Intermediate Input Imports, Domestic Input Use and Firm-level Outcomes: Evidence from Survey Data

Published date01 August 2020
Date01 August 2020
Subject MatterArticles
05FTR920467_ncx.indd Article
Intermediate Input
Foreign Trade Review
55(3) 320–336, 2020
Imports, Domestic
© 2020 Indian Institute of
Foreign Trade
Input Use and Firm-
Reprints and permissions:
level Outcomes:
DOI: 10.1177/0015732520920467
Evidence from
Survey Data
Suryadipta Roy1
Imported intermediate inputs, that is, parts and materials sourced from abroad
and used to make products either consumed domestically or in producing
exported goods are a growing force in world trade. We test for the relative
effect of imported intermediate inputs and domestic inputs in promoting foreign
exports and various forms of domestic sales, using firm-level survey data. Imported
intermediate inputs are found to be associated with higher overall sales, foreign
exports and larger sales to multinational companies domiciled in the home
country. On the other hand, domestic inputs are not found to have statistically
significant positive effect on any firm-level outcomes. Since exporting firms are
usually more productive than domestic firms, the results point towards the
salience of outsourcing inputs in supporting firm productivity and the importance
of policymaking in facilitating trade in intermediate inputs across countries.
JEL: F10, F12, F23
Intermediate imported input, domestic inputs, foreign exports, domestic sales,
firm-level survey data
The global economy has experienced substantial fragmentation of production
activities via trade among affiliated parties or through arms-length trade in the
1 Phillips School of Business, High Point University, High Point, NC, USA.
Corresponding author:
Suryadipta Roy, Phillips School of Business, 210, High Point University, High Point, NC 27262, USA.

Roy 321
recent years. Following massive improvements in information and communications
technology, firms are increasingly able to source production inputs away from their
home countries. This trade in parts, components and accessories encourages the
specialization of different economies, leading to ‘trade in tasks’ that adds value
along the production chain. Recent evidences point towards the benefit of imported
or foreign intermediate inputs for the growth and development of economies (e.g.,
Amiti & Konings, 2007; Bernard, Jensen, Redding, & Schott, 2007; Halpern,
Koren, & Szeidl, 2015; Seker, 2012). An explanation has been that embedded
technology or quality transmitted via imported intermediate inputs contributes to
improvements in firm-level productivity and increases export competitiveness.
Models of firm-level heterogeneity predict that within sectors, more productive
firms source their inputs from foreign countries. This has been made possible due to
trade liberalization and lowering of input tariffs that, in turn, affect productivity.
Existing empirical literature in this area has mainly studied the role of intermediate
inputs on firm-level outcomes for producers belonging to a single country. For
instance, Bas (2012) exploits episodes of trade liberalization from Argentina to
draw a causal effect of imported inputs on firm-level exports. Similarly, Bas and
Strauss-Kahn (2014) provide empirical evidence on the direct and indirect channels
via which importing more varieties of intermediate inputs increases export scope for
a sample of French firms. Feng, Li, and Swenson (2016) document the positive
effect of imported intermediate inputs on exports for Chinese manufacturing firms.
Aristei, Castellaniy, and Franco (2013) investigate the existence of a two-way
relationship between importing and exporting for a sample of 1,085 firms belonging
to the East European and Central Asian republics. Their results suggest that greater
importing activities by firms lead to higher exports.
In contrast to the above-mentioned studies, this article uses production and
input data for firms belonging to countries with varying levels of trade liberalization
and economic development, to study the effect of imported intermediate inputs and
domestically sourced inputs on overall sales, direct exports and domestic market
sales. Our baseline regression results are based on data for a sample of over 17,000
firms from a maximum of 59 countries spanning 23 different sectors over the
period 2001–2005. Intermediate input imports by firms are found to be associated
with higher total sales, direct exports, domestic sales to parent/affiliate firms and
domestic sales to other firms employing more than 300 workers (large domestic
firms). We do not find any statistically significant positive effect of domestically
sourced inputs on any firm-level outcome. Instead, greater use of domestic input
is associated with significantly lower volume of total sales, domestic sales to state-
owned enterprises, domestic sales to multinational enterprises (MNE) and large
domestic firms. Results from an instrumental variable (IV) specification where we
use firm-specific differences in fixed trade costs to instrument for changes in input
use mostly support these results from the baseline model. Based on the IV model,
on average, 1 per cent increase in the use of imported inputs is found to generate
more than 10 per cent increase in direct exports, and approximately 5 per cent
increase in total sales and domestic MNE sales. Similar increase in the use of
domestic input is found to reduce direct exports by 10 per cent and total sales and
domestic sales to MNE-s by 5 per cent.

Foreign Trade Review 55(3)
The article adds to the empirical literature on the significance of foreign
intermediate inputs in increasing firm’s output and promoting greater exports. The
article is motivated by the following question: if access to foreign inputs is
associated with higher firm-level productivity, does it also help firms to register
greater domestic sales? Moreover, we are also interested in exploring the dif-
ference between the effect of imported inputs and domestic inputs on firm
performance. Existing literature has highlighted three main channels through
which foreign intermediate inputs resulting from trade liberalization affect firm
performance: (a) reduction of factor costs of production; (b) access to new
imported input varieties; and (c) access to better-quality inputs. Since greater
availability of imported inputs increase firm productivity, it should help firms
to increase foreign exports as well as to increase sales in the domestic market.
Our findings on the positive effect of imported inputs on firm-level exports are in
line with the exiting literature (see, e.g., Bas, 2012; Bas & Strauss-Kahn, 2014;
Feng et al., 2016). However, while the primary focus of these articles has been on
the role of imported inputs in promoting greater exports, the current article also
investigates the effect of imported inputs on domestic sales by firms. We also
document substantial heterogeneity in the effect of imported intermediate inputs
on various components of domestic sales in contrast to Bas (2012), who found
input trade liberalization in Argentina to be accompanied by higher domestic
sales. For instance, we do not find any statistically significant effect of imported
inputs on domestic sales to the government sector or to state-owned enterprises.
Moreover, the positive effect of intermediate imported inputs on domestic sales to
parent/affiliate firms and large domestic firms based on the baseline model is not
found to be robust in the IV model. Comparing the effect of foreign inputs and
domestic inputs, our regression results consistently support the primacy of foreign
intermediate inputs over domestically sourced inputs in promoting firm-level
productivity and output.
Data Sources and Summary Statistics
Data on input use, domestic and foreign sales by firms have been obtained from
the Productivity and Investment Climate Survey (PICS) database, which is a
precursor to the World Bank Enterprise Surveys (WBES) data sets available from
the World Bank. The database comprises responses by a cross-section of more
than 40,000 firms, mostly from developing and transition countries, covering the
manufacturing, services, agribusiness and construction industry between 2001
and 2005. The sampling methodology for the PICS is stratified random sampling
where the strata are firm size, business sector and geographic region within a
country. Stratification by firm size divides the population of firms into three strata:
small firms (5–19 employees), medium firms (20–99 employees) and large firms
(100 or more employees). The survey covers a broad range of business environment
topics, including information about firm size, sector, geographic location and
performance measures. While some countries have been included in more than

Roy 323
one round of the survey, the data set is a cross-section of firms across industries
located in different countries.
Summary statistics for all variables used in the study are reported in Table...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT