices?. Agricultural Economics, 39, 431–441. Granovetter, M. (1985). Economic action and social structure: The problem of embeddedness. American Journal of Sociology, 91, 491–510. Granovetter, M. (1992). Economic institutions as social constructions: A framework for analysis. Acta Sociologica, 35, 3–11. Haglund, D. (2011). Blessing or curse?: The rise of mineral dependence in low and middle income countries. Oxford, UK: Oxford Policy Management. Hengsdijk, H., & Langeveld, J. W. A. (2009). Yield trends and yield gap analysis of major crops in the world. Wageningen, NL: Wettelijke Onderzoekstaken Natuur & Milieu. Holland, M., de Koning, F., Morales, M., Naughton-Treves, L. (2013). Complex tenure and deforestation: Implications for conservation incentives in the Ecuadorian Amazon. World Development. In press. International Institute for Applied Systems Analysis (IIASA). (2011). Global Agro-Ecological Zoning, GAEZ 3.0. Accessed. Jones, G., & Corbridge, S. (2010). The continuing debate about urban bias: The thesis, its critics, its influence and its implications for povertyreduction strategies. Progress in Development Studies, 10, 1–18. Lerner, A. M., & Eakin, H. (2011). An obsolete dichotomy?: Rethinking the rural–urban interface in terms of food security and production in the global south. The Geographical Journal, 177, 311–320. Lewis, W. A. (1954). Economic development with unlimited supplies of labour. The Manchester School, 22, 139–191. Lipton, M. (1977). Why poor people stay poor: Urban bias in world development. Cambridge, MA: Harvard University Press. Lipton, M. (2009). Land reform in developing countries: Property rights and property wrongs. London, UK: Routledge. Olken, B. (2009). Corruption perceptions and corruption reality. Journal of Public Economics, 93, 950–964. Okonta, I., & Douglas, O. (2001). Where vultures feast: Shell, human rights, and oil in the Niger Delta. San Francisco, CA.: Sierra Club Books.

Organization for Economic Cooperation and Development (OECD) (2011). Price volatility in food and agricultural markets: Policy responses. Paris, France: OECD. Pardey, P., Beintema, N., Dehmer, S., & Wood, S. (2006). Agricultural research: A growing global divide?. Washington, DC: Agricultural Science and Technology Indicators Initiative, International Food Policy Research Institute. Pegg, S. (2009). Is there a Dutch disease in Botswana?. Resources Policy, 35, 14–19. Petermann, A., Guzman, J. I., & Tilton, J. E. (2007). Mining and corruption. Resources Policy, 32, 91–103. Ross, M. L. (2004). What do we know about natural resources and civil war?. Journal of Peace Research, 41, 337–356. Ross, M. L. (2012). The oil curse: How petroleum wealth shapes the development of nations. Princeton, NJ: Princeton University Press. Rudel, T. K. (1995). Do property rights make a difference?: Open access, informal social controls, and deforestation in the Ecuadorian Amazon. Human Organization, 54, 187–194. Scott, J. C. (1985). Weapons of the weak: Everyday forms of peasant resistance. New Haven, CT: Yale University Press. Swidler, A. (1986). Culture in action: Symbols and strategies. American Sociological Review, 51, 273–286. Tilman, D., Balzer, C., Hill, J., & Befort, B. (2011). Global food demand and the sustainable intensification of agriculture. Proceedings of the National Academy of Sciences, 108, 20260–20264. http://dx.doi.org/ 10.1073/pnas.1116437108. Timberlake, M., & Kentor, J. (1983). Economic dependence, overurbanization, and economic growth: A study of less developed countries. The Sociological Quarterly, 24, 489–507. Timmer, C. P. (2007). How Indonesia connected the poor to rapid economic growth. In T. Besley, & L. J. Cord (Eds.), Delivering on the promise of pro-poor growth: Insights and lessons from country experiences (pp. 29–58). New York, NY: Palgrave MacMillan. United States Agency for International Development (USAID). (2007). Land Tenure and Property Rights Regional Report. Vol. 2.3—West Africa. Washington, DC. Van de Walle, N. (2001). African economies and the politics of permanent crisis, 1979 to 1999. Cambridge, England: Cambridge University Press. van der Ploeg, F. (2011). Natural resources: Curse or blessing?. Journal of Economic Literature, 49, 366–420. Waters, W. F. (1997). The roads of many returns: Rural bases of the informal urban economy in Ecuador. Latin American Perspectives, 24, 50–64. Weber, M. (1968). Economy and society: An interpretive sociology. New York, NY: Bedminister Press. Wunder, S. (2004). Oil wealth and the fate of the forest: A comparative study of eight tropical countries. London, UK: Routledge. Wunder, S. (2005). Payments for environmental services: Some nuts and bolts. CIFOR Occasional Paper 42(1), Bogor, Indonesia.

APPENDIX A. SUPPLEMENTARY DATA Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.worlddev.2013.05.015.