Economic Convergence, Capital Accumulation, and Income Traps: Empirical Evidence

• Published date: 01 March 2020
• Author: Cyn‐Young Park,Rogelio V. Mercado
• Date: 01 March 2020
• DOI: http://doi.org/10.1111/roiw.12398

© 2018 Internationa l Association for Res earch in Incom e and Wealth

26

ECONOMIC CONVERGENCE, CAPITAL ACCUMULATION, AND

INCOME TRAPS: EMPIRICAL EVIDENCE

by Cyn-young Park

Asian Development Bank

rogelio V. MerCado, Jr*

Southeast A sian Central Ba nks Research an d Training Centre,the SEACEN C entre

This paper examines the factors that increase the likelihood of economic transition to higher income

status, thereby attempting to answer the question of why some economies move to a higher income

country group while others do not. Using a quintile income distribution approach, we identify 62 econ-

omies that moved to a higher quintile income group in each decade from 1960s to the 2010s out of a

sample of 182 economies. Our findings show that higher physical and human capital growth and oil

revenues are significantly associated with a greater probability of transitioning to higher quintile income

group, although their effects vary not only across income groups within a sample period but also across

different periods. Our results indicate that economies that have attained substantial capital accumula-

tion (either physical or human, or combination thereof) and/or are blessed with natural resources have

avoided income traps and demonstrated a successful and often steady transition to higher income

groups.

JEL Codes: O10, O11, O40

Keywords: capital accumulation, economic convergence, middle-income trap

1. introduCtion

Economic convergence in per capita income has been studied by economists

for several decades. The neoclassical growth model implies that countries with low

per capita income would grow faster than countries with high per capita income,

and over time will converge independent of their initial conditions. However,

empirical evidence on this “absolute convergence” has been elusive as considerable

variations in per capita income between different countries have persisted. The

consensus in development economics now points to either “conditional conver-

gence”, where countries converge to their country-specific long-run equilibria

based on their structural characteristics independent of their initial conditions, or

“club convergence”, where income levels of countries converge to one another in

The authors are grateful to the guidance of D.S. Prasada Rao and to comments and suggestions of

anonymous reviewers. The views expressed in this paper are those of the authors and do not necessarily

reflect the views and policies of the Asian Development Bank or its Board of Governors or the govern-

ments they represent; as well as those of the South East Asian Central Banks Research and Training

Centre or its member central banks/monetary authorities.

*Correspondence to: Rogelio Mercado, Jr., Senior Economist at the Southeast Asian Central

Banks Research and Training Centre (the SEACEN Centre) (rogelio.mercado@seacen.org).

Cyn- Young Park is a Director at the Asian Development Bank. Email: cypark@adb.org.

Review of Inc ome and Wealth

Series 66 , Number 1, March 2020

DOI : 10.1111 /roi w.123 98

bs_bs_banner

Review of Income and Wealth, Series 66, Number 1, March 2020

27

© 2018 Internationa l Association for Res earch in Incom e and Wealth

the long-run provided that their initial conditions and structural characteristics are

similar.1

One would have thought that the debate on economic convergence in the 1980s

and 1990s has been settled by now. But since Gill and Kharas (2007) introduced the

term “middle-income trap” to describe selected East Asian economies that had not

reached high-income level status by the 2000s despite their remarkable growth rates

in the previous decades, a renewed policy debate on how to avoid or escape the

middle-income trap has emerged. It is important to stress that both economic con-

vergence and income traps are conceptually related as whether economies surpass

or get stuck in an income trap may simply depend on the economic convergence

path they are following. For instance, economies that escape the middle-income

trap could exhibit a higher economic convergence path, while those that get stuck

could exhibit a lower economic convergence path. Questions remain as to what

enables economies to follow any specific economic convergence path.

A growing literature has attempted to explain the phenomenon and find evi-

dence of a middle-income trap. Eichengreen et al. (2011 and 2014) argue that once

countries reach a certain level of income, they tend to experience a sudden growth

slowdown. While they suggest that the inflection levels could be $10,000–$11,000

and $15,000–$16,000 (at 2005 US dollar purchasing power parity), they argue that

“slowdowns are less likely in countries where the population has a relatively high

level of secondary and tertiary education and where high-technology products

account for a large share of exports.” Aiyar et al. (2013) also define a middle-in-

come trap as a sudden deceleration in growth. The authors argue that countries in

the middle-income category, like some Asian and Latin American emerging market

economies, are disproportionately more likely to experience growth slowdowns.

They examine factors that cause a sudden drop in growth rates, such as the role of

institutions, demographic structures, infrastructure, the macroeconomic environ-

ment, and output and trade structures.2 Implicit in these studies is the notion that

most countries find it difficult to achieve a higher convergence path as their growth

rates slow.

Felipe et al. (2012) study historical transitions across income groups to iden-

tify economies caught in the middle-income trap, those that take longer than the

median number of years to cross the high-income threshold. In a follow-up study,

Felipe et al. (2017) refute the idea of a middle-income trap as a permanent state

and argue that economies simply differ in their pace of transition from middle to

high-income. Im and Rosenblatt (2015) examine historic transition phases in the

cross-country distribution of income using both absolute and relative income

thresholds. Their analysis of the transition matrix offers little support for the

notion of a middle-income trap, as they find no systemic pattern for a

1See Barro (1991), Barro and Sala-i-Marin (1992), Barro (1996), Cohen (1996), Galor (1996),

Romer (1990), Sachs and Warner (1995), Solow (1956), Solow and Samuelson (1953), and Swan (1956)

for the literature on economic convergence and endogenous growth theory.

2Eichengreen et al. (2014) and Bulman et al. (2014) consider sources of the middle-income trap

from growth perspective, such as diminishing returns to capital and total factor productivity. In con-

trast, Gill and Kharas (2007 and 2015) and Aiyar et al. (2013) look into structural characteristics of

economies—such as demographics, infrastructure, institutions, and economic structure—in relation to

the sources of growth slowdowns.

Review of Income and Wealth, Series 66, Number 1, March 2020

28

© 2018 Internationa l Association for Res earch in Incom e and Wealth

middle-income economy to stay in the same income bracket.3 Han and Wei (2017)

also reject the middle-income trap by using a transition matrix analysis, similar to

Im and Rosenblatt (2015). They argue that relative growth depends on different

country fundamentals and policy factors such as the size of the working-age pop-

ulation, financial development, and macroeconomic stability.

However, there are noticeable differences in what researchers directly or indi-

rectly refer to as a middle-income trap.4 These differences pertain to both the defi-

nition of “trap” and what constitutes a “middle-income” country.5 More

3Using individual country transitions to and from various income levels for a sample of economies,

the authors find that, on average, middle-income countries have a 30 percent probability of moving to

a higher income group.

4Most studies suffer from the limitation of somewhat arbitrary assumptions in setting thresholds

for middle-income economies in addition to a range of empirical problems such as: sample selection

bias; data availability and consistency; and measurement and specification errors. See Gill and Kharas

(2015) and Glawe and Wagner (2016) for an extensive literature review on the concepts, methodologies,

and causes of the middle-income trap.

5Semantics aside, a “trap” refers to a position or situation from which it is difficult or impossible to

escape (Merriam-Webster Online). Based on this definition, a “trap” is the “difficulty” or “impossibil-

ity” of middle-income economies from moving up to high-income status. A rejection of the idea cer-

tainly implies that it is not impossible for middle-income economies to reach high-income status.

However, it disregards that fact that most middle-income countries are finding it difficult to do so.

Figure 1. Cross- Country Per Capit a Income Distri bution.

Notes: Values refer to t he number of econom ies belonging t o a per capita inco me range. Values for

2010s are average fi gures from 2010 to 2014. Data are base d on decade average v alues of real GDP

per capita a t constant US dollar pu rchasing power pa rity.

Source: Da ta based on Penn World Tables 9.0 acc essed Januar y 2017. [Colour figur e can be viewed

at wileyonlinelibrary.com]

020 40 60

Frequency

05000 10000150002000 0

1960 pcgdp

020 40 60

Frequency

0100002000030000

1970 pcgdp

020 40 60

Frequency

0100002000030000

1980 pcgdp

020 40 60

Frequency

010000200003000040 000

1990 pcgdp

020 40 60

Frequency

020000400006000080 000

2000 pcgdp

020 40 60

Frequency

020000400006000080 000

2010 pcgdp