6.2.1 - Critical Realism
Critical realism is the philosophical approach to science adopted in this study. The approach was pioneered by the work of Bhaskar [1975, 1986], and has recently been actively linked to the field of economics, most notably in the work of Lawson [1997, 2003].
The central ontological argument of critical realism is that a reality exists; however, it is
152 both complex and deep, and therefore, it cannot be reduced to our surface interpretation of it based on immediate experiences. Critical realists conceive of the world as being layered, observing a distinction between the empirical world of experience, and the real world of underlying causal powers, also known as generative mechanisms [Walters and Young, 2001]. Events that take place at the level of the actual cannot be explained solely on the basis of empirical evidence or event regularity, but rather they have resulted from the underlying processes that operate at the level of the real.
Bhaskar [1975] argues that social reality is an open and evolving system comprised of intentional individual agents, who, though being constrained and influenced by the social structures surrounding them, are capable of shaping the structures in which they are embedded. As such, there is an ongoing interaction between structure and agency. Given the open and transitive nature of social reality experiment is rendered impossible, and thus, purely empirical explanations are not adequate for theoretical constructions of this complex reality [Lawson, 2003]. The objective of the critical realist in carrying out economic research is therefore to develop ways of uncovering the causal structures and generative mechanisms, which govern the flux of events in an essentially open world [Lawson, 2003].
There are notable similarities between the critical realist conception of the individual and that characterised in the institutional economics of Commons [Lawson, 1996]. In particular the interaction between agency and structure is representative of the mutual understanding that the two approaches share. Indeed, critical realists have also observed strong similarities between their own ontological position and those adopted by prominent academics within various heterodox economic traditions, such as Post-Keynesian economics and institutional economics [Lawson, 2003, 2006; Davidsen, 2005]. Indeed, based on Lawson’s [2006] assessment of heterodox economics, it may be suggested that the strands of NIE and GVC developed in this study are not inimical to heterodox approaches, thus rendering them suitable for a critical realism methodology.
153 In association with heterodox economic traditions critical realism has also been highly critical of the ontological stance adopted in orthodox165, or ‘deductive’ economics.
Arguments against the ontological basis behind orthodox economics have been led by Lawson [1997, 2003]. These arguments are built upon what critical realists believe to be the ‘epistemic fallacy’ [Bhaskar, 1975] occurring within orthodox economics, through which questions of being [ontology] have been reduced to questions of knowing [epistemology]. Whilst critical realists search for generative mechanisms operating at the
‘real’ or deep level of an ‘open’ social reality, deductive economists highlight event regularity as their mode of explanation [Lawson, 2003, pg 5]. It is argued that the deductive method based on event regularity is only applicable in analysis of ‘closed’
systems, and not the open social reality envisaged by critical realists.
Critical realists have also been active in their critique of neoliberalist economic policies within the developing world [Jones, 2001; Morgan, 2003; Dow and Dow, 2005]. Critical realists argue that the influence of orthodox economic thinking has led to the proliferation of flawed generalisations based on the principle of universality [Lawson, 2003]. As a result,
‘the issue of policy credibility, which would seem to lend itself to context specific analysis, has instead been addressed with externally designed policy rules’ [Dow and Dow, 2005, pg 1131]. In light of this, one of critical realisms strongest attributes is its objective to achieve a deeper understanding of the context in which causal structures operate. Indeed, context is fundamental to the approach of critical realism based on the belief that the outcome produced by a particular causal mechanism is highly dependent on the context in which it is situated [Sayer, 1992; Pawson and Tilley, 1997].
Building on the problems associated with the Washington consensus, the methodology for this study is focused on achieving a deeper understanding of the partial model. Both the institution of the Cocobod and the developing country environment are unique systems, for which the deductive approach of orthodox economics appears to be ill-suited. The cocoa market in Ghana is widely influenced by a variety of social, political, historical and
165 This may be likened to the neo-classical economics that informed the Washington Consensus.
154 economic factors. As such, it appears that critical realism may represent the ideal ontological grounding from which to investigate the partial model. Indeed, critical realisms belief in a complex social reality and commitment to a context specific understanding, present a suitable platform from which to build knowledge of this progressive economic phenomenon.
6.2.2 - Critical Realist Methodology
The methodological approach employed in this study has attempted to work within the guidance of a critical realist ontology; however, at the epistemological level the value of empirical evidence has not been completely refuted. This is considered a moderate interpretation of the epistemological beliefs espoused within critical realism. There is an acceptance that knowledge claims must come about through a give and take between both abstract theorising and concrete evidence about the phenomenon being examined.
Using evidence in this way has helped to preserve the relevance of research findings for audiences within both the academic community and those working at a policy level.
Indeed, whilst there is active debate within critical realist literature on the best way to treat empirical evidence [Walters and Young, 2001; Downward et al, 2002], this study has attempted to employ empirical evidence where appropriate in qualification of the generative mechanisms established through the process of abstraction. This approach builds upon Davidsen’s [2005] attempt towards a critical realist inspired economics in which systematic ontological reflection will become an integral part of the scientific process166.
Methodologically critical realism is operationalised through the process of ‘retroduction’.
Bhaskar [1986, p11] describes retroduction as the process by which an argument ‘moves from a description of some phenomenon to a description of something which produces it or
166 Davidsen himself draws direct inspiration from Cruikshank [2003], and refers to Lawson’s work on critical realism in economics throughout. However, it must be acknowledged that Davidsen does not support several of Lawson’s criticisms of mainstream economics.
155 is a condition for it’. As such, retroduction can be seen as the central logic of inference associated with a critical realist epistemology [Downward et al, 2002]. Retroduction is an iterative process, which begins with a critique of prior theory developed around the phenomena being studied. From this critique, it is possible to identify potential generative mechanisms that may help to explain the phenomenon in question. The existence of such mechanisms in practice must then be ascertained through the collection of data on the concrete phenomena. Throughout data collection, Yeung [1997] observes that it is essential to use interactive interviews, enabling a deeper level of investigation than that offered by quantitative methods.
After data collection the process of iterative abstraction is used to ‘obtain knowledge of the real structures or mechanisms which give rise to, or govern the flux of the real phenomena’
[Lawson, 1989, pg 69]. This is the most well known method used by critical realists to discover and conceptualise generative mechanisms [Yeung, 1997]. Despite the emphasis placed on abstract theorising into deeper causal powers, empirical evidence can play an important role in critical realist research [Downward et al, 2002]. Reference to appropriate empirical evidence is required throughout the iterative process of retroduction, in order to bolster any claims to knowledge and ensure that research output maintains a link with the concrete phenomena being examined. However, regularities are not treated as explanation, but instead they are seen as ‘demi-regs167’ [Lawson, 1997], or partial event regularities, which are used to help identify generative mechanisms.
The process of retroduction relies on continuous data collection, where fresh input is used to revise or reaffirm generative mechanisms identified in the initial abstraction. As such, the process moves in an iterative fashion until the point at which the alleged generative mechanisms are robust and powerful enough to explain the phenomenon. This process is considered highly appropriate for analysis of the ever changing Ghanaian cocoa market.
Both at the domestic and international levels, commodity markets are in a constant state of flux creating a high number of emerging issues which, in their immediacy, may appear to
167 Demi-reg’s indicate the occasional, but less than universal, actualisation of a generative mechanism [Lawson, 1997].
156 define the market at any one period of time. Indeed, examples of market shocks, media hype, industry speculation, and one off instances have emerged within data collection. In this light, Miles and Huberman [1994] warn about the risk of generalising from non-representative events when carrying out causal analysis. As such, the level of critical reflection required throughout the retroductive process has helped the analysis go beyond the surface level of immediate observations.
Investigating the phenomenon of the partial model itself has also required commitment to a deeper level of understanding. Given that there is limited past research on partial models on which to assess the research findings, the process of retroduction has been necessary to critically assess the strength and validity of generative mechanisms identified throughout the study. Indeed, the level of critique inherent within the process of retroduction may help to establish the credibility within research findings required to transfer the lessons of the partial model to commodity markets elsewhere in sub-Saharan Africa.