Why is causal explanation critical in/to economic geography?

Abstract

This intervention elaborates on why causal explanation can serve as an indispensable building block towards robust theory development in economic geography. It argues for the critical importance of causal explanation in the subfield’s intellectual development and to its wider appeal to the social sciences. First, I show how this vital importance is premised on explanations that uncover the causal mechanisms of economic events, practices and processes that make things happen in society and space. Put differently, explanation needs causal connections as its necessary condition of explanatory power and practical adequacy. Its empirical operation is grounded in contextual contingencies and place-based specificities in an economic-geographical world characterized by complexity, multiplicity and emergence. Second, I explain why causal explanation represents a necessary step towards pragmatic research in economic geography. Our socio-spatial interventions can be better developed if we have a clearer sense of why and how carefully theorized causal mechanisms interact with contingent contexts to produce specific events and outcomes in the space-economy. Framed in this double hermeneutic sense of being both vital and pragmatic, causal explanation is critical in/to economic geography.

Wanted – explanatory theory!

As a large and expanding subfield in critical human geography, economic geography has seen its fair share of reflections on methodology and approaches to explanation over the past two decades in this millennium. The first decade was dominated by debates on the kind of theory that should matter for economic geography (Amin and Thrift, 2000; Yeung, 2001). This call for the ‘cultural turn’ came after what Scott (2000) had termed ‘the great half century’ of innovative research in the field. Much ink was spilled on what this ‘new economic geography’ might be and how we should go about practicing it (see Yeung, 2003, 2005). By the turn of the last decade, leading economic geographers such as Barnes and Sheppard (2010) had argued for a more pluralistic and catholic approach to method and explanation in economic geography (cf. Sheppard, 2011). In all fairness, the 2010s saw another even more intense wave of methodological debates in the field that made their appearances in leading handbooks (Barnes et al., 2012; Clark et al., 2018) and teaching texts (e.g. Barnes and Christophers, 2018; Coe et al., 2020). Even more ink is now spilled on such methodological reflections and disciplinary introspection in recent work (Bathelt et al., 2017; Gong and Hassink, 2020; Hassink et al., 2019a; James et al., 2018; Martin, 2021; Peck, 2023; Rosenman et al., 2020; Yeung, 2019a, 2023a).

In this disciplinary context and as a practitioner involving in such debates, I suspect economic geography might well be subjected to greater introspective dialogues and exchanges than other subfields in critical human geography (e.g. cultural geography and political geography). So why bother with yet another intervention in this Exchanges forum of a leading field journal that is now over half a (great) century old? In fact, another even older economic geography journal would be publishing its centennial issue now (2024) – a good testimony to the century-long history of economic-geographical research. In this piece, I am primarily concerned with making a ‘positive’ (not positivist!) and mostly conceptual case for constructing explanations in economic geography. My focus is on explaining why causal explanation can serve as an indispensable building block towards robust theory development in critical human geography – this is part and parcel of the key focus in my forthcoming monograph Theory and Explanation in Geography (Yeung, 2024). After elaborating on what causal explanation might look like in theory development, I examine why causal explanation is critical – in a double hermeneutic sense of being both vital and pragmatic – in/to economic geography. To reduce its somewhat detached ‘feel’, this paper will introduce two illustrative examples of geographical research to back up some of my main claims. For an obvious reason of space constraint, I can only be very brief on them.

Causal explanation and theory development

Taking cues from my recent work (Yeung, 2019b, 2023a) and some sympathetic critics (e.g. Strauss, 2019; Whiteside, 2019), I eschew the deterministic grand theories of everything in structuralist approaches and the discursive styles of open-ended theorizing in many ‘post-something’ theories. Instead, I argue for the epistemological importance of mid-range theories grounded in mechanism-based causal explanations. In this conception, causal explanation is both necessary and important for understanding why and how economic-geographical phenomena take place. But such explanation cannot be based on Humean law-like ‘mechanical’ regularities in empirical events and temporary sequences, as often practised in scientific explanation. It is often this kind of ‘scientific’ mechanism-based thinking and language that gives rise to the common perception and even dismissive reading of mechanism as too ‘machine-like’, that is, mechanical, technical or even macho-mechanical-technical.

In the social sciences, the analytical focus should be much more on causal processes and events leading to the identification and specification of (causal) mechanisms. By emphasizing causal powers and dynamic processual unfoldings in the socio-spatial world, this non-deterministic conception of (social) mechanisms should give economic geographers a reasonable sense of relief or even immunity from the allergic reaction to the above scientific-mechanical view of mechanisms. Two key ‘components’ often constitute a mechanism: (1) material objects: entities or parts that are stable bearers of properties or causal powers; and (2) actualized activities: interactions or operations that are more than just emergent potential or capacities and produce real changes in society and space. A mechanism therefore cannot be a pure idea or an abstract object, as in logic, mathematics and general linguistics (see more in Yeung, 2024: ch.5).

Indeed, explanation in my approach should be based on carefully theorized causal mechanism(s) that specifies the agency and pathway(s) for ‘making things happen’ in diverse economic-geographical events and phenomena – the causal powers of agency generate answers to the ‘why’ question and the pathways or mechanisms through which these powers produce outcomes help address the ‘how’ question. Simply put, explanation needs causal connections as its necessary condition of explanatory power and practical adequacy. The empirical operation of this causal explanation must also be grounded in contextual contingencies and place-based specificities in an economic-geographical world characterized by complexity, multiplicity and emergence (see my recent empirical application in Yeung, 2022).

This more ‘open systems’ view of causal explanation in economic geography fundamentally disrupts and deviates from the deductive-nomological or universal ‘covering law’ model of scientific explanation based on constant empirical conjunctions or invariant regularities of events. Explanation in economic geography demands the specification of causal mechanisms. These causal theories are mid-range in nature because their explanatory power within certain substantive domains of real-world research is premised on the specification of the causal mechanisms necessary for particular empirical events/outcomes in these domains and the contingent historical-geographical contexts in which such causal mechanisms become efficacious. In short, mid-range theories in economic geography require both causal explanations and specific contexts of operationalization to work; their explanatory capacity goes beyond individual cases and subject experiences, and yet they are not about planetary laws and universal generalizations. This insistence on theory and explanation in context can also be a useful epistemological strategy to avoid the sort of Cartesian desire for explanation, finality and (accurate) representation that has been much critiqued in some critical approaches (see further discussion in Yeung, 2024).

Despite the inherent limits to our knowledge production, my sense is that we need more robust causal explanations of an ever more complex and uneven geographical reality. Such explanatory theories represent only one critical step towards the lofty goal of accounting for the enormous complexities of uneven and unjust socio-spatial outcomes. To me, describing, contextualizing, situating and empathizing with subject experiences and identity formation in socio-spatial phenomena can only go so far in normative terms. Without the explanatory clarity of causal mechanisms at work, these descriptive procedures and processual approaches might not be sufficient in helping us understand, change and transform such uneven and unjust socio-spatial realities. Causal explanation should therefore go beyond interpreting, understanding, accounting, experiencing, making sense, critiquing, interrogating, (re)thinking, contextualizing and so on of diverse economic events, practices and processes to uncover their causes that really make things happen in society and space.

By way of the first illustrative example, let me refer to Peck’s (2023) recent call for the practice of conjunctural methodologies and their application to his ongoing collaborative study of China’s post-reform economic transformation and its recent state-led initiative in developing the Guangdong-Hong Kong-Macao Greater Bay Area (see also Yeung, 2023a). Following his earlier work in Peck (2017), the explanatory relevance of this conjunctural analysis can be summarized into several key ‘ingredients’ consistent with my call above: (1) midlevel concepts and theories (i.e. mid-range theory), (2) articulation of causality and causal codetermination (i.e. causal mechanisms) and (3) situational analysis, historicization, thick theorization as context-rich explanation (i.e. context-specificity). Taking this causal and explanatory approach to theory, he notes that mainland China’s actually existing political-economic transformations cannot be abstracted purely at the national scale (i.e. ‘methodological nationalism’). There are many other (re)formulations of such transformations at different spatial scales, from the body all the way to the neighbourhood, provincial, national and even macro-regional and global scales. To ‘operationalize’ such a midlevel approach to China’s variegated spatial economy grounded in unique socialist-market mechanisms, Peck advises us to engage in the study of intra-national regional (re)configurations and instituted mechanisms of political-economic change.

In sinology and social science, the Communist Party state in China has been well theorized as the causal mechanism of economic transformation through its adoption of marketization and neoliberalizing practices (Lim, 2019; Tan, 2021). Exercising its control over ideology, polity, economy and society through coercive and even violent means, the Party state confronted its own crisis of legitimacy since the late 1980s through further economic reform and marketization during the 1990s. This politics of reinventing the market and remaking the economy with Chinese characteristics, such as the consolidation of state-owned enterprises alongside the proliferation of entrepreneurial township and village enterprises and private firms, was well illustrated in many social science studies as orchestrated by the Party state for its own regime reproduction, political stability and modernist nation-building (Fuller, 2016; Huang, 2022; Naughton and Tsai, 2015). As such, it was the Party state that activated and turned the (contingent) process of neoliberalization in China – identified by Harvey (2005) and others – into a specific and necessary set of social relations for the (re)production of geographical uneven development. This marketization is further shaped by China’s unique political-economic structures and socio-spatial contexts, such as the highly corporatist nature of the local state, the extensive role of inter-personal networks in everyday life, the land-based logic of development and the continual domination of state-owned enterprises.

Overall, my argument for robust mid-range theory and causal explanation necessitates a conceptually precise and valid distinction between process and mechanism. This non-deterministic and mechanism-based approach to causal explanation and theory development can serve as a productive zone of engagement with other reflexive and critical approaches in human geography. Geographical theory can be explanatory in nature, but its explanatory power depends on the identification and specification of generative mechanisms connecting causal powers with specific events and outcomes in economy and space. This epistemological approach has good potential in making important contributions to theory development in the wider social sciences.

In short, my synthetic approach towards causal explanation in economic geography necessitates a careful examination of (1) the nature of mid-range theories and mechanism-based explanations; (2) the role of causal powers practiced by economic actors through relationality or relational geometries in making things happen in economy and space; (3) the analytically significant difference between mechanism in causal explanation and process in processual thought; and (4) the relevance of a process-based comparative method for such kind of explanatory theory development. My epistemological position thus views theory not as an abstract device for open-ended discursive critiques and/or ontological meandering, but more importantly as a causal explanation of life-changing struggles, persistent inequalities and uneven outcomes in the space-economy. In this position, an abstract or ‘theoretical’ critique is not necessarily a theory, nor must a theory be only critical in its core tenet(s). Rather, a theory must be explanatory and its explanandum must be specified. Explanatory theory therefore entails a different kind of normative position in economic geography because causal explanations are a necessary step towards understanding how socio-spatial interventions can be better developed.

Why is causal explanation critical (in/to economic geography)?

My gesture for non-deterministic mid-range theory and mechanism-based causal explanation is premised on several normative concerns in our politics of theorizing, the importance of socio-spatial contexts in geographical accounts and the necessary yardstick of practical adequacy. Here, I note that many critical theories in human geography are ontological in their discursive orientation and often become theoretical critiques of other theories, leading to what social theorist Bourdieu (1988: 774) terms ‘theoretical theory’. Other social scientists and philosophers of social science, such as Swedberg (2014: 15) and Schatzki (2019: 24), have also cautioned that such theoretical theories or abstract theories found in different ontological or philosophical debates are practically inadequate in accounting for social change in a material world. Taken together, I argue that economic geographers ought to recognize the inherent dangers of excessive ontological lock-in or abstract theorization in certain philosophical quarters that underpin contemporary critical theories (e.g. Marxism, poststructuralism and others).

In such critical terms then, causal explanations are vital to political action because they allow us to do things with them and, in doing so, ‘make things happen’ through our interventions in politics and practices (cf. the ‘so what’ question so often raised in relation to other kinds of descriptive explanations and accounts in critical theories). In this sense, causal explanations are critical in economic geography’s intellectual development. But to ensure the vitality of causal explanations, the underlying necessary causal relationships should be manipulatable such that their absence or alternations might (not ‘must’ in the sense of Humean generalized laws in the natural sciences) lead to different outcomes in diverse economic-geographical contexts. This ‘might’ in causality refers to the possibility of counterfactuality in answering the what-if-things-had-been-different question and thus the possibility of these causal relationships rendered knowable through such counterfactual thinking and explanations.

While causal relationships (might) exist in economy and space, ‘discovering’ them is a human-manipulative practice and explanation that is causal in nature requires distinctive epistemic interventions argued here. But why should we be bothered with causal explanations and knowledge? Here, I concur with philosophers of science Woodward (2003: 30) and Glennan (2017: ch.8) that causal explanations offer practical payoffs because more normative interventions can be made possible with such explanatory knowledges. Nevertheless, prediction is not what causal explanations are about – predicting future outcomes are not the inherent value of causal explanations nor causal theories. Prediction can be done via knowledge of Bayesian probability-based correlations alone without much causal enquiry of the kind recommended here. As such, we do not need to know why certain things happen in order to predict when they will happen again. But we do need causal explanations of these happenings to make sure that they will not likely happen again or they will happen differently again. This is possible only if we exploit our causal knowledge to intervene successfully in the underlying relationships of these happenings in the space-economy.

On policy-relevant research in economic geography, let me bring in the second brief example of the geographical work on regional innovation and regional diversification. Often couched in the form of evolutionary economic geography, this body of mid-range theory work at the regional and sub-national scales has helped specify certain key causal mechanisms of regional change, such as path dependency, industrial branching and related variety (Boschma, 2022; Boschma and Frenken, 2018; Boschma and Martin, 2007; MacKinnon et al., 2009, 2019; Rodríguez-Pose, 2013, 2021). Conceptually, path dependence is viewed as a causal mechanism through which past pathways influence the present and the future path development. Causal explanation is the raison d’être of this kind of evolutionary economic geography when path dependence is used to explain causal outcomes in the location behavior of firms, the spatial evolution of industries and networks, the co-evolution of firms, technologies and territorial institutions and the convergence or divergence in spatial systems. The explanatory emphasis is not placed on predicting regional change, but rather on uncovering the specific causal mechanisms that can be ‘manipulated’ through policy interventions, such as the European Union’s Smart Specialization programme (Balland et al., 2019; McCann and Ortega-Argilés, 2015; Whittle and Kogler, 2020).

As a meso-level unit of analysis in theory development and in explaining regional evolution, the study of networks has also been well established in evolutionary economic geography (and other approaches in economic geography). Here, networks are crucial in understanding knowledge diffusion and innovative activities within industrial clusters and regional economies (Boschma et al., 2017; Grillitsch, 2019; Hassink et al., 2019b). As well argued in Boschma (2022: 127), evolutionary economic geography takes a unique understanding of networks and their evolving complexity within and beyond regional economies in order to identify new growth paths and possibilities for different regional economies. This causal explanatory approach is particularly relevant in the post-pandemic era of deglobalizing, decoupling or de-risking overdependency on specific national economies (e.g. China in manufactured goods and the United States in technology products/services) and/or macro-regions (e.g. the EU in advanced manufacturing and Northeast Asia in semiconductors and electronics; see empirical analysis in Yeung, 2016, 2022).

Still, why is causal explanation critical to economic geography too? This points to the pragmatic potential of causal explanations that, in critical double hermeneutic terms (i.e. more-than-vital), justifies the practical adequacy of our work and the discipline’s survival need in the wider and rapidly changing social sciences. Sayer (2015: 106) has made it easy to understand the practical adequacy of our explanatory theories and knowledges because ‘the fact that we can successfully do so many things through our practical interventions in the world suggests that the knowledge informing those interventions has at least some “practical adequacy”’. This consideration goes some way to address the ‘so what’ or ‘why bother’ question and to identify an explanatory theory’s relevance for important pragmatic goals in economic geography. In philosopher Hacking’s (1999: 20; 95) terms, we should unmask the underlying functions served by existing ideas and understandings in established order and raise our collective consciousness to make the world a better place. Our theories need to engage with the practical adequacy of explanation for making possible our interventions and transformations in a material world. Explanatory theories are necessary for practical efforts in addressing the real-world issues of massive poverty, inequality, abuses of human/civil rights and so on faced by many in the Global South and, increasingly, in the Global North too, including recently heightened geopolitical tensions!

Looking forward

The search for good theories by inference to the best explanation remains elusive even in the natural sciences. To me, explanatory reasoning in economic geography is but one form, albeit a very critical one, of abductive and inductive accounts of human beliefs, activities and events and economic-geographical changes. In this approach to explanation, there is no fundamental truth and transcendental reality to be explained. Rather, there are pragmatic and realistic causal explanations of particular happenings and events in economy and space – they might not be the ‘best’ and the most unfalsifiable in a scientific epistemological sense (e.g. rationalism and logic thinking), but they are practically adequate and perhaps even partial, in normative terms. This sort of explanation I seek is similar to philosophers of science Khalifa et al.’s (2018: 84) notion of ‘explanatory pluralism’ through what they term ‘thick bundles’ of relations, some of which are not necessarily explanatory (e.g. representation, prediction, analogy and embedding).

Just about half a century ago, Harvey (1969: 486) ended his Explanation in Geography by the grand statement that ‘By our theories you shall know us’. Today, it is not exaggerating to claim that economic geography has been characterized by lots of ‘theories’ but perhaps too few good explanations of diverse economic-geographical phenomena, particularly in the post-pandemic 2020s (see Yeung, 2023b, 2024). Geographers are now perhaps better known for our theoretical theories than our explanatory power. This Exchanges piece has forwarded a hopefully positive enough dialogue on what economic geography’s explanatory mission might be and how we can achieve it better in the next one to two decades. This epistemological task is particularly necessary and timely in the present turbulent world in which intellectual critiques and academic theories seem to have lost their public appeal and trust in many democracies. Revisiting and constructing (causal) explanations in economic geography can be a way forward to rebuild the analytical rigour, responsible practice and public relevance of our discipline.