Methodological reflections on ‘land grab’ databases and the ‘land grab’ literature ‘rush’

Abstract

This article focuses on two issues. First, it critically engages with the emerging global databases on large-scale land deals and the ‘data’ on global trends so far circulated, paying special attention to their accuracy and reliability, and the extent to which they represent an instance of ‘false precision’. The usefulness of these datasets and some of the possible side effects of their abuse or bad use are also critically explored. Second, the article considers some important methodological problems that may lie behind the fast-growing research on ‘land grabs’. In particular, some of the untested assumptions implicit in many of the existing studies as well as the problematic dichotomies that often influence researchers' interpretations of available evidence will be critically discussed, emphasizing the need for greater ‘reflexivity’ in current ‘land grab’ research. The paper concludes that the need for methodological rigour is not a luxury, but rather politically and tactically crucial from the point of view of those who campaign against real dispossession and exploitation. There is no shortcut for good quality evidence.

Keywords:

• land grabs
• methods
• evidence
• bias

Notes

This is a revised and expanded version of a paper presented at the Plenary Session Roundtable on Methodologies: Identifying, Counting and Understanding, International Academic Conference on Global Land Grabbing II, organized by the Land Deals Politics Initiative (LDPI), Cornell University, 17–19 October 2012. I would like to thank Saturnino ‘Jun’ Borras for inviting me to contribute to this forum. I have also benefited from the very useful comments, criticisms and suggestions of several people, including Saturnino ‘Jun’ Borras, Marc Edelman, Christopher Cramer and John Sender, as well as two anonymous referees. I am solely responsible for the analysis, and any errors, in the final paper.

¹An online debate on these claims showed how superficial and casual the empirical analysis contained in the Oxfam report was. See Collin (2013).

² A necessary caveat is in order. While GRAIN is a small non-governmental organisation (NGO) with very limited resources and mainly focused on food-related issues, the Land Matrix is a product of the International Land Coalition in partnership with other organizations, including an impressive list of large NGOs, international financial institutions like the World Bank and IFAD (International Fund for Agricultural Development), intergovernmental organizations like FAO (Food and Agriculture Organization) and activist networks. Moreover, the Land Matrix is an ongoing initiative while the GRAIN database was a one-off initiative designed to kick-off the counting of land deals.

³In the course of the conference, in different debates, it was also pointed out that terms like ‘land rush’ were more appropriate and less ideologically-charged than ‘land grabs’ whose main advantage was polemical punch (see Li 2012). The definitions of the term may vary but they usually refer to large-scale land transactions (usually above 200 or 1000 hectares) with some criteria to qualify as ‘grabs’ (see Oxfam 2012, box 4). In fact, the term ‘land grab’ is still widely used and is indeed preferred among activist networks and NGOs. I will use the term in this paper because this is what is most widely used and understood, although it remains a very problematic term. See also a recent discussion of the problems with current definitions and some alternative options (Borras et al. 2012).

⁴On the idea of ‘killer facts’ and how to create them, see Green (2012).

⁵The speed of growth of academic literature on this topic is astonishing. The number of published articles and special issues in well-known journals since 2009 is remarkable. Journal editors (particularly in the Journal of Peasant Studies and the Journal of Agrarian Change) and academic referees are being literally flooded by numerous submissions on ‘land grabs’ and related topics.

⁶The literature ‘rush’ on ‘land grabs’ is analogous to the literature ‘rush’ on ‘China in Africa’.

⁷See Thomas and Mohan (2007) for an excellent collection of essays on the art and limitations of fast fact-finding.

⁸To access papers, especially by younger scholars, see http://www.cornell-landproject.org/activities/2012-land-grabbing-conference/papers/

⁹See various contributions to this debate on Reddy's website, http://www.sanjayreddy.org/ A similar argument could be made about the precision of estimates of the global number of malnourished children (Svedberg 1999).

¹⁰See also caveats about possible biases as presented by the Land Matrix portal (Land Matrix 2013) on http://landportal.info/landmatrix

¹¹See also a similar situation in other fields, e.g. Cramer and Goodhand (2011), on the way this production of ‘factoids’ or institutionalised facts happens in the field of violent conflict.

¹²The Land Matrix ‘encourages citizens, researchers, governments, and companies to provide data (Land Matrix 2013) ’: http://landportal.info/landmatrix.

¹³A ‘land grab’ is supposed to do all of the following: violate human rights; fail to consult affected people; ignore requirements for proper consent; and happen in secret. It is certainly not clear that a great proportion of the deals reported in GRAIN or Land Matrix fulfil all these criteria. Moreover, evidence on the criteria is in itself difficult to obtain and/or verify. See also Borras et al. (2012) on the problems with some of these definitions.

¹⁴ Comparisons between global positioning system (GPS) measurements and farmers' responses to questionnaires about the size of the plots they cultivate also reveal large discrepancies, especially in the case of larger farmers (Carletto et al. 2011).

¹⁵Thus some of the emerging generalisations about ‘land grabs’, which essentially refer to these contexts, may not work in other regions, like Latin America (Borras et al. 2012).

¹⁶See Bräutigam's post (Bräutigam 2012). See also Hanlon (2011) on Mozambican land grab myths.

¹⁷See Rinat (2013).

¹⁸I discovered this in the roundtable at the Cornell University conference.

¹⁹Duncan Green is Oxfam GB's Senior Strategic Adviser and runs a very popular blog: www.oxfamblogs.org/fp2p/

²⁰Study also related to Deininger (2011).

²¹Many researchers will obviously also search for particular cases because of a previously existing professional interest in or knowledge of a particular area.

²² See Datt et al. (1999).

²³ There are some organizations that have contributed much to this kind of ‘grey’ literature in the form of general and country research reports, sometimes with dubious or problematic research methods. The Oakland Institute is a good example of this. See reports in http://www.oaklandinstitute.org/land-deals-africa-publications

²⁴One frequently cited study (Schoneveld et al. 2010) actually produces very little information on employment, much of it based on assumptions about labour needs per hectare and without a clear explanation of the sampling methods used. The study focuses on multiple aspects and essentially gathers information from geographical remote sensing data, employers and key informants, as far as one can tell.

²⁵This also refers to another study that concludes that ‘jobs tend to be unskilled, short-term and small in number relative to the size of the investment’ (Vermuelen and Cotula 2010, 912).

²⁶See Cramer et al. (2008) and Oya (2013) for a discussion of methodological challenges in rural surveys of wage employment and different findings. See also Maertens et al. (2011) for related quantitative evidence that contradicts the typical claims on jobs in the ‘land grabs’ literature.

²⁷Some of these debates are noted in Akram-Lodhi and Kay (2010), Bernstein (2010), and Borras et al. (2012, 2013).

²⁸See Borras et al. (2012) for a more elaborate discussion of the inadequacy of relying only on land size as a measure of scale. For example, flower farms are a good example of large capitalist agribusiness on relatively small size plots given the capital invested per hectare.

²⁹The latest Oakland Institute report on Ethiopia presents this kind of argument about the influx of workers in the new farms (OI 2011, 37).

³⁰Cotula (2012) reminds us that in the land grab databases and literature, domestic ‘grabbers’ are not given the same degree of attention as foreign ‘grabbers’. See also Cramer et al. (2008) for comparative survey evidence on this question in the context of Mozambique.

³¹A related call is made by Li (2011).

³²See Hanlon's reflection on the exaggerations of land grab data in Mozambique (Hanlon 2011). See also Collier and Venables (2011).

Additional information

Carlos Oya is Senior Lecturer in Political Economy of Development, Development Studies Department, School of Oriental and African Studies, University of London. He has done primary research mostly in Sub-Saharan Africa, particularly in Mozambique, Senegal, Mauritania, Uganda and Ethiopia, focusing on the political economy of agrarian change, capitalist accumulation, rural wage labour and poverty. He is also co-editor of the Journal of Agrarian Change. Email: [email protected]