Brazil and China: Two Routes of Economic Development?

• DOI: http://doi.org/10.1111/rode.12175
• Published date: 01 August 2016
• Date: 01 August 2016

Brazil and China: Two Routes of

Economic Development?

Laura Policardo, Lionello F. Punzo, and Edgar J. Sánchez Carrera*

Abstract

We look at two emerging economies, Brazil and China, and propose an evaluation of their recent develop-

ment in terms of growth performance and the evolution of income inequality. Our analysis, therefore, is

related to the literature on their recent vast growth and also to the much debated Kuznets curve and

theory. However, we claim that neither the growth nor the Kuznets approach capture recent relevant phe-

nomena characterizing such countries’ dynamics: namely, the presence of at least two distinct growth

models. Cointegration analysis and empirical evidence seem to corroborate our interpretation. They also

offer some further insights. We surmise that while contributing to press for a further re-examination of con-

vergence in the light of the issue of income distribution, such findings have interpretative relevance and

policy implications for other less developed countries.

1. Introduction

After a lengthy and on the whole probably inconclusive debate, research has started

to look for supplementary variables to explain the all-pervasive evidence of diver-

gence in per capita income among developed countries as well as between them and

less developed countries (LDCs) and more recently emerging economies. An exten-

sive literature has blossomed on the causality relationship between inequality

(however, measured) and economic performance, as measured by say, per capita

gross domestic product (pc GDP). Recently, this debate has merged with a debate on

the so-called “middle-income trap” (in particular, in relation to some of the emerging

economies).

Generally in a cross-country framework, we have seen the (re-)discovery of analy-

ses positing the existence of a functional relation between growth and inequality. It is

natural to think of the time honored Kuznets curve (1955) with its associated

dynamical hypothesis and prediction, a relation whose peculiarity is to describe a

causal association that reverses at the (high or low) level of a country’s development.

In more recent years, a shared attitude has been the rejection of the very existence (or

of the “end”, Palma (2011)) of a K curve constructed from cross-country analysis.

Hereafter, we take instead a time series approach to study the evolution of pc GDP

(levels and growth rates) vs income inequality (through a Gini index) in the two

economies, China and Brazil. (A companion paper by Risso et al. (2013) examines

México.)

An econometric exercise shows that in both countries pc GDP and income inequal-

ity do stand in a cointegrated, long-term relationship. However, while vector

* Sánchez Carrera: Faculty of Economics, Autonomous University of San Luis Potosí, México. Av.

Pintores S/N, Fracc. Burocratas del Estados, CP 78263, San Luis Potosí, SLP, México. E-mail:

edgar.carrera@uaslp.mx. Policardo: Department of Economics, University of Siena, Italy. Punzo: Depart-

ment of Economics and Statistics, University of Siena, Italy (also affiliated to INCT/PPED, UFRJ Rio de

Janeiro, Brazil).

© 2015 John Wiley & Sons Ltd

Review of Development Economics, 20(3), 651–669, 2016

DOI:10.1111/rode.12175

autoregression (VAR) Granger causal relationships indicate that China’s economic

growth “predetermines” or “Granger-causes” income inequality (perhaps, one could

say “álaKuznets”), in Brazil the causal relationship seems to be reversed. Moreover,

in China such a relationship is positive (such as in Risso and Sánchez Carrera (2012))

while in Brazil it is negative. Then, for each country, we construct, analyze and

compare the qualitative behavior of curves in a plane with (pc GDP, Gini index) on

the coordinate axes, a K-plane. Obviously, such curves are not true Kuznets curves,

owing to their time series origins, though we may call them “Kuznets-like”.

We suggest a re-interpretation of the experiences of the two countries, where the

key notions distinguish between two growth models: a fast growth with income con-

centration, “investment” supported model (prevailing in China, at least until recently)

and a more moderate growth with income redistribution, “consumption” supported

model (of Brazil), as well as the possibility of a model switch.1

Section 2 briefly surveys relevant literature. Section 3 goes through the steps of an

econometric exercise to discuss empirical evidence for China and Brazil focusing on

the existence of stable long-run relationships between inequality and growth of the

type as implied by cointegration and Granger causality. The aim of section 4 is to

sketch some stylized facts of Brazil’s and China’s recent development in a K-plane,

and to introduce the notion of growth models and model switch. Section 5 puts the

two countries vis-à-vis with one another and summarizes our results, while section 6

outlines some conclusions.

2. The Literature: A Brief Review

In recent years, considerable efforts have been spent on understanding the differen-

tial growth experience of various countries in a hunt for the perfect policy recipe for

success. Especially emerging economies, but also some LDCs, have been scrutinized

with the tools of growth (rather than development) theories. One key issue that has

come up is whether a growth goal is achievable through redistributing wealth or

otherwise letting wealth and income concentrate. It is still subject to debate.

The dynamic process creates new resources as well as modifying their distribution.

This was very clear to the theorists of economic development seeking explanations

and consequences in major structural changes. Ever since Kuznets and Lewis, theo-

retical constructs about the effects of performance on income distribution have

focused on several basic mechanisms. In a bold generalization from limited cross-

country evidence, Kuznets (1955) maintained that an increase in inequality is inevi-

tably associated with certain phases (“stages”) of the development process—that is,

distributional inequality would increase as the economy progresses from an agrarian

to an industrial structure, only to decline later on with the transition having been

accomplished and the country entered into the club of the developed ones.2Kuznets’

theory came to be discussed within the growth literature and then within the empirics

of cross-country convergence.

As seen by this more recent literature, the issue came to be synthetized in a single

central question: is it growth that causes inequality, or else can income inequality be

its engine? Thus, one has to inquire not only about the direction but also the sign of a

unidirectional causality relation. In contrast to the development approach focusing on

structural change, springing from the growth and convergence debate, the more

recent contributions have been searching for a stable, unidirectional relation between

those variables, generally through a cross-country approach similar to Kuznets’s own.

Opposite, often-controversial results were produced.

© 2015 John Wiley & Sons Ltd

652 Laura Policardo, Lionello F. Punzo, and Edgar J. Sa

´nchez Carrera