Navigation – Plan du site

AccueilNumérosHors-sérieNouvelles questions, nouveaux objetsInterdisciplinarity and the futur...

Nouvelles questions, nouveaux objets

Interdisciplinarity and the future of development studies after the 2019 Nobel Prize in economics

Cristiano Lanzano, Cecilia Navarra et Elena Vallino
p. 315-329

Résumés

Le prix Nobel 2019 pour l’économie a consacré l’hégémonie de l’œuvre de A. Banerjee, E. Duflo et M. Kremer sur l’économie du développement. La place qu’ils consacrent aux expériences de terrain et aux essais randomisés contrôlés (RCT) s’adapte bien à l’accent sur les politiques « fondées sur les preuves » et sur l’évaluation qui a dominé l’aide au développement dans la même période. Nous nous inspirons ici des débats générés par l’attribution du prix pour réfléchir sur l’avenir des études du développement. Alors que la posture empirique de ces économistes aurait pu stimuler le dialogue interdisciplinaire, leur approche expérimentale a eu tendance à marginaliser des approches et des méthodes concurrentes – surtout celles qui caractérisent l’anthropologie et les autres sciences sociales qualitatives. Cette dynamique a renforcé les tendances « impérialistes » dans la discipline économique, auxquelles on devrait opposer un engagement renouvelé vers le pluralisme dans le champ des études du développement.

Haut de page

Texte intégral

A debated award: RCTs in development economics and in the development sector

  • 1 The actual name of the prize is the Sveriges Riksbank Prize in Economic Sciences in memory of Alfr (...)

1In 2019, the Nobel Prize in economics1 was awarded to development economists A. Banerjee, E. Duflo and M. Kremer. The award was highly debated. On the one hand, it was welcomed by many as an opportunity for a renewed visibility of the themes of development and poverty reduction in the global South. On the other hand, it consecrated the awardees’ decade-long hegemony in development economics.

2The Royal Swedish Academy of Sciences (2019) motivated the choice with the argument that the three recipients had “introduced a new approach to obtaining reliable answers about the best ways to fight global poverty”. The reference here is especially to the application to development economics of Randomized Control Trials (RCTs), an experimental method to evaluate the impact of policy interventions. RCTs are based on the idea of having a “treatment group” on which a certain policy is tested, and a “control group” that is as similar as possible to the former, and only differs from it in the fact of not being exposed to the policy. This methodology is borrowed from the medical sciences, where it has been often used to test the effects of medical treatments and drugs. It is particularly associated with “evidence-based medicine”, a paradigm which emerged in the early 1990s, expanding the approach of clinical epidemiology – the application of epidemiological findings and methods to the clinical treatment of individual patients – and relying to a significant extent on systematic reviews, meta-analyses and, in particular, RCTs (see Lambert, 2006; Solomon, 2015).

3The three 2019 Nobel awardees championed the application of RCTs in development economics, which earned them the nickname of randomistas. Beginning in the 1990s, through M. Kremer’s work in specific domains, and being applied more broadly during the following decade thanks to the lead taken by A. Banerjee and E. Duflo, the influence of this method expanded radically in economic research. In 2003, A. Banerjee and E. Duflo founded, together with their colleague S. Mullainathan, the Abdul Latif Jameel Poverty Action Lab (J-PAL), dedicated explicitly to randomized impact evaluations and to addressing critical questions relating to poverty (Kvangraven, 2020a). The approach became increasingly popular in research, in policy evaluations and in education curricula in development, up to the point at which the Sveriges Riksbank Prize Committee argued that the “experimental research methods now entirely dominate development economics” (Kvangraven, 2020b). E. Duflo and A. Banerjee have stated, on various occasions, that many failures in development policies in the past were the consequence of an ideology-driven development research and agenda (Banerjee, 2005; Banerjee and Duflo, 2011). They claim that the use of experiments may help to detach the researcher from ideological debates, and to produce more objective findings. The method would solve the bottleneck of unknown confounding factors that impede a clear understanding of the relationship between independent and dependent variables. In fact, the Nobel laureates, along with many other scholars, claim that by conducting randomized interventions over many units (for example households, hospitals, schools), the causal impact of the intervention may be extrapolated (Kvangraven, 2020a; Deaton and Cartwright, 2018).

4It is widely recognized that this method has gained a hegemonic position in development economics in recent years (Akram-Lodhi, 2014) and has often been defined as a “gold standard” (Cartwright, 2011; Ravallion, 2020). A. Banerjee, E. Duflo and M. Kremer (2016) themselves argue that RCTs have altered the academic discipline of development economics and provide metrics to quantify their appearance in academic publications, conferences and impact evaluations. RCTs have been highly successful in different domains of development economics (Van der Meulen Rodgers et al., 2020; Barrett and Carter, 2020), including in agricultural (de Janvry and Sadoulet, 2019; De Brauw and Hoffmann, 2020), health, education (Banerjee and Duflo, 2011) and conservation studies (Asquith, 2020).

5At the same time, the randomista approach has received considerable criticism, including from fellow economists. Firstly, part of the criticism deals with the types of questions that it is possible to address through RCTs. The method appears more adapted to answering localized, micro-level questions dealing with incentives at individual level in the fight against poverty, rather than to addressing questions of large-scale structural transformations of a society (Stevano, 2020). Secondly, many scholars observe that, far from being atheoretical as they are presented, RCTs do not substantially depart from neoclassical assumptions on the maximizing behaviour of individuals, and this has consequences on the experimental design. Indeed, while behavioural economics is often represented as an innovation with respect to neoclassical theory, there are limited traces of its heterodox aspects in the RCT framework (Fine et al., 2016). Thirdly, once RCTs have identified the underlying causal mechanisms leading to the final result, they are not able to explain them (“why” the cause leads to the effect): this limitation may represent an obstacle to the application of their results in other contexts, e.g. the “external validity” problem (Rao et al., 2017; Deaton and Cartwright, 2018).

6The question of the external validity of RCTs has been discussed from early on, and the randomistas have variously positioned themselves with respect to it, sometimes acknowledging the challenges of generalization, and sometimes defending it by proposing technical solutions (see Donovan, 2018). N. Kabeer (2020) analyses some of E. Duflo’s work where she appears to derive general claims from experimental evidence on gender inequalities in India. N. Kabeer argues that the possibility to generalize is also limited by the strong small-scale focus of the study, which disregards broader dynamics. In her work, E. Duflo uses evidence that service jobs have a positive impact on women’s empowerment to conclude that gender inequalities in India are declining. N. Kabeer objects that these jobs are rarely available to women in India and that, on the contrary, women’s participation in the labour market in other sectors is declining despite economic growth. Similarly, E. Duflo generalizes the findings of an RCT that found the same rates of immunization between boys and girls in the Indian state of Rajasthan, concluding that there are no everyday gender inequalities in India in access to healthcare: this conclusion, N. Kabeer observes, fails to acknowledge that, according to the existing literature, Rajasthan represents an exception in this regard.

7One may wonder why, despite the highlighted criticism, the RCT methodology has become so popular in development economics, development studies and international aid. F. Bédécarrats et al. (2020) analyze the supply and demand sides of the arena in which RCTs have become hegemonic. On the supply side, the randomistas deliver an actually innovative “scientific business model”, easy to market and reproduce. On the demand side, the academic environment in economics has produced a favorable ground for RCTs due to a weakened focus on general social structures and power relations, and an increased attention to the micro-foundations of macroeconomics and to quantification efforts in social sciences. RCTs also draw on the success of behavioural and experimental economics, which has gained high status in economics since the Nobel Prize was awarded to D. Kahneman (in 2002), V. Smith (in 2002) and R. Thaler (in 2017).

8At the same time, the aid sector has experienced profound changes, too. At a time when the aid industry has been polarized by the general debate on aid effectiveness – opposing some scholars such as J. Sachs, advocating for more aid, to others, such as W. Easterly and D. Moyo, raising attention to the adverse effects of aid – the emergence of RCTs has shifted the focus to a smaller scale. It has encouraged the search for evidence on what is proven to work and what is not in more circumscribed contexts (Abdelghafour, 2017; de Souza Leão and Eyal, 2020). Simultaneously, the growing influence of “evidence-based policymaking” (see also below) and of “new public management” has called for results-oriented approaches in the delivery of public services, and RCTs have turned out to be the perfect tool for this purpose. Such approaches have been reflected in major international conferences and in guidelines on development assistance since 2005 (Bédécarrats et al., 2020). For the same reasons, disciplines and approaches based on less reproducible practices – for example, qualitative or mixed research methods – or less aligned with the political imperative to quantify, compare and rank, became increasingly marginalized (see below).

9The crisis of the traditional Official Development Assistance (ODA) and the subsequent emergence of decentralized aid (see Bierschenk et al., 2000) has played a role in reinforcing the hegemony of RCTs. The success of the latter reflects a decreased trust in large-scale transformative policies in the international aid community (Pritchett, 2020). With the decreased intensity of the ODA financing, governments have become imbricated in more complex arenas and diverse coalitions of development actors, including NGOs, private businesses and civil society organizations: these actors are oriented to work at a smaller scale and have more urgency to prove the impact of their work (de Souza Leão and Eyal, 2020). The withdrawal of the state as a large planner led to a “thinking small” approach (Stevano, 2020) and to an effort for the delivery of “humanitarian goods” (Redfield, 2012) limited to a given space and time in order to address urgent but narrow questions in poverty, health and education. RCTs gained particular success in a phase where it became evident that poverty persisted during the structural adjustment phase despite the trust in “trickle down” channels: in this phase, programs aimed at poverty reduction gained relevance, while still leaving broadly unchallenged the orientation of macroeconomic policies. As G. Chelwa (2020: 4) has argued, in the post-structural adjustment era, international financial institutions “required governments in Africa to prepare Poverty Reduction Strategy Papers (PRSPs) that were to articulate how governments would protect the welfare of the poor”, while still holding “the view that statist policies were to blame for the crisis. […] It is into this milieu that today’s mainstream version of development economics was born”. The possibility of keeping the focus on the micro level, disconnected from macro factors (as we will discuss later on), is likely to have played a role in the success of RCTs in the donor community.

A failed promise of interdisciplinarity?

10The relationship between RCTs and an intrinsically interdisciplinary field such as development studies is a complex one. Valuing empirical data collection in the field, and allowing the relaxation of some behavioural assumptions of neoclassical economics, the approach heralded by the randomistas could potentially create spaces of interaction with scholars from other disciplines. However, as argued by N. Abdelghafour (2017) in this journal, the position that has been accorded to RCTs within development economics and development practices has ended up marginalizing alternative research methods. Moreover, the centrality of RCTs in development economics has tended to corroborate reductionist explanations of poverty, to reinvigorate methodological individualism at the expense of analysis of structural factors, and to abstract from both context and explicit discussion of theoretical foundations of development questions.

11On the one hand, according to N. Kabeer (2020), the development of RCTs has encouraged development economists to shift from their reliance on econometric modelling using large databases to greater efforts to collect their own first-hand field data. While it should be noted that this renewed attention of development economists to data collection in the field is not exclusive to the randomistas, they contributed to it and are commonly perceived as empiricist enthusiasts in the public debate. E. Duflo, for example, is reported to have declared: “I love everything about [fieldwork]. It is the only way, when you work on development, to get an intuitive sense of how people really live their lives” (quoted in Donovan, 2018: 38). This could potentially favour encounters with other disciplines that rely more traditionally on fieldwork, such as anthropology. In practice, this has happened much less than expected, and only in the rare cases when it has been deliberately pushed for. A positive example is presented by V. Rao et al. (2017), where an in-depth ethnography conducted alongside an RCT (on participation and democracy in the Indian state of Karnataka) provided important insights into the mechanisms of change – where RCTs by themselves are quite silent – and could explain why the RCT itself found no effect of the intervention analysed.

12A second potential pathway to interdisciplinarity could be the crossing paths of the evolution of behavioural economics and RCTs. The “behavioural” approach nuances some rigid assumptions on human behaviour and especially the assumption of instrumental rationality that used to characterize standard microeconomics. A relaxation of these assumptions was not primarily brought about by the application of RCTs to economics. It was, rather, part of a broader change within the discipline, which incorporated the idea of “boundedly rational” individuals (first formulated by H. Simon, 1957), whose decision-making processes cannot be modelled as if they were perfectly informed “utility maximisers” and fully capable of calculations. The development of this approach included an extensive use of experiments – mostly borrowed from social psychology – and randomistas largely built on that background. These approaches now have an important space in development economics, especially since the World Bank devoted the 2015 World Development Report to them (World Bank, 2015). Instead of assuming perfectly rational and self-interested individuals, this publication acknowledges that people’s decisions are influenced by social norms and shared mental models. Applying this principle to development issues, the report argues that “poverty is not only a deficit in material resources but also a context in which decisions are made. It can impose a cognitive burden on individuals that makes it especially difficult for them to think deliberatively” (World Bank, 2015: 13). This statement exemplifies the ambivalent role of behavioural economics – and, relatedly, of RCTs – with respect to interdisciplinarity: while it allows for a more articulate understanding of economic decisions, it also leads to a specific reductionist perspective, interpreting poverty as a problem of behaviour rather than as a result of structural factors. The question then becomes “why do the poor act (seemingly) irrationally?” and “what incentives can make the poor act less irrationally?”, instead of asking what factors make them poor. Following the critique of B. Fine et al. (2016), the particular approach to behavioural economics adopted by RCTs still draws substantially on the rational agent, framed as the benchmark that boundedly-rational individuals should be pushed to reach. This signals a certain continuity, rather than rupture and innovation, with respect to the rational agent model.

  • 2 N. Kabeer discusses some of E. Duflo’s works, which establish a link between individual preferences (...)

13Indeed, a major critique that has been addressed to randomistas is their methodological individualism – which also characterizes neoclassical economics – and the reduction of poverty questions, which are exclusively addressed to the individual behavioural dimension. This critique is developed by S. Stevano (2020), who does not necessarily see a problem in RCTs trying to address “small questions” (meaning local, specific), but criticizes their refraining from searching for “big answers”. Analysing an RCT-based study on nutrition in Morocco, S. Stevano argues that the authors, preoccupied with understanding why the poor do not spend more on food (and why, when possible, they buy better-tasting instead of higher-nutrient food), fail to acknowledge the role of the food industry in shaping tastes, and the rising prices of healthy food. Similarly, N. Kabeer (2020) argues that the randomistas’ focus on individual characteristics and behaviours leads them – in continuity with neoclassical economics – to treat preferences as random and idiosyncratic to the individual, rather than derived from entrenched social constructions.2

14One of the elements of success of RCTs is their claim to use very little theory, i.e. to require minimal assumptions and to be able to operate with little prior knowledge. This claim is meant to position the experiments “outside the social sciences”: it limits de facto the possibility of interdisciplinary dialogue, which presupposes – in order to be productive – identifying and making explicit a common area of definitions, theoretical approaches and concepts. As A. Deaton and N. Cartwright (2018: 2) write, “this is an advantage when persuading distrustful audiences, but it is a disadvantage for cumulative scientific progress, where prior knowledge should be built upon, not discarded”. This attitude also limits, at least according to some critics, the capacity of RCTs to respond to questions about “why results happen”: V. Rao et al. (2017) argue, for example, that impact evaluations have the ability to provide relatively precise estimates of size of the impact of an intervention, but tell us very little about where those impacts come from.

15To be sure, the “insularity” of development economists and their alleged reluctance to engage in interdisciplinary collaboration have been brought up before, as P. Hill’s early critical contribution demonstrates (Hill, 1986). However, the “experimental turn” in development economics may have confirmed, and accentuated, pre-existing tendencies. As RCTs became the dominant methodology both in academic and in applied development economics, their proponents have argued in favour of their “methodological superiority”, thus dismissing findings from non-experimental studies. Several scholars do indeed criticize the delegitimization of other research approaches. Paradoxically, the rare positive examples of collaboration with other disciplines have shown that other methods – and especially qualitative analysis – can allow a better understanding of the results of RCTs, for example in those cases where no impact is found in the experimental setting (Rao et al., 2017).

Evidence and context in anthropology

16An explicit goal – and, according to their proponents, a major selling point – of RCTs is to “standardize by removing the background noise that makes regions and datasets incomparable” (Brives et al., 2016: 370). As N. Kabeer (2020) also observes, RCTs are structurally indifferent to historical and contextual specificity. Their very ambition is to reduce “uncertainty” (Donovan, 2018) by explaining what works in development: in other words, to identify a “clean” causality between a policy and an effect, by ruling out all confounding factors, including elements that could be context-specific. All these factors are “absorbed by the control group” (Abdelghafour, 2017).

17This emphasis on identifying the crude causal links in development and social change by avulsing them from their political-economic and socio-cultural context on different scales raises, understandably, sensitive epistemological issues not only for economists, but also for other social scientists and especially for anthropologists. Ultimately, it questions the ways in which different disciplines incorporate empirical work and relate to “truth”, “objectivity” and (social) “facts” – however defined. In anthropology, feminist, postmodern and postcolonial approaches have variously challenged the neutrality of the knowledge produced by the discipline, and put the authority that scholars draw from practising ethnography under scrutiny, from as early as the 1970s. However, anthropologists started to reflect more systematically on the notion of “evidence” in their discipline only a few decades later (and to a surprisingly limited extent, at least according to V. Dominguez, 2013), in consequence of the push towards “evidence-based policy”.

18Reflecting on the relations between knowledge and evidence, K. Hastrup (2004) reassesses the case for abandoning traditional notions of objectivity, but maintains that anthropologists are nonetheless entitled to the aspiration of “getting it right” and to the pursuit of knowledge that, like any knowledge, is necessarily reductive and selective. Confronted with the challenge of theorizing from their situated observation point, most anthropologists do not give up the ambition of establishing connections and associations between the phenomena they study, and of identifying patterns in the social and cultural realities with which they engage personally in the field – that is, of producing knowledge. Yet, intimately linked to ethnographic fieldwork as it is, the production of anthropological knowledge is essentially relational. M. Engelke (2008) observes that it is precisely the explicitly intersubjective and personal nature of ethnographic methods that poses the question of how anthropologists turn their experience of fieldwork into objects of evidence.

19This question is particularly relevant when anthropologists are, in conversation with other disciplines or in interaction with other professional settings, expected to contribute more or less directly to decision-making with some form of empirical knowledge that is accepted as sufficiently solid and “evidentiary”. The involvement of anthropological expertise in legal procedures, for example, confronts anthropologists with the different and often conflicting mechanisms of validation of the “truth” that they are supposed to represent or with which they must interact. This may take place in relation to the testimonies of asylum seekers struggling to obtain a residence permit from state authorities (see Good, 2007; Fassin, 2012), to accusations of witchcraft brought to court during a formal legal dispute (see Fancello and Bonhomme, 2018; and the other contributions in the same issue), or in many other instances. But it is probably in the subfield of medical anthropology that this reflection has developed most substantially. The emergence, since the early 1990s, of “evidence-based medicine”, with its reliance on systematic reviews, meta-analyses and RCTs, has triggered various reactions. While criticizing the evidence-based turn, H. Lambert (2009) argues that it provides, for the first time since the “crisis of representation” of the 1980s, an opportunity for anthropologists to move beyond their deconstructivist assumptions and engage with the criteria for the validation of evidence that characterize their work while remaining largely implicit. S. Ecks (2008) provides a similar argument and provocatively formulates a proposal for an “evidence-based medical anthropology” that is equipped to question, when necessary, the epistemology of evidence-based medicine and the causal relations it identifies.

20The relational character of the knowledge produced through fieldwork is all the more evident in the anthropology of development and social change. As D. Mosse (2014) argues, reflecting on a rural development programme among Adivasi groups in western India, this relationality is highlighted in the confluence of the knowledge produced by the different actors – scholars, experts and development workers, “local” farmers whose agricultural knowledge was elicited through participatory schemes – and by the selective and strategic use that these make of the programme. Turning back to the question of context, D. Mosse underlines how development programmes, and their failure, often revolve around mechanisms of disembedding knowledge from its original context (such as the professional trajectories shaping the knowledge of official experts, or the networks of obligations and relations constituting Adivasi cultivation practices) and re-embedding it in new social and institutional situations:

Knowledge claimed as universal […] is actually embedded, albeit in unacknowledged ways, in the particular prejudices and structures of the originating policy-making institutions, and has to lose (or hide) its context and history to become relevant as international development policy. (Mosse, 2014: 518)

21Partly because of the particular aptitude of ethnography for capturing “excess” or surplus information (a point that both K. Hastrup and M. Engelke re-elaborate from V. Das, 1998), anthropologists can claim to be well equipped to provide understanding of the social and cultural contexts within which development projects, medical treatments and other policy interventions occur. Conversely, they often lament the elision of context produced by the push toward evidence-based policy and its methods – such as RCTs. In recent years, many development anthropologists, inspired by R. Rottenburg’s (2009) fictionalized ethnography of development aid in East Africa, have reflected on the “traveling models” that dominate development interventions in the era of standardization and new public management (Bierschenk, 2014; Mosse, 2014; Olivier de Sardan et al., 2017; see also Le Meur in this issue). In his last book, J.-P. Olivier de Sardan defines travelling models as “any standardized institutional intervention […] aimed at producing social change, and relying on a ‘mechanism’ and on ‘dispositifs’ supposed to possess intrinsic properties that enable them to generate this change in various implementation contexts” (Olivier de Sardan, 2021: 26, our translation). The promotion of performance-based financing to improve healthcare services or the creation of cash transfer programmes modelled on the Brazilian and Mexican early cases are examples of these travelling models, whose trajectories consist of a continuous process of decontextualization and recontextualization. Methodologies of evaluation (turned into research methods) such as RCTs are both a product and a driver of this tendency, relying on a similar principle of “testing the intrinsic effectiveness of a mechanism, while neutralizing all the effects of the contexts of implementation” (ibid.: 71, our translation). To be sure, evidence-based policies and RCTs do acknowledge the contexts of intervention or of experimentation, incorporating them in the form of quantitative variables, statistics and socio-demographic data. This “structural context”, in J.-P. Olivier de Sardan’s terms (“representational context” according to C. Brives et al., 2016), is needed, to a certain extent, precisely to ensure comparability and clearly identify causal relations. What this operation leaves out entirely is context in a broader and more dynamic understanding of the notion: the “pragmatic” (or “interactional”) context constituted by the perceptions, strategies, practices and interactions of the actors and groups involved in the conception, translation, appropriation and implementation of a policy.

22In the critical view of these scholars, then, the contexts of policy interventions are not simply the prerogative of qualitative methods and anthropological research, to be used at best in feasibility studies or to explain away possible failures and implementation gaps. Context is itself an irreplaceable component of evidence, if by evidence we mean an empirically sound knowledge of how development processes do work in practice, and not only of how they are imagined to work.

Beyond the “imperialism” of economics

23As we have seen, the hegemony of RCTs in development economics appears to be in substantial continuity with some characteristics of the economic discipline, or at least of its dominant (or “mainstream”) approaches. On the one hand, RCTs rely on the methodological individualism of economics, which still largely inspires economics even after the discipline has been substantially influenced by the “behavioural shift”. As said above, rationality still remains the benchmark against which behaviours (and especially the behaviours of poor people) are identified. According to M. Fana and L. Giangregorio (2019), the interpretation of poverty as a fundamentally behavioural issue is in continuity with L. Robbins’ epistemological approach to economics, which he defined as the discipline studying human behaviour as a relationship between ends and scarce means which have alternative uses. This early “formalistic” turn can be seen as a milestone of the “imperialistic” tendencies of economics (tendencies that can be traced back to A. Smith, according to R. Marchionatti and M. Cedrini, 2017, and to the marginalist revolution, according to B. Fine and D. Milonakis, 2009), marking a distance from the other social sciences and attempting to use the tools of the natural sciences. This turn was not without opposition; among others, J.M. Keynes maintained that economics belongs to the human sciences – intended as those disciplines that deal with human beings in their social environment, including their motives and expectations, with values and introspection (Marchionatti and Cedrini, 2017).

24Relatedly, the criticism spurred by the ambition of RCTs to provide a technical method for identifying causal relations without building theoretical assumptions or developing any knowledge of the context, which we evoked in the previous sections, is not entirely new in economics. An earlier controversy involved, again, J.M. Keynes, who engaged with J. Timbergen in a debate on econometrics at the end of the 1930s. J. Timbergen argued that econometrics could discover “what causes are operative and how strongly each of them operates” (Garrone and Marchionatti, 2007: 7). Conversely, J.M. Keynes argued that only if “the economist has correctly analysed beforehand the qualitative character of the causal relations” (Keynes, quoted in Garrone and Marchionatti, 2007: 11) can the econometrician then address their quantitative importance and estimate how strongly each variable operates. J.M. Keynes did not oppose econometric work per se. Rather, he criticized the attempts at statistical inference without any prior effort at evaluating the suitability of the economic material for making such inferences and at selecting the relevant factors of a phenomenon, by taking the characteristics of the historical world into account (Garrone and Marchionatti, 2007) – and therefore considering qualitative and contextual knowledge.

  • 3 The notion of imperialism contributes in other ways to the critique of economics. With the increase (...)

25The imperialism of economics3 is usually defined as a form of disciplinary expansionism that pushes for the application of economic approaches to human behaviour to territories that lie outside the traditional domain of the discipline of economics (see Mäki, quoted in Marchionatti and Cedrini, 2017). For example, A.N. Sindzingre (2016) describes how, with the decline of Marxism and other “grand” theories started in the 1980s, economic anthropology gradually became marginalized, while its traditional objects of study tended to be absorbed by an economic science increasingly dominated by the neoclassical paradigm and the mathematization of its language. E.P. Lazear (2000: 99-100) explains the successful expansion of economics beyond its traditional borders with “a rigorous language that allows complicated concepts to be written in relatively simple, abstract terms. The language permits economists to strip away complexity. Complexity may add to the richness of description, but it also prevents the analyst from seeing what is essential”. What we have tried to argue here is that the (presumed) essential character of certain forms of knowledge, as opposed to the inessentiality of others, is not a given; instead, it reflects theoretical assumptions – and power relations in knowledge production – that should be made explicit.

  • 4 For example, articles in the American Political Science Review cite the top 25 economics journals m (...)

26M. Fourcade et al. (2015) underline both the insularity and the rigid hierarchy that characterize economics as enabling factors in explaining why economics tends to perceive itself as “superior” to other social sciences. Economists have started to consider topics that are more traditionally associated with other disciplines, yet cross-disciplinary citation patterns continue to offer evidence of its relative insularity.4 This reflects the fact that economics still appears to give lower value to interdisciplinarity than other social sciences (Marchionatti and Cedrini, 2017). However, as also argued by D. Rodrik (2015), there is a certain degree of heterogeneity within economics that should not be disregarded, as also highlighted in recent years by repeated calls for the diversification of curricula.

27As we observed earlier, many of the criticisms addressed to RCTs are in reality connected to broader challenges in the relationship between the discipline of economics and other social sciences in the field of development studies – and with the need to recognize economics more explicitly as part of the social sciences. We do not argue in favour of the dismissal of RCTs: however, in line with Keynes’ argument about the emergence of econometrics, we underline that they are partial tools that cannot replace other methods – including qualitative analysis – but only complement them (as in the proposal by R.M. Gisselquist, 2020, to integrate RCTs and case studies more systematically).

28Similarly, we argue that the debate around RCTs should lead to a stronger call for a truly democratic exchange between economics and the other social sciences (Marchionatti and Cedrini, 2017). If a polyphonic and interdisciplinary conversation on development and social change can resume, it will require increased theoretical eclecticism (Chelwa, 2020) and methodological pluralism.

Acknowledgments

29The authors would like to thank some colleagues and friends – particularly Andrea Ghione, Paola Minoia, Pietro Ortoleva and Sara Romanò – for participating with different views in the informal discussions about the Nobel Prize that led to the idea of this article, and the two anonymous reviewers for their feedback. Obviously, all shortcomings are our own.

30C. Lanzano thanks the Nordic Africa Institute (Uppsala, Sweden). C. Lanzano’s work is also supported by the Belmont Forum and NORFACE Joint Research Programme on Transformations to Sustainability (via the research project “Gold Matters. Sustainability Transformations in Artisanal and Small-scale Gold Mining”, grant number 462.17.201, co-funded by DLR/BMBF, ESRC, FAPESP, ISC, NWO, VR, and the European Commission through Horizon 2020); and by Vetenskapsrådet (via the research project “Soft infrastructures: Labour mobilities across and between secondary cities in West Africa”, project number 2018-04602).

Haut de page

Bibliographie

Abdelghafour N., 2017, “Randomized Controlled Experiments to End Poverty? A Sociotechnical Analysis”, Anthropologie & développement, n° 46-47, pp. 237-264.

Akram-Lodhi H., 2014, “Review of Poor Economics: A Radical Rethinking of the Way to Fight Global Poverty”, The Journal of Peasant Studies, n° 41(3), pp. 426-429.

Asquith N., 2020, “Large-scale randomized control trials of incentive-based conservation: What have we learned?”, World Development, n° 127, 104785.

Banerjee A., 2005, “New Development Economics and the Challenge to Theory”, Economic and Political Weekly, n° 40, pp. 4340-4344.

Banerjee A., Duflo E., 2011, Poor Economics: A Radical Rethinking of the Way to Fight Global Poverty, New York, Public Affairs.

Banerjee A., Duflo E., Kremer M., 2016, “The Influence of Randomized Controlled Trials on Development Economics Research and on Development Policy”, paper presented at The state of economics, the state of the world, The World Bank, 8-9 June 2016.

Barrett C.B., Carter M.R., 2020, “Finding our balance? Revisiting the randomization revolution in development economics ten years further on”, World Development, n° 127, 104789.

Bédécarrats F., Guérin I., Roubaud F., 2020, “Controversies around RCT in Development. Epistemology, Ethics, and Politics”, Working Paper DT/2020/15, DIAL, Paris.

Bierschenk T., 2014, “From the Anthropology of Development to the Anthropology of Global Social Engineering”, Zeitschrift für Ethnologie, n° 139, pp. 73-98.

Bierschenk T., Chauveau J.P., Olivier de Sardan J.P. (eds.), 2000, Courtiers en développement. Les villages africains en quête de projets, Paris, Karthala.

Brives C., Le Marcis F., Sanabria E., 2016, “What’s in a Context? Tenses and Tensions in Evidence-Based Medicine”, Medical Anthropology, n° 35(5), pp. 369-376.

Cartwright N., 2011, “A Philosopher’s View of the Long Road from RCTs to Effectiveness”, The Lancet, n° 377(9775), pp. 1400-1401.

Chelwa G., 2020, “Pop Developmentalism in Africa”, CODESRIA Bulletin, n° 1, pp. 3-5.

Das V., 1998, “Wittgenstein and Anthropology”, Annual Review of Anthropology, n° 27, pp. 171-195.

Deaton A., Cartwright N., 2018, “Understanding and Misunderstanding Randomized Controlled Trials”, Social Science and Medicine, n° 210, pp. 2-21.

De Brauw A., Hoffmann V., 2020, “The influence of the 2019 Nobel Prize winners on agricultural economics”, World Development, n° 127, 104793.

Dominguez V., 2013, “On the Paradoxes of Evidence in Contemporary Anthropology”, American Anthropologist, n° 115(4), pp. 643-645.

Donovan K., 2018, “The rise of the randomistas: on the experimental turn in international aid”, Economy and Society, n° 47(1), pp. 27-58.

Ecks S., 2008, “Three propositions for an evidence-based medical anthropology”, Journal of the Royal Anthropological Institute, N.S., pp. S77-S92.

Engelke M., 2008, “The objects of evidence”, Journal of the Royal Anthropological Institute, N.S., pp. S1-S21.

Fana M., Giangregorio L., 2019, “La povertà dei Nobel per l’economia”, Jacobin Italia, 16 October 2019, https://jacobinitalia.it/la-poverta-dei-nobel-per-leconomia/ (accessed 4 August 2021).

Fancello S., Bonhomme J., 2018, “L’État et les institutions face à la sorcellerie”, Cahiers d’études africaines, n° 231-232, pp. 573-591.

Fassin D., 2012, Humanitarian Reason: a Moral History of the Present, Berkeley, University of California Press.

Fine B., Johnston D., Santos A.C., Van Waeyenberge E., 2016, “Nudging or fudging: The World Development Report 2015”, Development and Change, n° 47(4), pp. 640-663.

Fine B., Milonakis D., 2009, From Economics Imperialism to Freakonomics: The Shifting Boundaries Between Economics and Other Social Sciences, London, New York, Routledge.

Fourcade M., Ollion E., Algan Y., 2015, “The superiority of economists”, Journal of Economic Perspectives, n° 29(1), pp. 89-114.

Garrone G., Marchionatti R., 2007, “Keynes, statistics and econometrics”, Working Paper n° 3, Centro di Studi sulla Storia dei Metodi dell’Economia Politica, Dipartimento di Economia Cognetti de Martiis, Università degli Studi di Torino.

Gisselquist R.M., 2020, “How the cases you choose affect the answers you get, revisited”, World Development, n° 127, 104800.

Good A., 2007, Anthropology and Expertise in the Asylum Courts, Oxon, Routledge-Cavendish.

Hastrup K., 2004, “Getting it right. Knowledge and evidence in anthropology”, Anthropological Theory, n° 4(4), pp. 455-472.

Hill P., 1986, Development Economics on Trial. The Anthropological Case for a Prosecution, Cambridge, Cambridge University Press.

Hoffmann N., 2018, “When are experiments corrupt?”, Journal of Contemporary African Studies, n° 36(4), pp. 532-552.

Janvry A. de, Sadoulet E., 2019, “Jumping the fence: How experimental research in agriculture has gone from lab to field”, University of California at Berkeley, November 2019.

Kabeer N., 2020, “Women’s Empowerment and Economic Development: A Feminist Critique of Storytelling Practices in ‘Randomista’ Economics”, Feminist Economics, n° 26(2), pp. 1-26.

Kvangraven I.H., 2020a, “Nobel Rebels in Disguise – Assessing the Rise and Rule of the Randomistas”, Review of Political Economy, n° 32(3), pp. 305-341.

Kvangraven I.H., 2020b, “Impoverished economics? A critical assessment of the new gold standard”, World Development, n° 127, 104813.

Lambert H., 2006, “Accounting for EBM: Notions of evidence in medicine”, Social Science and Medicine, n° 62, pp. 2633-2645.

Lambert H., 2009, “Evidentiary truths? The evidence of anthropology through the anthropology of medical evidence”, Anthropology Today, n° 25(1), pp. 16-20.

Lazear E.P., 2000, “Economic Imperialism”, The Quarterly Journal of Economics, n° 115(1), pp. 99-146.

Le Meur P.Y., 2021, « Industrie minière et fabrique voyageuse du développement », Anthropologie & développement, this issue.

Marchionatti R., Cedrini M., 2017, Economics as Social Science. Economics imperialism and the challenge of interdisciplinarity, London, New York, Routledge.

Mosse D., 2014, “Knowledge as Relational: Reflections on Knowledge in International Development”, Forum for Development Studies, n° 41(3), pp. 513-523.

Olivier de Sardan J.P., 2021, La revanche des contextes. Des mésaventures de l’ingénierie sociale, en Afrique et au-delà, Paris, Karthala.

Olivier de Sardan J.P., Diarra A., Moha M., 2017, “Travelling models and the challenge of pragmatic contexts and practical norms: the case of maternal health”, Health Research Policy and Systems, n° 15(60), pp. 71-87.

Pritchett L., 2020, “Randomizing Development: Method or Madness?”, in F. Bédécarrats, I. Guérin and F. Roubaud (eds.), RCT in Development: A Critical Perspective, Oxford, Oxford University Press, pp. 79-107.

Ravallion M., 2020, “Highly Prized Experiments”, World Development, n° 127, 104824.

Rottenburg R., 2009, Far-Fetched Facts. A Parable of Development Aid, Cambridge, MA, MIT University Press.

Solomon M., 2015, Making Medical Knowledge, Oxford, Oxford University Press.

Souza Leão L. de, Eyal G., 2020, “Searching under the streetlight: A historical perspective on the rise of randomistas”, World Development, n° 127, 104781.

Stevano S., 2020, “Small development questions are important, but they require big answers”, World Development, n° 127, 104826.

Rao V., Ananthpur K., Malik K., 2017, “The Anatomy of Failure: An Ethnography of a Randomized Trial to Deepen Democracy in Rural India”, World Development, n° 99, pp. 481-497.

Redfield P., 2012, “Bioexpectations: Life technologies as humanitarian goods”, Public Culture, n° 24/1(66), pp. 157-184.

Rodrik D., 2015, Economic Rules: The Rights and Wrongs of the Dismal Science, New York, London, Norton.

Royal Swedish Academy of Sciences, 2019, “Press Release: The Prize in Economic Sciences 2019”, 14 October 2019, https://www.nobelprize.org/prizes/economic-sciences/2019/press-release (accessed 4 August 2021).

Simon H., 1957, Models of Man, New York, Wiley.

Sindzingre A.N., 2016, “L’économie peut-elle absorber les autres sciences sociales ? La pertinence des concepts de l’anthropologie”, Afrique Contemporaine, n° 2, pp. 157-171.

Van der Meulen Rodgers Y., Bebbington A., Boone C., Dell'Angelo J., Platteau J.P., Agrawal A., 2020, “Experimental approaches in development and poverty alleviation”, World Development, n° 127, 104807.

World Bank, 2015, The World Development Report 2015: Mind, Society, and Behavior, Washington DC.

Haut de page

Notes

1 The actual name of the prize is the Sveriges Riksbank Prize in Economic Sciences in memory of Alfred Nobel”.

2 N. Kabeer discusses some of E. Duflo’s works, which establish a link between individual preferences of women and outcomes in terms of children’s welfare. In a feminist perspective, it is not the preferences of women per se that drive the empirical associations, but the socially assigned roles that has to be seen in context.

3 The notion of imperialism contributes in other ways to the critique of economics. With the increase of critical analyses of the power relations characterizing the global politics of knowledge production – especially in area studies and development – and the multiplication of calls for the decolonization of academia, the hegemony of the randomistas has invigorated discussions on the Eurocentrism and the ethics of economic research in countries of the global South. N. Hoffmann (2018), for example, describes a controversial policy experiment in Kenyan schools, whose theoretical and ethical fallacies evoked, to a certain extent, the power asymmetries and the objectification that characterized colonial science (see also the special issues consecrated by the CODESRIA Bulletin in 2020 to African critical perspectives on development economics and on the 2019 Nobel prize). However, for the purpose of our argument, here we limit ourselves to a specific understanding of “imperialism” as the expansionism of economics at the expense of other disciplines.

4 For example, articles in the American Political Science Review cite the top 25 economics journals more than 5 times as often as articles in the American Economic Review cite the top 25 political science journals (Fourcade et al., 2015).

Haut de page

Pour citer cet article

Référence papier

Cristiano Lanzano, Cecilia Navarra et Elena Vallino, « Interdisciplinarity and the future of development studies after the 2019 Nobel Prize in economics »Anthropologie & développement, Hors-série | -1, 315-329.

Référence électronique

Cristiano Lanzano, Cecilia Navarra et Elena Vallino, « Interdisciplinarity and the future of development studies after the 2019 Nobel Prize in economics »Anthropologie & développement [En ligne], Hors-série | 2021, mis en ligne le 01 juin 2022, consulté le 22 juin 2025. URL : http://journals.openedition.org/anthropodev/1322 ; DOI : https://doi.org/10.4000/anthropodev.1322

Haut de page

Auteurs

Cristiano Lanzano

Senior Researcher, The Nordic Africa Institute, Uppsala (Sweden); cristiano.lanzano[at]nai.uu.se

Cecilia Navarra

Cecilia Navarra is a policy analyst in the European Parliamentary Research Service (EPRS), the internal research service and think tank of the European Parliament (Brussels, Belgium). She is writing in a personal capacity and any views expressed do not represent an official position of the Parliament; cecilia.navarra[at]gmail.com

Elena Vallino

Research Fellow, Politecnico di Torino, Turin Center on Emerging Economies, Collegio Carlo Alberto (Italy); elena.vallino[at]polito.it

Haut de page

Droits d’auteur

CC-BY-4.0

Le texte seul est utilisable sous licence CC BY 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search