Aggregate Demand and Demand Leakages in a Post-Keynesian Investment Function: A Dynamic Panel Data Analysis for Developed and Developing Countries

ABSTRACT

We investigate how multiple components of the profit rate differently impact the investment decisions of developed and developing countries in the medium and long run. First, we propose a modification of Bleckers’ (2016) accelerator model, mainly by decomposing the profit variable and incorporating the New Developmentalism theory of access to demand. Second, we use the ICIO tables of OECD and other sources to create a database with 59 countries from 2000 to 2018 and test our hypothesis using the dynamic GMM methodology. Results suggest that demand growth — the accelerator component — is an important determinant of investment in all countries regardless of their development stage. However, while increasing access to demand in both the domestic and foreign markets has shown to be crucial for investment decisions of developing countries, raising the labor share has a pronounced importance on investment decisions in developed countries.

KEYWORDS:

• Investment function
• accelerator principle
• profit rate
• New Developmentalism

JEL CODE:

• E12

Acknowledgements

Without implicating them in any way for the contents of the paper, I would like to acknowledge, with thanks, comments on an earlier draft of this paper by Bresser-Pereira, Nelson Marconi and Guilherme Magacho. I would also like to express my gratitude for very helpful comments and suggestions from two anonymous referees of this journal.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 It is important to take into account that leakages may be of two kinds: Losing market share of domestic demand and losing market share of foreign demand. In the former, local producers are facing competition from external producers through the import of inputs and final products. In the latter, they are losing the capacity to capture markets abroad.

2 For the ND theory, maintaining a competitive exchange rate is the most important macroeconomic variable that defines the capacity of the state-of-the-art businessmen to access global demand (Bresser-Pereira 2012, 2014, 2020).

3 Classical economists such as Ricardo and Marx, even though through different rationalities, argued that profit rate is the main variable that determines investment decisions (Bresser-Pereira 1991; Chraki 2022). This view is abandoned by the neoclassical economists, who put the interest rate at the center of macroeconomic theory and investment determination. Even after Keynes, mainstream economics — including the neoclassicals and New Keynesians — kept the interest rate at the center of the investment function. For example, in Hicks' (1937, p. 149) model investment depends exclusively on the interest rate. Similarly, Samuelson (1948) also put the interest rate as the main determinant of investment, leaving net productivity (marginal efficiency of capital) as a given variable.

4 Following Glyn (1997) and Sawyer (2019), Basu and Das (2017) highlight that some works of Kalecki predate Robinson on the role of realized profits as a key determinant of investment because it could impact expectations ease financing constraints.

5 A final and more recent extension of the BM investment function to include all components of the profit rate has been developed by Foley and Michl (1999) and tested by Basu and Das (2017, 2018). For the authors, there is no reason to rule out the capacity to capital ratio in the equation, which, combined with the profit share and the capacity utilization rate, composes the profit rate.

6 There are critics of this interpretation arguing that as several studies that estimated equations using the structural approach did not find cointegration relations, this method may also be capturing short-run results. Rolim (2021, pp. 416–418) shows this is the case in many studies of the US economy, and argues that both aggregative and structural approaches are more prone to be capturing short-run effects. Teixeira, Missio, and Dathein (2022) review this debate in detail and investigate the demand regimes in Brazil using both aggregate and structural approaches.

7 Blecker (2016) lists different econometric issues of previous studies and asserts that previous findings are very sensitive to various aspects of their specifications, such as data frequency and lag lengths, functional forms (linear or non-linear), transformation of the variables, the included control variables and the methodology to control for endogeneity in the models.

8 This is so because of how the equation is specified. See Fazzari et al. (1988, p. 164) and Chirinko, Fazzari, and Meyer (1999, p. 61).

9 The author presents the equation with different notations as he makes the distributed lags of each variable explicit.

10 Indeed Fazzari, Hubbard, and Petersen (1987) have shown that firms constrained to raise funds externally are more sensitive to movements in cash flow since internal finance becomes the main source of financing investment.

11 The ND argues that both supply-side and demand-side factors influence the producers’ capability to access demand. However, while supply-side elements — such as appropriate capabilities, infrastructure, etc. — are accepted by the economic literature, demand-side factors — notably a competitive exchange rate — are not. For example, Ribeiro, McCombie, and Lima (2016, 2017) argue that once the effect of a currency devaluation on income distribution is included in the analysis, the role played by the exchange rate on economic growth is ambiguous. Moreover, Bottega and Romero (2021) and Magacho, Ribeiro, and Rocha (2021) have questioned the role played by the exchange rate in industries with higher technological content, highlighting the role of non-price competition for these industries. Hence, ND theory has centered its theoretical and empirical research on analyzing the relationship between the exchange rate and investment, structural change, and economic growth (Bresser-Pereira 2012, 2014, 2020, 2021; Bresser-Pereira, Araújo, and Peres 2020; Marconi et al. 2021; Marconi, Porto, and Araujo 2022). Nevertheless, there are studies that have focused on understanding the need for complementarity between macroeconomic and industrial policies (Bresser-Pereira, (Guzman, Ocampof, and Stiglitz 2017; Ocampo 2020)and Rugitsky, 2018; Nassif, Bresser-Pereira, and Feijo 2018).

12 Now all variables are entered in the first difference.

13 The significance of the coefficients in equations 9–12 is calculated using the command nlcom in Stata which uses the Delta Method for estimating the Standard Errors. In its essence, the delta method expands a function of a random variable about its mean, usually with a one-step Taylor approximation, and then takes the variance. A detailed explanation with examples can be found in Oehlert (1992) and in Stata forum https://www.stata.com/support/faqs/statistics/delta-method/)

14 With a longer time horizon of analysis, from around the 1990 to 2016, Bresser-Pereira, Araújo, and Peres (2020) found that capital openness negatively impacted the performance of the manufacturing sector measured as the share of the value added of manufacturing to total value added of the economy. However, the authors did not find significant effects of capital openness on GDP per capita.

Additional information

Funding

This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brasil (CAPES) – Finance Code 001.