Markups of Exporters and Importers: Evidence from Hungary

• Author: Balázs Muraközy,Cecília Hornok
• DOI: http://doi.org/10.1111/sjoe.12292
• Date: 01 July 2019
• Published date: 01 July 2019

Scand. J. of Economics 121(3), 1303–1333, 2019

DOI: 10.1111/sjoe.12292

Markups of Exporters and Importers:

Evidence from Hungary*

Cec´ılia Hornok†

Kiel Institute for the World Economy, DE-24105 Kiel, Germany

cecilia.hornok@ifw-kiel.de

Bal´azs Murak¨ozy

Institute of Economics (MTA KRTK KTI), Budapest, HU-1097, Hungary

murakozy.balazs@krtk.mta.hu

Abstract

In this paper, we study the relationship between different proxies of firm-level markups and

trade status, using balance sheet information linked to detailed trade data from Hungary between

1995 and 2003. Wefind t hat importing is stronglypositively correlated with markup measures,

both across and within firms. We argue that this correlation can reflect three channels: self-

selection, higher physical productivity resulting from access to a larger variety of inputs, and

quality upgrading based on high-quality imported intermediate inputs. We present evidence for

the relevance of the third channel by showing that importers’markup premium is higher when

inputs arrivefrom developed countries, and that importing is cor related with higher-quality(price-

adjusted revenue) exports. We find no robust evidence for exporter premium when controlling

for importing. We argue that the non-existent exporter premium might result from the stronger

competition in export markets relative to domestic markets.

Keywords: Detailed trade data; markup premia; quality; self-selection

JEL classification:D22; D24; F14; L11; L60

I. Introduction

A large body of empirical work has shown that trading firms are larger

and more productive than non-traders (Bernard and Jensen, 1995). More

recent literature has found that firms importing intermediate inputs are

also more productive, larger, charge higher prices, and pay more for their

inputs (Kugler and Verhoogen, 2009; Halpern etal., 2015). A growing body

of research has started to study the relationship between trade status and

markups, a key measure of competitiveness. De Loecker and Warzynski

*Weare grateful for the financial suppor t of the Firms, Strategyand Performance Lendulet Grant

from the HungarianAcademy of Sciences. Wethank L´aszl´oHalpern,Mikl´os Koren,G ´abor B ´ek´es,

and YanpingLiu for their valuable comments.

†Also affiliated with the Kiel Centre for Globalization.

C

The editors of The Scandinavian Journal of Economics 2018.

1304 Markups of exporters and importers: evidence from Hungary

(2012) and Bellone et al. (2014) have provided evidence that exporters also

have higher markups than non-exporters. Marin and Voigtl¨ander (2013), in

contrast, found no markup premia for exporters.

With this paper, we aim to contribute to this literature by relating both

export and import status to different measures of firm–year level markups,

using Hungarian balance sheet and disaggregated trade data. Importantly,

it turns out that handling exporting and importing in a symmetric way

is important, because the exporter markup premium disappears when we

take into account importing. While our preferred proxy for markups is the

one proposed by De Loecker and Warzynski (2012), we use a number of

other measures to validate our findings, including the price–cost margin,

the return on equity, and the profit margin.

We find robust and consistent evidence for a markup premium of

importers in the different specifications and with different measures. We

present a simple variable markups model following Melitz and Ottaviano

(2008) and Antoniades (2015) to incorporate the possible channels that

might drive this relationship. One channel is the self-selection of more

physically productive firms into importing under a fixed cost of importing

(such a fixed cost is also assumed in the outsourcing literature; see Antras,

2015). A second channel is the access to a larger variety of intermediate

inputs, which can increase the firm’s physical productivity (Halpern et

al., 2015). Finally, impor ting high-quality intermediate inputs might help

firms to upgrade their quality level (Kugler and Verhoogen, 2009, 2012;

Atkin et al., 2015). If increased quality rotates out the demand curve,

then importing firms face a less elastic demand and will charge higher

markups.

Our database does not allow us to quantify the importance of these

mechanisms directly, and we are only able to provide indirect evidence

about the role of different channels. Our first approach compares premia

estimated from cross-sectional and panel regressions. The cross-sectional

importer premium should include all three channels while within-firm and

event study evidence might isolate the two latter channels. In terms of

our preferred markup measure, we find a cross-sectional premium of about

4–5 percent compared to 2–2.5 percent within-firm. Therefore, self-selection

can explain about half of the cross-sectional premium.

We also present evidence for the relevance of the quality channel,

using a similar strategy to Bas and Strauss-Kahn (2015). In particular,

we show that the importer markup premium is larger when the imported

intermediates arrive from developed countries, which are likely to specialize

in higher-quality intermediate good production. We also use finely

disaggregated trade data to demonstrate that when a firm starts to import

intermediate products, the measured quality (price-adjusted revenue) of

its exported products increases. Similarly to the markup premium, this

C

The editors of The Scandinavian Journal of Economics 2018.