Empirical stock-flow consistent models: evolution, current state and prospects

Pierros, Christos Article Empirical stock-flow consistent models: evolution, current state and prospects European Journal of Economics and Economic Policies: Intervention (EJEEP) Provided in Cooperation with: Edward Elgar Publishing Suggested Citation: Pierros, Christos (2025) : Empirical stock-flow consistent models: evolution, current state and prospects, European Journal of Economics and Economic Policies: Intervention (EJEEP), ISSN 2052-7772, Edward Elgar Publishing, Cheltenham, Vol. 22, Iss. 3, pp. 301-316, https://doi.org/10.4337/ejeep.2024.0136 This Version is available at: https://hdl.handle.net/10419/333441 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/ European Journal of Economics and Economic Policies: Intervention, Vol. 22 No. 3, 2025, pp. 301–316 First published online: September 2024; doi: 10.4337/ejeep.2024.0136 Journal compilation © 2025 Edward Elgar Publishing Ltd © 2025 The Author Research Article Received 14 March 2024, accepted 2 May 2024 This is an open access work Growth models, growth strategies, and power blocs in Turkey and Egypt in the twenty-first century Ali Rıza Güngen Social Sciences, Columbia College, Canada Ümit Akçay Institute for International Political Economy, Berlin School of Economics and Law, Berlin, Germany [email protected] Analysis of the growth patterns in the Global South in the twenty-first century suggests there is room for authoritarian states to search for new growth models. Authoritarian states, such as Turkey and Egypt, benefited from global financial circumstances in the early 2000s and experienced shifts in growth strategies in the 2010s, suppressing political space further. Our main research question, thus, is focusing on what the main domestic political economy causes of these growth strategy and model changes are. To explain the changes in growth strategies and models amid the strength of reinforced authoritarian regimes in these two countries, we employ a hybrid research strategy, tying growth model changes to conflicts within the power bloc. We argue that in the mid-to-late 2010s, peripheral goods producers gained the upper hand in Turkey, while a military takeover in Egypt was followed by the promotion of exports and new investments. We also contend that power bloc reconfigurations in the last decade and the rise of new growth strategies both in Turkey and in Egypt aimed to change previous domestic demand-led demand and growth models. Keywords: comparative political economy, growth models, growth strategies, Turkey, Egypt JEL codes: B52, E65, E66, F43, O43, P52 1 INTRODUCTION Authoritarian states in Turkey and Egypt rejuvenated themselves in the 2010s. This was a development contrary to the widespread expectation that when faced with deep economic crises and brewing social discontent, authoritarian regimes are less likely to maintain their power. This study elaborates on the growth models of Turkey and Egypt in the twentyfirst century. Despite significant differences regarding export capacity and macroeconomic indicators, political economic developments converge in various aspects in these two countries. Moreover, the authoritarian regimes in both Turkey and Egypt maintained their power while increasingly suppressing the political space in the 2010s (Tuğal 2016). We describe authoritarianism as a set of practices that isolates key policy-making processes from democratic oversight and excludes large groups such as working classes, ethnic minorities or subaltern groups from institutional politics (Salgado 2022). From a critical political economy perspective, authoritarian practices cannot be conceived as clearly cut from Research Article This isan open access work Received 8 December 2022, accepted 7 November 2023 European Journal of Economics and Economic Policies: Intervention, Vol. 21 No. 1, 2024, pp. 151–171 First published online: April 2024; doi: 10.4337/ejeep.2024.01.09 Journal compilation © 2024 Edward Elgar Publishing Ltd © 2024 The Author Empirical stock–flow consistent models: evolution, current state and prospects Christos Pierros* Researcher, Labour Institute of the General Confederation of Greek Workers (INE GSEE), Athens, Greece The paper explores the evolution of empirical stock–flow consistent (SFC) models, emphasising their structure, scope and number of financial and physical assets. Three types of models can be identified. The New Cambridge type is characterised by the aggregation of households, firms and banks into one aggregate private sector. The Godley–Lavoie type, termed after the impact of their collective work, treats the main institutional sectors separately. The third type, despite being largely heterogeneous, is marked by higher complexity. The paper argues that the structure of the models should vary according to the research question at hand, as higher complexity is ensued by augmented discrepancies between insample projections and actual data, especially in the financial domain. Despite this trade-off, several aspects of the models need to be improved, even at the expense of having more complicated structures. The paper provides some indications towards the direction that these improvements ought to take. Keywords: empirical stock–flow consistent, stock–flow consistent modelling, empirical macrostructural models JEL codes: E10, E12, E17 1 INTRODUCTION Despite that the first empirical stock–flow consistent (SFC) model was developed in the late 1970s, SFC models gradually rose to prominence only very recently for reasons already discussed in the literature (Caverzasi/Godin 2015; Nikiforos/Zezza 2018; Zezza/Zezza 2019). The structure and the behavioural framework of these models have evolved from simple to much more complex according to the research questions they are trying to address, the dynamics they are trying to capture and the specificities of the economy under consideration. Contrary to other types of models, in which the actual values are plugged into generic structures,1 the task of applying an SFC model in an actual economy is much more demanding. Since their initial development in the late 1970s, three classes of SFC models can be identified. The New Cambridge (NC) approach follows firmly the methodology developed primarily in the Cambridge Economic Policy Group (CEPG) in the late 1970s and 1 . In example, according to Cherrier et al. (2023), the systematisation and formalisation of Dynamic Stochastic General Equilibrium (DSGE) models is such that their tractability and portability is ensured nowadays with minimum effort. * Email: [email protected]. A shorter and quite different version of this paper appears in Byrialsen, M., Valdecantos, S., and Raza, H. (eds.), Empirical Macroeconomic Models: An Applied Approach, London, UK: Routledge. European Journal of Economics and Economic Policies: Intervention, Vol. 22 No. 3302 Journal compilation © 2025 Edward Elgar Publishing Ltd © 2025 The Author early 1980s. The second type is largely influenced by the seminal work of Godley/Lavoie (2006) (henceforth GL), and it has a larger structure as compared to the first type and is able to assess rigorously a larger set of policies. Models that belong to the third type have a highly complex structure as they aim to capture more complicated transmission channels and mechanisms that underline the function of actual economies. For this reason, they are termed High Complexity (HC) models. The paper provides a systematic review of the empirical SFC models, based on their structure. The models taken under consideration capture the majority of an economy’s features, and their behavioural framework is either estimated econometrically or calibrated according to actual data. The discussion takes into account 28 empirical SFC models, the vast majority of which are estimated econometrically, while two are of reduced form. It is important to note that there is a growing number of models that are estimated–calibrated for multi-country regions instead of single countries (for example Jacques et al. 2023). In order to make the review more comprehensive, the focus is laid on models that are applied to single countries.2 The paper argues that, on the one hand, the risk of poor performance is greater, with the more complex being its structure. Simply put, there is a trade-off between complexity and the goodness of fit. On the other hand, some specific aspects are overly simplified and need to be further developed. The actual choice of the level of complexity should depend on the nature of the research question, data availability and the features of the economy at hand. The remainder of the paper is organised as follows. Section 2 discusses some particular features of empirical SFC models, while Section 3 is dedicated to the description of each of the three types of SFC models. Section 4 highlights some issues that ought to be addressed in the future, and Section 5 concludes. 2 MAIN FEATURES The term stock–flow consistency has steered a debate on which models ultimately fall within this category. Nikiforos/ Zezza (2018) and Zezza/Zezza (2019) indicate a set of stock–flow consistency principles that characterise the SFC models, which, however, are not meant to define them accurately. Dos Santos (2006) and Ehnts (2019) argue that a major characteristic of the SFC models is that the behavioural framework is not only rooted on but also constrained by a solid accounting structure, as the latter reduces the degrees of freedom in the construction of the behavioural framework (Taylor 2004 cited in Lavoie 2022: 294). Going a step further would require a specific focus on what Zezza/ Zezza (2019) mention as the stock-to-flow feedback principle. SFC models can be identified through an evolutionary view on the nature of the balance sheet, which, in essence, is the cumulative outcome of decisions taken in the past. It registers in quantitative terms the historical route or the evolutionary path of the individual, sector or the economy under examination and constrains the potential future outcomes. This comes about through the integration of balance sheet items in the behavioural equations. The dynamic multiplier is a typical example of this, as the introduction of the lagged net wealth effect on consumption3 reflects in a consistent manner the lasting impact of the accumulated past decisions on the current and future outcomes. 2 . An exception is made for the global model of Cripps/Izurieta (2014) as it is estimated econometrically. 3 . See Godley/Lavoie (2006: ch.3). Empirical stock–flow consistent models: evolution, current state and prospects 303 Journal compilation © 2025 Edward Elgar Publishing Ltd© 2025 The Author This might not be a unique characteristic of the SFC models, but it helps in distinguishing them from the vast majority of other types of models, such as the DSGE models in which the balance sheet is stripped off from its evolutionary aspects and merely adjusts gradually to some predetermined equilibrium values.4 On the contrary, in SFC models the balance sheet constitutes one of the main reasons for the presence of path dependency. The trajectory that is governed by the accumulation of assets and liabilities is, practically, a sequence of short-run outcomes that dictate future flows (e Silva/Santos 2011; Zezza/Zezza 2019). The rest defining features of the SFC models, whether theoretical or empirical, are rooted mainly in the Post-Keynesian tradition. 2.1 Structure and scope Canelli et al. (2021) divide the empirical SFC models into two groups, the theory-driven and the data-driven. In this paper, I take a different route and divide models according to the complexity of their structure, the number of institutional sectors and the amount of financial and physical assets. This further allows dividing empirical SFC models into three types, namely, the NC type, which aggregates households, firms and banks in one sector; the Godley and Lavoie type, which treats separately the standard institutional sectors; and the HC type, which is characterised by great heterogeneity. It is logically coherent to argue that more complex structures are usually theory-driven, though this need not always be the case. For instance, the model for the Netherlands (Meijers/Muysken 2022) has a complex financial structure as the Dutch economy is heavily financialised, while the model for Colombia (Godin et al. 2023) has a large set of financial assets due to the importance of borrowing in foreign currency. In both cases, the complexity is imposed by the actual circumstances and not by the theory. Nonetheless, higher complexity bears a cost in terms of the model fitting actual data. Not many cases reviewed in this paper provide in-sample projections. From the available data, it occurs that, in their vast majority, models fit quite well data of the real side of the economy but miss the financial side. This could be attributed to a number of reasons: (a)the use of different data sets that are not entirely consistent with national accounts data; (b) difficulties in modelling capital gains, which are of higher frequency, more stochastic and volatile; (c) higher sensitivity of more complex models to parameter changes; (d) closures and assumptions on residuals that are usually associated with the behaviour of the financial sector; and (e) behavioural frameworks, such as Tobinesque rules and accommodative banks, which are not suitable for the economy at hand. Table 1 presents some main characteristics of the three types of empirical SFC models. The majority of them has a highly complex structure, and from the median year of publication, it seems that HC models are more recent and preferable. The average number of sectors and financial and physical capital assets ensure the validity of the division applied in this paper. On average, the number of assets is doubled between the NC and GL models. The same is the difference between the GL and HC models. Regarding the sectors, on average NC models utilise a minimum set, the GL models make use of the five standard 4 . For instance, in the former DSGE model of the European Commission (Ratto et al. 2009), any discussion about money holdings is ruled out by assumption as portfolio choices adjust to the interest rate rule. In both versions of the DSGE model of the ECB (Gomes et al. 2012, Bokan etal. 2018), the balance sheet of non-liquidity constrained households is also consisted of statecontingent securities, which absorb the discrepancies from agents’ optimal decisions on other asset holdings and have no further impact on their spending behaviour, or their portfolio choice. The stock-to-flow feedback effect is muted. European Journal of Economics and Economic Policies: Intervention, Vol. 22 No. 3304 Journal compilation © 2025 Edward Elgar Publishing Ltd © 2025 The Author institutional sectors, while models that are more complicated integrate slightly more sectors, either in the production or in the financial domain. Table 2 presents the scope(s) of each type of model as reported by the respective author(s). The scope of the model usually defines its structure. For instance, mediumterm projections are carried out through NC models, while financialisation and green transition dynamics necessitate the examination of more complex dynamics and transmission mechanisms. Therefore, HC models are more suitable for this task. The GL models have a comparatively more generic structure and fit data well; thus, these models are multi-purposed. 2.2 Empirical form and estimation In the majority of the empirical SFC models, the behavioural equations are estimated econometrically. The most typical empirical form is the error correction, as it is fully compatible with the underlying mechanism of SFC models, according to which stock–flow norms do not deviate for a long time away from their long-term values and the steadystate growth.5 For instance, in the Levy model for Greece (Papadimitriou et al. 2013) the private expenditure, which is a typical feature of the NC approach, is estimated econometrically and depends on its long-term relation with the disposable income of the private sector, its lagged accumulated financial wealth, household credit and capital gains, as well as on the respective error correction terms. The formulation is very similar in estimated– calibrated HC models. For example, in the continuous-time model for Colombia (Godin et al. 2023), households first decide on a target level of consumption, based on their disposable income, wealth and new credit. The target is equivalent to the long-term relation in the NC model. Then households adjust their current consumption to the targeted one, which is also equivalent to the error correction term in the NC model. 5 . For a detailed discussion, see Zezza (2013). Table 1 Empirical stock–flow consistent models by type and main characteristics Type Number of models Median year of publication Average financial assets Average capital assets Average sectors NC 7 1999 4 1 3 GL 8 2019 6 2 5 HC 13 2022 11 4 6 Note: NC, New Cambridge; GL, Godley and Lavoie; HC, High Complexity. Table 2 Empirical stock–flow consistent models by type and scope Type IM MTP FMF FID FIN GT NC 5 3 1 1 … … GL 3 1 7 … … … HC 3 … 6 1 2 5 Note: IM, external or sector imbalances; MTP, medium-term projections; FMF, fiscal, monetary or financial operations; FID, functional income distribution; FIN, financialisation; GT, green transition. Empirical stock–flow consistent models: evolution, current state and prospects 305 Journal compilation © 2025 Edward Elgar Publishing Ltd© 2025 The Author In this constellation, a negative shock is ensued by a gradual adjustment towards the target or the long-term value, with the speed of adjustment depending also on the magnitude of the balance sheet effect, for example, the presence of path-dependency. This does not imply that the long-term coefficients act necessarily as attractors in an equilibrium sense, but rather as long-term tendencies, which, however, could be subject to structural changes at any point in time. Figure 1 presents two examples of opposite cases, namely the wealth-to-income ratios of Germany and Greece for the period between 1995 and 2022. In the case of Germany, the ratio is fairly stable around its trend, implying that the error correction term could justify any gradual adjustment to the long-term structure. In the case of Greece, the presence of a structural break is clearly evident in 2007. The trend before the break has been negative, while after 2007 the stock–flow norm fluctuates around zero, not in a homoscedastic manner. SFC models do not impose equilibrium conditions, and for this reason, they are able to trace imbalances and unsustainable conditions. In order to do so, there is a requirement for a constant re-evaluation of stability conditions. The researcher ought to carefully balance between the stability of the model, for example, respond to Lucas’ critique through cointegrating relations, and the stability of the economy under examination, for example, avoid ergodicity that might be imposed by the cointegrating relations, which in the real world are always subject to change. A final note in the estimation of empirical models relates to whether the introduction of shocks, in order to carry out policy assessments, ought to take place in isolation of other dynamics. The common practice in all models discussed below is to introduce shocks after the estimation of a baseline scenario. In this case, the possibility that unobserved dynamics affect the policy outcome is particularly high. In order to address this issue, one would first have to estimate a potential steady state and then introduce the shock. Nonetheless, for the estimation of the steady state the researcher would either have to arbitrarily impose Note: Net financial wealth excludes the assets and liabilities of the central bank. Source: Own elaboration of Eurostat data. Figure 1 Net financial wealth to disposable income ratio of the private sector in Germany and Greece (1995–2022) European Journal of Economics and Economic Policies: Intervention, Vol. 22 No. 3306 Journal compilation © 2025 Edward Elgar Publishing Ltd © 2025 The Author long-term conditions to the current state of the economy or impose other ad hoc rules. In addition, a shock in the current state of the economy might be biased due to the presence of unobserved dynamics, but the policy outcomes are fairly consistent with the conditions that prevail at the specific point in time, in which the shock has been introduced. Simply put, a shock at time t yields a biased outcome, due to the presence of unobserved dynamics. However, assuming the implementation of the policy at time t in a real economy its overall performance would depend also on the unobserved dynamics. The actual choice on whether to assume a steady state or not depends on whether the researcher aims to highlight the properties of the economy or its performance at a specific time period. 3 EMPIRICAL SFC MODELS The presentation of the three types of models pays particular attention on how the structure of each type of model is closely connected with its scope, as stated by the authors. Due to space limitation issues, the discussion is restricted to the description of groups of models and not on a model-to-model basis, although, in some instances, it is less convenient to do so. Additionally, the section examines the nexus between complexity and the goodness of fit, conditionally to the availability of in-sample projections that allow such a discussion. A few remarks are in place. First, the performance of each model in the in-sample projections largely depends on whether the simulation is dynamic or static. In the former case, the model makes use of the first-period data and provides projections based on the estimated coefficients and the exogenous values. In the static simulation, the model makes use of the previous period’s data. In the dynamic simulation, residuals are accumulating and the overall performance is expected to be poorer as compared to the static simulation. Unfortunately, in the vast majority of the papers, the authors do not provide any relevant information. Second, ad hoc rules are employed in order to assess the goodness of fit. High performance occurs when projections follow the same trend with actual data and their discrepancy is trivial. Medium performance is characterised by either a discrepancy between the projected and actual data that does not exceed 2 per cent of GDP or the projections’ trends follow a different path from the actuals. The model performs poorly if both trends are different from those of actual data and the discrepancy is large. Finally, the following section briefly presents some notable closures applied in each model. 3.1 Mark I: New Cambridge The main feature of the NC approach is the aggregation of households, non-financial and financial corporations in one sector. In most cases, this is reflected in an aggregate private expenditure function, which treats consumption and investment as one single variable. The latter typically depends on the aggregate private sector disposable income, net financial wealth and other financial variables.6 Given this formulation, one could obtain stable medium-term stock–flow relations, which translate to a stable net financial wealth-disposable income ratio.7 This approach was initiated in the CEPG during the 1970s and has 6 . For further details on the explanatory variables, see for example Cripps et al. (1976). 7 . For a discussion on the necessity of this formation on behavioural grounds, see Fetherston/ Godley (1978). Empirical stock–flow consistent models: evolution, current state and prospects 307 Journal compilation © 2025 Edward Elgar Publishing Ltd© 2025 The Author been criticized in both the Keynesian (Bispham 1975) and the Monetarist and Neoclassical (Blinder 1978) domains.8 Table 3 summarises the empirical SFC models of the NC type. In the majority of the models, the self-reported scope is to trace imbalances, based on the evolution of the sector financial balances. In addition, most models are used for medium-term projections. Concerning the latter case, the light structure in conjunction with the comparatively small number of behavioural equations implies that the degrees of freedom are not restricted significantly and that the model is expected to be less sensitive to parameter changes. Models with such a standardised form include the model for the UK (Cripps/Godley 1976), the model for the US (Godley 1999) and the two models for Greece (Papadimitriou et al. 2013; Pierros 2021), even though the one model for Greece (Pierros 2021) transforms the private expenditure function in order to account for the impact of functional income distribution on economic performance. The rest models are variations of the NC approach as they have separate consumption and investment functions, but in the balance sheet of the model, the private domestic entities are treated as one sector. The slightly different structure of these variations is also evident in their scope, which is either more generic or more suitable to examine specific instability issues, such as the volatility of the exchange rate in the case of Argentina (Valdecantos 2022). Not many models provide in-sample projections. Both the models for Greece and the one for Argentina have a very good fit in terms of national account data. Regarding financial data, the one model for Greece (Pierros 2021) does not provide in-sample projections, while the other (Papadimitriou et al. 2013) has a very good fit when estimating financial variables at historic costs but not a great one concerning market costs. According to the authors, this is in large due to the combined use of different data sets. The model for Argentina provides in-sample projections for savings as a proxy for the replication of 8 . For a detailed exposition of the debate regarding the private expenditure function, see Maloney (2012) and Mata (2012). Table 3 Empirical NC models Authors Country Scope Fit Notable closure Real Financial Cripps/Godley (1976) UK IM/MTP N/A N/A Private sector financial wealth Davis (1987a,b) UK FMF N/A N/A Private sector financial wealth Godley/Zezza (1992) Denmark FMF Good N/A Private sector wealth Godley (1999) USA IM/MTP N/A N/A Private sector financial wealth Papadimitriou et al. (2013) Greece IM/MTP N/A Adequate/ Good Private sector net external debt Pierros (2021) Greece FID Good N/A Private sector net external debt Valdecantos (2022) Argentina IM Good N/A Private sector money holdings European Journal of Economics and Economic Policies: Intervention, Vol. 22 No. 3308 Journal compilation © 2025 Edward Elgar Publishing Ltd © 2025 The Author financial data. The fit is excellent but it does not replicate actual financial variables, in a strict sense. Finally, in almost all cases the main closure of the model is related to the wealth of the private sector. This follows from the very simple balance sheet that the majority of the NC models utilise. In this respect, the private sector wealth absorbs all the discrepancies. There are minor variations in this set-up. Wealth can be either financial or total as in the case of Godley/Zezza (1992) and further, it could be total net financial wealth or vis-à-vis the external sector. A notable extension is the case of the model for Argentina, which has a richer structure in terms of the balance sheet. In this particular case, money holdings clear the portfolio allocation of the private sector. 3.2 Mark II: Godley and Lavoie The publication of the book by Godley/Lavoie (2006), which currently serves as the major textbook on theoretical SFC models, has exposed analytically how the stock–flow relations affect the behaviour of each of the five standard institutional sectors and their interaction. This gave rise to a new generation of empirical SFC models, aiming to capture dynamics emanating from richer structures as compared to the first generation. In terms of the demand components, consumption and investments are now determined separately, but in advance so are the balance sheet and the financial balance of households and firms. However, structures that are more detailed are equally more demanding in terms of data. Thereby, data availability constraints become more binding and the difficulty of constructing from whom-to-who matrices becomes much higher. Almost all models, presented in Table 4, have a rather more generic macroeconomic structure and thus they are able to assess a large variety of fiscal, monetary and financial operations. The only two models that examine very specific research questions are the model for Moldova (Le Heron/Yol 2019) and the model for Iceland (Raza et al. 2019). Both are in the reduced form. They are clearly theory-driven, in the sense that first the structure is developed in isolation to the actual economy and then the model is calibrated according to actual data. Additionally to their structure, almost all models of the GL type are also heavily influenced by the behavioural framework of Godley/Lavoie (2006). This is evident in the Tobinesque portfolio allocation of wealth, the behaviour of the banking sector, which holds an accommodative stance, as well as in the central bank operations, which clear the bond market. In the Icelandic model, this role is held by the foreign sector. A notable exception is made in the case of the model for Italy (Passarella 2019) and the model for Campania (Canelli et al. 2021), which, so far, is the only empirical SFC model at sub-national level. Both models make use of the ‘Other financial assets’ item of the financial accounts as the variable that absorbs all discrepancies in the portfolio allocation of each institutional sector. This resembles the role that state-contingent securities hold in the DSGE models. However, a detrimental difference is the presence of the stock-toflow feedback mechanism in the two SFC models that is the balance sheet effect in the behavioural equations. Note that in the case of the UK model (Gudgin et al. 2015), there is a lack of information regarding the closures of the model. As in the case of the NC models, there is limited provision of in-sample projections. The models for the UK and Campania (Canelli et al. 2021) have an excellent fit regarding the real side of the economy, while the model for Moldova misses the actual values in some specific variables and for a limited period. In broad terms, it projects consistently real data. Concerning the financial side, only the Italian models provide in-sample projections. The model of Passarella (2019) replicates the net lending/borrowing positions of all Empirical stock–flow consistent models: evolution, current state and prospects 315 Journal compilation © 2025 Edward Elgar Publishing Ltd© 2025 The Author Espagne, E., Ngo-Duc, T., Nguyen, M.-H., Pannier, E., Woillez, M.-N., Drogoul, A., Huynh, T.P.L., Le, T. T., Nguyen, T.T.H., Nguyen, T. T., Nguyen, T. A., Thomas, F., Truong, C.Q., Vo, Q. T., Vu, C. T. (2021): Climate Change in Viet Nam; Impacts and Adaptation. A COP26 Assessment Report of the GEMMES Viet Nam Project, Paris: Agence Française de Développement. Fetherston, M. J., Godley, W. A. (1978): “New Cambridge” macroeconomics and global monetarism: some issues in the conduct of UK economic policy, in: Carnegie-Rochester Conference Series on Public Policy, 9, 33–65. Godin, A., Yilmaz, D., Andrade, J., Barbosa, S., Guevara, D., Hernandez, G., Rojas, L. (2023): Can Colombia cope with a Global Low Carbon transition?, AFD Research Papers, No 285. Godley, W. (1999): Seven unsustainable processes, Levy Economics Institute NY, Special report. Godley, W., Zezza, G. (1992): A simple stock flow model of the Danish economy, in: Themes in Modern Macroeconomics, London: Palgrave Macmillan UK. Godley, W., Lavoie, M. (2006): Monetary Economics: An Integrated Approach to Credit, Money, Income, Production and Wealth, Berlin: Springer. Gomes, S., Jacquinot, P., Pisani, M. (2012): The EAGLE. A model for policy analysis of macroeconomic interdependence in the euro area, in: Economic Modelling, 29(5), 1686–1714. Gudgin, G., Coutts, K. J., Gibson, N. (2015): The CBR Macro-economic model of the UK Economy (UKMOD), Centre for Business Research, University of Cambridge, Working Paper, No 472. Hermitte, R. B., Cagnazzo, A., Favero, C. A., Felici, F., Macauda, V., Nucci, F., Tegami, C. (2023): ITFIN: a stock–flow consistent model for the Italian economy, in: Economic Modelling, 119, 106113. Jackson, T., Victor, P. A. (2020): The transition to a sustainable prosperity – a stock–flow-consistent ecological macroeconomic model for Canada, in: Ecological Economics, 177, 106787. Jacques, P., Delannoy, L., Andrieu, B., Yilmaz, D., Jeanmart, H., Godin, A. (2023): Assessing the economic consequences of an energy transition through a biophysical stock–flow consistent model, in: Ecological Economics, 209, 107832. Kalecki, M. (2013): Theory of Economic Dynamics, London: Routledge. Kinsella, S., Tiou-Tagba Aliti, G. (2012): Towards a stock flow consistent model for Ireland, available at SSRN 2011462 Lavoie, M. (2022): Post-Keynesian Economics: New Foundations, Cheltenham, UK and Northampton, MA: Edward Elgar Publishing. Le Heron, E., Yol, N. (2019): The macroeconomic effects of migrants’ remittances in Moldova: a stock–flow consistent model, in: European Journal of Economics and Economic Policies: Intervention, 16(1), 31–54. Maloney, J. (2012): The treasury and the new Cambridge school in the 1970s, in: Cambridge Journal of Economics, 36(4), 997–1017. Mata, T. (2012): Godley moves in mysterious ways: the craft of economic judgment in post-war Britain, in Contributions in Stock-Flow Modeling: Essays in Honor of Wynne Godley, London: Palgrave Macmillan UK. Mazier, J., Reyes, L. (2022): A stock flow consistent model for the French economy, in: MacroModelling, Economic Policy and Methodology: Economics at the Edge, London: Routledge. Meijers, H., Muysken, J. (2022): The macroeconomic implications of financialisation on the wealth distribution-a stock-flow consistent approach, UNU-MERIT, Working Paper, No 2022-035. Michelena, G. (2021): Un modelo dinámico Stock Flujo para la economía Argentina, Central Bank of Argentina, Economic Research Working Papers, No 89. Miess, M., Schmelzer, S. (2016): Extension of the Empirical Stock–Flow Consistent (SFC) Model for Austria: implementation of several asset classes, a detailed tax system and exploratory scenarios, Institute for Advanced Studies Vienna, Research Report, No 58. Nguyen, T.T.H., Espagne, E., Reyes, L., Tran, T.A.-D. (2023): Policy responses to cope with COVID-19 in Viet Nam: an empirical stock-flow-consistent approach, Asian Development Bank Institute (ADBI), Working Paper, No 1359. Nikiforos, M., Zezza, G. (2018): Stock–Flow Consistent macroeconomic models: a survey, In Veneziani, R. and Zamparelli, L. (eds.), Analytical Political, Economy, NJ, USA: John Wiley & Sons. European Journal of Economics and Economic Policies: Intervention, Vol. 22 No. 3316 Journal compilation © 2025 Edward Elgar Publishing Ltd © 2025 The Author Papadimitriou, D. B., Zezza, G., Nikiforos, M. (2013): A Levy Institute Model for Greece, Levy Economics Institute NY, Technical Paper, May 2013. Passarella, M. V. (2019): From abstract to concrete: some tips for developing an empirical stock– flow consistent model, in: European Journal of Economics and Economic Policies: Intervention, 16(1), 55–93. Pierros, C. (2021): Assessing the internal devaluation policy implemented in Greece in an empirical stock‐flow consistent model, in: Metroeconomica, 72(4), 905–943. Ratto, M., Roeger, W., Veld, J. (2009): QUEST III: an estimated open-economy DSGE model of the euro area with fiscal and monetary policy, in: Economic Modelling, 26(1), 222–233. Raza, H., Gudmundsson, B. R., Zoega, G., Byrialsen, M. R. (2019): Crises and capital controls in small open economies: a stock–flow consistent approach, in: European Journal of Economics and Economic Policies: Intervention, 16(1), 94–133. e Silva, A.C.M., Santos, C.H.D. (2011): Peering over the edge of the short period? The Keynesian roots of stock–flow consistent macroeconomic models, in: Cambridge Journal of Economics, 35(1), 105–124. Valdecantos, S. (2021): Grasping Argentina’s Green transition: insights from a stock–flow consistent input–output model, Aalborg University Business School, Working Paper, No 4. Valdecantos, S. (2022): Endogenous exchange rates in empirical stock–flow consistent models for peripheral economies: an illustration from the case of Argentina, in: Journal of Post Keynesian Economics, 45(4), 636–666. Yilmaz, D., Ben-Nasr, S., Mantes, A., Ben-Khalifa, N., Daghari, I., Godin, A. (2023): Climate change, loss of agricultural output and the macro-economy: the case of Tunisia, AFD, Research Papers, No 286. Zezza, G. (2013): Income distribution and borrowing: tracking the US economy with a ‘New Cambridge’ model, in Brancaccio, E., Fontana, G. (eds), The Global Economic Crisis, London: Routledge. Zezza, G., Zezza, F. (2019): On the design of empirical stock–flow consistent models, in: European Journal of Economics and Economic Policies: Intervention, 16(1), 134–158. Zezza, F., Zezza, G. (2022): A stock-flow consistent quarterly model of the Italian economy, in Macroeconomic Modelling, Economic Policy and Methodology, London: Routledge.