The Brazilian Economic Regimes and the Possibility of a Regime Switch

ABSTRACT

The Brazilian experience of economic growth with social inclusion from 2004 to 2010 contrasts with the period between 2010 and 2015, in which economic growth was lower despite of the maintenance of the income transfer policies. Based on the Kaleckian theoretical model, we analyze how changes in the economic context altered the response of each aggregate demand component to shifts in income distribution, reducing the likelihood of wage-led regimes and possibly leading to a regime switch. While relevant to explaining specific economic contexts, the possibility of regime switches has been largely overlooked by the literature so far. The empirical estimations based on a SVAR model with an estimated structural break for the period between 2003 and 2015 confirm a reduction in the likelihood of a wage-led regime and suggest that wage-led economic regimes became profit-led after the break in 2010. Thus, the higher wage share had a positive impact on the Brazilian economy, but such effect weakened when the underlying economic conditions changed.

KEYWORDS:

• Distribution
• demand-led regimes
• growth
• regime switch
• Brazil

JEL CODES:

• E12
• E22
• E25

Acknowledgements

The authors are grateful to Rafael Ribeiro, Rosangela Ballini, Marc Lavoie, and two anonymous referees for helpful comments and suggestions. This work was supported by the Brazilian National Council for Scientific and Technological Development (CNPq) under grant #131651/2015-3 (LNR). This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior — Brasil (CAPES) — Finance Code 001.

Disclosure Statement

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

Notes

1 The model developed in this section is largely based on Hein (2014, ch. 6-7). The author combines key contributions in the Kaleckian literature to derive an open economy model with positive saving from workers and a post-Kaleckian investment function. This comprehensive approach is interesting for our article’s aim as it includes the most important relations that determine an economy’s regime.

2 For a detailed analysis of the relations derived by this model, see Hein (2014).

3 The process of manufacturing production being replaced by imports started after the liberalizing reforms at the beginning of the 1990s due to the currency appreciation and the higher international competition (Laplane and Sarti 1997; Laplane, Coutinho, and Hiratuka 2003; Coutinho and Belluzzo 1996), characterizing a deindustrialization process (Palma 2005). Moreover, investment was concentrated in industries less exposed to the international competition and intensive in natural resources and scale (Hiratuka and Sarti 2017), inducing a regressive specialization in the country (Coutinho 2003).

4 The increase in the wage share in 2015 should be analyzed carefully, as it may be partially due to the countercyclical behavior of this variable. Since there was a negative real minimum wage growth rate (reducing the strength of the redistributive process) and negative output growth rate in this year (Table 1), it is likely that this cyclical behavior played a relevant role. This behavior is related to the presence of overhead labor and labor hoarding and is discussed in more detail by Lavoie (2014, ch. 3).

5 The sample has 52 observations. The main restriction to the sample size is given by the data on the income shares in Brazil, leading to a sample size close to that of most of the other studies on the Brazilian demand and accumulation regimes.

6 As a SVAR model captures short-run relations, the growth rate of GDP is an adequate proxy to capacity utilization. The use of this proxy has the advantage of avoiding the use of capital stock estimations or the HP-filter, which leads to spurious dynamics in the data (Hamilton 2018).

7 The wage share was obtained by dividing the aggregate wages by the value added and assuming that the mixed income is split between labor and capital income in the same proportion as the remaining income. The quarterly wage share was obtained following Bastos (2012), who suggests applying the Denton (1971) interpolation method using the quarterly wage share given by the quarterly aggregate nominal wages (IBGE 2022b) divided by the nominal value added (IBGE 2022c) as the indicator variable and using the wage share from the annual national accounts (IBGE 2022a) as the benchmark variable.

8 Unit root tests are available upon reasonable request.

9 The latter assumption reflects the immediate effect of a lower share of overhead labor income following increases in capacity utilization (Lavoie 2014, ch. 3). Empirical evidence supporting this effect is presented by Cauvel (2022) and Rolim (2019).

10 Due to space limitations, the model outputs are not reported and are available upon reasonable request.

11 We are unable to estimate the model for the subperiod after the break (between 2010q2 and 2015q4) due to the small sample size obtained for this subperiod.

12 The contrast between the results presented in figure 2 and 3 is partly explained by the fact that the model for the subperiod between 2003q1 and 2010q1 obtain different contemporaneous parameters in the structural decomposition.

Additional information

Funding

This work was supported by CNPQ [grant number 131651/2015-3] and CAPES [grant number 001].