Economic Growth and Wage Stagnation in Peru: 1998–2012

• Date: 01 May 2015
• Published date: 01 May 2015
• Author: Carlos Urrutia,Peter Paz
• DOI: http://doi.org/10.1111/rode.12145

Economic Growth and Wage Stagnation in Peru:

1998–2012

Peter Paz and Carlos Urrutia*

Abstract

From 1998 to 2012, the Peruvian economy exhibited rapid growth. Moreover, the composition of the labor

force improved in terms of education and experience, two variables that are typically associated to higher

human capital. The average worker in 2012 had a higher level of education and was one and a half years

older than in 1998, reflecting the impact of the demographic transition. However, the average real wage was

roughly constant. We show that a decline in the wage premium for education, and to a minor extent for

experience, is responsible for the lack of growth in the average real wage. Had these two premia remained

constant throughout the period of analysis, average labor earnings would have increased by about 2.6% per

year, of which 0.7 percentage points are accounted for by the changes in the composition of the labor force

in terms of age and education. We explore the role of the relative supply of workers with different levels of

human capital as an explanation for the decline in the wage premium for education. Finally, we analyze the

implications of these findings for some macroeconomic variables, as earnings and wage inequality, the labor

share and total factor productivity.

1. Introduction

In the last two decades, the Peruvian economy exhibited rapid growth. We focus on

the period 1998–2012, in which aggregate labor productivity (measured as gross

domestic product (GDP) per worker) grew at an annual rate of 2.5%. Moreover, in

this period the labor force became more experienced and better educated, responding

to demographic trends and partly to increases in the coverage of public education.

The average worker in 2012 had completed a higher level of schooling and was one

and a half years older than in 1998. These two trends are typically associated to

increases in the average level of human capital.

However, in the same period real labor earnings per worker and the average real

hourly wage remained basically constant. We document this fact using the main Peru-

vian household survey, Encuesta Nacional de Hogares (ENAHO). If anything, we

observe a modest growth in the average wage of 0.2% per year between 1998 and

2012 (or 0.6% between 2002 and 2012), well below the rate of growth of labor produc-

tivity. Moroever, average labor earnings actually decreased at an annual rate of

−0.12%.1

This evidence leads to an obvious question: “Why has the average real wage in

Peru remained stagnant in a period in which the composition of the labor force both

in terms of age and education improved and labor productivity increased?” In this

* Urrutia: Centro de Investigación Económica, Instituto Tecnológico Autónomo de México (ITAM), Av.

Camino Santa Teresa No. 930, Colonia Héroes de Padierna, México, DF 10700, Mexico. E-mail: currutia@

itam.mx. Paz: Banco Central de Reserva del Perú. Email: peter.paz@bcrp.gob.pe. We would like to thank

the editor of this special issue, two anonymous referees, and the organizers and participants at the Dynam-

ics, Economic Growth and International Trade (DEGIT) XVIII Conference in Lima, Peru. We also thank

Slvio Rendon, Hugo Ñopo Hernando Zuleta for their comments and Michelle Infanzón for helpful advice

on the data analysis. All mistakes are our own.

Review of Development Economics, 19(2), 328–345, 2015

DOI:10.1111/rode.12145

© 2015 John Wiley & Sons Ltd

paper, we attempt to provide an explanation based on a careful analysis of the data of

wages at a disaggregated level. Our analysis is mostly descriptive and to a large extent

we let the data speak for itself. Still, issues of aggregation are considered in the

context of an economic model, albeit a simple one.

The starting point of our investigation is the observation that there is substantial

heterogeneity across age groups and educational levels in terms of the evolution of

labor earnings and wages. Young and uneducated workers have experienced substan-

tial earnings growth, while workers in the age and education categories that grew

more during these years (35–44 age group, secondary and higher/non-college educa-

tion categories) featured either no wage growth or significant real wage losses.2

Therefore, the wage premia for education and experience decreased significantly

during the period.

To assess the impact of changes in these two premia on the average wage we postu-

late a simple production function in which aggregate output is produced using physi-

cal and human capital. The key assumption is that the human capital of workers of

different age and education are perfect substitutes. The mapping between the individ-

ual levels of human capital and the observed relative wages is derived assuming

profit-maximizing firms and perfect competition in the labor market. In this setup, we

show using a simple decomposition that average labor earnings and the average wage

would have increased by about 2.6% per year had these two premia remained con-

stant. This is roughly the rate of growth of average labor productivity; in this sense,

the decline in the wage premium for education, and to a minor extent for experience,

accounts for the lack of growth in average labor earnings and the average real wage.

According to our decomposition, changes the composition of the labor force in

terms of age and education accounts for 0.7 percentage points of the rate of growth of

the average labor earnings. The direct impact of improvements in education and the

demographic trends on wages are then positive and significant. However, this impact

is more than offset by the changes in the market return for these levels of experience

and education whose relative supply grew more in the period. The second question

that we ask in the paper is whether there is a connection between the increase in the

relative supply of educated workers and the decline in the educational wage premium.

For this, we perform an alternative decomposition assuming imperfect substitutabil-

ity between the input of workers with different levels of human capital. In this alterna-

tive setup, relative wages respond endogenously to changes in the relative supply of

educated workers. Our quantitative analysis show that the observed changes in the

composition of the labor force in terms of education account now for a 2% decrease in

the average wage, keeping the relative levels of human capital constant. Hence, a case

can be made for the increase in the supply of educated workers being responsible for a

large fraction of the decline in the wage premium and the gap between the growth of

the average wage and the growth of aggregate labor productivity in Peru.3

These findings have important implications for some macroeconomic variables, as

the labor share, inequality, and total factor productivity (TFP). We show how the

stagnation of the real wage in Peru is associated to a decline in the aggregate labor

share of income. Also, we link the decline in the educational wage premium in Peru

to an important reduction in earnings and wage inequality. Finally, we explore the

impact on the growth of measured TFP (or the Solow residual) of the changes in the

returns to age and education. We illustrate how growth accounting exercises that con-

sider only the changes in educational attainment (not on the wage premia) as a

measure of human capital accumulation would grossly underestimate the contribution

of TFP to economic growth in Peru.

GROWTH AND WAGE STAGNATION IN PERU 329

© 2015 John Wiley & Sons Ltd