House price cycles, housing systems, and growth models

Kohler, Karsten; Tippet, Benjamin; Stockhammer, Engelbert Article House price cycles, housing systems, and growth models European Journal of Economics and Economic Policies: Intervention (EJEEP) Provided in Cooperation with: Edward Elgar Publishing Suggested Citation: Kohler, Karsten; Tippet, Benjamin; Stockhammer, Engelbert (2023) : House price cycles, housing systems, and growth models, European Journal of Economics and Economic Policies: Intervention (EJEEP), ISSN 2052-7772, Edward Elgar Publishing, Cheltenham, Vol. 20, Iss. 3, pp. 461-490, https://doi.org/10.4337/ejeep.2023.0121 This Version is available at: https://hdl.handle.net/10419/284344 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ House price cycles, housing systems, and growth models* Karsten Kohler** Economics Department, Leeds University Business School, UK Benjamin Tippet*** Department of International Business and Economics, University of Greenwich, UK Engelbert Stockhammer**** Department of European and International Studies, King’s College London, UK The paper provides a framework for theorising the role of house price cycles in national growth models. We synthesise Minskyan approaches with comparative political economy (CPE) by arguing that institutions influence the extent to which countries experience what we call ‘house-price-driven growth models’. First, we argue that house price dynamics have been undertheorised in existing growth models analysis. Finance-led models can be properly understood only against the background of rising house prices that stimulate consumption through wealth effects and investment through construction. Second, we identify behavioural and Minskyan theories of housing cycles as suitable frameworks to theorise the impact of housing on growth. However, this literature does not provide an analysis of cross-country differences in housing cycles. Third, drawing on the CPE literature on housing systems, we argue that factors such as private homeownership and mortgage-credit encouraging institutions can explain differences in the intensity of housing cycles. We provide preliminary empirical support for this framework from a cross-country analysis. Our results show strong cross-country heterogeneity in the intensity of housing cycles. Countries with more intense house price cycles also tend to exhibit more volatile business and debt cycles. Homeownership rates and mortgage-credit encouraging institutions are positively correlated with the volatility of house price cycles. Keywords: post-Keynesian economics, comparative political economy, growth models, housing, house price cycles JEL codes: E32, O57, R21, R31, B52 1 INTRODUCTION Housing markets and household debt have received much attention in analyses of the 2008 Global Financial Crisis. Since then, specialised literatures have been growing in * This work was supported by the Leverhulme Trust under Grant RPG-2021-045 (‘The Political Economy of Growth Models in an Age of Stagnation’). The paper has benefited from comments by two anonymous referees, participants of the ‘Frontiers in Growth Regimes Research’workshop at IPE Berlin as well as the 26 th FMM conference in Berlin in October 2022. The usual disclaimers apply. ** Corresponding author. Email: [email protected]. *** Email: [email protected]. **** Email: [email protected]. Received 7 November 2022, accepted 30 May 2023 Research Article This is an open access work European Journal of Economics and Economic Policies: Intervention, Vol. 20 No. 3, 2023, pp. 461–490 First published online: August 2023; doi: 10.4337/ejeep.2023.0121 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd The Lypiatts, 15 Lansdown Road, Cheltenham, Glos GL50 2JA, UK and The William Pratt House, 9 Dewey Court, Northampton MA 01060-3815, USA both comparative political economy (CPE) and heterodox economics that theorise unstable financial dynamics in the household sector. However, the extent to which household debt dynamics are linked to housing markets varies in these debates. In the growth models approach to CPE, housing sits within what is variably called the ‘finance-led’,‘debt-driven’ or ‘consumption-led’growth model (Baccaro/Pontusson 2016; Ban/Helgadóttir 2022; Reisenbichler/Wiedemann 2022). For the USA, the UK and Spain, the role of house prices and the construction industry as a driver of growth is widely acknowledged. However, the terms ‘consumption-led’growth models and ‘debt-led consumption boom’suggest a focus on consumption and its funding rather than on housing. Furthermore, many studies focus on household debt rather than houses price cycles. In general, housing cycles do not form a core feature in the conception of these growth models (Wood/Stockhammer 2020). In heterodox economics, theoretical models of booms and busts in the housing market have been developed, drawing on behavioural and post-Keynesian/Minskyan approaches. In behavioural models, speculative economic actors are boundedly rational and prone to herd behaviour. Expectations about house price appreciation temporarily overshoot, leading to volatile cycles in house prices (Dieci/Westerhoff 2012, 2016). In post-Keynesian and Minskyan approaches, asset price booms spill over to the real economy as they stimulate debt-financed private spending (Nikolaidi/Stockhammer 2017). Booms thus come with a build-up of financial fragility that prepares the bust. While Minsky himself was mostly concerned with corporate debt and equity prices, more recent work has applied his framework to household debt and housing markets (Zezza 2008; Caverzasi/Godin 2015; Ryoo 2016). While these models provide a rigorous explanation of the endogenous nature of cycles in housing markets and economic activity, they are typically abstract and do not specify which countries or growth models are more likely to develop intense housing cycles. In particular, the focus has been on within-country rather than cross-country analysis. A stream within the CPE literature studies varieties of ‘residential capitalism’, focussing on cross-country differences in housing institutions such as homeownership and household debt ratios (Schwartz/Seabrooke 2008; Johnston/Kurzer 2020). This literature argues that certain financial and housing institutions are conducive to mortgage credit expansion and can thus explain cross-country differences in the relevance of mortgage debt. However, this literature has not explicitly investigated the role of institutions for house price cycles as its focus is typically on debt and political outcomes. Thispaperarguesthatwhileeachofthesestreamsofresearchcontainsimportant insights, due to their separation, these literatures individually fall short of providing a coherent account of the role of housing in growth models (henceforth GMs). Our contribution is, firstly, to critically review the theoretical role of housing in the existing GM literature. We conclude that it lacks a coherent foundation to theorise what we call ‘house-price-driven GMs’. Secondly, we provide a stylized framework for integrating housing into the GM perspective and, thirdly, offer some supportive cross-country evidence for this framework. At the theoretical level, we synthesise the Minskyan approaches to boom–bust cycles in housing markets with the CPE literature on housing institutions. Our synthetic approach can be summarised as follows. First, house prices are central to the macroeconomics of growth processes and thus need to be a key feature of GM analysis. Second, house prices are driven by speculative behaviour that generates Minskyan endogenous cycles. House price dynamics affect consumption, residential investment as well as financial instability and thereby translate into business cycles. Third, we argue that housing institutions can explain cross-country differences in the intensity of such housing cycles. Drawing on the CPE literature on housing systems, we consider private homeownership and 462 European Journal of Economics and Economic Policies: Intervention, Vol. 20 No. 3 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd mortgage-credit encouraging institutions as country-specific factors that potentially favour the emergence of house-price-driven GMs. We support our argument with empirical data from a cross-country dataset of 32 OECD countries. We employ turning point analysis to establish some stylized facts on house price cycles across countries. Our results suggest strong cross-country heterogeneity in the intensity of cycles. Countries with more intense house price cycles also tend to exhibit more volatile business and household debt cycles. We present preliminary evidence that homeownership rates and, to a lesser extent, mortgage-credit encouraging institutions, are positively correlated with the volatility of house price cycles, suggesting that institutional structures indeed matter. We conclude that the Minskyan approach provides a macroeconomic link between CPE analyses of housing systems and the growth models approach. Institutions determine the extent to which housing is treated as a speculative asset that favours unstable growth processes. This implies that political efforts to curb the cyclical dynamics of finance-led growth models should focus on the re-regulation of housing markets and of mortgage finance. Public housing can further reduce the share of housing available for speculation. The paper is structured as follows. Section 2 critically reviews the role of housing in the GM approach and argues in favour of putting house price cycles at its centre. It then introduces behavioural and Minskyan approaches to speculative house price cycles. Finally, it links these approaches to the CPE literature on housing systems. Section 3 provides a concise statement of our framework for synthesising these approaches. Section 4 establishes some stylized facts on house prices dynamics across countries and provides preliminary evidence in support of our theoretical framework. Section 5 concludes. 2 HOUSING IN GROWTH MODELS, HETERODOX MACROECONOMICS AND CPE 2.1 Growth models: debt-driven consumption-led growth or house-price-driven growth? Differences in growth dynamics across countries are a key topic in both CPE and Kaleckian macroeconomics. In the 2000s, CPE was dominated by the Varieties of Capitalism (VoC) approach (Hall/Soskice 2001). VoC builds on neo-institutionalist theory and analyses how different institutional configurations provide a comparative advantage to firms. It identifies institutional sources of microeconomic efficiency that allow different economic models to persist in a globalised world economy. Drivers of aggregate demand, in particular real estate booms and household debt, have not featured in first-generation VoC analyses due to its focus on corporate finance institutions and their implications for competitiveness. In the early 2010s, Kaleckian macroeconomists developed the notion of ‘demand regimes’ to capture country-specific macroeconomic regimes (for example, Lavoie/Stockhammer 2013a). In this approach, the formation of aggregate demand plays a key role, in particular through functional income distribution. Building on the model in Bhaduri/Marglin (1990), it was argued that demand regimes can either be wage-led or profit-led, depending on whether the stimulating effect of an increase in wage shares on consumption outweighs the potentially negative effect on investment. Lavoie/Stockhammer (2013b) argue that in the neoliberal era, demand regimes are in principle wage-led, but due to falling wage shares, other growth drivers have taken centre stage: debt-driven and export-driven growth. Hein/ Mundt (2013) identify export-driven GMs based on the respective contributions of net exports to GDP growth, and for debt-driven models they use the growth contribution of House price cycles, housing systems, and growth models 463 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd private consumption combined with information on the change in borrowing by the household sector. In 2016, Baccaro/Pontusson (2016) published a highly influential article that made the case for introducing Kaleckian macroeconomic analysis of demand regimes into CPE. They consider post-war capitalism as wage-led and the post-1980s as different forms of profit-led regimes. They analyse four country cases for the post-1980 period and distinguish between export-led (Germany and Sweden), what they call ‘consumption-led’(UK) and a failed model (Italy). What defines these growth models is an institutional and political structure that favours a strong dynamism of a specific component of aggregate demand relative to the others. Baccaro/Pontusson’s(2016)‘growth models perspective’has become widely used in CPE and inspired various follow-up studies (for example, the edited volume by Blyth et al. (2022), Behringer/van Treeck (2019), Hein et al. (2021), Kohler/Stockhammer (2022) and O’Donovan (2021)). To understand the role of housing and house prices in the growth models approach, we must take a closer look at the analysis of debt-led growth models in the GM approach. The precise labels differ: Baccaro/Pontusson (2016) use ‘consumption-led’growth; Hein/Mundt (2013) use the category of ‘debt-driven consumption boom’; 1 and Lavoie/ Stockhammer (2013b) speak of debt-driven growth. We want to highlight three theoretical weaknesses in these concepts: first, the question of whether consumption growth hinges solely on credit expansion or whether house price growth is a key factor in stimulating credit creation and consumption demand. Second, the question of how accurate it is to conceive of debt-driven GMs as predominantly consumption-led. Third, the question of cyclicality of the debt-driven model and of house prices in particular. The first issue is about the determinants of consumption. Following Baccaro/Pontusson (2016), a wide range of CPE scholars use the term ‘consumption-led growth models’ (often with adding ‘credit driven’, for example, Reisenbichler/Wiedemann 2022). In Baccaro/Pontusson (2016) as well as in Hein/Mundt (2013), this is based on a GDP growth decomposition that identifies consumption as the largest GDP component in terms of its growth contribution. However, for much of the institutionally oriented CPE literature, causal chains in economic relations remain underspecified, thus there is some ambiguity about what drives consumption. For some authors (for example, Reisenbichler/Wiedemann 2022) it is clear that housing is central to credit growth, and that credit growth is central to consumption and economic growth. For others like Baccaro/Pontusson (2016), housing only features in a single footnote, but through most of their analysis one could substitute ‘consumer credit’for ‘household debt’. In heterodox economics, causal relations are clearer as arguments are formulated as models that specify the relevant determinants. Below we will propose a Minskyan interpretation of the debt-driven growth model based on housing cycles. The main alternative to this (in heterodox economics) is the emulative-behaviour theory of consumption. Emulative consumption posits that lower-income households try to emulate the consumption levels of richer peer groups (Frank 2014). For example, households may look to the income decile above their own. Microeconomically, that means that consumption not only depends on the household’s own income, but also on the consumption (or income) of some reference group. Macroeconomically, it means that widening income inequality would lead to increasing consumption. It also means that some households will have to debt-finance their consumption. The theory is attractive to heterodox macroeconomists because it presents a feasible microeconomic alternative to mainstream economics and because it can 1. Hein et al. (2021) speak of a ‘debt-led private demand boom’to include the effects on private investment. 464 European Journal of Economics and Economic Policies: Intervention, Vol. 20 No. 3 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd explain some stylized facts of the US economy in the period before the Global Financial Crisis (GFC): rising income inequality paired with dynamic consumption growth and rising household debt. Kapeller/Schütz (2015) integrate the argument into a Minskyan model of consumer debt cycles and Behringer/van Treeck (2019) into a GM analysis of the United States. Belabed et al. (2018) develop the argument empirically in an open economy setting. Prante et al. (2022) present simulation results for a two-country macro model that effectively generates debt-driven as well as export-driven GMs. What all these models share is an understanding of the consumption-led GM as driven by consumer credit. They get by without mortgages and without a housing market. The emulative consumption hypothesis is plausible in terms of its microeconomic analysis of consumption, but it is deficient as a theory of contemporary household debt. It may explain why households want to consume beyond their current income, but it does not explain why households would be confident in their ability to take on and service debt. Even if one grants that households are ignorant of the long-term financial implications of their consumption spending, the question arises why financial institutions would lend to households. Banks would only do so if the household can offer appropriate collateral. The most important financial asset of households is real estate, and in fact most household debt is mortgage debt, not consumer debt. In short, emulative consumption behaviour can explain why households want to borrow more, but it cannot explain why banks are willing to lend to them. For that they need collateral, which establishes a centrality of house prices. A second issue is the question of how central consumption growth is in ‘credit-driven consumption-led economies’. The origin of the concept of debt-led growth is the preGFC boom and, methodologically, GDP growth decompositions. Indeed, in that period in the USA and UK, consumption growth constituted a large share of overall GDP growth. More generally, Ban/Helgadóttir (2022) identify a high share of consumption in GDP as a key characteristic of financialised GMs. However, this runs the danger of overstating the significance of consumption in the context of rising real estate prices. Investment seems to respond as least as strongly to changes in house prices as consumption (for example, Stockhammer/Wildauer, 2016; Stockhammer/Novas Otero 2022). Investment is a smaller share in GDP than consumption (in advanced) economies, but it is also more volatile. In some economies, namely Ireland and Spain, residential investment has played an important role in the booms before the GFC. The role of residential investment as an (autonomous) driver of economic activity has recently also been highlightedinsomeoftheSraffiansupermultiplier literature (Fiebiger/Lavoie 2019; Teixeira/Petrini 2023). The third issue is to what extent one regards house price cycles as due to specific circumstances or as systemic (endogenous) features. To illustrate, Reisenbichler/Wiedeman (2022: 236) write: ‘the very forces that bring about growth can also have destabilizing effects when financial actors engage in risktaking behaviors, disrupt existing markets with new financial innovations, or exploit regulatory loopholes in pursuit of profit’. This seems close to the position that financial crises are due to exogenous shock. In contrast, a Minskyan approach, which we introduce more systematically below, argues that financial cycles are endogenous (Kohler/Stockhammer 2022). In summary, the GM approach lacks a satisfactory explanation of the drivers of private demand in finance-led growth models. While emulative consumption may be a contributing factor, we argue that it cannot convincingly explain the rise of mortgage debt in many finance-led models, which is strongly linked to house price dynamics. We therefore posit that housing finance and house prices dynamics are central to understanddebt-ledGMs. House price cycles, housing systems, and growth models 465 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd 2.2 Heterodox macroeconomics: endogenous booms and busts in housing and economic activity House price dynamics and their cyclical nature have received somewhat more attention in macroeconomics, both mainstream and heterodox. In the mainstream literature, the cyclicality of house prices has mostly been acknowledged in empirical research on ‘financial cycles’(Claessens et al. 2012; Borio 2014; Rünstler/Vlekke 2018; Strohsal et al. 2019; Schüler et al. 2020). Three key stylised facts about financial cycles have been established. Firstly, credit and property prices follow each other closely, whereas equity prices exhibit more idiosyncratic dynamics. Secondly, the duration and amplitude of financial cycles is larger than that of conventional business cycles (that is, cyclical changes in GDP). While conventional business cycles tend to last up to 8 years, the average length of financial cycles is roughly 16 years (Borio 2014). However, business cycles also exhibit a medium-term frequency that is closely correlated with financial cycles (Rünstler/Vlekke 2018). Thirdly, recessions associated with house price busts are longer and deeper than other recessions (Claessens et al. 2012). This suggests a strong link between property prices and economic activity. What drives these financial cycles? Borio (2014: 186) suggests that cycles are caused by endogenous forces such that ‘the boom sows the seeds of the subsequent bust’. However, the mainstream financial cycle literature has not yet put forward a coherent theoretical framework for modelling such endogenous cycles. The absence of a coherent theoretical underpinning for financial cycles may partly stem from the difficulty involved in modelling endogenous fluctuations within a neoclassical framework (Borio 2014). In a neoclassical world, house prices are assumed to be determined by an arbitrage relationship between owner-occupied and rental housing (see Duca et al. 2021). In equilibrium, the rates of return (adjusted for costs) from owning and renting a house will be equal. House prices are then given by expected house price appreciation plus future rent income, discounted by the so-called ‘user cost’of housing (typically the sum of the interest rate, the rate of housing depreciation, and property-related taxes). If agents form rational expectations about future house prices, any shock will induce a rapid adjustment of house prices towards a stable path that leads the housing market back towards its fundamental equilibrium (see Dieci/Westerhoff 2016 for a neat discussion). While house prices may thus exhibit some temporary overshooting in response to shocks, there are no periodic cycles. 2 The intrinsic stability of housing markets in neoclassical and New Keynesian models contrasts with behavioural and post-Keynesian/Minskyan frameworks. In these heterodox approaches, housing markets are a source of instability that can generate endogenous booms and busts –with severe spillovers to the real economy. This heterodox literature can broadly be grouped into behavioural and post-Keynesian/Minskyan approaches. In behavioural asset pricing models, cycles are generated by the speculative behaviour of investors (Hommes 2006; Dieci/He 2018). Unlike in neoclassical models with a single representative agent, there is behavioural heterogeneity. Agents rely on simple behavioural rules (or heuristics) rather than rational expectations to anticipate future prices, and the prominence with which these rules are used changes over time. In a common setup, two types of rules are considered. Extrapolative rules assume that past trends continue and thus tend to destabilise price dynamics. Regressive rules expect prices to revert to 2. Qualitatively similar results prevail in more complex dynamic stochastic general equilibrium models, where fluctuations in housing markets are driven by exogenous shocks. For example, in Iacoviello/Neri’s (2010) estimated DSGE model of the US housing market, the 1998–2005 boom is mostly explained by technology shocks, while monetary policy shocks account for most of the subsequent bust. 466 European Journal of Economics and Economic Policies: Intervention, Vol. 20 No. 3 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd their mean values. The interaction between these two types of expectation-formation rules is a key source of asset price dynamics. Dieci/Westerhoff (2012) introduce such a mechanism into a simple housing model. The demand for houses is decomposed into real demand and speculative demand. Real demand is decreasing in house prices, but speculative demand is increasing in future expected prices. Expectations are then driven by the interplay of extrapolative and regressive rules. For house prices close to the equilibrium, the majority of agents are optimistic and expect further house price inflation. This leads to a boom during which house prices become overvalued. Construction responds to price increases. There is thus an interaction between demand and supply whereby (speculative) increases in house prices induce an expansion of the housing stock, which feeds back negatively into house prices. This will eventually induce some agents to expect a correction of house prices. As more and more agents switch to such regressive expectations, house prices embark on a downturn. Once house prices are close to their fundamental value, extrapolative expectations take over again and push prices below equilibrium. As a result of this interplay between extrapolative and regressive expectations as well as the housing stock, endogenous fluctuations emerge. 3 While behavioural models offer rich analyses of endogenous cycles in housing markets, they typically do not analyse the effects of asset price cycles on aggregate economic activity. By contrast, post-Keynesian and Minskyan approaches consider the interaction of house price dynamics with private debt and business cycles (Zezza 2008; Charpe et al. 2011: ch. 11; Caverzasi/Godin 2015; Ryoo 2016; Teixeira/Petrini 2023). Charpe et al. (2011: ch. 11) introduce a housing market into a high-dimensional Keynes–Metzler–Goodwin model. Housing dynamics are driven by the interplay of rent and the housing stock. During the boom, rising rents have expansionary effects on output via an increase in construction. It is assumed that worker households consume in excess of their income, and thus go into debt. The increase in debt over time depresses consumption, while the increase in the housing stock reduces the return on housing from rents, which reduces in residential investment. This brings the cycle to an end. However, house prices are not analysed explicitly. Similar dynamics are at play in Zezza’s (2008) and Teixeira/Petrini’s (2023) stock-flow consistent models with a housing market. In Zezza (2008), poor households can buy or rent houses from rich households. Poor households can also take out debt to finance consumption based on the emulation approach discussed above. Rich households invest in houses based on expected capital gains, forming extrapolative expectations. During house price booms, the consumption of rich households increases through wealth effects. This induces emulation effects by poor householdswhogointodebt,whichsowsthe seeds for the downturn. In Teixeira/Petrini (2023), only capitalist households engage in debt-financed residential investment. In both models, a shock to (expected) house prices triggers a sustained housing boom that stimulates residential investment, followed by a bust. Unlike in Charpe et al. (2011) and Zezza (2008), where emulative consumption is a key mechanism, Ryoo (2016)’s model introduces a collateral channel into a Minskyan house price model. House price inflation relaxes collateral constraints and induces households to take on more credit to finance consumption. The build-up of debt during the boom ultimately weighs on consumption due to an increasing debt-service burden. To maintain their desired housing-to-consumption ratio, households reduce their demand 3. Dieci/Westerhoff (2016) provide an extension of this framework in which the equilibrium house price is derived from a user cost framework, facilitating the comparison of the behavioural with a neoclassical approach. House price cycles, housing systems, and growth models 467 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd for housing accordingly, which depresses house prices and brings the boom to an end. Collateral-based consumption effects can also be found in the stock-flow consistent model by Caverzasi/Godin (2015). 4 The aforementioned models are all theoretical, which reflects a noticeable imbalance between theoretical and empirical work on housing in these traditions. While there are empirical studies in the behavioural approach, they have so far focused on stock markets (Lof 2012; Chiarella et al. 2014; Hommes and in ’t Veld 2017) and foreign exchange markets (Westerhoff/Reitz 2003; de Jong et al. 2010) rather than on housing. Gusella/ Stockhammer (2021) provide evidence that house price dynamics are consistent with momentum trader models based on aggregate data for the USA, UK and France. Empirical work in the Minsky tradition is sparse and has focussed on household debt rather than house prices (Palley 1994; Kim 2016). In sum, heterodox approaches highlight the intrinsic instability of housing markets. In behavioural models, house prices are prone to endogenous boom–bust cycles due to speculative behaviour. Post-Keynesian and Minskyan approaches theorise the effects of house prices on macroeconomic dynamics, via either residential investment (Charpe et al. 2011; Zezza 2008) or consumption through collateral constraints (Caverzasi/ Godin 2015; Ryoo 2016). In this way, Minskyan approaches fill the gap in the theory of consumption-led GMs discussed in the previous section. However, a limitation of these macroeconomic theories is that they have little to say about differences across countries. Certain parameters will determine models’dynamics such as the existence and intensity of cycles, and these parameters indirectly capture structural features of the model economy. 5 The theoretical literature typically offers little discussion of country-specific factors, in particular what institutions would influence the existence and intensity of housing cycles. This means these approaches cannot be readily utilised for comparative analyses. 2.3 CPE on housing institutions across countries The CPE literature on housing often has a focus on distributional and political (rather than economic) outcomes, but also analyses cross-country differences in institutions such as homeownership rates and household debt levels (see Johnston/Kurzer 2020 for an overview). While it has not explicitly analysed speculative behaviour in housing markets, it identifies institutions that shape different housing systems. We argue that it provides a useful resource to build an understanding of differences in housing cycles across countries and the role of institutions therein. Early work in CPE focused on the level of homeownership as a key characteristic of housing systems (Harloe 1995; Kemeny 1995). It was argued that regulated housing markets arose in the Nordic and European core countries, where governments invested in rent-controlled and abundant social housing, which in turn competed with the private rental sector to push up standards and disincentivised homeownership. By contrast, profit-based housing markets with high owner-occupier rates arose in Anglo-Saxon 4. Different from some of the literature discussed in the previous section, the emulation effect in their model impacts households desired consumption (and thus credit demand), but credit supply is based on the leverage ratio of household, which depends on house prices. 5. In recent behavioural models, Martin et al. (2021, 2022) and Schmitt/Westerhoff (2022) examine the role of housing taxes, monetary policy and rent controls, respectively, on the amplitude of their model’s cycles. 468 European Journal of Economics and Economic Policies: Intervention, Vol. 20 No. 3 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd intheNordiccountries(Swedenbeingtheexception), with average slopes of around 7.6 per cent. This pattern changes substantially in the second period (1995–2019). While house price cycles did not uniformly become more intense, there is much greater variance across countries, partly due to the greater sample size, with slopes for individual countries ranging from less than 2 per cent per year (Finland) to staggering 23 per cent (Estonia) (see Figure A2 in Appendix 1). Country groups that exhibit moderate house price cycles comprise the European Core and now also East Asia as well the Nordic countries. Southern Europe on average has a lower slope but exhibits strong heterogeneity, with Greece and Spain undergoing substantially more volatile cycles than Portugal and Italy. The Anglo-Saxon countries experience more intense house price cycles in this period with an average slope of 5.7 per cent per year. At the top are the countries at the eastern periphery of the European Union: the Visegrád+ countries and, exhibiting the most extreme cycles, the Baltics. Taken together, the stylized facts presented in this section support the notion of regular boom–bust cycles in real house prices. Countries with more intense booms in house prices also undergo more intense busts. There are substantial differences across countries both in the timing of cycles and in their intensity. Some country groups such as the European core have consistently exhibited relatively stable housing markets, whereas Southern Europe and Anglo-Saxon countries tend to have more volatile ones. Eastern Europe as a whole exhibits the most unstable house prices in recent decades, with slopes up to five times larger than in the European core. 4.2 House price cycles and macroeconomic dynamics The Minskyan approach implies that asset prices are closely correlated with business cycles via their effects on private demand and debt. Figures 4 and 5 plot the slope in house prices against several macroeconomic indicators. Importantly, bivariate correlations in the scatter plots should not be interpreted as causal estimates of the effect of house prices on economic activity. Instead, they represent stylized facts whose consistency with the Minskyan framework is to be assessed. Even more relevant for our comparative analysis is the ability of scatter plots to visualise cross-country patterns. Figure 4 plots the average slope of house price cycles against the slope of real GDP for the period 1995–2019. Figure 5 does the same with the slope of the household debt to GDP ratio. Countries with volatile boom–bust cycles in housing markets, such as the Baltics, the Anglo-Saxon countries Ireland and the UK, and the Visegrád countries Hungary and Poland, also exhibit intense business and debt cycles. 11 These findings corroborate the Minskyan view that cycles in asset markets, in this case housing, are closely linked to cycles in economic activity and private debt (Zezza 2008; Ryoo 2016). Our analysis suggests that some Anglo-Saxon, Baltic and Visegrád countries may be characterised as house-price-driven growth models. 12 11. The result with HHD are not driven by the extreme value for Ireland. When excluding Ireland, the coefficient is still positive and has a p-value of 0.06. 12. Figures A3 and A4 in Appendix 3 further display the correlation between the average slope of house price cycles and the average relative GDP growth contributions of investment as well as construction. The relationship is positive and statistically significant, suggesting that residential investment may be a key driver of the relationship between house prices and economic activity, supporting the argument that finance-led GMs are not just driven by consumption but also by (residential) investment (Zezza 2008; Charpe et al. 2011). House price cycles, housing systems, and growth models 475 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd 25 20 15 10 Slope HPR 5 0 1.5 BEL JPN DEU ITA AUT FIN SWE SVK HUN POL ISL IRL LTU KOR LVA EST GRC ESP CZE USA DNK NZL GBR SVN AUS CAN FRA PRT NOR NLD CHE coeff= 2.02; p-val= 0.01; R2= 0.51; obs= 31 2.5 3.5 4.5 Slope real GDP 5.5 6.5 765432 Notes: Slope: amplitude/duration (% change per year). HPR: log of real house prices; real GDP: log of real gross domestic product; HHD: household debt to GDP ratio (in %). Figure 4 Average slope of house price cycles (HPR) against slope of real GDP, 1995–2019 0 2 4 6 8 10 24 Slope HHD HUN POL GRC ESP IR L GBR SWE CZE NLD NOR USA NZL DNK PRT AUS KOR CAN CHE DEU FRA AUT ITA BEL FIN JPN coeff= 0.77; p-val= 0.00; R2= 0.32; obs= 25 Slope HPR 68 Notes: Slope: amplitude/duration (% change per year). HPR: log of real house prices; HHD: household debt to GDP ratio (in %). No HHD data for Iceland, Slovakia, Estonia, Latvia, Lithuania, Croatia and Slovenia. See Appendix 1 for details on the data. When excluding Ireland, the coefficient is in the bottom panel is still positive and has a p-value of 0.06. Figure 5 Average slope of house price cycles (HPR) against slope of household debt to GDP ratio (HHD), 1995–2019 476 European Journal of Economics and Economic Policies: Intervention, Vol. 20 No. 3 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd 4.3 House price cycles and institutions Based on the CPE literature on housing, we argue that institutions are likely to impact the relevance of these housing cycles across countries. Homeownership rates and financial institutions that encourage mortgage credit creation are the two most common ones highlighted in this approach (Schwartz/Seabrooke 2008; Fuller 2015; Johnston/Kurzer 2020). Figure 6 shows that the homeownership rate exhibits a remarkably tight relationship with the intensity of housing cycles. Countries with high homeownership rates, such as Ireland, Spain and the Eastern European states, tend to exhibit much more volatile cycles than those with low homeownership rates, such as the German-speaking countries. Figure 7 displays the link between cycle intensity and an index of mortgage credit encouragement developed in Fuller (2015). The index takes on higher values for countries with institutions that are more conducive to mortgage credit expansion, considering interest rate restrictions, capital gains taxes on the transfer of households’assets, the typical loan-to-value ratio, mortgage subsidies and the size and type of the secondary debt market. There is a positive and statistically significant link but the relationship is weaker compared to the homeownership rate, and the R-squared of 7 per cent is rather low. Anglo-Saxon countries like the US, Ireland and the UK exhibit high scores along with peripheral countries such as Iceland and Lithuania and underwent intense housing booms in the last decades. By contrast, core European countries such as Austria and Germany exhibit more restrictive credit institutions. This suggests a potential role for 40 CHE DEU AUT KOR NLD DNK SWE CZE USA GBR GRC POL IRL LVA EST ISL HUN LT U HRV SVK ESP SVN NOR PRT ITA BAL CAN FIN AUS FRA 0 5 10 15 20 25 50 60 70 80 90 coeff= 0.16; p -val= 0.00; R2= 0.23; obs= 30 Homeownership rate Slope HPR Notes: Slope: amplitude/duration (% change per year). See Appendix 1 for details on the data. Figure 6 Average slopes of house price cycles (HPR) against average homeownership rate, 1995–2019 House price cycles, housing systems, and growth models 477 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd mortgage-credit encouraging institutions, but further research is needed to assess the robustness of this relationship. In sum, these two institutional factors alone certainly do not fully explain the crosscountry variation in house price cycles, but they do provide elements of an answer to the question why some countries are more prone to develop unstable house-price-driven GMs. First, countries with high homeownership rates and thus large private housing markets are more prone to exhibit volatile housing cycles. Second, financial institutions conducive to mortgage credit expansion might further increase the volatility of cycles. 5 CONCLUSION This paper has argued for assigning a central role to house price dynamics in the analysis of growth models. House price growth is key for debt-driven models, whereas emulative consumption behaviour is at best an amplifying factor. While the existence of booms and busts in housing markets has been noted previously, especially in the literature on finance-led growth models (Ban/Helgadóttir 2022; Reisenbichler/Wiedemann 2022), the key role of house prices as a cyclical driver of both consumption and investment has not been fully appreciated (Wood/Stockhammer 2020). Similarly, the CPE literature 0 0 –5 –4 –3 –2 –1 123 SVN POL HUN SVK GRC LVA EST IRL LTU ISL US A NLD ESP DNK NZL PRT CAN FIN JPN Mortgage encouragement index BEL CHE DEU ITA AUT FRA KOR AUS SWE NOR GBR CZE 5 10 15 20 25 coeff= 0.52; p-val= 0.03; R2= 0.07; obs= 31 Slope HPR Notes: Slope: amplitude/duration (% change per year); Mortgage encouragement index is taken from Fuller (2015) and is based on interest rate restrictions; capital gains on the transfer of households’ assets; the typical loan-to-value ratio; mortgage subsidies; and the size and type of the secondary debt market. See Appendix 1 for details on the data. Figure 7 Average slopes of house price cycles (HPR) against mortgage credit encouragement index, 1995–2019 478 European Journal of Economics and Economic Policies: Intervention, Vol. 20 No. 3 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd on housing has studied the varied rise of mortgage debt across countries but has not systematically analysed house price cycles, nor their implications for the growth models perspective (Schwartz/Seabrooke 2008; Fuller 2015; Johnston/Kurzer 2020). We have argued that behavioural and Minskyan approaches provide the missing link between the growth models approach and the CPE literature on housing (see also Stockhammer/Wolf 2019). Behavioural approaches highlight the role of speculative dynamics in driving endogenous housing cycles (Dieci/Westerhoff 2012, 2016). The Minskyan approach emphasises the relevance of such asset price dynamics for finance-led growth (Zezza 2008; Charpe et al. 2011; Caverzasi/Godin 2015; Ryoo 2016). Like other asset prices, house prices (in a liberalised economy) are prone to cyclical dynamics. Such booms and busts are not merely accidents, but a systemic feature of private housing markets. Basedonthisintegratedtheoreticalframework, we have proposed the notion of a house-price-driven growth model. House price booms may drive growth for temporary episodes of around 8 years on average, yet subsequent busts render this type of growth model intrinsically unstable. Our cross-country analysis showed that countries are not equally susceptible to these cycles. In recent decades, it was mostly Anglo-Saxon, some southern European and above all Eastern European countries that exhibited houseprice-driven growth models. By contrast, core European, Nordic and East Asian countries are characterised by more stable housing markets. Institutions identified in the CPE literature on housing systems matter; arguably those that encourage speculative dynamics in housing markets. We presented preliminary evidence that countries with comparatively large private housing markets exhibit much more volatile house prices. In addition, less regulated mortgage markets may be more likely to undergo speculative bubbles. Our approach provides an understanding of why some countries are more likely to exhibit unstable finance-driven growth models. Combined with country studies from the CPE literature, this perspective brings in a role for politics. Bohle (2014, 2018) analyses the privatisation of the public housing stock in Estonia, Latvia, and Hungary as a key institutional change that created the environment for the spectacular housing bubbles these countries underwent in the years before the GFC. High levels of private homeownership were accomplished via Right-to-Buy policies combined with tax exemptions that incentivised tenants to buy property. Similar developments took place one or two decades before in Iceland and Ireland (Bohle 2018) as well as the US and UK (Ryan-Collins 2021; Reisenbichler/Wiedemann 2022). By contrast, most core European and some Nordic countries exhibit low homeownership rates and less volatile housing cycles, partly due to a greater role for social housing (Kholodilin et al.2022). The fact that institutions that favour housing cycles are the result of policy also implies that action can be taken to curb cycles. Ryan-Collins (2021) suggests a number of measures, such as imposition of a land value tax that disincentivises speculative purchases of land, tenant protection laws that make the rental market more attractive and stricter macroprudential regulation of mortgage credit creation. In addition, an active policy of social housing or the promotion of non-profit housing associations can help transform housing from a speculative asset that generates macroeconomic instability into a good that provides shelter and stability. The exact role of housing institutions will require further research. We considered the homeownership rate and mortgage credit encouraging institutions as imperfect measures that only partially capture the prevalence of speculative demand. Our empirical results showed that they only explain a portion of house price cycle intensity. Future research ought to examine institutions that more directly incentivise speculative behaviour, such as taxes and regulations on the sale of houses and the realisation of capital gains. House price cycles, housing systems, and growth models 479 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd What do our arguments mean for future research on finance-led growth models? Are all finance-led growth models house price-driven? The growth models approach provides institutionally and historically specific analyses of growth models. In this spirit, our argument is also historically contingent: in many advanced economies, house prices have been the key variable to understand finance-led growth over the past decades. In theory, one could equally imagine a share price-driven growth process (Boyer 2000). However, as housing wealth is now widely distributed (much more widely than other forms of financial wealth) and real estate is a commonly used collateral, housing has become macroeconomically important (relative to corporate debt and equity). Is the house-price-driven growth model only about cycles or does it also allow for sustained increases in house prices? Speculation may also give rise to a secular increase in the level of real house prices: by treating real estate as a financial asset class, new investors enter the housing market which increases housing demand and thus the price for housing. In addition, government intervention in the downturn of the housing cycle (for example, by subsiding mortgage-financed housing demand) may reduce the amplitude of busts compared to booms. Whether a secular trend in house prices can constitute a stable long-run growth driver is another question that will require further research. REFERENCES Baccaro, L., Pontusson, J. (2016): Rethinking comparative political economy: the growth model perspective, in: Politics & Society, 44 (2), 175–207. Ban, C., Helgadóttir, O. (2022): Financialization and growth models, in: Blyth, M., Baccaro, L. and Pontusson, J. (eds), Diminishing Returns. The New Politics of Growth and Stagnation, New York: Oxford University Press, 351–374. Behringer, J., van Treeck, T. (2019): Income distribution and growth models: a sectoral balances approach, in: Politics & Society, 47(3), 303–332. Belabed, C.A., Theobald, T., van Treeck, T. (2018): Income distribution and current account imbalances, in: Cambridge Journal of Economics, 42(1), 47–94. Bhaduri, A., Marglin, S. (1990): Unemployment and the real wage: the economic basis for contesting political ideologies, in: Cambridge Journal of Economics, 14(4), 375–393. Blyth, M., Pontusson, J., Baccaro, L. (2022): Diminishing Returns. The New Politics of Growth and Stagnation, New York: Oxford University Press. Bohle, D. (2014): Post-socialist housing meets transnational finance: foreign banks, mortgage lending, and the privatization of welfare in Hungary and Estonia, in: Review of International Political Economy, 21(4), 913–948. Bohle, D. (2018): Mortgaging Europe’s periphery, in: Studies in Comparative International Development, 53(2), 196–217. Borio, C. (2014): The financial cycle and macroeconomics: what have we learnt?, in: Journal of Banking & Finance, 45, 182–198. Boyer, R. (2000): Is a finance-led growth regime a viable alternative to Fordism? A preliminary analysis, in: Economy and Society, 29(1), 111–145. Caverzasi, E., Godin, A. (2015): Financialisation and the sub-prime crisis: a stock-flow consistent model, in: European Journal of Economics and Economic Policies: Intervention, 12(1), 73–92. Charpe, M., Chiarella, C., Flaschel, P., Semmler, W. (2011): Financial Assets, Debt, and Liquidity Crises: A Keynesian Approach, Cambridge, UK; New York: Cambridge University Press. Chiarella, C., He, X.-Z., Zwinkels, R.C.J. (2014): Heterogeneous expectations in asset pricing: empirical evidence from the S&P500, in: Journal of Economic Behavior & Organization,105, 1–16. Claessens, S., Kose, M.A., Terrones, M.E. (2012): How do business and financial cycles interact?, in: Journal of International Economics, 87(1), 178–190. 480 European Journal of Economics and Economic Policies: Intervention, Vol. 20 No. 3 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd Dieci, R., He, X.-Z. (2018): Heterogeneous agent models in finance, in: Hommes, C., LeBaron, B. (eds), Handbook of Computational Economics, Elsevier, 258–320, doi: https://doi.org/10.1016/ bs.hescom.2018.03.002. Dieci, R., Westerhoff, F. (2012): A simple model of a speculative housing market, in: Journal of Evolutionary Economics, 22(2), 303–329. Dieci, R., Westerhoff, F. (2016): Heterogeneous expectations, boom-bust housing cycles, and supply conditions: a nonlinear economic dynamics approach, in: Journal of Economic Dynamics and Control, 71, 21–44. Drehmann, M., Borio, C.E., Tsatsaronis, K. (2012): Characterising the financial cycle: don’t lose sight of the medium term!, BIS Working Papers, No 380. Duca, J.V., Muellbauer, J., Murphy, A. (2021): What drives house price cycles? International experience and policy issues, in: Journal of Economic Literature, 59(3), 773–864. ECB (2018): Real and Financial Cycles in EU Countries: Stylised Facts and Modelling Implications, LU: Publications Office. Fiebiger, B., Lavoie, M. (2019): Trend and business cycles with external markets: non‐capacity generating semi‐autonomous expenditures and effective demand, in: Metroeconomica, 70(2), 247–262. Frank, R.H. (2014): Expenditure cascades, in: Review of Behavioral Economics, 1(1–2), 55–73. Fuller, G.W. (2015): Who’s borrowing? Credit encouragement vs. credit mitigation in national financial systems, in: Politics & Society, 43(2), 241–268. Gusella, F., Stockhammer, E. (2021): Testing fundamentalist–momentum trader financial cycles: an empirical analysis via the Kalman filter, in: Metroeconomica, 72(4), 758–797. Hall, P.A., Soskice, D. (eds) (2001): Varieties of Capitalism. The Institutional Foundations of Comparative Advantage, New York: Oxford University Press. Harding, D., Pagan, A. (2002): Dissecting the cycle: a methodological investigation, in: Journal of Monetary Economics, 49(2), 365–381. Harloe, M. (1995): The Peoples Home: Social Rented Housing in Europe and America, John Wiley & Sons. Hein, E., Meloni, W.P., Tridico, P. (2021): Welfare models and demand-led growth regimes before and after the financial and economic crisis, in: Review of International Political Economy, 28(5), 1196–1223. Hein, E., Mundt, M. (2013): Financialization, the financial and economic crisis, and the requirements and potentials for wage-led recovery, in: Lavoie, M., Stockhammer, E. (eds), Wage-led Growth. An Equitable Strategy for Economic Recovery, Basingstoke: Palgrave Macmillan, 153–186. Hommes, C., in ’t Veld, D. (2017): Booms, busts and behavioural heterogeneity in stock prices, in: Journal of Economic Dynamics and Control, 80, 101–124. Hommes, C.H. (2006): Heterogeneous agent models in economics and finance, Chapter 23 in: Tesfatsion, L., Judd, K.L. (eds), Handbook of Computational Economics, Elsevier, 1110–1146, doi: https://doi.org/10.1016/S1574-0021(05)02023-X. Iacoviello, M., Neri, S., (2010): Housing market spillovers: evidence from an estimated DSGE model, in: American Economic Journal: Macroeconomics, 2(2), 125–164. Johnston, A., Kurzer, P. (2020): Bricks in the wall: the politics of housing in Europe, in: West European Politics, 43(2), 275–296. de Jong, E., Verschoor, W.F.C., Zwinkels, R.C.J. (2010): Heterogeneity of agents and exchange rate dynamics: evidence from the EMS, in: Journal of International Money and Finance,29(8),1652–1669. Kapeller, J., Schütz, B. (2015): Conspicuous consumption, inequality and debt: the nature of consumption-driven profit-led regimes: conspicuous consumption, inequality and debt, in: Metroeconomica, 66(1), 51–70. Kemeny, J. (1995): From Public Housing to the Social Market: Rental Policy Strategies in Comparative Perspective, London: Routledge. Kholodilin, K., Kohl, S., Müller, F. (2022): The Rise and Fall of Social Housing? Housing Decommodification in Long-Run Perspective, Max Planck Institute for the Study of Societies, MPIfG Discussion Paper 22/3, No 22/3. Kim, Y.K. (2016): Macroeconomic effects of household debt: an empirical analysis, in: Review of Keynesian Economics, 4(2), 127–150. House price cycles, housing systems, and growth models 481 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd Kohler, K., Stockhammer, E. (2022): Growing differently? Financial cycles, austerity, and competitiveness in growth models since the Global Financial Crisis, in: Review of International Political Economy, 29(4), 1314–1341. Labat-Moles, H., Summa, R. (2023): A supermultiplier demand-led growth accounting analysis applied to the Spanish economy (1998-2019), in: European Journal of Economics and Economic Policies: Intervention, advance access, doi: https://doi.org/10.4337/ejeep.2023.0115. Lavoie, M., Stockhammer, E. (eds) (2013a): Wage-Led Growth. An Equitable Strategy for Economic Recovery, Basingstoke, UK: Palgrave Macmillan. Lavoie, M., Stockhammer, E. (2013b): Wage-led growth: concept, theories and policies, in: Lavoie, M. Stockhammer, E. (eds), Wage-led Growth. An Equitable Strategy for Economic Recovery, Basingstoke, UK: Palgrave Macmillan, 13–39. Lof, M. (2012): Heterogeneity in stock prices: a STAR model with multivariate transition function, in: Journal of Economic Dynamics and Control, 36(12), 1845–1854. Martin, C., Schmitt, N., Westerhoff, F. (2021): Heterogeneous expectations, housing bubbles and tax policy, in: Journal of Economic Behavior & Organization, 183, 555–573. Martin, C., Schmitt, N., Westerhoff, F. (2022): Housing markets, expectation formation and interest rates, in: Macroeconomic Dynamics, 26(2), 491–532. Morlin, G.S., Passos, N., Pariboni, R. (2022): Growth theory and the growth model perspective: insights from the supermultiplier, in: Review of Political Economy,1–26. Nikolaidi, M., Stockhammer, E. (2017): Minsky models: a structured survey, in: Journal of Economic Surveys, 31(5), 1304–1331. O’Donovan, N. (2021): Demand, dysfunction and distribution: the UK growth model from neoliberalism to the knowledge economy, in: The British Journal of Politics and International Relations, 136914812110580. Palley, T.I. (1994): Debt, aggregate demand, and the business cycle: an analysis in the spirit of Kaldor and Minsky, in: Journal of Post Keynesian Economics, 16(3), 371–390. Prante, F., Hein, E., Bramucci, A. (2022): Varieties and interdependencies of demand and growth regimes in finance-dominated capitalism: a Post-Keynesian two-country stock–flow consistent simulation approach, in: Review of Keynesian Economics, 10(2), 264–290. Reisenbichler, A., Wiedemann, A. (2022): Credit-driven and consumption-led growth models in the United States and United Kingdom, in: Blyth, M., Pontusson, J., Baccaro, L. (eds), Diminishing Returns. The New Politics of Growth and Stagnation, New York: Oxford University Press, 213–237. Rünstler, G., Vlekke, M. (2018): Business, housing, and credit cycles, in: Journal of Applied Econometrics, 33(2), 212–226. Ryan-Collins, J. (2021): Breaking the housing–finance cycle: macroeconomic policy reforms for more affordable homes, in: Environment and Planning A: Economy and Space, 53(3), 480–502. Ryoo, S., (2016): Household debt and housing bubbles: a Minskian approach to boom-bust cycles, in: Journal of Evolutionary Economics, 26(5), 971–1006. Schmitt, N., Westerhoff, F. (2022): Speculative housing markets and rent control: insights from nonlinear economic dynamics, in: Journal of Economic Interaction and Coordination, 17(1), 141–163. Schüler, Y.S., Hiebert, P.P., Peltonen, T.A. (2020): Financial cycles: characterisation and real-time measurement, in: Journal of International Money and Finance, 100, 102082. Schwartz, H. (2008): Housing, global finance, and American hegemony: building conservative politics one brick at a time, in: Comparative European Politics, 6(3), 262–284. Schwartz, H.M., Seabrooke, L. (2008): Varieties of residential capitalism in the international political economy: old welfare states and the new politics of housing, in: Comparative European Politics, 6(3), 237–261. Stockhammer, E., Novas Otero, A. (2022): A tale of housing cycles and fiscal policy, not competitiveness. Growth drivers in Southern Europe, New Political Economy, advance access, doi: https:// doi.org/10.1080/13563467.2022.2141702. Stockhammer, E., Wildauer, R (2016): Debt-driven growth? Wealth, distribution and demand in OECD countries, in: Cambridge Journal of Economics, 40(6), 1609–1634. Stockhammer, E., Wolf, C. (2019): Building blocks for the macroeconomics and political economy of housing, in: The Japanese Political Economy, 45(1–2), 43–67. 482 European Journal of Economics and Economic Policies: Intervention, Vol. 20 No. 3 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd Strohsal, T., Proaño, C.R., Wolters, J. (2019): Characterizing the financial cycle: evidence from a frequency domain analysis, in: Journal of Banking & Finance, 106, 568–591. Teixeira, L., Petrini, G. (2023): Long-run effective demand and residential investment: a Sraffian supermultiplier based analysis, in: Review of Keynesian Economics, 11(1), 72–99. Van Gunten, T., Navot, E. (2018): Varieties of indebtedness: financialization and mortgage market institutions in Europe, in: Social Science Research, 70, 90–106. Westerhoff, F.H., Reitz, S. (2003): Nonlinearities and cyclical behavior: the role of chartists and fundamentalists, in: Studies in Nonlinear Dynamics & Econometrics, 7(4), 1–15. Wood, J., Stockhammer, E. (2020): House prices, private debt and the macroeconomics of comparative political economy, Post Keynesian Economics Society, PKES Working Paper, No 2005. Zezza, G. (2008): U.S. growth, the housing market, and the distribution of income, in: Journal of Post Keynesian Economics, 30(3), 375–401. House price cycles, housing systems, and growth models 483 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd APPENDIX 1: DATASET Variable definition Variable abbreviation Source(s) Sample start Notes Real house price index (log) HPR Constructed from four datasources: (1) BIS (2) OECD (3) ECB (4) Palacin & Shelburne (2005) 1970, except for: AUT (2000), GRC (1997), ISL (2000), PRT (1988), ESP (1971), KOR (1975), CZE (2000), SVK (2005), EST (2002), Lat (2006), HUN (1999), LTU (1999), HRV (2002), SVN (2005), POL (2000) (1) BIS data for all countries and periods where available; (2) For data where thereisOECDdatabutnoBISdata, extrapolate the BIS data back using the growth rates of the OECD series; (3) For Latvia and Czech Republic for years from 2000–2008 & 2000–2006 respectively, the data has been interpolated using the growth in the real house price level series in ECB data. (4) For Hungary, Estonia and Poland the data has been interpolated using annual Palacin & Shelburne (2005) data on residential property prices for the years 1999q1–2007q1, 2002q1–2005q1 and 2000q1–2005q1 respectively. Real gross domestic product (log) GDP OECD Quarterly National Accounts 1995, except for: ITA (1996), NLD (1996), CZE (1996) (continues overleaf ) East Asia European Core Nordic Southern Europe VisegrádþAnglo-Saxon Baltics Individual countries Japan, South Korea Austria, Belgium, France, Germany, the Netherlands, Switzerland Denmark, Finland, Norway, Sweden Greece, Italy, Portugal, Spain Czech Republic, Hungary, Poland, Slovakia, Slovenia Australia, Canada, Ireland, New Zealand, UK, USA Estonia, Latvia, Lithuania Croatia, Iceland 484 European Journal of Economics and Economic Policies: Intervention, Vol. 20 No. 3 © 2023 The Author Journal compilation © 2023 Edward Elgar Publishing Ltd