1. Introduction
When asked nowadays about “comparative institutional advantage,” many scholars of the Varieties of Capitalism theory might shrug. Other crucial concepts from this strand of political economy have obviously influenced the debate much more than this one, for example: complementarity (Amable, Reference Amable2016; Crouch et al., Reference Crouch, Streeck, Boyer, Amable, Hall and Jackson2005), path dependence (Bennett and Elman, Reference Bennett and Elman2006; Djelic and Quack, Reference Djelic and Quack2007; Kay, Reference Kay2005), or skill specificity (Busemeyer, Reference Busemeyer2009; Emmenegger, Reference Emmenegger2009; Streeck, Reference Streeck, Busemeyer and Trampusch2012). This is certainly a little bitter because a second look at the subtitle of the edited 2001 volume indicates that Peter A. Hall, David Soskice, and the other contributors had the intention of making an institutional theory of comparative advantage the next big thing (Hall and Soskice, Reference Hall and Soskice2001b). Now – many people read Varieties of Capitalism as an attempt to show that “continental” (Crouch and Streeck, Reference Crouch and Streeck1997) or “Rhénan” (Albert, Reference Albert1991) regimes of production and accumulation could keep up with their liberal counterparts in terms of performance. Hall and Gingerich's (Reference Hall and Gingerich2004, Reference Hall and Gingerich2009) econometric article about the U-shaped correlation between the degree of coordination and macroeconomic performance indicators is seminal and iconic in this regard. Some have interpreted this as an apologetic plea for continental, social democratic modes of economic governance. Wolfgang Streeck (Reference Streeck2005) speaks of “subversive functionalism”: Hall and Soskice would defend social democratic public policy against tendencies toward liberalization, albeit unfortunately using arguments about efficiency instead of (pertinently) arguments about social citizenship and participation. They would rightfully oppose the suggestion that national political economies were disposed to converge toward a neoliberal model but they would unjustly reproduce the rationalist ontology on which neoliberal dispositions tended to rest in the first place (Streeck, Reference Streeck, Granovetter and Swedberg2011). In fact, Hall and Soskice generally explicitly rejected the “convergence thesis” that pertained to comparative political economy of their times. But it seems that they were interested in an even stronger proposition, namely that “coordinated” regimes outperformed liberal ones in certain regards. They suggested that, in a sense, these regimes were comparatively much better equipped to produce “incremental innovation.” This can be taken to mean that, if these political economies were allowed (or forced, for that matter) to converge with a more liberal line, incremental innovations would be practically doomed to vanish from the landscape of global trade.
This is certainly a strong and appealing claim, which drove home an intuition that had opaquely been around comparative political economy for years. But what has become of it? What follows is sort of an obituary for the theory of comparative institutional advantage. It represented a certain ambition among proponents of the Varieties of Capitalism approach, but it was also necessarily doomed to fail given the strong claims it made in terms of institutional theory. The next section discusses the origins of the concept and shows how surprisingly central it was in early discussions. The section that follows traces its straight decay, which occurred when the theory was faced with theoretical developments in comparative political economy as well as strong empirical and methodological criticism. A third section draws some melancholic conclusions and shows that the theory stands for the spirit of the discipline of comparative political economy of the millennium more than any other.
2. Origins of the concept
Evidently, the concept was among the major buzzwords during the early stages of developing the Varieties of Capitalism theory, especially around the WZB in Berlin where major contributors to the approach gathered regularly and fostered the approach, such as David Soskice, Sigurt Vitols, Steven Casper, and Mark Lehrer. As early as 1988, David Soskice had suggested the term “comparative institutional advantage” in a conference speech in Venice, proposing to explain the differences in unemployment patterns across countries through the effects of national institutional frameworks (Soskice, Reference Soskice, Brunetta and Dell'Aringa1990). He and others then continued to use the idea quite regularly in early contributions that went on to set the tone for the 2001 volume (e.g. Carlin and Soskice, Reference Carlin and Soskice1997; Fioretos, Reference Fioretos1996; Soskice, Reference Soskice and Zapf1994, Reference Soskice, Kitschelt, Lange, Marks and Stephens1999; Soskice et al., Reference Soskice, Hancké, Trumbull, Wren, Müller, Bihn, Delsen and de Jong1998; Vitols et al., Reference Vitols, Casper, Soskice and Woolcock1997). In 1997, Mark Lehrer picked it up in the title to his doctoral dissertation (Lehrer, Reference Lehrer1997), a comparative study of Western European airlines that eventually led to a contribution to the edited volume of 2001 (Lehrer, Reference Lehrer, Hall and Soskice2001).
The main idea behind the theory was basically transaction cost theoretic (cf. Riordan and Williamson, Reference Riordan and Williamson1985), namely that national institutional frameworks differed consistently by the specificity of the assets in which they incentivized actors to invest (Hall and Soskice, Reference Hall, Soskice, Hall and Soskice2001a). Hall and Soskice proposed that the more uniform – or complementary (passim) – these incentives were within an institutional framework, the more stable was the institutional framework. As they distinguished only between specific or general skills (passim), it is no surprise that they ended up with just two ideal types of institutional frameworks: the infamous “liberal–coordinated” dichotomy. Note how the label “liberal” is a rather late development, as David Soskice still distinguished between coordinated and “uncoordinated” regimes as late as 1999 (Soskice, Reference Soskice, Kitschelt, Lange, Marks and Stephens1999). Anyway, Hall and Soskice suggested that liberal frameworks favored investments in general assets across the entire political economy and that coordinated frameworks stimulated investments in specific assets. Now, borrowing the distinction between radical and incremental “paces” of innovation that was pertinent in neo-Schumpeterian thought (e.g. Freeman, Reference Freeman1984), Hall and Soskice tried to demonstrate that the incentives that liberal frameworks set were ideally suited to foster radical innovation, in contrast to coordinated frameworks that ideally governed incremental innovation. This is essentially a variation of an argument that had been around since much earlier in neo-Schumpeterian “innovation systems” research (Blättel-Mink and Ebner, Reference Blättel-Mink and Ebner2009). Protagonists of this approach such as Richard Nelson (Reference Nelson1993), Bengt-Åke Lundvall (Reference Lundvall1995), Charles Edquist (Reference Edquist1997), and Keith Pavitt (Reference Pavitt1999) had long been arguing that structural conditions, including institutional frameworks, constituted “innovation systems” which determined the innovation patterns of a national (or regional) political economy. At the time, the “innovation systems” approach was much less conceptually rigid than that of Hall and Soskice. It was influenced strongly by evolutionary economics (Nelson and Winter, Reference Nelson and Winter1982) and was much more open to informal structures and knowledge networks, while Hall and Soskice, with Hall especially rooted deeply in political science, focused quite strongly on formalized governance institutions. Still, Hall and Soskice seem to have been more inspired by the “innovation systems” approach than they openly concede.
Two connected conclusions follow from this: First, that national frameworks would each converge toward one of the two models over time; and second, that the innovative activities dominant within these frameworks would also converge toward the respective “pace” of innovation in the meantime. The more a national political economy profited from these fits between institutions, and between the institutional framework and innovative activities, the more would the returns from these fits rise over-proportionally as time went by. These increasing returns would lock the political economies into paths on which their later developments would depend and the national political economies would hence specialize. The result, ideally, would be a stable division of labor in international trade across a set of national political economies that, in turn, would be neatly divided into coordinated and liberal tokens. Hall and Soskice proved their argument briefly on patent data from Germany and the US.
Hall and Soskice (Reference Hall, Soskice, Hall and Soskice2001a: 57) were convinced enough of the effects of comparative institutional advantage that they even suggested that business would pursue what they called “institutional arbitrage”: they envisioned mobile capital roaming the globe freely looking to invest its respective assets exactly wherever the institutional framework best suited its goals. Firms could choose their location by the match between their innovative strategies and the transaction cost structures that the different institutional frameworks imposed. Similarly, multinational corporations could branch out across production regimes depending on which kind of innovation each of their branches was supposed to pursue. Hence, firms should disperse across production regimes based on a conscious, rational decision to save transaction costs by choosing the right institutional environment. Thus, “strategy follows structure” (Hall and Soskice, Reference Hall, Soskice, Hall and Soskice2001a: 15), less through strong structural determination in the sense that “strategy results from structure” (although one might think that this is what Hall and Soskice still had in mind sociologically when they proposed this formula), than spatially, in that strategic assets trail structural incentives.
3. Life-cycle of the concept
But what happened to the concept? Is it just out of fashion or has it been refuted? It seems that no conclusive criticism has been raised against it that would explain its disappearance, at least not more severely than of other concepts of the approach, such as path dependency or institutional complementarity. Or has it?
Indeed, in early critiques of the approach, comparative institutional advantage was among the crucial issues that were discussed. In 2003, shortly after the edited volume appeared, Comparative European Politics published three critiques by Mark Blyth, Robert E. Goodin, and Matthew Watson (Blyth, Reference Blyth2003; Goodin, Reference Goodin2003; Watson, Reference Watson2003). While Blyth focused on the capacity of coordinated market economies (CMEs) to adjust to accelerating trends of post-industrialization and Goodin doubted the viability of the parsimonious distinction between (liberal market economies) LMEs and CMEs altogether, Watson aimed his quite fierce objections mainly at the idea of comparative institutional advantage. He criticized Hall and Soskice for simply assuming, rather than proving, that territorial nations were reasonable objects of study. Similarly, they suggested, instead of testing, that national political economies were coherent entities. For Watson, they unreasonably favored conceptual modeling over openness for empirical manifestations. Moreover, as early as 2003, Watson pointed out that “comparative advantage,” institutional or otherwise, was politically created and sustained. Hall and Soskice overlooked the political contingency of the comparative institutional advantages they took to explain patterns of specialization across political economies. As a result, they tended to explain institutional frameworks by the “advantages” they conferred – an explanation that Watson rejected as functionalist and teleological. In sum, Watson labeled these perspectives typically “Ricardian” (passim), referring to David Ricardo's (Reference Ricardo1817) methodology and epistemology. A few pages further on, Hall and Soskice placed a reply in the same journal where they addressed the issues that the commentators had raised (Hall and Soskice, Reference Hall and Soskice2003). Towards Watson specifically, they relativized their functionalism, saying that they had never intended to explain the evolution of the institutional frameworks they discussed. Yet they still proposed bluntly that policy makers as well as firms could indeed identify advantages and disadvantages of alternative institutional frameworks, concluding thus that these advantages did in fact count as at least one of a few explanations for the persistence of such frameworks. Beyond that, they countered Watson's points mainly on methodological rather than theoretical grounds and left many details of Watson's extensive argument untouched.
Around the same time, being openly skeptical especially about Hall and Soskice's epistemology, Bruno Amable presented a taxonomy of five, not two, types of capitalism that he embedded in a more materialist theory and tested quite extensively using a variety of quantitative instruments (Amable, Reference Amable2003). He picked up readily on the theory of comparative institutional advantage, appreciated its major propositions, and also showed that his numbers fulfilled his expectations fairly well. In particular he presented a remarkable correlation between the liberality of the “market-based” (passim) institutional framework and various indicators for innovative success in information and communications technology (ICT). Vitols and Engelhardt (Reference Vitols and Engelhardt2005), unaware of Amable's results, used a similar intuition to explain why innovation policy had been so unsuccessful in fostering the New Economy in Germany. They argued that ICT was characteristically a radically innovative industry with a strong entrepreneurial disposition. The few isolated policy measures to foster it had simply been much too weak and narrow to effectively correct the path of specialization that corresponded to the enduring comparative institutional advantage of the German national institutional framework, which it had been following for too long already. They were especially concerned with the lack of risk and venture capital in Germany, which they said effectively impeded most high-tech industries from developing there.
Besides that, it seems that the community rapidly lost interest in the concept. One of the few yet also rigid receptions of the concept was Casper and Whitley's (Reference Casper and Whitley2002) comparative research paper on biotechnology and ICT firms, in which they continued the work that Casper had started earlier together with others (Casper et al., Reference Casper, Lehrer and Soskice1999). It appeared in Research Policy two years later (Casper and Whitley, Reference Casper and Whitley2004) and again as a collaborative chapter in Whitley's monograph Business Systems and Organizational Capabilities (Whitley, Reference Whitley2007), albeit, interestingly, stripped of Hall and Soskice's vocabulary. In this latest version, Casper and Whitley merely spoke of “more liberal,” in contrast to “more coordinated,” political economies without however changing much else from the original text from 2002. The fact that they practically just transplanted an article on varieties of capitalism into Whitley's book as an illustrative case study of his business systems theory (Whitley, Reference Whitley1999) supports the intuition that these two separate approaches to political economy have in fact more in common than they used to admit. Anyway, Casper and Whitley left aside Hall and Soskice's original orientation on transaction costs and argued instead that LMEs and CMEs specialized because they governed the risks of innovative activity differently. They also refined the object of the theory, in that they distinguished between subsectors of the industries in question by the pace at which they innovated, whereas Hall and Soskice had ascribed these paces to entire sectors. By and large, this led Casper and Whitley to conclude that comparative institutional advantage explained the developments of ICT and biotech industries (and their subsectors) rather well across their cases.
In their time, Casper and Whitley's results collided with some “empirical evidence against Varieties of Capitalism's theory of technological innovation,” to quote the title of Mark Zachary Taylor's empirical study of 2004 (Taylor, Reference Taylor2004). This study is one of the central critiques of Hall and Soskice, reflected by its appearance in Hancké’s infamous reader of 2009 (Taylor, Reference Taylor and Hancké2009). Taylor demonstrated quantitatively on patent data that he could only confirm the patterns of specialization that the theory of comparative institutional advantage predicted for Germany and the US. As soon as he repeated the regression on various CMEs and LMEs clustered together, the resulted distribution of radical and incremental innovations across clusters was surprisingly erratic. His cluster of CMEs even tended to be more radically innovative than the cluster that comprised all LMEs except for the US which he had assembled out of pure curiosity. He quite pessimistically concluded, “Under these conditions VOC [Varieties of Capitalism] theory has only marginally more predictive power than random chance” (2004: 617). However, Taylor pointed out that patent data might not be the most adequate operationalization for innovative success, as patenting innovations was potentially endogenous to both the innovation in question and the juridical regime (Crouch, Reference Crouch2005b). Not every successful innovation might need to be patented and not every legal framework might allow patenting all innovations equally. Thereby, he relativized not only his own study but also Hall and Soskice's original empirical illustrations. Five years later, Akkermans, Gastaldi, and Los (Reference Akkermans, Castaldi and Los2009) reprised and scrutinized Taylor's critique in a rigid quantitative analysis. Broadly in line with the classical predictions, their results showed that LMEs did in fact tend to specialize in chemicals and electronics while CMEs tended to specialize in machinery and transport equipment. Yet the patterns of incremental and radical innovations deviated strongly from the connection between sectors and patterns of innovation that Hall and Soskice had proposed. Akkermans et al. drew the ambiguous conclusion that Taylor had been too strict in rejecting the theory of “comparative institutional advantage,” and that it could in fact under certain conditions be demonstrated empirically, but also that it was much more complex and multidimensional than Hall and Soskice had originally thought.
Steven Casper, however, continued further along the line of argument that he and Whitley had opened (e.g. Casper and Matraves, Reference Casper and Matraves2003; Casper and van Waarden, Reference Casper and van Waarden2005). He has always rather positively defended the explanatory potential of comparative institutional advantage (Casper, Reference Casper2009) against criticisms of many sorts. For example, in a debate that took place in Socio-Economic Review, Herrmann convincingly showed that firms’ structures and innovative strategies followed pertinent institutional incentives far less stringently than Hall and Soskice had presumed because firms could “circumvent” institutional constraints (Herrmann, Reference Herrmann2008). A little later and in the same journal, Lange (Reference Lange2009) presented a similar argument on the discretion, or “leeway” (passim), that firms could make use of within different institutional frameworks. Just a little earlier, Sébastien Lechevalier had criticized the practice of too easily identifying a “representative firm” (Lechevalier, Reference Lechevalier2007), saying that it underestimated the effective room for variation in organization structuring and the heterogeneity of firm structures within a specific institutional framework. This could be turned into more than one sort of criticism against the theory of “comparative institutional advantage,” empirical, methodological, or theoretical, but it definitely raised doubts about the theory in general and its stringency in particular. The debate continued further when Casper responded (Casper, Reference Casper2009), mainly with methodological concerns, and Herrmann replied yet again (Herrmann, Reference Herrmann2010). She continued to make the criticism that the major proponents of the theory of comparative institutional advantage set out to show that production regimes specialized in certain sectors, rather than in innovative strategies. After they uncritically assumed, instead of proving, a typical innovative strategy and pace for each of these sectors, they could then easily argue how much the institutional frameworks of the production regimes favored or incentivized these strategies. However, a more fine-grained analysis comparing the strategies in neighboring firms from a single sector would find that in fact their strategies differed, at times strongly, even within one and the same national production regime. Herrmann concluded that production regimes did perhaps specialize on certain types, or paces, of innovation, but much less than expected because of the effects of institutional frameworks. In short, there might be a comparative advantage but not really an “institutional” one. Matthew M. C. Allen has raised similar arguments (Allen, Reference Allen2004; Allen et al., Reference Allen, Funk and Tüselmann2006), saying that Hall and Soskice significantly underestimated variations in innovative strategies and organizational preferences on a number of levels. He explains this in detail in his 2006 book, the only systematic discussion of comparative institutional advantage that has reached book length (Allen, Reference Allen2006).
Moreover, Beyer (Reference Beyer2006, Reference Beyer, Pfau-Effinger, Sakač Magdalenić and Wolf2009) found no evidence for institutional arbitrage in the long term. This is no surprise given Herrmann's or Lange's conclusions that strategies would not follow structures very closely in any case. It is however remarkable that Beyer's studies are among just a very few on institutional arbitrage. Similarly, and with regard to multinational businesses, Morgan and Kristensen (Reference Morgan and Kristensen2006) demonstrated that the subsidiaries of multinational corporations only rarely adapted to and exploited the institutional frameworks of the production regime in which they were located. It was much more common that they imported the corporate cultures and strategies of their respective parent companies. This immediately disqualifies the institutional framework as the rational incentive for locating the subsidiaries there to begin with. Rugman and Verbeke (Reference Rugman, Verbeke, Collinson and Morgan2009) furthermore indicated that institutional arbitrage might help save transaction costs but that integrating different branches from different production regimes also created transaction costs in itself.
Recently, some authors have revived and also redefined the concept briefly but without leaving a lasting impact on the debate. For example, continuing calculations they had obviously started as early as 2004 (Paunescu and Schneider, Reference Paunescu and Schneider2004), Schneider and Paunescu published a longitudinal quantitative study in 2012 that tended to support Hall and Soskice's claims surprisingly well in the light of the criticisms made in the meantime (Schneider and Paunescu, Reference Schneider and Paunescu2012). They conceded that Hall and Soskice had underestimated how fuzzy and dynamic the empirical patterns of LMEs and CMEs actually were. But they also pointed out that the more a national framework resembled an LME, the more it would specialize in radical innovations, and vice versa with CMEs. In other words, national institutional frameworks might not fall neatly into any of the categories, but if they did – or to the degree that they did – the specialization occurred as predicted by the theory of comparative institutional advantage. They had already presented this argument a little earlier together with Schulze-Bentrop (Schneider et al., Reference Schneider, Schulze-Bentrop and Paunescu2010). Here, their numbers also showed that firms did in fact practice institutional arbitrage. However, it is worth noting that for them, “institutional arbitrage” meant local firms drawing on foreign institutions to compensate for the shortcomings that they were seeing in their own respective institutional environment. This definition might seem different from that of Hall and Soskice (Reference Hall, Soskice, Hall and Soskice2001a), who expected firms to actually move into more favorable environments. Allen and Whitley's (Reference Allen, Whitley, Lane and Wood2014) argument echoed this definition, although they seem to have been oblivious to Schneider et al.’s suggestions. Anyway, Allen and Whitley reprised Casper and Whitley's (Reference Casper and Whitley2004) focus on risk sketched above, and argued that firms profited from internationalization because it gave them the opportunity to draw on foreign institutions if their own institutional environment did not govern their risks adequately. Far outside the “LME/CME” dichotomy, Friel (Reference Friel2011) argued similarly yet again when he showed that Arcor, a Latin American producer of candy, strategically combined available institutions to build its own comparative institutional advantage within Argentina's “hierarchical market economy” (cf. Schneider and Soskice, Reference Schneider and Soskice2009). It stands out that, here, the concept of comparative institutional advantage is molded to mean a more or less contingent combination of regulatory elements that are favorable for a single firm rather than denoting national patterns of specialization in international trade. About the last who mention the concept, then, are Alison Johnston et al. (Reference Johnston, Hancké and Pant2014) who borrow the term to denote how different institutional endowments enabled certain European political economies to cope, or disabled them from coping, with the financial crisis of the late 2000s. Aidan Regan (Reference Regan2014) reprised the concept in passing when he used it to explain that the Irish government refused to accept certain post-crisis measures from the EU because it feared they would damage the comparative institutional advantages of the Irish political economy. However, the concept remains marginal and Regan and Johnston do not even mention it in another article on the topic where, among other things, they draw their arguments together (Johnston and Regan, Reference Johnston and Regan2016).
4. Conclusion
So what has become of the concept? Of course, certain institutions incentivize some innovations more than others, otherwise there would be little sense in “innovation policy” (Borrás and Edquist, Reference Borrás and Edquist2014; Flanagan and Uyarra, Reference Flanagan and Uyarra2016). Yet they do not constitute the basis for enduring “comparative advantages” in a Ricardian sense, or, subsequently, for a balanced global division of labor in international trade. Hence, those scarce times that researchers still refer to comparative institutional advantage today, they seem to use it as a buzz word that more generically stands for the ways national policies govern innovations, rather than with the Ricardian innuendo that Hall and Soskice and others had in mind in the late 1990s. One major reason seems to be that institutions are much more endogenous and dynamic than the conditions that constitute comparative cost advantages in Ricardo's (Reference Ricardo1817) perspective. Greif and Laitin (Reference Greif and Laitin2004) have argued most concisely that the “rules of the game” of the political economy are in fact “quasi-parameters” (passim) of the game that may change if the game repeats often enough, much like most of recent institutional theory has argued that institutions are in fact objects of political-economic strategies as often as they govern them (Mahoney and Thelen, Reference Mahoney and Thelen2015). In contrast, the strong, Ricardian theory of comparative institutional advantage makes much stronger claims to path dependency and lock-in: claims that were very common in institutional theory around the end of the 1990s (Arthur, Reference Arthur1994; Pierson, Reference Pierson2000). The same holds for complementarity in that, typical for the context and period where it was formulated, the theory implies much stronger ties of complementarity between different institutions of the same framework than recent institutional theory would say exist (Crouch, Reference Crouch, Morgan, Whitley and Moen2005a; Kenworthy, Reference Kenworthy2005). Last but not least, the Varieties of Capitalism approach, a “firm-centered approach” in its own words, limits its perspectives to the supply side of the political economy. Only recently have scholars begun to take the demand side into account in comparative political economy (Baccaro and Pontusson, Reference Baccaro and Pontusson2016), with various conclusions for international trade. It remains open for now what their distinction between different “growth models” (passim) could mean to a theory of “comparative institutional advantage.”
All in all, the theory of comparative institutional advantage is somewhat dead. It has lost its Ricardian impetus, by which Hall and Soskice sought to show that the global division of labor in international trade vitally needed coordinated market economies and, hence, social-democratic economic policy. What remains is quite a transformed, much more generic version of the theory that neatly blends in with orthodox neo-Schumpeterian perspectives (Hanusch and Pyka, Reference Hanusch and Pyka2006) about how institutions, mediated by many other factors, influence innovation in national innovation systems (Blättel-Mink and Ebner, Reference Blättel-Mink and Ebner2009). However, the speed with which the discipline lost sight and focus of the theory of comparative advantage obscured how central it must have been to Hall, Soskice, and the others. It is this very theory where all of the other elements of the original theory of Varieties of Capitalism cumulate, like parsimonious dichotomy, institutional complementarity, and path dependence – and this is why the theory represents the zeitgeist of the edited volume and its related works so brilliantly. It represents the functionalism, structural determinism, methodological nationalism, and political optimism of the approach (Streeck, Reference Streeck, Granovetter and Swedberg2011) better than any other of its propositions and is therefore an intriguing witness to the wishful perspectives that were pertinent in comparative political economy at the time. At the same time, of course, its decay stands for the critical developments that institutional theory and comparative political economy have undergone in the meantime in each of these regards.
But is the concept of comparative institutional advantage actually dead? Drawing all these pieces together, Michael A. Witt and Gregory Jackson have presented an intriguing theory of “institutional comparative advantage” (sic) only recently (Witt and Jackson, Reference Witt and Jackson2016). They are skeptical of the common suggestion that stability and success actually stem from constellations wherein institutional elements share one and the same coordinating or liberal pattern exclusively. Instead, they pick up most recent debates about “hybrids” (Campbell et al., Reference Campbell, Hall and Pedersen2006; Campbell and Pedersen, Reference Campbell and Pedersen2016), as well as the argument that institutional frameworks are fragile rather than stable if they are too unambiguous (Kenworthy, Reference Kenworthy2005). They also understand complementarities and incentive compatibilities as the result of contingent political struggles (Amable, Reference Amable2016; Amable and Palombarini, Reference Amable and Palombarini2008). Their fuzzy set QCA makes it seem as if some constellations – hybrid, temporary constellations that combine “liberal,” market-oriented institutional elements in one area with “coordinated” ones in other areas – might indeed have institutional comparative advantages with regard to certain innovations. They write that actually combining “opposing principles” (passim) to build capabilities and flexibilities might be the crucial condition for political economies to cope with the challenges of globalized capitalism and to develop novel capacities in terms of technological innovation. On the other hand, of course, this would revive the concept of comparative institutional advantage completely at the expense of parsimony and the rigor of the binary distinction with which the Varieties of Capitalism approach amazed its spectators. Still, perhaps we may be looking forward to Institutional Comparative Advantage, a sort of Comparative Institutional Advantage – Mark II, so to speak: contingent, agonistic, and contested, but still temporarily and locally powerful and hence an important aspect of institutionalist political economy.
Acknowledgements
I would like to thank Jürgen Beyer and Birgit Pfau-Effinger for fruitful and patient discussions on the topic (although I took quite some time to write them up), as well as Christine Trampusch, Anita Engels, Lukas Dylus, Malte Bersdorf, Christopher Grages-Karabiner, Hannah Sabrina Hübner, Alexander Ebner, and the participants of my 2017 class on the concept of subversion in institutional theory.
