Reframing deindustrialization

ABSTRACT

Mapping and analysing the processes of deindustrialization is core to the project of regional economic development and arguably, economic geography. The evolution in the field itself often follows an iterative path adding insights on specific regions and industries that have a long history at the core of regional economic analyses. Students are trained in the field by conducting empirical work – data collection and original analysis – on the experiences of regions and industries as they work through industrial transformations. Canonical articles on iconic industries have become the shared language of this multidisciplinary field focused on the spatial distribution of economic activity and how (and why) it changes over time. This article comments on Andy Pike's recent ‘Coping with Deindustrialization in the Global North and Global South’ which shifts our attention back to theorizing deindustrialization rather than simply documenting it.

KEYWORDS:

• Deindustrialization
• economic geography
• regional development
• industrial restructuring

This article refers to:

Coping with deindustrialization in the global North and South

Highlights

• Many of our expectations of the deindustrialization process, derived from the experiences in the Global North, do not seem to map onto the experiences of cities and regions in the Global South – we have some sense of what matters – institutions, governance, firm strategies, labour markets, (global) production networks, competition and trade, infrastructure, technology and innovation, and product (profit) cycles – but we do not know why these factors and forces seem to be recombining in different ways in the Global South than they did (and are doing) in the Global North.

• Whether it is in the Global North or the Global South deindustrialization is destabilizing because it erases the development gains of an emerging middle class through job loss and wage erosion everywhere it occurs. And ultimately, it is the jobs that industrialization brings, the jobs that allow access to economic opportunities to empower and enable people and places, that matters so very much.

• We see that the imbalances between labour supply and demand, skill mismatch, and wage stagnation are the outcomes of deindustrialization. Often we struggle to cope with these challenges through some ad hoc combination of national, local, and regional policies variously targeted at industries, firms, institutions, and individuals. Perhaps the policy learning from the experience of the Global North is that pretending that deindustrialization is rare, and that piecemeal policy approaches are sufficient, is a fairy tale well past its expiration date.

1. A twenty-first century approach to studying deindustrialization

Mapping and analysing the processes of deindustrialization is core to the project of regional economic development and arguably, economic geography. Students are trained in the field by conducting empirical work – data collection and original analysis – on the experiences of regions and industries as they work through industrial transformations. Canonical articles on iconic industries have become the shared language of this multidisciplinary field focused on the spatial distribution of economic activity and how (and why) it changes over time.

Andy Pike’s call to shift our attention back to theorizing deindustrialization rather than simply documenting it serves an important purpose. The evolution in the field itself often follows an iterative path adding insights on specific regions and industries that have a history at the core of regional economic analyses. Examples include the auto industry in Detroit, aerospace industry in Seattle, the film industry in Los Angeles, and a variety of electronics industries in older industrial regions as well as rural areas (Bailey & De Propris, 2014; Christopherson & Clark, 2007; Christopherson & Storper, 1989; Gray, Golob, & Markusen, 1996; Katz, 1996; Pike, 2005). All of these empirical efforts, often conducted by different scholars in different cities and regions over different times develop the layered foundation for theory building in the discipline. As Pike points out, the trouble with excavating these carefully laid strata for theory-building is that they tell us something about the highly contingent (socially, politically, geographically) historical past and only (hopefully) point us in the direction of the right variables to observe and analyse in a newly regionally variegated and globally integrated world.

These constraints not withstanding we do know where to look. And, the framework of ‘Geographical Political Economy’ (GPE) is one effort to organize an approach to that looking necessitated by the reality that we are now focused on many more places. At the core of Pike’s article is the assertation that the processes of deindustrialization are continuous and, also, varied. And further, that many of our expectations of the deindustrialization process, derived from the experiences in the Global North, do not seem to map onto the experiences of cities and regions in the Global South (Hassink, 2020). Again, we have some sense of what matters – institutions, governance, firm strategies, labour markets, (global) production networks, competition and trade, infrastructure, technology and innovation, and product (profit) cycles – but we do not know why these factors and forces seem to be recombining in different ways in the Global South than they did (and are doing) in the Global North (Bryson, Clark, & Vanchan, 2015; G. Clark, Feldman, Gertler, & Wójcik, 2018; Pike, Rodríguez-Pose, & Tomaney, 2006).

Certainly one explanation is time; the world is simply different now. This is not your grandfather’s deindustrialization process. Another explanation is space; there are fundamental differences in factor conditions in the Global South. Another explanation, perhaps harder to accept, is that we are not finished with deindustrialization in the Global North and, as a consequence, we are not done analysing and theorizing deindustrialization because it isn’t over yet. In some sense, we do not yet know what we are talking about.

The motivation behind Pike’s proposed GPE approach is to gain ‘a clearer understanding of the institutional architectures and conditions in particular geographical and temporal settings and widening of the range of institutions and policies considered beyond those typically found in the Global North’ (Pike, 2020). This is a call to develop greater specificity (Hassink, 2020). I would offer this friendly amendment to the design of geographical political economy approach: a return to work. Over and over again we see that the imbalances between labour supply and demand, skill mismatch, and wage stagnation are the outcomes of deindustrialization (J. Clark & Bailey, 2018). We end up, often if not always, struggling to cope with these challenges through some ad hoc combination of national, local, and regional policies variously targeted at industries, firms, institutions, and individuals. Perhaps the policy learning from the experience of the Global North is that pretending that deindustrialization is rare, and that piecemeal policy approaches are sufficient, is a fairy tale well past its expiration date.

2. Leaving deindustrialization behind?

And this returns us to Pike’s framing of deindustrialization around the debate about ‘left behind’ places and the ‘populist’ movements appearing in Europe and the US since the 2008 global recession. These debates are often referred to in shorthand by nods to UK’s Brexit or the Trump election in the US. As Pike says, ‘Deindustrialization has been at the heart of the renewed concern with social and spatial inequalities following the global financial crisis from 2008. Interest has grown in so-called “left behind” places, acutely affected by globalization, economic, and technological change’ (Pike, 2020). Electoral analyses by place coupled with protest movements like Gilets Jaunes point to something like a working-class discontent manifest in demands for increasingly reactionary responses to the broader provisioning of public services (publicly subsidized health care as an emblematic example) (Dijkstra, Poelman, & Rodríguez-Pose, 2020). But these same electoral analyses indicate, at least in the US, not that deindustrializing cities and regions are trending en masse towards some conservative populism but rather that cities show a distinctly different political orientation from their suburbs and the less populated regions of the states in which they are located. The phenomenon of dense ‘blue islands’ sitting in sparsely populated ‘red seas’ characterizes the US electoral landscape in almost every state.

And, the rise of movements like Black Lives Matters point again to the reality that the ‘discontent’ of the ‘left-behind’ people and places is not only an issue of industrial restructuring but also how power and privilege has been allocated and maintained across decades of institutional design (Wainwright & Kim, 2008). Perhaps this is where the Global South is particularly instructive as Latin American countries have seen significant swings back and forth from left-wing to right-wing political movements in recent decades with the consequent volatility in approaches to investment and disinvestment in regional industrial development (see Brazil, Mexico, Bolivia) (Wainwright, 2008). Many of these swings have depended, in part, on whether and to what extent indigenous and underrepresented voices have been heard or suppressed and the validity of the representation process (e.g. elections).

So, the discussion about a ‘revenge of left-behind places’ (and populations) sounds discordant in some contexts. Indeed, some places may be experiencing the fits and false starts of deindustrialization but many of these places in the Global North have not, in fact, been politically disenfranchised. And further, some of these places were not particularly industrial but rather remain rooted in agricultural production, both decidedly rural and heavily subsidized. In the Global North political and economic systems have been designed and maintained specifically so that the white working-class matters disproportionately. In the US, one need only look to the electoral college, congressional redistricting practices, or the flow of federal dollars back to underpopulated places, for evidence of the endemic structural efforts to make ‘left behind places’ consistently receive more than their ‘fair share’ of economic and political power not less.

Now, that does not mean that the communities experiencing deindustrialization have not been left behind by the processes of industrial transformation and the decades of failed policy interventions that have botched reinvestment in labour markets and effective responses to skill biased technological change; they certainly have (Chapple, 2015; Doussard, 2013; Lowe & Wolf-Powers, 2018). This is simply to say that not understanding how to effectively use political power to change economic circumstances is not the same thing as not having political power.

And further, we know from the long study of urban economies and urban agglomerations that over specialized small and mid-sized cities are vulnerable to deindustrialization in ways that that seem less extreme to larger and more diversified regional economies. That said, there are exceptions to these general trends and no prescription for success (or for that matter, failure) seems to hold across all spaces and across time. Hence Pike’s call for a focus on theorizing deindustrialization and moving beyond describing it resonates.

3. Deindustrialization and the development path

Ultimately, we want to understand deindustrialization because we want to do something about it. In the end, this is about managing change in order to continue along a particular path of development. The promise of industrialization was (and is) the broad benefits it has brought to places that were, in their time, left behind. Although Pike does not discuss this in detail, the assumed goal here is not extractive wealth building but the sort of wealth creation for communities that plots a path both out of poverty and beyond economic and political exploitation (Gibson-Graham, 2013). We still do not have a good example of another development model that produces such rapid and broad benefits.

This fear of deindustrialization’s ability to erode development faster than the industrialization process built it up underscores the urgency in Pike’s call for a GPE approach that can further ‘cross-national policy learning.’ The phenomenon of South to North policy transfer is well under way and mapping these policy exchanges is increasingly of interest to economic geographers. For example, the transfer of policy models such as conditional cash transfer programmes (CCT) and participatory budgeting from Latin and South America to North America have caught the attention of policy analysts and academics (Baker, McCann, & Temenos, 2020; Peck & Theodore, 2015; J. Clark, 2018, 2020). Such learnings will require leaning on scholars from the Global South and ‘newly industrialized economies’ who know and understand the nuances of deindustrialization in varied contexts (Bailey, Coffey, & Tomlinson, 2007; Bailey, Lenihan, & De Ruyter, 2016). As just one example, the industrialization processes in South Korea are incredibly instructive because of the institutional and governance structures carefully developed in tandem with the industries that came to dominate the country’s development (Lee, Kim, & Wainwright, 2010; Park, 2014; Sonn, 2007; Sonn & Lee, 2015). The South Korean case presents policy learnings in a way that the more laissez faire and ex post facto interventions into deindustrialization processes in the US or UK simply cannot.

The anxieties about deindustrialization are tremendous in the academic literature and in the political experience of the Global North (Bluestone & Harrison, 1982). Simply put, once a community adjusts to the quality of life of a $50 per hour autoworker job it is devastating to exchange that for the standard of living supported by $25 an hour often for that same work, sometimes even located in the same plant. Or, alternatively and arguably worse, for a $15 an hour job as a cashier at the big box retailer down the street. It turns out that it is easier to ride the wave up than on the way down. These waves are part of an economic instability significant enough to cause political instabilities as well (Block, 2018; J. Clark, 2019). It also highlights that we are perhaps leaving people, more so even than places, behind.

Pike points clearly to why it is so important that we learn to cope with deindustrialization:

Manufacturing still matters because of its generative role as the flywheel of economic growth. Compared to services, manufacturing has higher potential for technological advances and innovation spill-overs that improve productivity, generate increasing returns to scale, foster backwards and forwards linkages in supply networks, and create relatively well-paid job opportunities especially for people with limited formal or intermediate level qualifications and skills. (Pike, 2020)

What matters about manufacturing, more than any other issue, is its ability to produce well-paid jobs at a moderate skill level as compared to the alternative.

What we do know about deindustrialization is that it disrupts the expected development path. Coping with deindustrialization means dealing with a two-steps-forward, one-step-back development trajectory that pulls people and places into a cycle of wins and losses instead of onto a progressive curve upward (Doussard, Peck, & Theodore, 2009; Hackworth, 2019). Whether it is in the Global North or the Global South deindustrialization is destabilizing because it erases the development gains of an emerging middle class through job loss and wage erosion everywhere it occurs. And ultimately, it is the jobs that industrialization brings, the jobs that allow access to economic opportunities to empower and enable people and places, that matters so very much.

Disclosure statement

No potential conflict of interest was reported by the author(s).