Components of Autonomous Demand Growth and Financial Feedbacks: Implications for Growth Drivers and Growth Regime Analysis

ABSTRACTThis paper presents a simple closed-economy model that is driven by the growth of two autonomous components of demand, namely government spending and rentiers’ consumption out of interest income. Using this model, it seeks to make two contributions. First, by focusing on the financial dynamics that arise from debt-financed government spending, we show that an underlying stock-flow interaction provides an endogenous mechanism which, under certain conditions, aligns the two autonomous growth rates in the long run. Hence, an economy can be driven by two autonomous components of demand, without one dominating the other in the long-run nor without policy changes being required to align the two growth drivers. Second, we prove that if two autonomous demand components are dynamically interdependent, then the relative size of their growth contributions may be misleading as a guide to classify growth regimes, both in the long-run equilibrium as well as during the traverse towards this equilibrium. Furthermore, we show that the relative growth contributions are economic policy contingent. We thus argue that interdependencies between autonomous growth components arising from financial stock-flow interactions should not be ignored in Sraffian supermultiplier growth decomposition exercises that aim to identify underlying growth drivers.KEYWORDS: Sraffian supermultiplier and endogenous credittwo autonomous growth driversdemand-led growth accountinggrowth regimesJEL CODE: E11E12E20E62 AcknowledgementsWe would like to use this opportunity to acknowledge and thank the reviewers who reviewed this article and aided in its publication.Disclosure StatementNo potential conflict of interest was reported by the author(s).Notes1 See Cesaratto (Citation2017), Cesaratto and Di Bucchianico (Citation2020), and Cesaratto and Pariboni (Citation2022) for detailed accounts of the close relationship between autonomous demand-led growth and post-Keynesian monetary circuit and endogenous money theory. In this context, Cesaratto and Pariboni (Citation2022, 306) also point out the close relationship of Sraffian supermultiplier models with Kalecki’s notion that external markets created by government deficits and export surpluses are drivers of growth: ‘The autonomous components of aggregate demand, which in the supermultiplier model drive growth and coincide with Kalecki-Luxemburg’s external markets, can be financed by purchasing power creation by banks, by accumulated wealth or foreign income. … By definition, external markets are fueled by debt creation, and the seed of financial instability and crisis can be traced here.’2 For a stock-flow supermultiplier model with two debt-financed autonomous expenditures (autonomous government and household consumption), and interrelations between sectoral autonomous demand growth and sectoral indebtedness, with possible emergence of financial fragility processes, see Pedrosa, Brochier, and Freitas (Citation2023).3 We do not mention government expenditure here, because we hold that in sovereign money economies, governments do not have to pre-fund their expenditures, although policy makers may have imposed some policy rules (government debt ceilings, debt brakes, balanced budgets etc.) which may impose a constraint on government expenditures, which may then interact with other autonomous expenditure components. In countries without sovereign money, government expenditures would have to be mentioned here, too.4 Freitas and Christianes (Citation2020) have two autonomous growth rates (government expenditures and capitalist consumption), but just assume that the two will be equal in the long run, without discussing their interaction. Brochier and Macedo e Silva (Citation2019) have presented an autonomous demand-led stock-flow consistent growth model driven by consumption out of wealth, with wealth being endogenous in the long run. They thus include financial issues, but only have one autonomous growth rate which is endogenous.5 In the journal version of his paper, Allain (Citation2022) has abandoned the distinction between ‘active’ and ‘passive’ components of autonomous demand in favour of ‘autonomous’ and ‘semi-autonomous’ demand components. He thus follows Fiebiger (Citation2018, 2020) and Lavoie and Fiebiger (2019), who have argued that ‘semi-autonomous’ may be a better terminology to indicate the independence of parts of demand from income generated in current production, which may be, nonetheless, through some other mechanism be related to the dynamics of production in the long run. They refer to Kalecki (Citation1968, 265–269), who is using ‘semi-autonomous’ for the autonomous part of capitalists’ consumption, and the effect of technical change and innovations on investment.6 For simplifying purposes, we are not considering class distinctions and its consequences for different propensities to consume. Therefore, we also do not explore the possible effect of different propensities to consume out of different income streams.7 Hence, this paper makes no further claims about the effects of changes of c nor i on demand and output.8 We ignore the trivial solutions where the long-run equilibrium values are zero.9 This has consequences for the level of output in the steady state (Equation 4), as a higher propensity to invest increases the level of output.10 See Freitas and Serrano (Citation2015, 14), who also elaborate on the interpretation of this stability condition. A similar stability condition is found in, for example, Hein and Woodgate (Citation2021) and Morlin (Citation2022).11 For an assessment of the empirical plausibility of the stability conditions (18) and (19) for the case of the United States, see respectively, Blanchard (Citation2019) and Haluska, Braga, and Summa (Citation2021).12 The level of disaggregation of autonomous components can be even more detailed, e.g., with the distinction between government consumption and investment, investment by state-owned enterprises, business investment in R&D, and consumption out of public transfers and public wages (Labat and Summa Citation2023). However, in none of the papers of the SSM demand-led growth accounting literature the consumption out of interest on public debt is included. We believe that empirically estimating this latter effect and including it in demand-led decompositions exercises for different countries, to check its contribution to growth, should be a desirable future development of this research agenda.13 We are simplifying the story, because exports can also be included as a second active autonomous driver for the Spanish economy.14 This is, of course, just a simplified stylized story, as autonomous consumption out of credit and residential investment can be included as important active autonomous drivers in the Brazilian case, too.15 Though it is, of course, possible that decreases in γ and/or increases in i may not be sufficient in size to change the dominant growth contribution. Though not illustrated here, one could equally examine the opposite cases wherein the policy shock is a rise in γ and/or fall ini. Unsurprisingly, the result in the post-shock period is, in qualitative terms, the mirror image of what is depicted in Figure 2, i.e. a rising government spending-debt ratio ρ, lower threshold value d, higher (and rising) growth contributions of government spending, as well as lower (and falling) growth contribution of autonomous expenditure.