Firm‐level export duration: The importance of market‐specific ownership linkages

• Date: 01 May 2020
• Author: Črt Kostevc,Katja Zajc Kejžar
• Published date: 01 May 2020
• DOI: http://doi.org/10.1111/twec.12939

World Econ. 2020;43:1277–1308. wileyonlinelibrary.com/journal/twec

|

1277

© 2020 John Wiley & Sons Ltd

Received: 12 December 2019

|

Accepted: 2 January 2020

DOI: 10.1111/twec.12939

SPECIAL ISSUE ARTICLE

Firm-level export duration: The importance of

market-specific ownership linkages

ČrtKostevc1,2

|

KatjaZajc Kejžar1

1School of Economics and Business, University of Ljubljana, Ljubljana, Slovenia

2VIVES, KU Leuven, Leuven, Belgium

KEYWORDS

export duration, FDI, production networks, product–market spells, sunk costs, survival

1

|

INTRODUCTION

During the recent global economic and financial crisis and its immediate aftermath, export success

has come to play an even greater role in the economic well-being of nations. With sagging domestic

demand, most economies on the EU's periphery have needed to rely on exports to boost their overall

economic performance. Continuing exporting success has hence been counted on to supplement drops

in domestic demand. While new exporters continually enter foreign markets, their initial contribution

to export volume is generally small, with incumbent exporters accounting for the vast majority of

export turnover. In particular, inexperienced exporters appear to be less equipped to deal with the un-

certainties of exporting, such as those associated with search costs, contract enforcement and access

to financing. The duration of export-market relationships is thus a key factor impacting the long-run

export performance of countries.

The issue of the prevalence of short-lived trade flows was brought to the attention of empirical

research relatively recently by Besedeš and Prusa (2006a, 2006b, 2007) and Brenton and Newfarmer

(2007). This line of research highlights the importance of the sustainability margin of exporting on top

of the traditional margins: intensive (growth in volume) and extensive (diversification of products and

markets). Since then, several studies confirm the very low survival rates of new trade flows using very

diverse datasets (Besedeš & Prusa, 2006a on the USA, Nitsch, 2009 on Germany, Hess & Persson,

2011 on the EU, Besedeš & Blyde, 2010 on Latin America, countries at different stages of develop-

ment, Brenton, Saborowski, & Von Uexkull, 2011; Fugazza & Molina, 2016; Besedeš & Prusa, 2011).

Recently, empirical studies have emerged that link longevity in export markets to participation

in global value chains (GVC). Obashi (2010) shows that the trade spells of East-Asian nations at the

product level are more stable for intermediate inputs than for final goods. Compared to machine-fin-

ished products, machinery parts and components are traded in longer-lasting and more stable rela-

tionships among East-Asian countries. Córcoles, Díaz-Mora, and Gandoy (2014) focus on automotive

industries and find the level of integration in international-scale networks reduces the risk of trade

interruptions. Córcoles, Díaz-Mora, and Gandoy (2015) further confirm that exports associated with

1278

|

KOSTEVC and ZaJC KEJŽaR

global production networks are more stable and enjoy higher survival rates than goods destined for

the consumer market.

The above-mentioned studies are based on aggregate country-product-level data where GVC par-

ticipation is determined based on the distinction between final goods and parts or components within

specific industries. To the best of our knowledge, there are no firm-level studies that explicitly account

for supply-chain trade status either through contract-based outsourcing relationships that give rise to

arm's length type of supply-chain trade or vertical integration that results in intra-firm trade trans-

actions. This paper aims to bridge two important empirical gaps in the literature: First, by exploring

transaction-level data (i.e. firm–product–destination) in analysing the duration of firm-specific trade

spells, and second, by complementing transaction-level trade data with information on firms' inward

and outward foreign ownership stakes. The former allows us to avoid the trap of aggregating a number

of issues pertinent to trade between firms (length of individual firm trade spells, firm-level determi-

nants of trade duration, substitutability of export spells at the firm level, etc.), while the latter permits

us to explore the ways in which market-specific ownership ties and vertical integration in international

production networks influence the longevity of firm–market–product-specific trade in a more direct

manner than has been done in the literature so far.

Several theoretical arguments in favour of longer duration of the export arrangements of firms with

ownership linkages to the destination/origin markets have been proposed in the literature. One rests

on Békés and Muraközy's (2012) findings on the exporter's endogenous choice between variable- and

sunk-cost trade technologies. The likelihood of enduring trade spells is higher for exporters opting for

sunk-cost trade technology characterised by an initial sunk-cost investment leading to subsequently

lower variable trade costs, which is typically the case for firms with outward and inward FDI in a

certain partner country.

Similarly, the positive impact of participation in global production chains and foreign ownership

on export survival may result from the relative size of sunk market-entry costs versus annual fixed

costs of exporting. In the exporter dynamics model by Albornoz, Fanelli, and Hallak (2016), the prob-

ability of surviving in an export market increases along with the ratio of the sunk to fixed export cost.1

The lower export-termination hazard for vertically integrated firms would therefore imply that verti-

cally integrated exporters in a certain market face a higher ratio of sunk to ffixed exporting costs

compared to non-integrated exporters. This is in line with Grossman and Helpman's (2002) notion of

fixed search costs associated with outsourcing where non-integrated buyers need to search for a suit-

able supplier and incur fixed search costs, suggesting lower fixed costs for vertical integration. In

choosing the optimal way to organise production firms, therefore weigh the costs of running a larger

and less specialised organisation against the costs associated with search frictions and imperfect con-

tracting in the open market.

Finally, in Grossman, Helpman, and Szeidl (2006) the use of complex strategies involving a mix

of FDI and exports sees a multi-product firm's exports positively correlated with FDI if there are

horizontal or vertical complementarities across product lines. The choice of integration strategy and

volume/composition of trade is dictated by fixed and variable production costs, idiosyncratic firm

productivity and relative size of the markets.

To test the role of the ownership linkages and trade complementarities for export duration, we

adopt the survival analysis approach on transaction-level data for the population of Slovenian firms in

the 2002–11 period, matched with detailed origin/direction of inward/outward FDI information and

1 A key finding is that the idiosyncratic market-profitability parameter does not affect the probability of survival upon entry as

firms compensate a higher ability to make profits in a specific market with a lower entry and exit value of the general

profitability process.

|

1279

KOSTEVC and ZaJC KEJŽaR

the firms' balance sheets. As an open CEE economy whose companies are extensively involved in

production networks,2 we believe Slovenia provides a suitable setting for a study of export duration.

The period considered in the survival analysis further supports testing whether and how the role of

FDI linkages changed during the last crisis period.

We find strong support for the positive role of ownership ties (either inward or outward) with

the exporting market on the duration of exporting spells. While certainly not conclusive, this offers

support to the notion that participation in production networks stabilises firm trade flows. The FDI's

export persistence effect is robust to several checks except for inward FDI through M&A where we

are not able to rule out the contribution of selection effect of MNEs acquiring or merging with those

incumbents predisposed to more persistent supply of intermediates. Further, we show that export-ter-

mination risk decreases with an increasing share of a particular product in a firm's exports, indicating

the importance of firms focusing on their core competencies for export survival. Lastly, we detect sev-

eral dimensions of positive export complementarities both across export products, export markets and

importing activity. Knowledge about the export market and a deeper insight into product performance

greatly contribute towards more durable product–destination-specific export spells.

The rest of the paper is organised as follows. Section 2 provides a literature review summarising

the factors and evidence of the persistence of firm export performance and adjustments to the export

product–market mix. Section 3 describes the transaction-level data and provides descriptive statistics

on export duration. In Section 4, we present empirical methodology for the survival analysis of firm

product–market export spells. Section 5 shows the estimates and discusses the results of the export sur-

vival specifications while Section 6 provides some robustness checks. Section 7 concludes the paper.

2

|

RELATED LITERATURE

Since the late 1990s, improved access to firm-level data has sparked a flurry of research looking into

the performance of exporting firms. The early studies primarily focused on the observed productivity

gap between exporters and non-exporters (Bernard & Jensen, 1997, 1999; Bernard & Wagner, 1997;

Clerides, Lach, & Tybout, 1998), finding that: (a) exporters are in the minority; (b) they tend to be

more productive and larger; and (c) they tend to export only a small fraction of their output. Later

research found corroborating evidence for a variety of different country datasets, at the same time

providing far more detail on the fundamental differences between exporters and firms that only sell

locally (see Greenaway & Kneller, 2007; Wagner, 2007, for surveys). Exporters were found to be

more innovative, more capital-intensive in production, to pay higher wages, have a better employee

skill structure, be less financially constrained, etc.

Another empirical regularity across a number of very diverse firm-level datasets has been the per-

sistence of exporter status (Andersson & Lööf, 2009; Bernard & Jensen, 2004). This fact is attributed

to the existence of substantial sunk costs of exporting, learning-by-exporting and/or firm heterogene-

ity (Roberts & Tybout, 1997; Timoshenko, 2015). Despite fast-growing empirical literature based on

different country datasets, evidence of firm-duration patterns in specific foreign markets with distinct

products remains scarce.

This is even more surprising given that the little evidence that exists on export duration suggests that

average export spells tend to be exceedingly short with very low initial survival rates for new exporters.

2 According to the WTO (2016), Slovenia is classified among the high-GVC (global value chain) participation economies

with a recorded GVC participation index of 58.7 in 2011, which is significantly above the average value for developed and

developing countries, that is 48.6 and 48.0, respectively, mostly on account of its strong backward participation.