Reducing Transport Costs Through Peasants´ Innovation In Peru. An Experience of Capability Stagnation
Javier M. Iguíñiz Echeverría Ramón Díaz Giannina Vaccaro1
ABSTRACT In this paper we will focus on one kind of freedom: “Economic facilities (in the form of opportunities to participate in trade and production).”(Sen 1999:11). One aspect of such freedom is access to markets. We carry out an approximation to one of the elements necessary to confront such a challenge, focusing our attention on one aspect of the workings of the market: peasants` innovation influencing transportation costs. Innovation here has the limited sense of crop substitution that increases the proportion of high valued products. Is there in Peru a more or less extensive process of switching to products with a higher value per unit of weight? The results are based on a subsample of the national household survey in Peru.
Peasants´ innovation Transport costs Capability approach Peru
Professor, Department of Economics, Pontifical Catholic University of Peru. [email protected]
, Associate Researcher, Institute of Peruvian Studies. [email protected]
, Ph.D. student associated to the NCCR LIVES. University of Geneva (UNIGE). Switzerland. [email protected]
Introduction In this paper we will focus on one kind of freedom: “Economic facilities (in the form of opportunities to participate in trade and production).”(Sen 1999:11). One aspect of such freedom is access to markets. We will carry out an approximation to some of the elements necessary to confront such a challenge, focusing our attention on one aspect of the workings of the market: peasants´ innovation influencing transportation costs. Furthermore, for Sen, “Ultimately, the challenge that the market systems have to face must relate to problems of equity in the distribution of substantive freedoms,” (2002: 526). In real life competition, and the distribution of freedoms is related to the difference in competitiveness among competitors2, and one aspect of such a difference is related to the diverse distance to the markets that each firm has to confront. Furthermore, the paper constitutes a first exploratory study of peasant´s innovation in Peru based on data from annual household surveys (Encuesta Nacional de Hogares, ENAHO) from 2004 to 2008. The higher the value per unit of weight the smaller the proportion of the transportation costs in the price of the product. " The key issue that needs to be addressed is how smallholders can switch to high value food commodities with minimum transaction costs and market risks" (Birthal, Joshi and Gulati, 2005: 2) Our paper deals then with a previous question: Is there in Peru a more or less extensive process of switching to higher value products? Contrary to the many case studies usually showing interesting and mostly successful experiences in this respect this paper is going to work with a sample of almost four hundred peasants that have been selected with the only criteria of having being surveyed in a national household survey five years in a row. In the first part we are going to approach geographical distance to the market as a restriction to freedom. In the second, we are going to show the importance attributed by experts to the change in patterns of cultivation towards high-value products. In the third, we present the sources of data and the method used in the paper. In the final part we show the results and we finish the paper with some brief concluding remarks.
This topic has been explored in Iguiniz (2009).
I.- Distance as a limit to freedom
The possibility to participate in trading, with all its advantages, limits, and risks widen the capabilities of individuals, families and firms. (Sen 1999, 2002) Such capability depends in part on the distance to the market which in itself depend on the characteristics of the good and service generated. As Nichols (1969: 1462) reminds: “With transport facilities of early America, it was generally uneconomical to move commodities of low value per unit of weight for long distances, and the inhabitants of each region had to become relatively selfsufficient. Long-distance trade, even if possible, was not practical-particularly for bulky commodities.” A common characteristic of landlocked countries and regions is their relative isolation of the market (Naudé 2007). This characteristic is ever-present to many producers within the Peruvian Andes and is one of the reasons for a very ample skepticism with respect to the possibilities of progress for the peasants of Peru and those of the world. In this paper we suggest the need for an exploration on the potential of small-scale agriculture to increase their access to the markets and to reduce the poverty of peasants. The Andean mountain range does not house great cities in its interior and geographically it is extremely rough. At the same time, there predominates small property and “minifundio”. It is said that the small scale of production and great distance to the market contribute to the narrowing of markets and to the reproduction of poverty. Farrington et al (2002:1) found that “rural poor face exceptionally high transactions costs and considerably risks in gaining access to production and trade opportunities” Under these and other conditions, peasants tend to react by diversifying the pattern of cultivation in order to protect against climate volatility and other risks. When we talk about diversification, we can mention 2 possibilities: “There are two sources of crop diversification. These are (i) area augmentation, and (ii) crop substitution.” (Joshi et al, 2003: 10).3 We are going to consider the second strategy as freedom enhancing.
"Agricultural diversification creates opportunities for achieving higher and more stable rural incomes through the more efficient use of resources and the exploitation of comparative advantage".( i et.al., 1992)
Due to the high costs of transportation of commodities, the great distance to markets is, no doubt, one of the factors that reduce the capacity of peasants to compete and progress.4 A classic answer has been to improve the roads and the means of transportation, as highways and trucks are normally decisive to send products to the main markets. In addition to this response, an increase in the scale of production also reduces transportation unit costs, but is not a possibility open to many. Cooperation amongst campesinos is also important to increase the volumes transported and to reduce transport unit costs. The core of the paper concentrates in an additional possibility, one that is potentially very important but not much present in literature. It consists of a change in the nature of the product and, in this way, augments the competitive potential by increasing the value of the product per unit of weight. In this paper we look for the existence of two ways of dealing with the problem: a) Diversifying to products of greater monetary value per unit of weight, and therefore lower transport unit costs, and b) Transforming agricultural products with the possibility that the manufactured result is also of a greater value by unit of weight than the original agricultural input (Krugman 1991, Iguíñiz 1998). This paper, therefore, has the objective of determining the existence or not of a diversifying behaviour used by peasant that is not purely defensive in the sense of increasing market isolation. This behaviour is valuable because, within the geographic and social context, ceteris paribus, it reduces the restriction on freedom that distance to the market can impose.5 We will, then characterize the activity of the peasant as innovative when, in its pattern of cultivation, he introduces an increasing proportion of products of higher unit value in the market. In other words, a greater proportion of these products in the portfolio of the peasants´ production will reduce the proportion of the transport costs in the final value of the product. As soon as this innovation expands the accessible market of the peasant, we can consider that it increases his competitive ability as well as his capabilities as an economic agent. With the present data we cannot conclude that the transformation of agricultural products 4
“Poor transport infrastructure limits market access for many farmers in the developing world” (DFID, 2005: 11) 5 Of course, such increasing capability may not translate into more sales or greater income; intermediaries could frustrate the process and discourage the innovating effort of the peasant. Other competitors in the same market could be relatively more productive.
results in an elevation of the value of the product by unit of weight. As such we have a partial answer to the one of the questions surrounding the effect of the transformation of agricultural products before their transportation.
II.- Relevance of the search: the economic importance of high/low value chains
Agriculture may be a very important activity to reduce poverty. "Income and crop diversification have been identified as essential strategies for raising income and reducing rural poverty" (Ibrahim, Rahman, Envulus, and Oyewole, (2009:84).6 Diversification can be understood as: “switching from low-value crop production to higher-value crops, livestock, and non-farm activities. Although “low-value crops” are sometimes defined in terms of the value per unit of weight, it is probably more useful to define them as crops that generate high economic returns per unit of labor or land. This definition focuses on diversification as a source of income growth and a potential means for poverty reduction" (Birthal, Joshi and Gulati, 2005: 30) The importance of value per unit of weight is evident once we recognize that “Most farmers are part of low-value market chains”.(Taylor et.al.2008: 43). “In this low-value chain equilibrium there is no incentive for the end buyer to offer higher prices to the processor, or the processor to the farmer.” (Taylor et.al. 2008, 43). Change in the type of products is considered very important although not easy to accomplish: “Broadly speaking, there are two ways in which agricultural markets can be used to help poor households meet the objectives of higher income and greater income stability. First, poor farmers can move from low to high value market chains. Many of the most captivating development success stories in recent years involve small farmers significantly increasing the value of what they produce by moving into new market chains. Second, they can increase their share of benefits in existing chains. These two challenges become one and the same for the world’s many small farmers who are unable to access any markets. 6
“Gallup et al (1997) reported that every 1% increase in per capita agricultural output led to a 1.61% increase in the incomes of the poorest 20% of the population”. (DFID 2005: 5)
Poor farmers face significant obstacles both to entering high-value chains and to strengthening their position within existing ones”. (p.10). But it is a hopeful road. “Agricultural diversification towards high-value commodities (such as fruits, vegetables, milk, meat, poultry, eggs and fish) is one of the most promising strategies to reverse the declining growth in agricultural sector. Demand for these commodities is growing with rising income, urbanization and globalization.” (IFPRI, 2007: ii) Recent market liberalization makes such trend more evident “At the same time that the public sector has withdrawn from markets, agricultural supply chains have bifurcated into high-value chains, supplying supermarkets, high-value processing and exports, and traditional chains leading to low-value markets and food processing” (Taylor et.al. 2008: 56). It is not enough to reform institutions related to intermediation “An example is the creation of a farmers’ cooperative that sells directly to the processor, or perhaps tries to augment the farmers’ market by identifying new buyers. Sometimes these projects are successful, at least temporarily, but rarely are they sustainable, because they do not address the basic problem, which is that the farmers remain part of low-value chains. After the development experts depart, the farmers’ cooperative and its members remain the weak link in this chain” (Taylor et.al. 2008: 44).
III.- Data and method The basic source of information used has been the National Household Survey (the Encuesta Nacional de Hogares or ENAHO), which is annually applied by the country’s official authority on statistics, the National Institute of Statistics and Informatics (the Instituto Nacional de Estadística e Informática or INEI). This survey has been designed with the purpose of recording the incomes and expenses of Peruvian homes in detail in order to measure monetary poverty at a departmental level, and within other relevant domains. Details of both expenses and incomes are collected on the basis of the application of several modules at both a household and individual level. As part of the collection of the income data, a special individual questionnaire has been applied to all individuals engaged in developed farming activities. The aspects collected include: i) Income from selling crops and 6
livestock, as well as the by-products of both. ii) Characteristics of the farm, such as the area of land worked on, volumes cultivated, volumes sold, and volumes of self-consumption, among others. iii) Data of livestock and forest activities. iv) Expenses resulting from farming activity. The ENAHO has a departmental representativeness for each year starting from 2001. In this year, the size of the sample of ENAHO reached the figure of 16,500 households and surpassed that of 20,000 households in the sample of 2006. In the years 2001 and 2002, the survey was only carried out in the months from October to December, while from the year 2003 onwards it has been carried out from January to December each year. Of concern to us in this project are the patterns of change in the importance of crops with the highest monetary value per weight unit. The information we require comes almost exclusively from the module of the independent farming producer. Additional information on age, gender, educational level attained, and monetary poverty conditions is obtained from the other modules of the survey. Given our main concern, it will be necessary to carry out a follow-up on the farmers themselves in successive periods, so that we can approach their conduct over time. Each year, one part of the ENAHO sample is of a panel type, in that the same household is revisited whenever the survey is carried out once again. We must point out that, strictly speaking, the panel deals with dwellings rather than households. This means that if a household changes its dwelling place, in the next edition of the survey, the interview will be carried out with the new household occupied by this dwelling rather than the household that had previously occupied it. In order to carry out this exploratory study, the panel sample of the years 2002-2006 created by the INEI itself was used, using the information of the names and surnames of the people surveyed (which data are not of public access). Within this period, it was common for some of the households interviewed in the year 2002 to disappear in the following years and be replaced in the following years by other households that occupied the dwelling selected. It is thus that we decided to work exclusively with the farmers that have been surveyed in all the years between 2002 and 2006. This leaves us with a total of 399 farmers that have been surveyed every year between 2002 and 2006. It is worth noting that this final sample does not offer any level of major representativeness. The results found must be understood as a case study and only describe what happened in the period under study with the farmers of the
sample. But we can say that since 122 farmers have less than one hectare, 174 have between one and five, and 102 more than five, we are dealing mainly with very small and small farmers, indeed peasants, . The focus of the present Project is basically descriptive. We let the data “speak for themselves” and on this basis, we shall try to reach conclusions regarding the existence or non-existence of changes in production patterns. Four main indicators have been used and are to be followed during the study period: i) The highest price per kilogram among the each farmer’s products. This price (in real terms) shows which crop is the most valuable. ii) The percentage that the volume of the crops with the highest price represents within the total volume of all crops produced. iii) The percentage that the value sold of the most valuable crop represents in relation to the total sales of all crops. For each of the indicators proposed, the average obtained in the year 2006 is compared with that of the year 2002. This comparison is subdivided into various levels or cuts that we regard as relevant. The cuts chosen are determined by the type of crop (cereal, fruit, industrial, mainly coffee and cocoa, also vegetable and tuberous) corresponding to the crop with the highest price per kilo, the educational level attained, the farmer’s mother language (indigenous – non-indigenous), gender, and the area of the farm. Since we are dealing with various cuts, and our sample is not a very large one, we shall only present the results in which at least a minimum of 30 observations have been used, and in which, moreover, there are statistically significant differences between the 2006 and 2002 data. The confidence intervals are obtained in the assumption that the data follow an ordinary distribution, without carrying out any adjustment to the data; each farmer carries the same weight within the same sample.
Once again, we must point out that we are concerned
with the changes that took place within the 2002-2006 period (whether they be increases or decreases) rather than with the levels observed, which means that we only present the results corresponding to significant variations in statistical terms.
IV.- Results The following three tables show the results of the descriptions obtained after having analysed the sample worked upon. Only significant variations are shown. 1.- Table 1 shows the estimates of the average prices, taking into account the product with the highest price per kilogram. With this indicator we detect if there are systematic increases in the prices of more valuable crops, this information in addition to other indicators will allow us to identify whether one can speak of changes in cropping patterns. The results for this indicator only show significant changes in some of the cuts made with the data. In all cases, whenever the 20022006 differences have been significant, there is a price increase – in real terms – of the most valuable crop (the prices have been deflated in both time and space, which means that they are expressed in soles of 2009 of the Lima Metropolitan Area). In the case of the crop type, the variation has only been significant in the case of industrial crops (mainly coffee and cocoa) and this has been positive. There have been no significant negative variations. Likewise, the breakdown by genders indicates that a significant price increase can only be perceived for women. The differentiation by mother tongue (which is our mode of approaching the indigenous condition) only shows an increase in the average price of the most valuable product in 2006 for non-indigenous farmers who speak Castilian; as for the educational level attained, a significant increase can only be seen for those farmers without any formal education. There is not, then, a general price increase or decrease.
Table 1: Estimate of the highest average price per kilogram per each farming unit stan yea avera d. r ge error
[Conf. Dif. Interv 2002- Ave2006/ al 95%] 2006 Ave2002
Crop type industrial industrial gender of the farmer Female Female indigenous by mother tongue non-indigenous non-indigenous educational level None None Source: ENAHO (Panel 2002-2006)
20 02 20 06
20 02 20 06
20 02 20 06 20 02 20 06
The data on production also shows that in the years considered, the total volume of the goods produced, and also those that were sold declined between 2002 and 2006. The median of those indicators fell 11% in the case of volumen produced and 39% in the case of the volume sold.. Then, the market context within such period was negative. We are looking for innovation in a negative market evolution. 2.- The second indicator presented is that showing the importance within the total volume of all products of the crop with the highest price per kilogram. Table 2 shows the results corresponding to the average proportion of the volume represented by the most valuable crop. Unlike the average of the prices presented in the previous table, here one can see that the cut by crop type shows various significant differences between the periods described. However, there are increases and decreases in the importance within the volume according to the crop.
The cereals and vegetables show a decrease in importance within the total volume produced; whereas for those producers whose most valuable crop lies within the group of industrial and tuberous crops, one can observe a growth in the importance of this crop within the total volume. Only in the case of industrial crops does this increase coincide with a significant rise in the prices. Among the remaining cuts proposed, only in the cut differentiating gender can one find an increase in the importance of the most valuable crop, albeit exclusively within the female group; there are no significant changes in the remaining cuts. Table 2: Importance of the most valuable crop within the total volume produced
stand Proportio . year n error Crop type (of highestpriced crop x Kg) cereal 2002 0.21 0.03 cereal 2006 0.14 0.03 industrial 2002 0.29 0.04 industrial 2006 0.37 0.04 vegetable 2002 0.11 0.02 vegetable 2006 0.06 0.02 tuberous or root 2002 0.09 0.04 tuberous or root 2006 0.19 0.10 Gender of farmer female 2002 0.12 0.04 female 2006 0.22 0.06 Source: ENAHO (2002-2006 panel)
Dif. [Conf. 2002 Interv Aver2006/Aver20 al 95%] 2006 02
216 216 198 198 259 259
0.16 0.07 0.22 0.30 0.07 0.03
0.26 0.20 0.36 0.44 0.15 0.09
3.- The importance of the most valuable crop within the total value of the sales of all crops is shown on table 3. For this indicator, the cut by crop type shows that there have only been significant changes in the period described in the case of those farmers whose most valuable crop lies within the vegetable and cereal group. In both cases, the importance of the most valuable crop within the total sales value has diminished for 2006, almost by 50%. In the case of women who 11
independently run a farm, we find a significant increase in the importance of the most valuable crop within the total sales value. There have been no significant variations for any of the other crop groups. Table 3: Importance of the most valuable crop within the total sales value
stan Proporti d. year on error Crop type (of highestpriced crop x Kg) cereal 2002 0.28 0.03 cereal 2006 0.16 0.03 vegetable 2002 0.16 0.02 vegetable 2006 0.08 0.02 gender of farmer female 2002 0.18 0.05 Female 2006 0.32 0.07 Source: ENAHO (2002-2006 panel)
[Conf. Interv 95 al %]
Dif. 200 2200 6
216 216 259 259
0.22 0.10 0.11 0.05
0.34 0.23 0.20 0.12
4.- The purpose of the three indicators presented is that of inquiring whether there has been a greater change favouring crops of greater value per weight unit, so that their transport costs amount to a smaller proportion of the final sales value. So far, no clear pattern has been found that may support a conclusive statement. Within this non dynamic context, the distinction by genders is the only one showing a consistent pattern of progress. An indicator that may give us some additional information is that of the number of total crops maintained by the farming units. However, for all the cuts proposed, the work with the estimated averages of this indicator show that the number of crops might have increased for 2006 in relation to 2002, under all the cuts made. This fails to give us additional information for our purpose of finding a typology of “progressive” farmers that tend to specialize in crops of greater value per weight unit.
An additional device is that of observing what crop types have gained in importance, and what crop types have lost it within the period under study. It is for this purpose that we have presented a transition matrix of the most valuable crop that is that with the highest value. The matrix shows that cereals have ceased to be the most valuable crops, whereas industrial crops (or those potentially suitable for industrial use), principally, and vegetable crops have gained in value. Table 4: Transition matrix of the most valuable crops in 2002-2006
2002 cere forag fru Vegeta Industri vegeta al e it ble al ble Cereal Forage fruit Vegetable industrial Vegetable others tuberous
Total Source: ENAHO (2002-2006 panel)
othe tubero Tot rs us al
42 1 5 10 11 50 5 5
1 0 0 0 1 0 0 0
4 0 3 1 3 4 0 0
3 0 0 2 2 2 1 0
6 0 5 0 66 7 0 1
24 0 5 1 20 64 0 7
1 0 1 0 0 1 0 0
7 0 2 0 9 11 1 0
88 1 21 14 112 139 7 13
As it is very difficult to try to establish a typology when there are no clear patterns in the descriptions presented, we shall use a regression to simultaneously control multiple characteristics and observe if there are any significant associations as regards the different specialisation patterns that have been constructed. Three estimations are presented, each one of them with different specialisation indicators. The first will adopt the value of 1, comparing the year 2006 with 2002; the most valuable crop has increased its importance within the total volume produced and within the sales value, it will be 0 in any other case. The second indicator contains the same two conditions but adds the fact that the number of crops maintained in 2006 is beneath that of 2002. Finally, the third indicator contains the first two but adds the condition that the price per kilogram of the most valuable crop for 2006 is superior (in real terms) to that established for 2006.
Table 5: Logit regressions for the different specialisation indicators 4 conditions Greater importance in total volume, in the sales value, fewer crops in 2006 and greater value per Kg for 2006 Education Primary secondary + Higher Mother tongue indigenous ++ gender: male Age 36 to 60 + over 60 + Poverty non-extreme poverty + non-poverty + Area of land 1 to 5 ha more than 5 ha + Number of crops --price of most valuable crop + Natural region Mountains + Forest + Most valuable crop type industrial ++ vegetable + Fruit Tuberous Vegetable Others Constant Pseudo R2 .2015761 Chi2 33.69788 N 332 Legend: +/- p<1 ++/-- p<0.05 Not significant Significant at 95
Greater importance in total volume, in the sales value and fewer crops in 2006
Greater importance in total volume and in the sales value
+ + +
+ -.0677292 36.71914 397
.1202946 42.02393 376 +++/--- p<0.01 Significant at 99
Given the objective of establishing a typology, we only present the direction of the association and whether this is or is not significant in statistical terms. Thus, only two or three signs will indicate a significant association, whether positive or negative. The results of the estimations show that the associated factors displayed as significant are the mother tongue, and the maintenance of a reduced number of crops, that is, the fact that a certain level of specialisation on the farmers’ part has already existed. Finally, the fact that the most valuable crop is among those potentially suitable for industrial use also favours specialisation. Surprisingly, there is no association with the educational level, the area of land worked on, which constitutes an approximation to the scale of production. V.- Concluding remarks Following the trajectory of the peasants in Peruvian agricultural sector during the period 2005-2008 we have not found a clear innovative behaviour related to the change in the pattern of cultivation towards products of higher value per unit of weight. One explanation emerging from the data we had access to is that during the short period studied the prices of the most valued products have changed very little. The price-incentive to change towards a greater proportion of production of such goods was not present. Furthermore, those years, both the volume produced and the total value sold declined. What kind of innovative behaviour emerges in moments where the market for the peasants is expanding is something that has to be researched. Contrary to our expectations, what we have registered in fact is a move away from specialization, something that corresponds with the literature on peasants´ risk aversion. From our limited, very specific, approach to capabilities we can say that viewing the behaviour of the sample of peasants as a whole there was not an expansion of the opportunities to reach more distant markets. When we divided the information in various types of subsamples there was not a statistically significant type of innovative evolution. Size of land, educational level or age of the peasant was not a factor differentiating peasants´ behaviour. Gender
appeared as an important differentiator of innovative practice but it was not powerful enough to change our general conclusion. All this suggests that scale of production and characteristics of the peasant may not be as important as appears in many studies. Further studies on this kind of innovation should consider also factors like location, participation in clusters and other associative efforts, public policy, phases of the economic cycle, degree of openness to foreign markets, etc.
References Birthal, P. S, Joshi, P. K and Gulati, A. (2005). Vertical Coordination in HighValue Food Commodities: Implications for Smallholders. Markets, Trade and Institutions Division (MTID) Discussion Paper No. 85, IFPRI, Washington, DC. Barghouti, Shawki; Garbus, Lisa; and Umali, Dina (1992). "Trends in Agricultural Diversification: Regional perspectives". World Bank Technical Paper Number 180. The World Bank. DFID (2005) “Growth and poverty reduction: the role of agriculture” A DFID policy paper. Published by the Department for International Development. December. Farrington, John; Christoplos, Ian; Kidd, Andrew; and Becckman, Malin. (2002) "Can extension contribute to rural poverty reduction? Synthesis of a six country study". Network paper No 123 Agricultural Research & Extension Network. ODI. July. ISBN 085003 602 X. Gallup, J., S. Radelet and A. Warner (1997) “Economic Growth and the Income of the Poor”. CAER II Discussion Paper No. 36. Harvard Institute for International Development, Boston MA. En: DFID policy paper. "Growth and poverty reduction: the role of agriculture". 2005. Ibrahim, H., Rahman, S.A., Envulus, E.E and Oyewole, S.O (2009) “Income and Crop Diversification Among Farming Households in a Rural Area Of North Central Nigeria” Agro-Science Journal of Tropical Agriculture, Food, Environment and Extension, Volume 8 Number 2 May 2009 pp 84 -89. IFPRI (2007) “Agricultural Diversification towards High Value Commodities. A Study in Food Surplus States in India with Focus on Andhra Pradesh and Punjab”. International Food Policy Research Institute (IFPRI). New Delhi, March. Iguíñiz, Javier (1998) “Transporte y viabilidad de la descentralización económica: marco elemental de análisis”. In: Plaza, Orlando editor, Perú.
Actores y escenarios al inicio del nuevo milenio. Lima: Fondo Editorial de la Pontificia Universidad Católica del Perú, pp. 111-138. Iguíñiz, Javier (2009) “On freedom to compete in the market: three approaches.” Paper presented to the HDCA Conference in Lima. INEI, Encuesta Nacional de Hogares, ENAHO, Lima. Joshi, P.K., Ashok Gulati, Pratap S. Birthal, and Laxmi Tewari (2003) Agriculture Diversification In South Asia: Patterns, Determinants, And Policy Implications. MSSD Discussion Paper NO. 57. International Food Policy Research Institute. February. Krugman, Paul (1991) Geography and Trade. Leuven, Cambridge (USA) and London: The MIT Press. Naudé, Wim (2007) “Geography and development in Africa. Overview and Implications for Regional Cooperation”. WIDER Discussion Paper, WDP 2007/03, September. Nichols, Jr. (1969). “Transportation and Regional Development in Agriculture”. American Journal of Agricultural Economics, Vol. 51, No. 5, Proceedings Issue December, pp. 1455-1463. Sen, Amartya (1999) Development as Freedom. New York: Anchor Books. Sen, Amartya (2002) Rationality and Freedom. Cambridge and London: The Belknap Press of Harvard University Press. Taylor, Edward J; Zezza, Alberto; and Gurkan, AliArslan (2008). "Rural poverty and Markets. Background paper for the IFAD Rural Poverty Report 2011". March 31.