Explaining Export Duration in Kenya

• Author: Eliud Moyi,Kamau Gathiaka,Majune Kraido Socrates
• Date: 01 June 2020
• Published date: 01 June 2020
• DOI: http://doi.org/10.1111/saje.12243

South African Journal of Economics Vol. 88:2 June 2020

doi: 10.1111/saje.12243

204

© 2020 Economic Society of South Africa

EXPLAINING EXPORT DURATION IN KENYA

MAJUNE KRAIDO SOCRATES*† , ELIUD MOYI‡ AND KAMAU GATHIAKA†

Abstract

This study establishes the hazard rate of exports from Kenya and identifies factors that explain

the duration of exports using a discrete-time random effects logit regression model. A difference-

in-differences estimator is used to assess the effects of AGOA. Export data between Kenya and

176 partners over 21 years (1995–2016) is used. We find that first-year survival rate is 39%.

The median duration of Kenya’s exports is 1 year. AGOA enhances export survival, especially for

apparels. COMESA also increases export survival but EAC has a dampening effect, even in SSA

region. Differentiated products unlike capital-intensive products improve export survival.

JEL Classification: F14, F15, C35, C41

Keywords: Export duration, export survival, intensive margin of trade, discrete-time models

1. INTRODUCTION

Trade survival is a relatively new concept in international trade literature whose interest is

rapidly growing. Mainstream international trade theory1 assumes that trade will persist

once established and for this reason often focuses on either trade creation or extensive

margin. Extensive margin is where a country expands exports by introducing a new prod-

uct in a new market, introducing an existing product in a new market or introducing a

new product in an existing market (Fugazza and Molina, 2016). In contrast, intensive

margin2 entails maintaining and increasing existing exports with existing partners, hence

export survival. Evidence suggests that developing countries perform well at the extensive

margin than at the intensive margin (Besedeš and Prusa, 2011; Brenton et al., 2012).

Their poor performance at the intensive margin makes them poor exporters than devel-

oped countries (Brenton et al., 2012).

Export survival is the consecutive number of years/months in which a country exports

nonzero values of a product to a trading partner (Brenton et al., 2012). It was first dis-

covered by Besedeš and Prusa (2006a; 2006b) and Sabuhoro, Larue and Gervais (2006)

in the context of international trade. Afterwards, a strand of product-level and lately

firm-level studies have emerged. They affirm that trade is short-lived in most countries

thereby justifying the need to study export survival, especially in liberalised economies

1 These are traditional theories of Absolute Advantage, Comparative Advantage and

Hecksher-Ohlin.

2 Definition of intensive margin is borrowed from Besedeš and Prusa (2006a) and Brenton et al.

(2012).

* Corresponding author: School of Economics, Main Campus, University of Nairobi, P.O. Box

30197,00100, Nairobi, Kenya. E-mail: mkraidosocrates@gmail.com

† School of Economics, Main Campus, University of Nairobi

‡ School of Economics, Middle Campus, University of Cape Town

South African Journal

of Economics

205South African Journal of Economics Vol. 88:2 June 2020

© 2020 Economic Society of South Africa

like Kenya. From a policymaking perspective, it is important to determine export dura-

tion and to explore factors that influence survival of exports. Once these factors are iden-

tified, policy interventions can be used to improve the conditions of potential exporters.

Raising export survival rates can also deepen existing relationships thereby enhancing

export growth. Therefore, raising export survival rates in Kenya to the levels observed in

other regions can produce fairly large increases in exports over the long run (Besedeš and

Prusa, 2011). Since studies have established a strong positive association between export

development, especially for manufacturers and accelerated growth in incomes, longer

export survival is likely to trigger economic growth gains.

Different cohorts of export relationships between Kenya and sub-Saharan Africa

(SSA), and the World,3 suggest that survival of Kenyan exports is low and declining with

the length of spell (see Table 1). Longer spells have a higher first-year and end-year sur-

vival rates indicating the importance of exporting experience. On average, 39.5% and

36% of exports from Kenya to SSA and the World, respectively, survive beyond their first

year. These rates are close to those established by export survival studies in Kenya which

range between 20% and 48% (Cebeci et al., 2012; Kinuthia, 2014; Fernandes et al.,

2016; Chacha and Edwards, 2017). The average export survival at the end of a spell is

26.5% for exports to SSA and 23% to the World. However, this rate can fall to as low as

10% by the thirteenth year (Kinuthia, 2014).

What explains this trend? A review of previous work in Section 3 identifies various

determinants of survival including product-specific factors (homogeneous or differenti-

ated) and exporter/importer-specific factors (market size, distance, trade agreements, ex-

perience, language, colonial history, exchange rates, fixed and sunk entry costs, quality of

institutions, value chain addition, servitisation, time zone4 and financial development).

Although previous studies have assessed the influence of regional trading blocs and recip-

rocal agreements on trade survival, non-reciprocal agreements (such as AGOA) have not

been analysed in Kenya.

3 SSA is chosen because it is the main market of exports from Kenya. Comparison is made with

the world which comprises of other countries that are major and minor export markets for Kenya.

4 Bista and Tomasik (2017) established that time zones have no effect on the intensive margin but

negatively affect the extensive margin.

Table 1. Duration of Kenyan exports to Sub-Saharan Africa and the world (mirror data of

HS-6 digit codes)

Spell

Sub-Saharan Africa World

Period (years)

Survival rate after 1st

year (%)

Survival rate at end

period (%)

Survival rate after 1st

year (%)

Survival rate at end

period (%)

1997–2016 20 47 32 42 26

2000–2016 17 40 26 35 21

2006–2016 11 39 24 35 21

2010–2016 7 32 24 33 24

Source: World Integrated Trade Solutions (2019).