P ost -K eynesian F inancialised G rowth M odels as an a lternative to v arieties oF c aPitalisM : (i n ) stability , i nstitutions and t axonoMic M ethod

This article evaluates the potential of the Post-Keynesian literature on growth models to gain influence over the Varieties of Capitalism approach within Comparative Political Economy. It shows that the future analytical strength of the latter approach depends, primarily, on the ability to consolidate macroeconomic principles consistent with a dynamic reality. On the other hand, Post-Keynesian macroeconomic foundations allow financialised growth models to capture the importance of power struggles in long-term growth as well as to integrate crises as recurrent and inherent phenomena to capitalist economies. That said, Post-Keynesian challenge to become a beacon within CPE lies on moving beyond country-based analyses towards the construction of a systematic association between, on the one hand, the institutional aspects shared between countries and, on the other hand, their classification on growth models.


introduction
There is a struggle to gain influence within Comparative Political Economy (hereafter, CPE), understood as the study of 'differences in institutions, policies, and economic outcomes across countries' (Stockhammer 2021, p. 2).The literature on Varieties of Capitalism (VoC), being the dominant view in the last two decades, proposes a 'firm-centred' and 'supply-sided' perspective, classifying countries as Coordinated (CME) or Liberal (LME) market-economies, depending on the type of institutional coordination.In contrast, the Post-Keynesian (PK) approach, which has built from Bhaduri and Marglin's (1990) demand-side wage-led (WL) and profit-led (PL) taxonomy, demands more attention (Stockhammer 2021, p. 3).During the last decades, the two perspectives have evolved in parallel; yet in a context of critical interaction between both approaches, theoretical and methodological debates have intensified (Baccaro and Pontusson 2016;2018;Hope and Soskice 2016;Stockhammer 2021;Hein et al., 2020;Kohler and Stockhammer 2021;Stockhammer and Kohler, 2022).The discussion has not so much revolved around the two original taxonomies, but around the growth models developed by some countries before and after the 2008 crisis.This exchange of views comprises mutual acknowledgements, criticisms and vindications that have led to an apparent stalemate.
The purpose of this article is to scrutinise the potential of each approach to gain influence within the CPE.Despite the many commonalities, there are marked theoretical divergences between the VoC and PK approaches.We argue these to be residing in two fundamental areas: (i) in the macroeconomic foundations -mainly the role of aggregate demand and system instability -and (ii) in the understanding of institutions -as being the outcome of the efficient interaction of rational agents or as rules that materialise power struggles.
We show that are precisely these divergences that mark the major challenges for each approach to gain dominance within CPE.In so doing, we suggest areas of future work that both approaches must fulfil if they are to cover existing substantial epistemological gaps.On the one hand, VoC literature lacks an explicit macroeconomic fundament; its future in CPE depends on consolidating macroeconomic principles consistent with a dynamic reality.Some authors have favoured linking it to the 3-equation model (Carlin and Soskice 2005), and yet the choice of this or another macroeconomic model will determine the ability of the VoC approach to explain phenomena such as endogenous financial instability or long-term stagnation affecting both LME and CME (or other secondary) types.As for now, VoC's exegetical strength remains truncated by the production of harmonious models that hardly explain diversity of dysfunctional growth models and a crisis ridden economic system.
In this respect, the PK approach has clear theoretical advantages: its proposal arises unambiguously from its own macroeconomic foundations; it captures the importance of power struggle in the determination of functional distribution which, in turn, is a fundamental factor of long-term growth.At the same time, it understands crises as recurrent and inherent phenomena to capitalist economies.That said, the possibility of establishing itself as an epistemic alternative within CPE depends, to a large extent, on the ability to embed the macroeconomic fundamentals within an institutional environment and develop a systematic and explicit relationship between the possible cultural, institutional and historical commonalities between countries, and their classification in relation to their growth model.
The article is structured as follows.The first section locates the reader by swiftly presenting both approaches as well as their contribution to CPE in detail.Later, we study the macroeconomic rationale of each.The third section contrasts the existing literature with the recent discussion between the leading authors of each approach and thus evaluates the relevance and implication of the coincidences and differences that have arisen among the multiple works.Section 4 concludes by summing up the main findings.

a literature review on the PK and voc ProPosals
In recent years, Post-Keynesianism and VoC have converged in the literature on growth models, coinciding in the main categories identified, the group of countries that make up the object of study, or in the typification of trade and financial relations between them.However, the theoretical path prior to the identification of growth models as well as the contributions to Comparative Political Economy differ between the two schools.

oriGins and nature oF PK and voc Growth Models
The literature on financialised growth models is based on the Post-Keynesian current, and comes from the debate around the effects of a variation in the level of wages on economic growth (eg; Rowthorn, 1981;Blecker, 1989;Bhaduri and Marglin, 1990;Stockhammer, 2021).The shift of some of the attention from these demand regimes to financialised growth models, especially since the 2010s, can be understood on the basis of an observation and the question that follows from it.If, as several empirical studies show (see Oyvat et al. 2020 for a compilation), most developed countries are wage-led, but wage share has been falling since the 1970s, how can we explain the relatively high growth experienced by some of them in the 1990s and 2000s, as well as the strong impact of the 2008 crisis?The answer lies in the profound transformation of demand patterns through increasing debt levels.
In this context at least two opposite growth models have emerged: the debtled growth model (DLG) and the export-led model (XLG)1 that have allowed wage-led countries to temporarily evade the contractionary effects of the fall in the wage-share (e.g; Stockhammer 2009;Hein and Mundt 2012;).In the DLG model, debt finances household consumption and even firms' investment; meanwhile XLG countries -like Japan, Germany, and other Eurozone core countries -accumulate growing trade surpluses.They are opposite, but still complementary models: the latter benefits from the debt of the former to increase its exports, while the endogenous debt expansion of the former is reinforced by the inflows from the latter.In a nutshell: Post-Keynesian literature considers that capitalist economies have significant common features that, depending on the country, lead to different growth paths (i.e.XLG, DLG).
In contrast to PKE, the VoC approach, based on Hall and Soskice's (2001) seminal work, proposes an agent centred classification of national economies comprising the supply side (Hope and Soskice 2016), where the firm's internal and external relations become the axis of analysis.It studies the different ways in which the firms coordinate with other players based on certain analytical dimensions.
Qualitative differences may crystallise into two stylised socio-economic functional models: in the liberal market economies (LME), firms' activities and collective bargaining are coordinated via organizational hierarchies and competitive market arrangements; in the coordinated market economies (CME), 'firms depend more heavily on nonmarket relationships to coordinate their endeavours with other actors and to construct their core competencies' (Hall and Soskice 2001, pp. 6-8).With the areas of study and the theoretical groups defined, the United States is the paradigmatic case of the group of 6 countries with features of an LME, while Germany would be the clearest reference among the total of 10 CME countries.A third, less detailed 'variety' would be the Mediterranean Market Economies (MME), characterised by non-market coordination in the sphere of corporate finance but more liberal arrangements in the sphere of labour relations (ibid., p. 21).
Despite the multiple critiques and subsequent extensions to the original scheme, there are still strong tendencies to maintain the theoretical benefits of problematizing the existing institutional variety to grasp a complex reality and simplify it.The most obvious evidence of the still wide support for the LME-CME taxonomy is the extensive literature of the last decade relating these types and two main growth models: the 'export-led' model (Germany and other 'Northern' European countries) and the 'domestic demand-led' model (United States, United Kingdom, Spain) financed with foreign capital.
Unlike the PK classification, however, each VoC type is linked to a growth model: while CME countries tend to adopt an export-led strategy, LME and MME countries would be prone to drive their growth with domestic demand (Hope and Soskice 2016).From this perspective, therefore, an explanation of imbalances -with special reference to the Eurozone -is offered by relating a country's demand patterns and the way its institutions are organised and articulated.For instance, price competitiveness (Hancké 2013;Höpner and Lutter 2014) and quality-based competitiveness (Iversen and Soskice 2012;Hall 2014;Iversen et al. 2016) appear frequently associated.The CME countries have mechanisms for wage coordination between employers and unions, vocational training systems and the implementation of interfirm research collaboration programs with which to promote high valueadded exports.Mediterranean countries (MME), however, do not have such institutionalised collaboration.Moreover, unlike LMEs, they have strong trade unions, but their large number prevents coordination as in CMEs (Hall and Soskice 2001;Hancké 2013;Johnston et al. 2014;Hall 2014;Hope and Soskice 2016).This divergence of models can be reinforced by the different economic policies: CME's less accommodative policies in response to adverse shocks reaffirm agents' commitment to wage moderation (Iversen et al. 2000;Soskice 2007;Hope and Soskice 2016;Hall 2018).

contributions to coMParative Political econoMy
Three essential aspects that shed light on the extent to which each of the two approaches contributes to CPE: a) the ability to translate complex cultural, historical and institutional realities into a clear classification, b) the interaction of the two growth models, c) capacity to assimilate and incorporate new patterns of the international economy, such as financial liberalization or monetary unification, within their theoretical framework.
Before the rise of the literature on financialised growth models in the 2010s, the PK contribution to CPE about demand regimes was already notorious: taking the distributional conflict between classes as a key element in the determination of aggregate demand, they proposed a dual classification applicable to any country.Based on this premise, several lines of research have been opened, among them: a line in which the factors behind the observed fall in the wage share from the 1980s in most developed countries are identified (e.g; Jayadev 2007;Onaran 2009;Lavoie and Stockhammer 2013;Kohler et al. 2019); a second one trying to determine empirically the demand regime of different countries (cf.Blecker 2016;Stockhammer 2017); and a third one suggesting characteristics -related to national culture, Welfare State configuration, financial regulation…-that lead aggregate demand components to react in a specific direction and amplitude to changes in functional distribution (Dutt 2011;Lavoie and Stockhammer 2013;Kapeller and Schutz 2015;Setterfield and Kim 2016;. Carvalho and Rezai 2016;Palley 2017;Stockhammer, 2021).Post-Keynesian Financialised Growth Models as an alternative to varieties oF caPitalisM: (in)stability ... revista de econoMía Mundial 65, 2023, 71-92 The path subsequently opened in relation to the study of growth models no longer proposes a classification based on the complex calculation of the effects on the level of output of changes in the functional distribution, but directly classifies the countries according to the superficial observation of the components of aggregate demand.However, it consolidates two key elements of Bhaduri and Marglin's (1990) proposal.First, it underlines the central role of distributional conflict in the swings of aggregate demand, already emphasised since Kalecki (1943).Second, contrary to VoC types, financialised growth models -as well as demand regimes -do not represent successful forms of capitalism in which to classify a few countries, but groups identified from the observation of patterns on the aggregate demand side, with which to classify any national economy and in which crisis or stagnation are foreseen phenomena.
This consolidation is achieved by introducing some elements.First, it introduces key elements to understand growth in financialised times.After all, the contribution of changes in income distribution to growth is estimated to be relatively modest (Stockhammer et al. 2009).Moreover, with most developed countries being wage-led, the net contribution of wage share changes in recent decades may have been negative.Besides, integrating financialisation into the analysis allows taking debt -frequently associated to wealth effects -as a key driver of growth until 2008 (Zezza 2008;Stockhammer 2012;Hein and Mundt 2012;Hein 2017;Stockhammer and Wildauer 2016).Finally, it also raises the problem of the unsustainability of the models when the financial cycle is over, as well as the characteristic interdependencies between countries before and after the Global Financial Crisis (GFC).
Regarding Varieties of Capitalism, the original proposal by Hall and Soskice (2001) does not include a detailed method to qualify the type of relationships in each of the five analytical dimensions proposed.Nor does it provide guidelines for ranking these dimensions in the event that the same institutional configuration and complementarity does not apply to all of them.The theoretical corpus has been built inductively, from an appreciable difference between continental European and liberal economies, and then applied tout court.This poses problems especially when trying to classify non-Western countries -like the case of China (Witt 2010; Fligstein and Zhang 2011), but also Western paradigmatic cases (Streeck 2009;and Piore 2016 for Germany and the United States).Perhaps because of this room for interpretation, the VoC literature has shown flexibility in integrating realities omitted from Hall and Soskice's original approach.In this respect, a number of authors have expanded and nuanced the original LME-CME, mainly with the aim of integrating different regions into the debate (e.g, Nölke and Vliegenthart, 2009).
The association with growth models has been a step forward for the VoC literature.As seen above, it is now able to associate different sets of institutions with demand patterns: CME with export-led growth and LME and MME with domestic demand-led growth.In addition, interactions between countries belonging to different models are clearer than in the original VoC proposal.
The VoC growth models were particularly relevant in explaining the 2008 crisis, describing the trade and financial relations between the main Western countries involved.More specifically, it offered a perspective on the uneven relationship between the countries in the centre and the periphery of the Eurozone.The VoC approach also integrates debt and financial stress into the dynamics between growth models before the crisis (Iversen and Soskice 2013;Johnston et al. 2013;Behringer and Van Treeck 2019) and draw a link to the spiral of 'over-exuberant' borrowing of LME and MME countries (Soskice 2009;Hall 2014).In the case of the Eurozone, however, instead of focusing on the new patterns of accumulation and the rise of debt (as the PK literature), the severity of the crisis in the 'South' is mostly explained as a result of the impossibility of depreciating the national currency due to the consolidation of the monetary union (Johnston et al. 2013;Höpner and Lutter 2014;Johnston and Regan 2016;Hall 2018).
By raising the problem of crises, the VoC approach is now open to reconsider that each variety is the result of (in)efficient interactions (Soskice 2009).And it is implicitly acknowledged that the efficiency of a VoC depends on the supranational regulatory framework; monetary union between CME and LME countries, for example, would be particularly advantageous for the former and detrimental for the latter (Hall 2018).

MacroeconoMic FundaMentals
Although succinctly, it can be seen that each approach underlies different ways of understanding the role of demand, the possibility and origin of crises and, ultimately, the growth drivers.In this section we show that these differences derive from the macroeconomic model in which they are (or are not) embedded.
VoC is micro founded and supply side focused, but it has not been built on a clear macroeconomic foundation.A posteriori, it has often been associated with the so-called 3-equation model (3EqM) (Carlin and Soskice 2005), consisting of the IS curve, the Phillips curve and the Monetary Rule.For instance, Hope and Soskice (2016) (being Soskice himself one of its co-developers) use the 3EqM explicitly.As Stockhammer (2021) suggests, the heads of VoC scholars were turned towards that model probably because the institutional focus of VoC can complement 3EqM's supply-side focus.In short, although long-run equilibrium is set from the supply side, the interaction of aggregate demand and supply -with consequences for employment levels, wages, inflation, and output, among others -is directly affected by the institutional setup of the national economy.This can be seen in the three main implications of 3EqM for VoC pointed out by Baccaro and Pontusson (2018): "(1) monetary and fiscal policies can have significant effects on the real economy…but only in the short-to medium-run": from the VoC perspective, for a [long run] high level of employment should insist more in making structural reforms of the liberal kind (i.e. a more flexible labour market) than in expansive monetary or fiscal policies."(2) there are important complementarities between aggregate demand management regimes and production regimes": in CME countries with few wagesetters, for example, an expansionary policy would have lasting real effects because of its ability to put downward pressure on real wages and, therefore, on the long-run equilibrium level.; "and (3) the different macroeconomic stances adopted by LMEs and CMEs are interdependent and, for the most part, complementary: (Iversen and Soskice 2012;Carlin and Soskice 2015;Iversen et al. 2016)".In this sense, the adoption of inflation targeting in both LME and CME allows the emergence and growth of bilateral trade imbalances.In LMEs, for example, if there is a positive aggregate demand shock, this will tend to generate a current account deficit, but an inflation-targeting central bank is unlikely to intervene by increasing interest rates so long as inflation remains subdued.In the same light, if coordinated wage bargaining in CMEs generates a depreciation of the real exchange rate and increasing exports the central bank will keep rates constant even if current account surpluses accumulate.
PKs' opinion about the merit(s) of the 3EqM is not unanimous, probably because it lies somewhere between New Keynesian (NK) and 'traditional' Keynesianism.In this regard, there seems to be some misunderstanding among PKs when equating the 3EqM with a fully-fledged NK model.It diverges from the canonical NK model, in which all agents are considered forward looking rational optimizers; the 3EqM is considered to be more 'realistic' and only the inflation targeting Central Bank is a forward looking optimizer (Hope and Soskice 2016, p. 583).Indeed, Lavoie (2015) admits that the 3EqM model includes relevant PK ideas: money supply is endogenous, inflation is the outcome of a distribution struggle between wages and profits, hysteresis effects as well as financial cycles may be included.
Apart from these coincidental elements, almost the entire literature that starts from financialization and seen above to study growth models reflects the fundamental elements of the PK theory (Lavoie, 2004).The clearest one -in contrast to the New Keynesians and also to 3EqM -is the principle of effective demand: aggregate demand is the main force determining output and the level of employment, also in the long run.In fact, the classification between countries is made mainly on the basis of the observation of the dynamism of the different components of aggregate demand, assuming a gradual adjustment of supply.The way debt is treated and its importance in the generation of demand -first -and instabilities -later -is also strongly consistent with two other basic principles of post-Keynesian economics.On the one hand, the analysis of national economies is dynamic and far away from any general equilibrium framework typical of the neoclassical school.On the other hand, crisis phenomena are endogenous to the nature of capitalism, intimately associated with the preceding financial expansion (Lavoie, 2022).
Summing up, disequilibrium in VoC models depends on market failures and institutional shortcomings rather than on intrinsic features of capitalist economies.This in turn assigns different roles to fiscal and monetary policy, considered to be stabilization instruments to dampen the business cycle for VoC models versus the necessary instruments to grant growth for Post-Keynesian models.

discussion: FroM eMPirical coincidence to diverGinG FundaMentals
The coincidence in the topics studied, in the categories used and even in some theoretical premises but also the fundamental macroeconomic theoretical divergences have served as an incentive for debate in recent years.In this exchange, each approach recognises certain coincidences and merits of the other.However, what underlies this 'dialogue' is the struggle for greater predominance in CPE, with each contribution highlighting only the virtues of its own approach over the others.
The trigger of the debate was the defence made by Baccaro and Pontusson (mainly 2016 but also 2018) of the necessity to include aggregate demand as a key analytical element to CPE literature.Applying their study to four countries -United Kingdom (LME), Sweden (CME), Germany (CME), and Italy (MME) -Baccaro and Pontusson consider that depending on institutional and structural factors, each country replaced the 'Fordist' pro-labour and productivity-seeking institutional arrangements until the 1970s by growth models driven by one (or more) of the components of private aggregate demand.Germany is considered a straightforward example of an export-led economy, the UK an economy which is debt-financed household consumption-led, Sweden is a 'hybrid' model able to combine internal and external demand thanks to the low price-sensitivity of its exports.Finally, the Italian case shows a failed model with persistent stagnation.
Baccaro and Pontusson's efforts have been recognised by leading authors of both streams of thought and they have even led to a degree of self-criticism on 'each side' (Hope and Soskice 2016;Stockhammer 2018;Stockhammer and Ali 2018;Hein et al. 2020;Stockhammer and Kohler, 2022).However, small concessions aside, each approach has been insisting on the analytical superiority of its method.This section is intended to contrast the controversies raised in the debate and their implications for the possibility of the PK gaining dominance and consolidating its position within CPE.Four aspects will be considered, starting from shared commonalities in empirical observation to finally expose their fundamental theoretical differences.

classiFyinG Growth Models
Recently, Hein et al. (2020) acknowledged the effort of Baccaro and Pontusson (2016) as 'an interesting attempt to introduce the PK notion of Post-Keynesian Financialised Growth Models as an alternative to varieties oF caPitalisM: (in)stability ... revista de econoMía Mundial 65, 2023, 71-92 macroeconomic demand and growth regimes into the CPE literature' .However, they also criticise such an 'attempt' .First, when talking about the evolution of wage-share during 'Fordist' and 'post-Fordist' periods, Baccaro and Pontuson (2016) 'have not clearly followed the distinction between wage-or profit-led demand and growth regimes, on the one hand, from pro-labour or pro-capital distributional policies and the resulting economic developments, on the other hand' (Hein et al. 2020, p. 4).Second, and related to the first point, Baccaro and Pontusson (2016) 2016) regarding the classification of Sweden is striking.However, equally remarkable is the divergence found in the classification method between Kohler and Stockhammer (2021) and Hein et al. (2020), both Post-Keynesians.The latter point out that, after the 2008 crisis, there was a general trend in OECD countries from 'debt-led private demand boom models' towards more export or weakly-export-led models.In contrast, according to Kohler and Stockhammer (2021), the improvement in the net financial position and trade balance of some countries is not a sign of a change of model, but the reflection of the contradictions of the debt-led pre-crisis growth model.
We should note two points regarding these coincidences and discrepancies.First, the export and debt-led models and their representative countries are found in more than these two currents (Stiglitz, 2007;Sinn, 2012;Flassbeck and Lapavitsas, 2013).Second, there is no concrete and consensual method to determine which country follows a debt-led, export-led or other intermediate model.Even the combined analysis of sectoral financial balances and the components of aggregate demand, as Hein does (Hein and Mundt 2012;Hein et al. 2020;Hein and Martschin 2020), could lead to ambiguous cases (Barredo-Zuriarrain 2019).Thus, the general observation of growth patterns or sectoral balances of a country or region may lead to coincidences between currents as well as to divergences within the same school (see Table 1 below, first row).In controversial cases, the coincidences or discrepancies are more related to the years, indicators and context analysed than to the theoretical approach employed itself.
One of the key contributions of Kohler and Stockhammer (2021) is the emphasis on growth drivers rather than growth components, so that we can 'broaden the theoretical foundations of growth models' .These growth drivers -competitiveness, asset price inflation, fiscal policy, etc. -are ultimately the factors that influence the growth of demand components.In this sense, Stockhammer (Stockhammer and Ali 2018;Kohler and Stockhammer 2021;Stockhammer 2021) criticised VoC for overemphasizing price-competitiveness and neglecting debt as a determinant of imbalances and growth in the Eurozone.However, it is worth noting that the PK literature provides examples of authors who have integrated price-competitiveness (Stockhammer 2012;Bibow 2013;Hein and Mundt 2012) and non-price competitiveness (e.g.Simonazzi et al. 2013;Gräbner et al. 2020) when explaining the Eurozone crisis.Here again there are points of coincidence not only between the PK and the VoC perspective, but also with others (e.g.Flassbeck and Lapavitsas 2013).
Notwithstanding particular coincidences, in PK literature, the key factor that marks pre-crisis growth and post-crisis stagnation is debt.In this sense, the difference -apparently of a minor order -between the VoC denomination of the 'domestic demand' model and the PK 'debt-led' is eloquent.For PK it is not just a matter of comparing the growth of the demand components, but of pointing to debt as an identifying feature explaining the difference and complementarity of two opposite models.How both approaches incorporate the role of debt in determining a growth model is not anecdotal because, as we show below, it poses a fundamental difference in how long-term stability is explained (see Table 1, second row).
Crucially, it is a great step forward to go from observing macroeconomic variables -growth components -to asking what drivers have led each one to evolve, as Kohler and Stockhammer (2021) propose, in a certain way.However, it is also worth asking why in each country one driver or another Growth drivers and triggers.has been 'activated' or not.What has led Spain, the United States or Ireland, for example, to base their growth during the first decade of the 21th century on leverage and not on export promotion?Why have households in Germany, affected by the same growing trend of inequality and financial liberalization as the former countries, not boosted their spending by means of increasing debt?
In short, what are the triggers, institutional or others, that favour a growth driver to emerge in a number of countries and not in others?From the VoC perspective, a systematic answer to these questions is provided by describing presupposed institutions of CME (export-led) and LME/MME (domestic demand-led) countries.A country is more likely to follow an export-led model if institutional coordination allows for it.In Iversen and Soskice (2012) and Johnston and Regan (2016), inter alia, political support for firms' mercantilist strategy depends, to a large extent, on the coalitions in power.On the other hand, also from the VoC, the possibility that households resort to debt in response to pressure on wages and rising inequality rests on different hypotheses, such as the type of regulation, the configuration of the Welfare State, workers' prospects (Carlin and Soskice 2009;Iversen and Soskice 2012;Behringer and Van Treeck 2019).
The PK approach does not differ much from VoC in terms of identifying the different triggers of growth drivers; all the factors outlined above are collected in the recent PK literature (e.g.Detzer and Hein 2014;Kapeller and Schutz 2015;Detzer 2016;Kohler and Stockhammer 2021).However, two points should be noted regarding the PK proposal.First, as their underlying theory is focused on aggregate demand and not on institutional analysis, the association between the growth model and institutional characteristics is not made on the basis of pre-established taxonomies but on the basis of country-based analysis.Only a few recent attempts point in the direction of linking the grouping of countries on the basis of common institutional characteristics with the type of growth models (e.g; Hein et al. 2020;Cárdenas et al. 2021).For instance, Hein et al. (2020) link Hay and Wincott's (2012) welfare-based taxonomy to growth models, finding a clear and differentiated relationship before the 2008 crisis between a country's welfare system and its growth model.Second, for most of the PK studies, the mention of growth drivers are not detailed investigations but brief references to other studies -some from the VoC perspective -which largely focus on the United States and Germany (e.g.Giavazzi and McMahon 2008;Van Treeck and Sturn 2012;Kumhof et al. 2012).This is surely the great challenge for PK to consolidate as an alternative within CPE.The classification and subsequent comparison of countries should not be based merely on the relative weight of the components of aggregate demand; not even on the specific study of the growth triggers of an isolated country.The comparative potential in terms of political economy will depend, to a large extent, on the ability to relate, on the one hand, growth patterns to, on the other hand, institutional, historical or cultural factors shared by groups of countries.Such a challenge is complex, especially in a volatile context as the current one, where national political realities change and periodic crises abruptly break apparently stable growth patterns.However, the very characteristics of such a challenge encourage us to think that a three-phase organization of such an endeavor is possible (see Table 2).
In a first stage, -(1) in Table 2 -empirical studies should keep further developing the relationship between growth models and growth drivers (in Table 2, column 1 and 2 respectively); that is, to deepen our understanding of the factors (such as real estate prices, export sophistication, monetary and fiscal policy, price competitiveness, etc.) that explain the dynamism of the different components of aggregate demand in each country.The existing theoretical research and case studies mentioned above must be accompanied by a line of work that is capable of linking growth drivers and components with groups of countries in specific periods.
Second -see (2) in Table 2 -, these drivers cannot appear in the analysis as exogenous elements, which may or may not be activated by chance in different countries, but must be studied as probable endogenous results of socioeconomic or heterogeneous cultural factors and institutional forms shared between countries.In other words, the association between these growth triggers (column 3 in Table 2) and growth drivers should be explored.In this field, of course, the PK literature can continue to draw on VoC research that relates forms of institutional complementarity and growth impacts.But the PK literature has its own institutionalist line that makes key contributions around forms of regulation and socio-cultural factors, on the one hand, and their impact on decision making, on the other (e.g; Eichner, 1985;Seccareccia, 1991;Lee, 1996, Prasch, 2005;Weller, 2015;Zalewski, 2018) Moreover, this institutionalist branch incorporates uncertainty and instability as elements specific to the international economy (Minsky, 1982;Whalen, 2020) and may therefore constitute a more consistent starting point than VoC research.
The simultaneous development of these first two phases or even their permanent dialogue from early stages would be advisable.Finally, the contributions resulting from both would allow, in a third phase -(3) in last row in Table 2), to consolidate the PK contribution as an alternative in the CPE, grouping countries based on the relationship between, on the one hand, the growth triggers -including the existing institutional diversity -that they share and, on the other hand, their classification in growth models.

instability and institutional eFFiciency
As we move from empirical observation to explanation and interpretation, coincidences give way to clear divergences.This is again evident when it comes to explaining the GFC.We have seen in the previous point Stockhammer's (2021) criticism of VoC for underestimating debt as the growth driver of domestic demand-led countries.In the same way Stockhammer adds that, to the extent that VoC is based on New Keynesian macroeconomics, it neglects debt as a crisis driver and financial instability.On this point, a clarification must be made: both debt and financial instability do already appear at the centre of the VoC interpretation of the GFC (Hall 2012;Hancké 2013).Indeed, Iversen and Soskice (2012, p. 42) draw on Minsky to explain how risks in the financial sector carry over into the 'real' sphere.However, the VoC explanations on the global crisis and especially that of the Eurozone, steer far from taking instability as an endogenous feature of the financial system.Although in each study there is no explicit mention of the macroeconomic principles on which it is based, there is a recurrent reference to 'inadequate regulation' and 'market failures' (Soskice 2009, Iversen andSoskice 2012;Hall 2014), which are traditional New-Keynesian postulates.
From the PK approach, just as debt was the main determinant for growth and pre-crisis imbalances, it is also the main explanatory factor for the crisis.This approach offers an explanation in which expansion and leverage are essentially endogenous phenomena attracting large gross flows of foreign capital (Barredo-Zuriarrain 2019).Unlike in the VoC literature, for Postkeynesians imbalances are a logical and relatively minor consequencecompared to the volume of debt -of the interaction between opposite growth models.Despite the disagreement between Kohler and Stockhammer (2021) and Hein et al. (2020) about the classification method, both articles agree on something fundamental: debt-led expansion brings with it the conditions to end up in a debt-burdened recession or stagnation.
This leads us to a fundamental question about the nature of institutions.Institutions do not mean the same thing for each approach.In the VoC literature, institutions appear as a 'set of rules' that help solving coordination problems among firms (Hall and Soskice, 2001, p.9).Thus CME and LME are both successful models of institutional complementarity.It is therefore difficult to explain severe recession phenomena from the VoC perspective (Bruff 2011;Jessop 2011).This is also why supranational regulatory elements are alluded Growth drivers and triggers.anti-crisis and lonG-run Growth Policies Baccaro and Pontusson's (2016) proposal to draw on PK economics for the sake of incorporating aggregate demand into CPE was replicated shortly thereafter by Hope and Soskice (2016).They consider the 'growth models approach as supporting recent developments in varieties of capitalism rather than undermining them' (ibid, p. 3).Moreover, they add that in order to include aggregate demand in CPE it is not necessary to resort to the PK approach; they remind that the 3EqM already incorporates demand and the first equation (the IS) shows precisely how important it is in determining output but in the short run.Besides, they point out that Post-Keynesian and neo-Kaleckian literature on growth regimes does not integrate either the interrelationship of wage demands with inflation -the Phillips Curve equation -or the importance for stabilization of macroeconomic policies (especially monetary policy) in the face of shocks -the Monetary Rule equation.In fact, to find references to the importance of demand policies from the VoC perspective, it is not necessary to resort to the 3EqM.Especially when studying the causes of and remedies to the Euro-crisis, several followers of VoC (e.g.Hall 2012;Iversen et al. 2016) refrain from disparaging the important role that a fiscal union and an expansive fiscal policy can play to overcome that crisis, nor do they minimise how important a banking union and the role of the central bank as lender of last resort is.
That said, it is true that the structural changes and interventions needed to overcome the GFC consequences (not only in the Eurozone), to minimise the probability of a financial crisis recurrence and to promote long run growth, differ considerably between both approaches.Coherently with the causes of the business and financial cycles, VoC followers advocate the implementation of counter-cyclical macroeconomic policies to combat the business cycle consequences and micro and macro prudential policies to avoid the occurrence of financial cycles (Iversen and Soskice 2012).Because any kind of crisis is due to exogenous shocks disturbing a stable long run equilibrium, there is no need for any other type of major changes in the economic system.Fiscal and monetary policy are, in any case, valid to close the output gap if efficiently managed but ineffective in moving output above its potential or unemployment below the NAIRU (Hope and Soskice 2016, p. 11).As for making the long run equilibrium grow in a sustainable way, this depends on productivity growth, technological upgrading and ultimately on the right institutional setup.The labour market is where the equilibrium level of employment (and its growth) is determined and therefore the rules and incentives governing the functioning of the labour market are of paramount importance.VoC scholars advocate for supply side (structural) policies making the labour and product market work with higher degrees of freedom and competition (Hall 2012, p. 365;2014, p. 1234).
There is a coherence already highlighted in section 3, regarding the analysis of the role of demand in the short-medium term between VoC and 3EqM.This has been made clear by the work in recent years of leading authors in both fields, such as Carlin and Soskice who are aware of the need to provide, albeit ex-post, a macroeconomic approach to the VoC perspective.However, the consolidation of such a link is a pending challenge for VoC, especially if it is intended to be predominant in the CPE.In any case, if such a link with 3EqM is affirmed, the treatment of aggregate demand is much closer to New Keynesianism than to the PK literature on demand regimes and financialised growth models.Stockhammer (2018) self-critically acknowledges the 'shortcomings of PK economics' in routinely incorporating the interventions of different social classes, as well as monetary and fiscal decisions, into the models.However, the author claims (Stockhammer 2018;2021), in agreement with Lavoie (2015) (also Blyth and Matthijs 2017; Schwartz and Tranøy 2019) that the PK approach gives more adequate macroeconomic foundations to CPE than New Keynesian economics.First, in PK, conflicts on factor distribution change the composition of GDP and affect inflation, but also have an impact on the level of aggregate demand.Second, aggregate demand does not 'only' affect the short run equilibrium level of output but it also stimulates economic growth in the long-run.Third, aggregate demand should be encouraged in a sustainable way.Allowing the financial system to keep the economy growing through debt financed consumption and mercantilist export-led growth models is not sustainable.Contrary to VoC, for which an ´artificial´ increase in wages (not driven by productivity growth) will have detrimental effects on short and longrun growth, the PK way to achieve healthy long-run growth is by implementing internationally coordinated pro-labour and redistributive policies which will keep demand growing along with output (Hein 2012;Kaufman 2012).

conclusion
CPE is full of similarities between the VoC and PK proposals, more than the respective scholars involved recognise: the identification of ideal-type growth-models, the reference countries of each model, the identification of growth and imbalance drivers and even the call for an expansionary fiscal and monetary policies in periods of crisis.But they differ markedly on fundamental issues, most importantly the underlying macroeconomic model as well as in the understanding of institutions.It is these divergences that highlight the contribution to the and from which challenges arise for each of the two approaches.
The future performance of the VoC approach within CPE depends to a large extent on the ability to associate its classificatory logic with a consistent macroeconomic model.If the proposal to support it in the 3EqM consolidates, so will the shortcomings pointed out by the PK economists, such as the difficulty to explain potential phenomena of endogenous financial instability or long-term stagnation.In this context, an opportunity opens for the PK literature on financialized growth models to consolidate itself within the CPE: it has a clear macroeconomic foundation, capable of integrating the role of demand as an engine of short and long-run growth, and the unstable nature of capital accumulation.That said, this is not a sufficient condition to establish itself as an alternative in CPE.It is true that the theorization of financialisation integrates institutional and regulatory aspects, where change and conflict are defining elements.However, in order to consolidate itself as an alternative in the CPE, the great challenge now for PK economics is to move from a countrybased analysis to a proposal for a systematic association between, on the one hand, institutional aspects shared between countries and, on the other hand, their classification in relation to their growth model.We have proposed a threephase work as good support point for such endeavour.The first two phases would consist of further developing the relationship between growth drivers and growth models; and to integrate these drivers as endogenous probable results of socioeconomic, cultural or institutional factors that we call growth triggers.Moving these research lines from a country-by-country study to a cross-country comparison would allow, in a third phase, to consolidate the PK contribution to CPE by grouping countries sharing grow triggers, drivers and similar growth patterns.
-and laterHope and Soskice (2016) -are confusing growth models with demand regimes(Hein et al. 2020; Stockhammer 2021, p.3)  2 .Third, Hein et al. (2020) criticise Baccaro and Pontusson's empirical method to claim that Sweden and Germany follow different models.Interestingly, an alternative estimate leads Hein et al. (2020) to the same conclusion as VoC authors Hope et al. (2016) a few years earlier in their own response to Baccaro and Pontusson (2016): a more correct estimate shows that Sweden is, like Germany, a clear export-led case with high non-price competitiveness.This coincidence between Hein et al. (2020) and Hope et al. (

table 1 .
Grounds For (dis)entanGleMent between PK and voc aPProaches

table 2 :
three-staGe worKinG ProPosal For a consistent PK contribution to cPe , to that the 'symbiotic' relationship between two VoC types becomes 'pathological' in the framework of a monetary union(Hall 2018).As opposed to VoC, the PK literature on growth models starts from financialisation and its effects on capital accumulation (see Table1, third row).Far from taking efficient models, it integrates the changing character of national economies as well as institutions arising from power struggles, which is a vision compatible with North's (1990) institutionalism or with the Regulation Theory, with which some PK scholars have already engaged into closer debate (e.g.Jäger and Springler 2015). to