Page 67 - 臺大管理論叢第33卷第1期

NTU Management Review Vol. 33 No. 1 Apr. 2023

modularity”. We firstly define task specificity as a focal activity performing only a
discrete, independent, and specific task at the offshore location (Contractor et al., 2010).
We then define project modularity as a focal activity performing high interdependence
across tasks (Baldwin and Henkel, 2015) and involving several tasks in an entire process

(Gooris and Peeters, 2016). To separate task specificity from project modularity, we
follow Elia et al. (2019) to use one item from the ORN survey as the conditional item:
Does/did this implementation involve one or more discrete tasks or entire processes?

This item offers binary results – Yes or No. If the answer to this item is “No,” the value
of task specificity is calculated based on the Likert 5-point scale score of the ORN item:
“The implementation requires personnel with company-specific knowledge involving
routines, procedures, products and services.” In contrast, if the answer to the conditional
item is “Yes,” the value of project modularity is calculated based on the score of another

ORN item: “Loss of synergy across firm activities.” According to Elia et al. (2019), the
low value of losing synergy across firm activities means high project modularity. For the
purpose of measurement, we reverse the coding of this item.

Control variables We include the control variables reflecting location, firm, industry,
and project characteristics in this study. First, we control the locational characteristics
such as “Turnover,” “Geographic proximity,” “Co-location (with a manufacturing plant),”
“Access to local market,” “Governance incentive,” “Cultural similarity,” and “Language
similarity” (Gooris and Peeters, 2016). These seven variables are measured by the ORN

survey questions using a Likert 5-point scale. We then control firm characteristics such
as “Nationality,” “Firm experience,” “Firm size;” we take a focal firm’s nationality into
consideration because several studies have found that firms from different nationalities

presented different offshoring patterns (Massini et al., 2010). We use a dummy variable
to measure the nationality of a focal firm. Specifically, the value equals 1 if a focal firm’s
nationality is US; otherwise 0. We control a focal firm’s past experience because recent
research has also found that a firm’s experience could affect its choice of offshore location
(Larsen et al., 2013). We measure firm experience by calculating the number of years

from the time of the focal firm’s first launch to the year of the focal project outsourced.
Since big firms have more resources to manage offshoring activities in different countries
(Massini et al., 2010), we control firm size by measuring the logarithm of the focal firm’s

employees in the home country. Besides, we also control “Industry” in this study since

59