Photo: Mel Watkins

As part of our special series celebrating the 50th anniversary of the publication of Mel Watkins’ classic article, “A Staple Theory of Economic Development,” we present the following commentary by Dan Ciuriak. Dan was the co-author of a provocative IRPP paper earlier this year on the need for a modern, revitalized industrial policy capacity in Canada. Here he imagines how the discovery of some future new staple might spur Canadian policy-making to think more effectively about how to make the most of our resource wealth.

A strict reading of Melville Watkins’ 1963 article, “A Staple Theory of Economic Growth” suggests that its relevance is entirely to history rather than to the modern world. As he writes: “The staple theory is presented here not as a general theory of economic growth, nor even a general theory about the growth of export-oriented economies, but rather as applicable to the atypical case of the new country” (143; emphasis added). The “new country” is an “empty land” that is being settled by newcomers. Such places of course do not exist (and indeed never did exist in reality (A key staple of the North American economy is maize or Indian corn, which was developed through millennia of domestication and selective breeding by indigenous peoples and spread through the Americas through trade in varieties. The fur trade was indeed trade and largely between indigenous economies exporting to Europe via intermediaries such as the Hudson’s Bay Company).

In this comment, I restate the theory in an updated context and discuss its relevance to Canada’s situation today. As a device I imagine the development of a brand new staple, and think through its implications in light of Watkins’ analysis.

Consider an existing small, open economy with an established palette of goods and services. A technological breakthrough allows the exploitation of previously unusable land to produce a new staple — in other words, a new and large extensive margin opens up (Modern trade economics emphasizes the distinction between the intensive margin (an expansion of trade in existing products) and the extensive margin (the introduction of new products into trade or alternatively of new firms or new destinations)). Let’s use switchgrass for biofuel production in Canada’s west to concretize the thought (One could use oilsands or diamonds as the example, which Watkins (2007) mentions as recent extensions of staples production; however, these are politically charged areas and raise complex additional issues related to optimal use of non-renewable resources and the appropriate use of revenues from the sale of such assets which deserve a separate focussed comment, including on the fiscal effects which Watkins (2007) mentions. I leave these issues for other comments and focus on a simpler example which in my view allows for a cleaner exposition of the main features of staples theory).

Due to breakthroughs in, say, genome sequencing and bioinformatics, it becomes possible to genetically modify switchgrass to generate a substantially higher biomass potential per acre; at the same time breakthroughs in microbial conversion of cellulose to ethanol make it possible to derive substantially higher net yields of ethanol for biofuel use per unit of input. A new and major staple is introduced with large export potential due to expanded ethanol content requirements in EU and U.S. biofuel regulations. At the same time, the economy’s industrial sector faces new potential in emerging products such as green energy technology (solar, batteries, etc.) driven by the same broad societal objective — clean energy. We now pick up the story in Watkins’ article.

Watkins sets out in rather general terms a multi-sector general equilibrium model with three factors of production: land, labour and capital. Since comparative advantage for the land-intensive economy lies in land-intensive production, namely staples, the expansion of available usable land shifts the economy’s comparative advantage towards land and thus to staples. The further economic development of our economy then is a process of diversification based on exploiting the abundant factor through exporting which generates a “spread effect.” The latter effect can be readily identified with the “spending effect” formulated by Corden and Neary (1982) in their exposition of “Dutch Disease”: income earned from export of a resource (staple) drives demand for and production of other products, including inputs to the staple sector and downstream production enabled by the staple production that is not directly exported. Watkins also mentions that the effectiveness of domestic entrepreneurship depends on the supply of capital and labour. This hints at the resource movement effect in Corden and Neary, although there is no clear reference to crowding out of other tradables. Overall, the staples theory can be characterized as a proto-Dutch Disease “resource curse” theory, albeit with an incomplete exposition of the implications for other tradables and non-tradables (Watkins (2007; 118) notes the need for a fuller treatment of non-traded services).

In our updated economy, the emergence of comparative advantage in switchgrass/biofuels would tilt the playing field away from other green energy technology. While going with comparative advantage results in unambiguous welfare gains for the economy in a static context, there may be something about the character of the staple that makes it less desirable in a longer-term growth context. Watkins emphasizes the importance of the character of the staple and the dynamics that its production induces, including in terms of exposing the economy to the boom/bust nature of commodities, influencing the quality of labour, and the scope for upstream and downstream linkages. However, while there is mention that the staple production will be subject to decreasing returns, the fact that the staples theory focuses only on the character of the staple and not on the character of the non-staple (because of the “new economy” context for which it was formulated), there is no hint of external economies or increasing returns in other sectors that would more powerfully make the case for policy to support the development of such sectors (in our case, alternative green energy technology). Interestingly, Watkins mentions Marshall McLuhan but not Alfred Marshall, who introduced the idea of industrial districts driven by learning externalities!

Watkins also sketches out a growth model. Factor markets are open and factors are mobile; accordingly, both capital and labour are free to flow into or out of the economy. Moreover, foreign direct investment is also free of restrictions and thus inflows bring technology and foreign entrepreneurship. Nonetheless, the pace of growth is dictated by the pace of exports of the staple(s). In part, this reflects the fact that most technology is imported and domestic technological advance is thus dependent on export earnings (domestic innovation is allowed for but its extent is not explained; there is a suggestion that it is mainly refinement of imported technology). In part, however, it reflects the influence of the staple itself on the nature of institutions in the economy.

The endogeneity of the quality of institutions to the nature of the staple anticipates the recent emphasis on the importance of the quality of institutions in economic growth and the tendency of rent-heavy economies to be plagued by corruption. The attractiveness of these insights to explaining the sharply divergent development trajectories in the resource-poor economies of East Asia versus the resource-dependent economies in Africa and Latin America is clear. However, the implications of this for our updated land-intensive economy would appear to be more subtle.

One imagines that the switchgrass revolution would generate an ongoing innovation dynamic in, say, Saskatoon’s biotech cluster. However, why an economy that is open to global capital markets should be constrained by export earnings, if there are good growth opportunities elsewhere, is left unclear by the staples theory. One needs additional Dutch Disease-type arguments to explain the crowding out of the innovation-intensive cluster in alternative green technology. The latter sector might have potentially much larger longer-run gains due to economies of scale, the development of related products based on the engineering learning-by-doing, and the “learning-by-exporting” effects that technology exporters would gain from interaction with sophisticated global production chains (For a recent survey of the literature on firm-level learning effects associated with export entry see Ciuriak (2013)).

Perhaps most importantly, as the modern firm-level trade literature emphasizes, the green technology firms that become exporters would be the more productive firms in their sector and their entry into export markets would be associated with a number of productivity-enhancing effects. These effects would include internal economies from adoption of process technology suited for global mass markets and from post-export-market-entry scale of production, but also industry-level productivity gains as market share is transferred from the lower productivity firms to the faster-growing high productivity firms (For an accessible, non-technical discussion of the new heterogeneous firm theory and the policy implications see Ciuriak et al. (2011)). These new exporter-starters would be Canada’s next generation Bombardiers and Blackberrys. This is perhaps the most important distinction between exploitation of a land-based extensive margin versus a technology/firm-based extensive margin, as it provides an insight into why land-rich economies might tend to grow more slowly in the longer-run growth than land-poor economies.

Watkins also touches on development theory. Interestingly, he articulates the idea that development is a process of diversification rather than specialization according to comparative advantage as an economy integrates globally — an idea that is nowadays generally credited to Imbs and Wacziarg (2003). However, Watkins’ dismisses the infant industry argument (in 1963 already, well before the economics profession had abandoned it!) without discussion. In particular, there is no comment on the role of the MacDonald tariff on Canadian inter-provincial trade or the role of the McKinley tariff on the New England economy in the late 1800s and early 1900s when New England took over from Manchester the title of the “workshop of the world” (see, e.g., Philadelphia’s period of industrial pre-eminence around 1880-1920). As a result, there is no obvious policy hook regarding what to do about reliance on staples.

* * * * *

Ultimately, Watkins’ (1963) theoretical development of the staples thesis had little traction in influencing mainstream economic growth theories. The reason may be, as suggested by Krugman (1995; 27), that “…the influence of ideas that have not been embalmed in models soon decays.” So while Watkins was able to think through in plain English issues such as the general equilibrium implications of staples exports, the implications of commodity price booms and busts as well as of a longer-term decline in the terms of trade, and the implications for labour markets (and by extension for outward migration) if the growth impetus from staples exports was insufficient to meet the job needs implied by natural population increase (assumed to be exogenous), the inferences lacked staying power because they were not embalmed in mathematics.

But as we revisit Watkins and recognize his insights as well as the incompleteness of the key arguments when transported to a current setting, it is also important to bear in mind the massive extent of development of economics in the past half century. Trade economics has gone through two revolutions — new trade theory (Krugman, 1979 etc.) and new new trade theory (Melitz, 2003 etc.). Development economics has been in almost constant revolution in the same period; in fact, if we go by Bradford DeLong’s count, I think we are now in the eighth revolution — what might be termed a “counter-reformation” against the neoclassical revolution (see DeLong, 2001). When Watkins wrote, the concept of a developmental state had not yet been articulated by Chalmers Johnson, we hadn’t had the Latin and African debt crisis of the 1980s, nor the African renaissance of the 2000s on the back of rising resource prices, nor the coming and going of the Washington Consensus (Williamson, 2004) and the (sort of) arrival of the Beijing Consensus (Williamson, 2012).

Accordingly, a modern reader visits this text with an enormously different perspective than its writer — and yet immediately recognizes many features. And it is those points whose relevance has survived the 50 most intense years of economic history ever experienced to which we need to pay attention. The enduring insight of the “staples thesis” is that what you do as a country heavily influences who you are — and that a country may wish to have an active hand in deciding exactly what it does for a living in the global economy.

Being a resource-driven economy has not prevented Canada from developing and achieving high levels of prosperity. In that sense, Canada has largely avoided the resource curse — or alternatively broken out of a staples trap. This does not mean that concern about the heavy role of staples in Canada’s product mix was entirely unwarranted — after all, Canada got to where it is thanks in part to a long history of industrial policy intervention. Indeed in 1963, even as Watkins was expounding his staples theory, Canada was creating the Department of Industry to address the competitiveness problems in domestic industry and was about to enter into the 1965 Auto Pact, a structured (rather than free) trade agreement with the United States to solve the scale economy problem in the auto sector. That being said, as the recent Dutch Disease debate in Canada demonstrates, there are still starkly divided views on whether Canada is sinking back into such a trap (I survey this debate in a working paper under development with a working title “The Great Canadian Dutch Disease Debate: Yes, No, Maybe So?” (hopefully forthcoming soon!)).

There are new extensive margins (alternatively, emerging/future/infant industries) opening up everywhere in the current innovation-intensive global economy. One need only look at the policy statements emanating from the major economies to identify those that are particularly attractive in terms of their characteristics. These are the new “battleground” sectors, with advanced manufacturing typically heading the list (see Ciuriak and Curtis, 2013).

In a world where comparative advantage is manufactured by government policy, a country that chooses “laissez faire” is in fact choosing a comparative advantage defined by foreign governments. One might think of a game of musical chairs where the stand-offish laissez faire player gets the seat left open after all the other players have chosen theirs. What staples theory tells us is that might not be optimal, depending on the specific characteristics of the economy at land-intensive extensive margins.

A list of references cited in this article is available here.