Exporting and the wage premium: The case of South African manufacturing firms

• Published date: 01 November 2021
• Author: Carli Bezuidenhout,Marianne Matthee,Neil Rankin
• Date: 01 November 2021
• DOI: http://doi.org/10.1111/rode.12810

Rev Dev Econ. 2021;25:2031–2051. wileyonlinelibrary.com/journal/rode

|

2031

© 2021 John Wiley & Sons Ltd

Received: 25 April 2018

|

Revised: 1 June 2021

|

Accepted: 21 June 2021

DOI: 10.1111/rode.12810

REGULAR ARTICLE

Exporting and the wage premium: The case of

South African manufacturing firms

CarliBezuidenhout1

|

MarianneMatthee2

|

NeilRankin3

1School of Economic Sciences and TRADE

Research Entity, North- West University,

Potchefstroom, South Africa

2Gordon Institute of Business Science,

University of Pretoria, Pretoria, South

Africa

3Department of Economics, Stellenbosch

University, Stellenbosch, South Africa

Correspondence

Carli Bezuidenhout, North- West University,

Potchefstroom, South Africa.

Email: carli.bezuidenhout@nwu.ac.za

Funding information

United Nations University World Institute

of Development Economics Research and

National Treasury; National Research

Foundation (NRF) of South Africa, Grant/

Award Number: 90709

Abstract

There is much literature to support the view that export-

ers (both developed and developing countries) pay higher

wages than nonexporters. While this so- called export wage

premium has also been found to be prevalent in South

Africa, it has not been thoroughly researched, with studies

to date having relied on cross- sectional sample data rather

than on the population of firms and workers. Using a newly

constructed employer– employee matched panel data set on

South African manufacturing firms, the study examined the

factors contributing to the export wage premium in these

firms— from firm, individual, and job characteristics (both

observable and unobservable) to firms’ distribution of

wages and export destinations (e.g., SACU- only [Firms ex-

porting only to Southern African Customs Union countries],

Africa- only [Firms exporting only to African countries], or

international [Firms exporting to both African and non-

African countries]). One of the key findings was that the ex-

port wage premium is not about being labeled an exporter.

It is, however, because of the “type of firm” (unobservable

firm characteristics) exporters are, the “type of workers”

(unobservable individual characteristics) they employ, and

the “type of jobs” (unobservable job characteristics) they

create. Policymakers should therefore be aware that simply

expanding the pool of exporters will not necessarily give

momentum to the export wage premium phenomenon.

Rather, policy measures should be aimed at increasing firm-

level productivity.

2032

|

BEZUIDENHOUT ET al.

1

|

INTRODUCTION

Exporters are known to pay higher wages than nonexporters (Bernard & Jensen,1999). There is ample

evidence of this in both developed and developing countries (Alvarez & López,2005; Bernard &

Jensen,2004; Bernard etal.,2007; Hahn,2005). Africa is no different. For example, Van Biesebroeck

(2005) studied the effect of export status on productivity in a range of manufacturing firms in nine

sub- Saharan African countries and found that exporters paid on average 40% more than nonexporters.

The so- called export wage premium is also in evidence in South Africa, as revealed in two studies by

Matthee etal.(2016) and Edwards etal.(2016), respectively.

The question should be asked: why do exporting firms pay higher wages? There could be a number

of reasons. First, exporting firms have different observable firm characteristics from nonexporting

firms (e.g., firm size, capital intensity, and output per worker) (see Bernard & Jensen,1997; Bernard

& Wagner,1997; Greenaway & Yu, 2004). Second, the observable individual characteristics of the

workforce (e.g., age and length of tenure) in exporting firms might differ from those in nonexport-

ing firms (Klein etal.,2013; Munch & Skaksen, 2008; Schank etal.,2007). Third, the export wage

premium might be due to unobservable firm and individual characteristics (e.g., Fafchamps,2009;

Munch & Skaksen,2008; Schank etal.,2007). Finally, the distribution of wages and the higher (on

average) wages paid by exporting firms might disguise the fact that exporting firms pay more at cer-

tain parts of the wage distribution, according to where in the wage distribution it matters most (Fu &

Wu,2013; Koenker & Hallock,2001).

Limited research has been conducted on the export wage premium in South Africa. The phenom-

enon was examined in a study by Rankin and Schöer (2013) using employer– employee matched data.

However, the study was limited to cross- sectional survey data. Furthermore, while the authors found

evidence of an export wage premium (even after controlling for firm and individual characteristics),

the premium differed by export destination. Among the findings were that the wages of firms export-

ing to the regional market of SADC (Southern African Development Community) were lower than

those of firms producing domestically, whereas the wages of firms exporting outside the region were

higher than those of both regional exporters and domestic producers.

This article helps to fill the gap in the literature by using newly available administrative data to

study the export wage premium among manufacturing exporters in South Africa. A panel data set

was constructed by merging three sets of administrative records collected by SARS (South African

Revenue Service), namely employee income tax data (IRP5 or PAYE), company income tax (CIT)

return data (ITR and ITR14), and customs transaction data. This new employer– employee matched

panel data set on South African manufacturing firms allows one to control not only for observable

(and unobservable) firm characteristics but also for observable (and unobservable) individual and

job characteristics (which constitutes a unique firm– individual match). The article examines whether

the export wage premium is due to observable or unobservable firm characteristics, individual or job

characteristics, distribution of wages, or actual exporting activities (the firm's export status).

KEYWORDS

employer– employee data, South African manufacturing exporters,

wage premium

JEL CLASSIFICATION

F10; F16