Exports of primary goods and human capital accumulation

• Author: Cem Karayalcin,Yulin Hou
• Published date: 01 November 2019
• DOI: http://doi.org/10.1111/roie.12428
• Date: 01 November 2019

Rev Int Econ. 2019;27:1371–1408. wileyonlinelibrary.com/journal/roie

|

1371

© 2019 John Wiley & Sons Ltd

1

|

INTRODUCTION

Since the 1960s, many developing countries have experienced export‐led growth. Export expansion

has increased the skill premium in some developing countries. For instance, with the integration into

the global economy, the return to skill and the percentage of the population with higher levels of

education grew markedly in China and India. In some countries, however, enrollment rates remain

still low at higher levels of education. For example, on average, only 11% of the population aged 15

and above in Latin America have enrolled in tertiary education in 2010.1

How does one account for

the higher level of human capital accumulation in some countries and persistent stagnation in others

Received: 9 October 2018

|

Revised: 7 April 2019

|

Accepted: 30 May 2019

DOI: 10.1111/roie.12428

ORIGINAL ARTICLE

Exports of primary goods and human capital

accumulation

YulinHou1

|

CemKarayalcin2

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

1Zhongnan University of Economics and

Law, Wuhan, China

2Florida International University, Miami,

Florida

Correspondence

Yulin Hou, School of Business

Administration, Zhongnan University

of Economics and Law, Wuhan, China

430073.

Email: yulinhou6@yahoo.com

Abstract

Many developing countries have experienced export‐led

growth in the last half century. This paper asks whether

the content of what economies export matters for human

capital accumulation. We construct a small open economy

model and find that expansion of primary exports can harm

human capital accumulation if the economy is initially al-

locating significant resources to the production of primary

goods. We then test this prediction empirically using Latin

American data over the period 1965 to 2010 and find ro-

bust evidence in support of the hypothesis that a shift to-

wards primary exports reduces human capital accumulation.

Given the importance of the latter for long‐run growth, our

results suggest a potential role for policy intervention.

JEL CLASSIFICATION

F16; J24; O12

1372

|

HOU and KaRaYaLCIn

during the export‐led growth period? Does the content of what economies export matter for human

capital accumulation?

This paper explores the trade–human capital nexus by focusing on changes in the exports of pri-

mary goods. We argue that expansion of primary exports can harm human capital accumulation if the

economy is initially allocating significant resources to primary goods production. Briefly, the effect of

international trade on educational acquisition depends on a country’s initial production specialization

pattern. When an economy has a strong comparative advantage in primary goods production, export ex-

pansion reinforces its traditional specialization patterns in the production of unskilled intensive goods.

The associated fall in the demand for skilled labor reduces the incentives to invest in human capital,

slowing down skill upgrading in the primary sector and transition to the modern sector. Therefore,

export expansion in primary goods can slow down the progress of human capital accumulation.

To understand the role of export structure in generating differences in human capital accumula-

tion, we develop a small open economy model of agriculture and manufacture sectors that lasts for

two periods. The analysis demonstrates that an exogenous increase in the relative world price of the

agricultural good can reduce the levels of human capital accumulation if the share of manufacturing

employment is initially less than a critical level. In other words, if the initial agricultural employment

and output are higher than a critical value, export expansion of agricultural goods may act as a positive

demand shock for unskilled labor. Naturally, this trend increases the relative return to unskilled labor

and reduces the incentives to invest in human capital. As a result, over time the share of skilled work-

ers in the labor force declines. This framework highlights an economy’s initial production composition

and exogenous trade shocks that result in the different levels of education acquisition.

The Latin America and the Caribbean (LAC) region provides a perfect setting to study the effects

of primary exports on human capital accumulation. Unlike most industrial countries and some de-

veloping countries, an important characteristic of LAC is the large share of primary goods in their

exports. Over the period from 1965 to 2010, the average share of primary exports in total merchandise

exports across LAC countries was around 59%.2

In particular, in the 2000s, some LAC countries

experienced an impressive trade boom in primary exports. For instance, according to the 2016 Latin

American Economic Outlook (Organisation for Economic Co‐operation and Development [OECD]/

Development Bank of Latin America [CAF]/United Nations Economic Commission for Latin America

and the Caribbean [ECLAC], 2015), in the period 2001 to 2010, mining and fossil fuels exports from

Latin America grew at an impressive 16% annually, followed by agricultural products at 12%.

In this paper, we document that the effects of export expansion in primary goods on human capi-

tal accumulation are negative and significant across countries in LAC over the period 1965 to 2010.

The results are robust to the consideration of many control variables and alternative measures of

educational attainment. Furthermore, to address the potential endogeneity of export structure, we use

resource discoveries and international primary goods export price index as instruments for primary

exports to provide further evidence of causality. The two‐sided least squares (TSLS) results are consis-

tent with our panel findings. Overall, these suggest that the specific characteristics of export structure

in LAC can explain the persistent stagnation of its human capital accumulation in the period from

1965 to 2010. Finally, we find evidence that a shift towards primary exports exacerbated the education

inequality in LAC over the same period.

This paper is related to a large literature addressing the trade–human capital nexus in a variety

of contexts. First, this paper provides evidence in support of models of trade with endogenous skill

acquisition (e.g., see Findlay & Kierzkowski, 1983; Foster & Rosenzweig, 1996; Galor & Mountford,

2008). These models predict that international trade will increase the return to the abundant factor

in a given economy. For example, Galor and Mountford (2008) argue that international trade has

played an asymmetrical role in the human capital accumulation between industrial economies and

|

1373

HOU and KaRaYaLCIn

nonindustrial economies. They suggest that in industrial economies, international trade has enhanced

the specialization in the skilled‐labor intensive goods production, which naturally raised the demand

for skilled workers and contributed to investment in education. However, in nonindustrial economies,

the expansion of international trade has generated less demand for skilled workers and fewer incen-

tives to invest in human capital.

Secondly, this paper complements studies that delineate the evolution in educational attainment

across countries. Easterly (2007) and Galor, Moav, and Vollrath (2009) argue that inequality in land

ownership is a key factor for human capital accumulation. Gallego (2010) shows that the degree of

democratization has played a positive role in the development of primary education, and political

decentralization has a significant impact in higher levels of education. Hendricks (2010) suggests that

the majority of the variation in educational attainment across countries is due to within‐industry vari-

ation rather than industry composition. Restuccia and Vandenbroucke (2013, 2014) find that wage and

life expectancy have played an essential role in explaining the trends in hours of work and educational

attainment of the United States over the period from 1870 to 1970.

Finally, this paper also contributes to the growing literature examining the effects of openness on edu-

cational attainment. In the context of India, Shastry (2012), Jensen (2012), and Oster and Steinberg (2013)

all find that the arrival of information technology service jobs had a positive influence on educational

attainment in India. Similarly, Heath and Mobarak (2012) provide evidence that new job opportunities in

garment factories in Bangladesh improved girls’ educational attainment. Atkin (2016) finds that export ex-

pansion in less‐skilled manufacturing in Mexico increased the school drop‐out rate during the period from

1986 to 2000. Li (2018) argues that high/low skill export shocks increase/decrease both high school and

college enrollment in China over the period from 1990 to 2005. In general, these studies focus exclusively

on specific sectors in a relatively small number of locations. One recent paper, by Blanchard and Olney

(2017), focuses on the role of types of sectoral growth in influencing educational attainment. The paper uses

panel data from 102 countries and argues that the change in the composition of a country’s exports is the

key demand‐side driver of education. We differ most importantly from Blanchard and Olney (2017) in that

we focus on exports of primary goods in LAC countries and highlight the initial production specialization

pattern. Furthermore, we use alternative instrumental variables, namely, resource discoveries and the inter-

national primary good export price index, to proxy the export expansion in primary goods. Finally, we pro-

vide evidence on the effect of primary exports on the education inequality and offer a potential explanation.

The rest of the paper is organized as follows. The next section introduces a small open economy

model. Section 3 presents data and the corresponding descriptive statistics. Section 4 provides the

empirical analysis. Section 5 pursues a variety of extensions. Section 6 concludes.

2

|

THE MODEL

We start by developing a small open economy model in this section to guide our thinking for the

empirical work that follows. We consider a small open economy that lasts for two periods and is

populated by a constant number of identical individuals. Competitive firms in the economy produce

two final goods: an agricultural good and a manufactured good, both of which are tradable. It is also

assumed that there is no international lending or borowing.3

We now turn to a detailed analysis of firm

and individual behaviors.

2.1

|

Firms

Firms produce the two final goods by employing labor as the only input. In the agriculture sector firms

use unskilled workers. We posit the following production function for agricultural good: