Understanding Manufactured Exporters at the Firm‐Level: New Insights from Using SARS Administrative Data

• Author: Marianne Matthee,Neil Rankin,Carli Bezuidenhout,Tasha Webb
• DOI: http://doi.org/10.1111/saje.12158
• Date: 01 January 2018
• Published date: 01 January 2018

UNDERSTANDING MANUFACTURED EXPORTERS AT THE

FIRM-LEVEL: NEW INSIGHTS FROM USING SARS

ADMINISTRATIVE DATA

MARIANNE MATTHEE*, NEIL RANKIN

†

,TASHA WEBB

†

AND CARLI BEZUIDENHOUT

‡

Abstract

Existing South African work on f‌irm-level data has been limited by access to large datasets that

track f‌irms over time. This paper overcomes this by analysing a new dataset of the population of

manufacturing f‌irms that are matched to their export transactions. South African f‌irm-level

exporting is similar to the stylised facts of f‌irm-level exporting found internationally. Moreover,

heterogeneity that exists within exporting is evident. Not only do exporters differ in terms of the

amount exported, but also in terms of the number of products and destinations they export too.

These in turn are related to f‌irm-level characteristics including productivity.

JEL Classif‌ication: F10, F14

Keywords: Heterogeneous f‌irms, exports, productivity, f‌irm-level data, multidestination,

multiproducts

1. INTRODUCTION

Internationally, the understanding of exporting at a disaggregated level has improved dra-

matically with the availability of f‌irm-level data. Initial work by Bernard et al. (1995)

and Bernard and Jensen (1999) showed that, contrary to the assumptions of many of the

“new” trade models where exporting f‌irms were homogeneous, exporting was relatively

rare and exporters were different to nonexporters across many dimensions. This, in turn,

led to the development of new models to explain this trade at the f‌irm level, such as

Melitz (2003). These types of models allow for f‌irm heterogeneity, often through differ-

ences in f‌irm-level productivity, which leads to heterogeneous export behaviour across

f‌irms. This is in accordance with the robust empirical evidence which shows that f‌irm-

level productivity is a strong correlate with entry and survival in the export market (see

Wagner’s 2007, 2011 reviews of this evidence). A subsequent wave of theoretical models

* Correspondin g author: Associate-Professor, School of Economics, Private Bag X6001,

North-West University, Potchefstroom, 2531. E-mail: Marianne.Matthee@nwu.ac.za

†

Department of Economics, Stellenbosch University

‡

School of Economics, North-West University

All the authors would like to thank the United Nations University World Institute of Devel-

opment Economics Research and National Treasury for funding support. Also, they would

like to thank the South African Revenue Service for the data. Finally, Marianne Matthee also

acknowledges support from the National Research Foundation (NRF) of South Africa (Grant

Number 90709). Any opinion, f‌inding and conclusion or recommendation expressed in this

material is that of the authors and the NRF does not accept any liability in this regard.

V

C2017 UNU-WIDER. South African Journal of Economics published by John Wiley & Sons Ltd on behalf of

Economics Society of South Africa.

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which

permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not

used for commercial purposes. doi: 10.1111/saje.12158

96

South African Journal of Economics Vol. 86:S1 January 2018

South African Journal

of Economics

have developed to explain the additional heterogeneity of export behaviour at the prod-

uctlevel(e.g.Bernardet al., 2011a,b).

Although a number of studies have looked at the micro-economics of exporting in

South Africa, this work has been hampered by a lack of access to the type of administra-

tive data used in other countries.

1

The lack of a “corpus” of academic research in this

f‌ield has also meant that the policy discussion around export policy has remained shal-

low. For example, the National Development Plan acknowledges the importance of

increasing exports but has very limited discussions of the types of f‌irms (rather than sec-

tors) where this export growth would come from (National Planning Commission,

2013). The availability of business-entity level data from the administrative records of

the South African Revenue Service (SARS) now means that the type of research that has

become commonplace internationally can be done on South Africa and this population

data overcomes the sampling issues that plagued previous research.

In this paper, we use the SARS data to highlight some “stylized facts” about South

African manufacturing exporters and to investigate the relationship between exporting

and productivity in more detail. This data shows that South African exporting at the

f‌irm level is similar to the stylised facts of f‌irm-level exporting found internationally: less

than a f‌ifth of f‌irms export in any given year; specialist exporting (exporting more than

half the value of total output) is very rare but total export value is dominated by a small

number of f‌irms; and exporters are larger, pay better and are more productive. There is

also a large degree of churn – entry and exit into the export market – but this does not

seem to translate into sustained exporting. Most of South Africa’s exporting growth

comes from expansion on the intensive margin – existing exporters expanding their

exports of existing products to existing destinations. The access to f‌irm and transaction

data shows the heterogeneity even within exporting. Not only do exporters differ in

terms of the amount exported, they also differ in terms of the number of products and

destinations they export to. These in turn are related to f‌irm-level characteristics includ-

ing productivity.

Multiproduct, and multidestination exporters contribute the most to total export

value. They are also bigger and more productive than other types of exporters. Productiv-

ity varies by both the number of products exported and the number of destinations

exported to. Single-product exporters, regardless of where they export to, have productiv-

ity levels that are similar to domestic f‌irms. Multiproduct exporters exporting within

Africa are only about 4% more productive, and multiproduct exporters exporting outside

of Africa are approximately 10% more productive compared to producers for the domes-

tic market.

Overall these results indicate that South African exporters are, in general, are very

similar in characteristics and behaviour to exporters in other countries where similar

research has been undertaken. Furthermore, they indicate that some of the earlier

results, such as the lack of an exporter premium, may have been driven by the char-

acteristicsofthesamplesusedinpreviousresearchortheinabilitytodifferentiate

the heterogeneous aspects of exporting behaviour like destinations and number of

products. These limitations have been overcome in this paper due to the use of pop-

ulation level administrative data.

1

A review of these papers is included the South African literature review section.

97South African Journal of Economics Vol. 86:S1 January 2018

V

C2017 UNU-WIDER. South African Journal of Economics published by John Wiley & Sons Ltd on behalf of

Economics Society of South Africa.