1 s2.0 S0969593123000367 Main

International Business Review xxx (xxxx) xxx
Contents lists available at ScienceDirect
International Business Review

journal homepage: www.elsevier.com/locate/ibusrev
Responses of FDI to geopolitical risks: The role of governance, information,

and technology☆
Adrien Bussy a, Huanhuan Zheng b, *
a
Swiss National Bank, Börsenstrasse 15, 8022 Zürich, Switzerland
b
Lee Kuan Yew School of Public Policy, National University of Singapore, 469C Bukit Timah Rd, 259772, Singapore
A R T I C L E I N F O A B S T R A C T
Keywords: We show that rising geopolitical risk and geopolitical uncertainty (GPR) deter foreign direct investment (FDI).
Geopolitical risk We further explore the role of governance, information, and technology in shaping the responses of FDI to GPR
Geopolitical uncertainty and document three main findings. First, effective governance in the destination market safeguards FDI against
FDI
GPR. Second, multinational corporations with better information gained from closer geographic, cultural, and
Governance
Policy
commercial ties choose to delay FDI in response to rising GPR instead of managing it. Third, FDI in more R&D-
Tie intensive industries is more resilient to GPR because intangible technology can more easily be transferred across
Technology borders.
Information
1. Introduction In response to the changing world order, do MNCs incorporate their

perceptions of geopolitical risk and uncertainty in their investment de
Political risk—“the risk that a sovereign host government will un cisions on top of general political risk? We seek to answer this question
expectedly change the rules of the game under which businesses oper by leveraging Caldara and Iacoviello (2022)’s GPR index, which cap
ate,” according to Butler and Joaquin (1998)—is undoubtedly an tures the international community’s perceptions of geopolitical risk and
important determinant of foreign direct investment (FDI) worldwide uncertainty “associated with wars, terrorist acts, and tensions between
(Busse & Hefeker, 2007; Click & Weiner, 2010; Darendeli & Hill, 2016; states that affect the normal and peaceful course of international re
Filippaios et al., 2019; Henisz & Zelner, 2005; Jensen, 2008). Not only lations” (Caldara & Iacoviello, 2022, p. 2), for 19 emerging markets
the level of political risk but also the perceptions of political risk affect (EMs).1
how multinational corporations (MNCs) organize their production net Drawing on the insights of Knight (1921, Chapter 7) and Pástor and
works (Giambona et al., 2017; Kobrin, 1979). There are different types Veronesi (2013), geopolitical risk is associated with a series of possible
of political risks, such as political instability, electoral uncertainty, known outcomes governed by a prior probability distribution (e.g.,
corruption, and geopolitical conflicts. The importance of each type to whether a war starts), and geopolitical uncertainty is associated with
investors has evolved over time and across countries. Recent geopolitical unpredictable outcomes with unknown probabilities (e.g., potential
events, such as Brexit and US-China tensions, have dominated the policy actions in response to the war that affect business). Geopolitical
headlines. As the frequency and significance of geopolitical conflicts risk encompasses political risk of an international and geographic na
escalate worldwide (Moyo, 2019), MNCs revise their perceptions and ture, but differs from political risk of domestic origin. Dealing with
change their investment behavior accordingly (Witt, 2019). geopolitical risk requires reconciliation and cooperation among multiple
Abbreviations: FDI, Foreign direct investment; GPR, Geopolitical risk and uncertainty; EM, Emerging market; MNC, Multinational corporation; R&D, Research and
development.
☆
We would like to thank the Senior Editor, Axele Giroud, and three anonymous reviewers for their helpful comments and suggestions over four rounds of revisions.
Donglai Luo and Mehtab Ahmed Jagil provided excellent research assistance. Financial support from an NUS startup grant and MOE Tier 1 grant is gratefully
acknowledged. The views expressed in this paper are those of the authors and do not necessarily reflect those of the Swiss National Bank.
* Corresponding author.
E-mail address: [email protected] (H. Zheng).
1
For alternative definitions of geopolitical risk and related discussion, please see Gray and Sloan (2014).
https://doi.org/10.1016/j.ibusrev.2023.102136
Received 23 May 2021; Received in revised form 9 March 2023; Accepted 20 March 2023
0969-5931/© 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
Please cite this article as: Adrien Bussy, Huanhuan Zheng, International Business Review, https://doi.org/10.1016/j.ibusrev.2023.102136
A. Bussy and H. Zheng International Business Review xxx (xxxx) xxx
parties instead of independent action by a local government, as in many investment. This implies that when geopolitical risk increases, MNCs
domestic political risk situations. Prior studies have typically explored would not cut FDI so long as returns are high enough. However, when
the impacts of specific instances of realized adverse geopolitical events, geopolitical uncertainty rises, MNCs should postpone FDI even if ex
often measured by binary variables—for example, interstate military pected returns are high and delay investing until the situation is clearer.
conflicts (Li, 2008; Li & Vashchilko, 2010) and international political Our empirical analysis shows that FDI drops in response to a rise in the
tensions (Gao et al., 2018; Kastner, 2007). However, geopolitical events GPR index even in markets with high economic growth that offer high
are diverse and threaten international business even if they do not returns to FDI, which suggests that geopolitical uncertainty is the main
materialize; after all, investment decisions are not only based on the driver of falling FDI.
realizations, but also on the perceptions of risk (Kobrin, 1979). For After establishing the negative impact of GPR on FDI, we proceed to
MNCs to make far-sighted and long-term investment decisions, they explore how different characteristics of the destination market, invest
have to update their beliefs about the probability that geopolitical ment itself, and the information set of investors, reshape MNCs’
conflicts will materialize (geopolitical risk) and the consequences of response to GPR. Good governance attracts FDI (see Beazer & Blake,
ever-changing geopolitical environments (geopolitical uncertainty) for 2018; Busse & Hefeker, 2007; Daude & Stein, 2007; Henisz, 2000;
their international business. It may be too costly and too late to retain Henisz & Zelner, 2005, among many others), and is associated with
existing business after the realization of geopolitical conflicts, since most effective and consistent policymaking (Busse & Hefeker, 2007), which
FDI is difficult to reverse. However, to the best of our knowledge, no cannot be easily influenced even in the presence of significant external
study has systemically investigated how MNCs allocate FDI in response shocks, such as adverse geopolitical events. Effective and high-quality
to perceptions of geopolitical risk and uncertainty, possibly due to the governance in the destination market is thus expected to reduce the
difficulty of measuring perceptions (Giambona, Graham, & Har policy uncertainty that accompanies volatile GPR by insulating policy
vey2017). Caldara and Iacoviello (2022)’s GPR index, which measures making from geopolitical pressures. Furthermore, a rapid and effective
perceptions of geopolitical risk and uncertainty, enables us to fill this response may not only shelter the private sector, but also facilitate the
gap in the literature. conflict resolution process (Sriram, Large, & Brown, 2009) and prevent
Caldara and Iacoviello (2022) construct their GPR index using the further escalation or conflict relapse (Hegre & Nygård, 2015). The
methodology of Baker, Bloom, and Davis (2016) based on textual smoother the resolution process, the sooner the uncertainty decreases
analysis of news coverage by leading newspapers of a wide range of and investment picks up (Colino, 2013). Thus, we expect effective
geopolitical events. The GPR index captures perceptions of both risk and governance in the destination market to mitigate the negative effect of
uncertainty in the context of acts and threats of geopolitical nature. A GPR on FDI, and find that effective governance and quality regulation
rich literature has documented how risk and uncertainty affect FDI. For safeguard FDI against GPR. Our findings are consistent with studies that
instance, risk increases transaction costs (Cuypers et al., 2021); distorts establish the role of good governance in mitigating political risks that
resource allocation (Col, Durnev, & Molchanov2018); and lowers prof are not geopolitical in nature (Jensen, 2008; Julio & Yook, 2016).
itability (Dunning, 2004). Uncertainty about economic policy (Baker In addition, we document that geographical, cultural, and economic
et al., 2016; Husted, Rogers, & Sun 2020); political elections (Cordero & proximity between source and destination markets, which allow MNCs
Miller, 2019; Jens, 2017; Julio & Yook, 2016); and legitimacy (Chan & to gain additional information about the host market (Björkman &
Makino, 2007; Darendeli & Hill, 2016; Henisz & Zelner, 2005); among Piekkari, 2009; Eden & Miller, 2004; Keller & Yeaple, 2013; Zaheer,
other factors, change the rules of the game, which require that MNCs 1995), exacerbate the negative effect of GPR on FDI. This is because
adapt how they play the game (Darendeli & Hill, 2016; Williamson, informed MNCs prefer to avoid markets that are plagued by geopolitical
1998). The GPR index incorporates perceptions of both geopolitical risk tensions due to existing uncertainty, rather than leverage their infor
and uncertainty but does not differentiate between them. We argue mation advantage to manage the risks created by these tensions. This
theoretically and show empirically that, relative to existing political risk result is consistent with that of Pástor and Veronesi (2013), whereby
measures, the GPR index incorporates additional information that is investors require a higher risk premium when their information is more
relevant for cross-border direct investment decisions. To facilitate dis precise and they become more certain about the existence of
cussion, we refer to geopolitical risk and uncertainty hereafter as GPR, uncertainty.
unless otherwise specified. Furthermore, we find that FDI associated with more intensive
Using data for 19 EMs from January 2003 to December 2019, we find research and development (R&D) is less responsive to GPR, which sug
that rising GPR at country level deters FDI at country, industry, and gests that technology embeddedness provides a shield against GPR. We
transaction levels, even after we control for a series of well-known risks. argue that this is because MNCs can leverage their closely knit internal
This impact is more pronounced when the GPR is high and volatile, organizational linkages to efficiently transfer, integrate, and build on
which suggests that MNCs are more responsive to GPR perceptions when technologies developed in various geographic locations (Zhao, 2006),
the geopolitical environment is more turbulent. This reflects the which protects R&D-intensive FDI against negative geopolitical shocks.
growing importance of GPR in shaping international business as Given the importance of R&D-intensive FDI for economic growth,
geopolitical tensions escalate worldwide, which has been well recog destination governments also have the incentive to prioritize the safety
nized by practitioners and policymakers. Our finding is consistent with of these FDIs upon outbreaks of geopolitical confrontations, which also
evidence that (perceived) risk and uncertainty increase costs and reduce mitigates the negative impact of GPR on FDI.
efficiency (see, for example, Baker et al., 2016; Buckley et al., 2007; This study furthers our understanding of geopolitical risk and un
Cuypers et al., 2021; Dunning, 2004). Note that our results hold after certainty by extending the political risk literature to the context of
controlling for country-level characteristics at annual frequency, escalating geopolitical divisions worldwide, exploring whether percep
including general and specific political risks and other macroeconomic tions of geopolitical risk and uncertainty add new information that shifts
risks. Our findings imply that GPR perceptions captured by the MNCs’ international investment behavior, and investigating potential
text-based GPR index incorporate additional information relevant for channels to mitigate the impacts of geopolitical risk and uncertainty on
cross-border investment decisions on top of frequently used risk international business.
measures. The paper proceeds as follows. Section 2 conceptualizes and con
Although the GPR index does not differentiate geopolitical risk from textualizes geopolitical risk and uncertainty, and develops four hy
uncertainty, we seek to uncover their relative importance based on the potheses regarding the effect of GPR on FDI and the roles of good
insights of Bernanke (1983) and Pástor and Veronesi (2013): An increase governance, information, and technology in shaping this effect. Section
in risk leads to a higher required payoff for the investment to be 3 describes the data and lays out the methodology. Section 4 empirically
worthwhile, while an increase in uncertainty causes delays in tests the hypotheses formulated in Section 2, and Section 5 concludes.
2
2. Theory and hypotheses decisions from third countries that are not directly involved in the
conflict. They also fail to consider how perceptions alone could influ
In this section, we first conceptualize geopolitical risk and uncer ence business decisions even if geopolitical conflicts do not materialize
tainty and compare them with existing measures of political risks. Then, (Giambona et al., 2017).3 It is therefore important to systemically
we adapt existing theories to develop hypotheses regarding the potential analyze the impact of geopolitical risk on FDI while taking into account
impacts of geopolitical risk and uncertainty perceptions on FDI. In light their dynamic features over time and heterogeneous roles across
of these theoretical accounts, we further investigate how (i) governance different economies.
in the destination market, (ii) information available to investors, and
(iii) the technology content of FDI may affect the impact of geopolitical 2.1.2. Measuring geopolitical risk and uncertainty
risk and uncertainty on FDI. Currently, geopolitical risk is incorporated in general measures of
political risk, such as the political risk rating from the International
Country Risk Guide (ICRG), which includes both internal and external
2.1. Conceptualization of geopolitical risk and uncertainty
conflicts (Busse & Hefeker, 2007; Click & Weiner, 2010); the number of
deaths related to political violence (Witte et al., 2017); and binary in
2.1.1. Geopolitical risk versus political risk
dicators of political crises (Huang et al., 2015), among others. These
Geopolitical risk is a subcategory of political risk that stems from
measures of political risk, which are mostly available at annual fre
conflicts and tensions between countries that disrupt international re
quency, are either binary or exhibit little variation over time, because
lations.2 While highly consequential, until recently adverse geopolitical
important political processes are often characterized by rare events such
events seldom occurred and mainly occurred in only a few countries.
as revolutions, coups, or wars (King & Zeng, 2001). Although re
The landscape has since changed as numerous geopolitical tensions
alizations of significant political events are rare, the threats of their
(such as the US-China trade war) frequently hit the headlines. Indeed,
possible occurrence and the associated risk and uncertainty are more
geopolitical risk has become one of the most impactful factors for
frequent and vary continuously with market participants’ perceptions.
business in the eyes of global executives (EY, 2020). However, it is un
The subjective perception of the risk and uncertainty associated with
clear whether geopolitical risk affects business like other types of po
political events is what affects managers’ investment decisions (Giam
litical risks. Moreover, it remains ambiguous whether multinational
bona et al., 2017; Kobrin, 1979). However, measuring perceptions is
managers can use their risk management experience and skills to deal
difficult (Giambona et al., 2017).
with such risk, which is highly context-specific and not limited to a
Recent developments in text analysis have considerably improved
particular market.
measures of perceptions, especially those related to policy uncertainty
Conceptually, geopolitical risk differs from political risk of a do
(Baker et al., 2016; Hassan et al., 2019; Husted et al., 2020). The
mestic origin, such as electoral uncertainty (Jens, 2017; Julio & Yook,
monthly GPR index developed by Caldara and Iacoviello (2022) pro
2016); legitimacy and institutions (Chan & Makino, 2007; Darendeli &
vides a real-time measure of perceptions of geopolitical risk by the
Hill, 2016; Henisz & Zelner, 2005); and internal frictions and conflicts
media, investors and policy makers. It is calculated as the number of
(Shi et al., 2017). For instance, MNCs may navigate political risk of a
articles that contain references to keywords related to nuclear, war, and
domestic origin by forming alliances with domestic firms to avoid
terrorist threats and acts, among other geopolitical events, divided by the
expropriation hazard (Henisz, 2000); developing linkages with local
total number of articles published in 11 national and international
politicians to enhance their legitimacy and sustainable competitive ad
newspapers. By following the methodology of Baker et al. (2016), Cal
vantages (Sun, Mellahi, & Wright, 2012); and establishing ties with
dara and Iacoviello (2022)’s GPR index essentially captures both risk
important social actors that are independent of the ruling government in
and uncertainty associated with geopolitical events. By incorporating
order to steer clear of political transitions (Darendeli & Hill, 2016),
both geopolitical risk and uncertainty, the GPR index differs from the
among others. A natural question is whether MNCs can protect their
geopolitical risk embedded in existing measures of political risk.4
investments from geopolitical risk in similar ways. Since resolving
geopolitical risk requires reconciliation and cooperation among multiple
parties instead of independent action by a local government, MNCs face 2.2. Hypotheses development
the dilemma of building good relationships with belligerent foreign
governments at the risk of losing the favor of local governments. Even if 2.2.1. The effect of GPR on FDI
political ties did help overcome geopolitical risk, MNCs would still need Political risk is an important determinant of FDI (Busse & Hefeker,
to establish close ties with not only local governments, but also their 2007; Click & Weiner, 2010; Darendeli & Hill, 2016; Filippaios et al.,
opponents, at the risk of alienating the local government. It is therefore 2019; Henisz & Zelner, 2005; Jensen, 2008). Different aspects of polit
challenging for MNCs to navigate geopolitical risk even if they have ical risks such as corruption (Cuervo-Cazurra, 2006); terrorism (Czin
good knowledge and experience of managing other forms of political kota et al., 2010); political violence (Witte et al., 2017); and political
risks. As such, whether existing tools and strategies can be applied to party tenure (Cordero & Miller, 2019) have been found to affect FDI.
manage geopolitical risk depends on local and global contexts as well as However, empirical evidence on geopolitical risk is mixed. While Gao
the business’s nature, which calls for additional investigation. et al. (2018) show that historical conflicts discouraged Japanese FDI to
Geopolitical risk encompasses political risk of international and China due to increased transaction costs and exchange hazards, Li and
geographic nature, such as interstate military conflicts (Li, 2008; Li & Vashchilko (2010) find no statistical evidence that a dyadic military
Vashchilko, 2010) and international political tensions (Gao et al., 2018;
Kastner, 2007). Prior studies focus on realized bilateral conflicts and
leave unexplored how geopolitical conflicts could affect business 3
For example, Li and Vashchilko (2010) find no statistically significant evi
dence that a dyadic military conflict or security alliance between source and
destination countries affect bilateral FDI flows. Kastner (2007) shows that
2
Political risk can be defined as “the risk that a sovereign host government conflicting political interests between source and destination countries deter
will unexpectedly change the rules of the game under which businesses oper their trade with each other, while Gao et al. (2018) document similar evidence
ate” (Butler & Joaquin, 1998). Geopolitical risk thus stems from unexpected whereby historical animosity deters FDI.
4
changes in the rules of the game in the host country that occur as a result of Monthly variation in the GPR index enables us to explore whether the role
tensions in international relations. Examples of the difference between political of geopolitical risk and uncertainty on FDI differs from that of general political
and geopolitical risk in the cases of Ukraine and Hong Kong are presented in risk, macroeconomic risk, or other slow-moving factors that are often recorded
supplementary materials and available upon request. at annual frequency by controlling for country-year fixed effects.
3
conflict or security alliance affect FDI.5 While these studies focus on Hypothesis 1b. Economic growth in the destination market weakens
realized geopolitical conflicts and examine their impacts on FDI between the negative effect of GPR on FDI.
countries, MNCs’ perceptions of risk could be more important than the
risk per se (Giambona et al., 2017). Perceptions of uncertainty are 2.2.2. Can governance safeguard FDI against GPR?
equally important, given that the consequences of geopolitical conflicts Rising GPR is often accompanied by higher policy uncertainty
have become increasingly unpredictable. Both perceptions of geopolit (Caldara & Iacoviello, 2022). When confronted with geopolitical con
ical risk and uncertainty could also disrupt the business of MNCs not flicts, some governments are under pressure to change policy (e.g., for
directly involved in geopolitical conflicts, which is not examined by nationalism or mercantilism reasons), while others may seize the op
previous studies. We seek to fill this gap in the literature using the GPR portunity to pursue their political agenda. For example, India banned a
index from Caldara and Iacoviello (2022), which captures the interna list of Chinese apps from top Chinese technology MNCs such as Alibaba
tional community’s perceptions of geopolitical risk and uncertainty and Tencent 2 months after the troop clash on the India-China border in
across 19 EMs. June 2020. In such a scenario, MNCs were not only hit directly by the
We are interested in the incremental effects of geopolitical perceptions geopolitical event, but also indirectly by the policy changes it triggered.
on FDI, in addition to general political risks. If all information about Good governance is associated with effective and consistent policy
geopolitical risk and uncertainty perceptions embedded in the GPR (Busse & Hefeker, 2007), which cannot be easily changed even in the
index were incorporated into existing measures of political risk, the presence of significant external shocks. Good governance therefore
correlation between the change in the GPR index and FDI should no mitigates policy uncertainty following geopolitical confrontations. We
longer be statistically significant after controlling for these risks. We are interested in whether good governance in destination markets can
argue that this is unlikely for two reasons. First, as documented above, safeguard FDI against GPR.
geopolitical uncertainty, which is relatively important for decision- Many studies have documented that good governance attracts FDI
making and underexplored in the literature, is not fully captured by (Beazer & Blake, 2018; Busse & Hefeker, 2007; Daude & Stein, 2007;
existing political risk measures. Second, existing measures are unlikely Henisz, 2000; Henisz & Zelner, 2005). Julio and Yook (2016) find that
to capture changes in geopolitical paradigms that lead MNCs to update stable governments not only attract FDI directly but also mitigate the
their beliefs about the probability of geopolitical events, such as wars negative effect of political uncertainty stemming from election cycles on
and deglobalization. We therefore expect the GPR index to contain FDI- FDI. Jensen (2008) shows similar evidence that high institutional
relevant information not captured by existing measures of political risk. quality mitigates the negative effect of political risk by lowering the risk
As a result, an increase in the GPR index is expected to deter FDI after of government expropriations and contract disputes. Conditional on
controlling for political risk. This leads to our first hypothesis: rising geopolitical tensions, good governance helps to reassure investors
that such threats will not realize, or that if they do, the government
Hypothesis 1a. GPR has a negative effect on FDI.
would intervene to mitigate their effects. Oh and Oetzel (2011) have
The GPR index does not differentiate geopolitical risk from geopo shown that good governance decreases MNCs’ disinvestment following
litical uncertainty, which may affect FDI differently. In the case of risk, major disasters through effective policy response. A rapid and effective
the probability of possible outcomes is known but not the actual reali response not only shelters the private sector, but also facilitates the
zation. However, in the case of uncertainty, neither the outcomes nor conflict resolution process (Sriram et al., 2009) and prevents further
their probability distribution are known (Knight, 1921, Chapter 7). This escalation or relapse of the conflict (Hegre & Nygård, 2015).6 The
implies that MNCs may maximize profit conditional on their exposure to smoother and faster the resolution process, the sooner uncertainty de
risk, but not to uncertainty (Friberg, 2015). Although risk discourages creases and investment picks up (Colino, 2013).
investments via a risk-return tradeoff (Pástor & Veronesi, 2013), high If government policy is independent of geopolitical pressure, MNCs
uncertainty motivates firms to delay investment and especially when it would not have to direct resources away from productive activities to
is costly to undo (Bernanke, 1983; Pástor & Veronesi, 2013). We can comply with changing policy. Moreover, if MNCs can do business as
then infer that an increase in geopolitical risk leads to a higher required usual despite growing GPR, they are less likely to cut FDI. We therefore
payoff for the investment to be worthwhile, while an increase in conjecture that good governance that reduces policy uncertainty asso
geopolitical uncertainty causes delays in investment. Although risk in ciated with geopolitical conflicts mitigates the negative effect of GPR on
creases the cost of investing abroadvia resources spent risk hedging FDI. This leads to our second hypothesis:
(Click & Weiner, 2010; Col et al., 2018), as long as fast economic growth
Hypothesis 2. Good governance in the destination market weakens
generates sufficiently high return to reward the risk, it could still attract
the negative effect of GPR on FDI.
growth- and profit-seeking FDI (Buckley et al., 2007; Dunning, 2004).
Uncertainty such as drastic regulation changes may jeopardize an in
2.2.3. Does information shape MNCs’ response to GPR?
vestment project, leading to unforeseen losses, and thus MNCs would be
MNCs face substantial information asymmetry when investing in
more concerned about survival in a worst-case scenario than profit
foreign markets. We are interested in whether MNCs with better infor
ability per se. Higher return prospects from fast economic growth may
mation gained through closer ties with destination markets choose to
thus address MNCs’ concerns regarding risk, but not regarding uncer
manage or avoid the negative effect of GPR.
tainty, in which case delaying investment until the situation becomes
It is well documented that geographic proximity and cultural
clearer is the optimal strategy. We may differentiate the roles of
geopolitical risk and uncertainty by testing whether fast economic
growth affects the impact of GPR on FDI. Therefore, if geopolitical risk is
6
driving our result, we should observe that fast economic growth in Most countries make attempts to conduct peace talks after the emergence of
destination markets mitigates the negative effect of GPR on FDI. geopolitical conflicts. To resolve conflict or reach a truce, a government needs
to (i) understand the needs of its nation and the conflicting party, (ii) come up
with an effective strategy acceptable to both parties, and (iii) convince the other
party that it can deliver what is agreed upon. Incapable governments could fail
in either stage, which undermines the truce probability and may even escalate
5
In a certain subsample and specification, FDI from high-income countries to the conflict (i.e., when governments do not deliver what they had committed).
low-income countries are negatively (positively) related to the dyadic military Good governments know how to seek common ground while putting differences
conflict (security alliance) at the 10% level, which is conventionally interpreted aside for the national interest and have the ability to convince the public and
as statistically insignificant. Please see Models 3 and 4 in Table 3 of Li and deliver commitment. Thus, more credible and effective governments are more
Vashchilko (2010). likely to reach a truce and resolve geopolitical conflicts.
4
similarity between source and destination markets mitigate the infor dominates. Recalling from our previous discussion that geopolitical risk
mation barriers faced by international transactions (Björkman & Piek can be compensated with financial returns but geopolitical uncertainty
kari, 2009; Rauch & Trindade, 2002). Similar cultures and institutions is difficult to manage, we expect that geopolitical uncertainty weighs
between source and destination markets facilitate information exchange more than geopolitical risk in investment decisions. Closer geographic,
and reduce information ambiguity (Eden & Miller, 2004; Simonin, cultural, and commercial ties between source and destination markets
1999). MNCs accumulate knowledge and experience about destination enable MNCs to better access information. This leads to our third
markets over time, which enables them to overcome the liability of hypothesis:
foreignness and assess information like locals (Chan & Zheng, 2019;
Hypothesis 3. Better information that originates from closer ties be
Peng & Beamish, 2014; Zaheer, 1995). Thus, closer geographic, cultural,
tween source and destination markets strengthens the negative effect of
institutional, and commercial ties between source and destination
GPR on FDI.
markets should enable MNCs to be more informed about the GRP in
destination markets. Prior studies have shown that geographic and so
2.2.4. Can technology increase resilience to GPR?
cial proximity directly affect FDI (Arikan & Shenkar, 2013; Guiso,
MNCs play an important role in transferring technology across bor
Sapienza, & Zingales 2009; Kastner, 2007; Makino & Tsang, 2011). We
ders. Technology in our context refers to factors that could improve
further investigate how they impact the effect of GPR on FDI through the
firms’ output given existing resources, which can be measured by
information channel.7
research and development (R&D), innovations, and related character
Low information asymmetry may enable MNCs to better interpret the
istics (Samaniego, 2010). EMs have attracted substantial
nature and consequences of geopolitical events. Pástor and Veronesi
technology-intensive FDI in recent decades because of their favorable
(2013) argue that political events lead investors to revise their beliefs
policies (Guimón et al., 2018); large pools of science and engineering
about the impact of the prevailing government policy (impact shock)
talents (Stephan, Silvia, & Arie, 2008); low wages (Demirbag & Glaister,
and about possible future policy actions the government might and their
2010); and upgraded technological capacities (Liu & White, 2001).
potential consequences (political shock). According to their theory, in
Technology-intensive FDI, whose spillover effects benefit domestic firms
vestors demand equity risk premiums for bearing these shocks and the
(Javorcik, 2004), is strategically important for policymakers’ pursuit of
equity premium that compensates political shocks is termed the “polit
economic growth. When there is an outbreak of geopolitical conflicts
ical risk premium.” In our context, geopolitical risk reflects impact
that threaten business operations and other economic activities, EMs
shocks by incorporating new information to update the probability
have incentives to dedicate their limited resources and capacity to pri
distribution of various known outcomes based on current policy, while
oritize the safety and prosperity of technology-intensive FDI (Narula &
geopolitical uncertainty captures political shocks by updating percep
Dunning, 1998). Thus, the technology embeddedness of FDI may pro
tions of unknown future political actions. Having better information
vide a shield against geopolitical risk and uncertainty.
about geopolitical risk effectively reduces the impact shock and there
From MNCs’ perspective, technology generates intangible assets that
fore the required risk premium. Thus, informed MNCs should be less
can be easily transferred across borders (Castellani & Lavoratori, 2020;
responsive to rising geopolitical risk they know more about, which is less
Castellani et al., 2013). MNCs distribute their technology, such as
harmful to them than to uninformed MNCs.
research and development (R&D), worldwide and integrate it within the
However, the story may be different when MNCs have more infor
firm for global applications (see the review by Papanastassiou, Pearce, &
mation about geopolitical uncertainty. Pástor and Veronesi (2013) show
Zanfei, 2020). These closely knit internal linkages and organizations
that investors require a higher political risk premium when information
allow MNCs to efficiently transfer, integrate, and build on technologies
is more precise, because they become more certain about the political
developed in various geographic locations, which protect MNCs against
decision-making (the uncertainty). Adapting this theorem to our
negative external shocks (Zhao, 2006). When threatened by the
context, MNCs demand a higher risk premium to compensate geopolit
outbreak of a geopolitical conflict, MNCs can then use their internal
ical uncertainty when they have better information. In an extreme case
organizations to protect their R&D outputs, i.e., by transferring intan
in which MNCs are perfectly sure that the policy would change (but not
gible assets to source markets or safer locations. Reverse knowledge
how), they would not invest immediately but instead wait until after the
transfer from EMs to source markets can happen in the initial stage of
policy change to make a decision. Thus, given the same investment
idea generation, the technological innovation phase, or the commer
returns, MNCs that are more informed about geopolitical uncertainty are
cialization phase (Von Zedtwitz et al., 2015). Technology-intensive FDI
less likely to invest. Precise information also enables investors to make
is associated with more R&D and intangible assets than is
informed choices and to act confidently on their information (Keller &
technology-light FDI, which should benefit more from MNCs’ interna
Yeaple, 2013; Zheng, 2020). When facing high uncertainty, it is optimal
tional organization of technology activities. Therefore, geopolitical
for firms to delay or avoid irreversible investments until the situation
conflicts should be less threatening to more technology-intensive FDI.8
becomes less uncertain (Bernanke, 1983). Thus, informed MNCs should
This leads to our fourth hypothesis:
be more responsive to rising geopolitical uncertainty in cutting FDI—in
an extreme case, they do not invest at all when geopolitical uncertainty Hypothesis 4. The technology-intensiveness of FDI weakens the
is sufficiently high. To summarize, when facing the same increase in negative effect of GPR on FDI.
geopolitical uncertainty, informed MNCs should reduce FDI more
We summarize our hypothesized roles of governance, information,
aggressively than their uninformed peers.
and technology in shaping the impacts of GPR on FDI in Appendix
Putting the above together, better information about geopolitical risk
Table A.2.
is expected to mitigate its negative impact on FDI, while more precise
information about geopolitical uncertainty is conjectured to magnify its
negative impact on FDI. Whether lower information asymmetry
8
strengthens or weakens the negative impact of rising GPR on FDI de One concern is that technology-intensive FDI may involve substantial fixed
pends on whether geopolitical risk or geopolitical uncertainty assets such as equipment and buildings that are costly to reverse, and thus
render them more sensitive to geopolitical risk. However, we find that the ratio
of fixed assets to total assets for the most technology-intensive sector (the
Chemical industry, as measured by R&D expenditure ratio) is 21%, which is not
7
Proximity could also facilitate forging alliances with economic actors in the greater than either the median or the mean of all sectors (28% and 21%,
destination market to mitigate the effect of GPR on FDI. However, as explained respectively). Moreover, in the case of a business shutdown, the net irreversible
in Section 2.1.1, this strategy is unlikely to bear fruit due to the nature of losses for technology-intensive FDI might still be lower than for other FDI,
geopolitical risk. Our focus is thus on the information channel. whose outputs cannot easily be shifted across borders.
5
Table 1 3.1.2. Geopolitical risk and uncertainty

Summary statistics. Geopolitical risk and uncertainty are measured by the GPR index of
Panel A: Summary statistics of key variables Caldara and Iacoviello (2022). This index reflects the automated
text–search frequency of articles related to geopolitical tensions in 11
Variable N Mean SD Min Median Max
leading international newspapers such as the New York Times, the
FDI (Billion 3876 1.298 2.187 0 0.482 21.792 Financial Times, and the Wall Street Journal. The algorithm–generated
USD)
log(1 + FDI) 3876 5.951 1.957 0 6.180 9.989
index is further audited by comparing it with a human–generated GPR
GPR 3876 100.467 38.510 21.748 93.226 366.719 index, which depends on manual reading and coding of selected articles.
log(GPR) 3876 4.544 0.363 3.080 4.535 5.905 The correlation coefficient of the algorithm–generated and human
Δlog(GPR) 3876 − 0.207 30.801 − − 144.891 –generated indices is as high as 98%. The GPR index incorporates both
(%) 151.937 0.972
threats and realizations of geopolitical events. Currently, a monthly GPR
Panel B: Geopolitical risk by market index is available for 19 EMs, which is normalized to enable compara
log (GPR) △log (GPR) (%) bility across markets.
Market N Mean SD Mean SD
3.1.3. Governance
Argentina 204 4.517 0.363 0.023 38.209
Governance essentially concerns institutional quality. We obtain
Brazil 204 4.605 0.274 − 0.125 29.187
China 204 4.615 0.258 0.073 17.838 country–level measures of governance from the Worldwide Governance
Columbia 204 4.274 0.387 − 0.563 36.125 Indicators (WGI). Based on a variety of surveys, this dataset compiles the
Hong Kong 204 4.524 0.334 0.202 25.356 perceptions of firms, institutions, experts, and households of the gov
Indonesia 204 4.215 0.379 − 0.602 35.169 ernment’s capacity to effectively formulate and implement good pol
India 204 4.423 0.193 − 0.149 19.941
Israel 204 4.463 0.196 − 0.313 18.612
icies. There are six dimensions of governance in the WGI. Government
South Korea 204 4.660 0.362 − 0.516 33.321 effectiveness captures the overall quality of public policy, policy’s inde
Mexico 204 4.694 0.221 0.213 21.780 pendence from political pressure, and the credibility of the govern
Malaysia 204 4.473 0.354 − 0.346 39.435 ment’s commitment to policies. Regulatory quality reflects the
Philippines 204 4.598 0.326 − 0.414 30.663
government’s capacity to formulate and implement regulations that
Russia 204 4.627 0.270 − 0.028 23.685
Saudi Arabia 204 4.617 0.264 − 0.373 23.838 permit and promote private sector development. Control of corruption
Thailand 204 4.520 0.405 − 0.348 37.940 measures efforts dedicated to curbing the practice of trading public
Turkey 204 4.754 0.327 − 0.016 28.484 power for private gain. Rule of law assesses the level of confidence in and
Ukraine 204 4.738 0.600 0.295 35.942 abidance by the rules of society. Political Stability and Absence of
Vietnam 204 4.620 0.299 − 0.389 35.373
South Africa 204 4.394 0.326 − 0.558 39.323
Violence/Terrorism estimates the reverse likelihood of politically moti
vated instability and violence, including terrorism, with a larger value
Panel A reports the summary statistics for foreign direct investment (FDI) and corresponding to greater peace. Voice and accountability quantifies the
GPR (perceptions of geopolitical risk and uncertainty) as well as their trans
extent of freedom of speech, the media, association, and elections. These
formation for 19 emerging markets (EMs) from 2003M1 to 2019M12. Panel B
measures of governance from different perspectives offer important in
summarizes log(GPR) and its first difference △log(GPR) for each EM.
sights into governments’ capacity to come up with effective policy
making to address geopolitical risk and uncertainty. Given that they are
3. Data and methodology highly correlated, we also aggregate these measures into a composite
governance index to explore their overall role in shaping the effect of
3.1. Data source GPR on FDI.
We consolidate multiple databases in this study. Appendix Table A.1 3.1.4. Information gained from geographic, cultural, and commercial ties
presents the definition and data sources for each variable. We describe MNCs from source markets that have closer geographic, cultural, and
them briefly below. commercial ties with destination markets face lower information bar
riers and should know more about the GPR in destination markets. We
3.1.1. FDI measure information by (i) geographic ties, denoted as Contiguitys,d, a
We obtain monthly transaction–level FDI data from fDi Markets and dummy variable that equals 1 if the source and destination markets
aggregate them by destination, source, and industry. This dataset re share common borders; (ii) cultural ties denoted as CommonLanguages,d,
cords real–time cross–border investments in building new subsidiaries which equals 1 respectively if the source and destination markets share
or expanding existing ones upon their announcement or opening. It common languages; and (iii) commercial ties captured by Experiences,d,t,
tracks and verifies FDI from actual investors to the ultimate destination, which is the log of 1 plus the cumulative investment amount from the
regardless of the funding channels and investment strategies—e.g., source market to the destination market since 2003M1, as well as
through offshore markets for tax efficiency. Treatys,d,t, a dummy that equals 1 if the source and destination market
We measure aggregate FDI flows in each destination, FDId,t, as total have established a bilateral investment treaty and 0 otherwise.
investments to destination market d in period t. Bilateral FDI, FDIs,d,t, is
the sum of all investments from source market s to destination market 3.1.5. Technological characteristics
d in period t. Bilateral industry–level FDI, FDIs,d,i,t, is the investments Following the trade and macroeconomic literature, we obtain in
from source market s to destination market d in industry i in period t. If dustry measures of technological characteristics, which are relatively
there is no investment record for a given specification, the value of FDI is stable over time and across countries (Ilyina & Samaniego, 2011;
set to zero. The logarithmic transformation of these FDI measures are Samaniego, 2010; Samaniego & Sun, 2015). Our key measure of tech
our main dependent variables. We also check the robustness of our nology is R&Di, the share of R&D expenditures in total capital expen
estimation results using alternative measures of FDI. Our sample is from ditures for industry i. We also use AssetFixityi, the ratio of fixed assets to
2003M1, the earliest period available from this dataset, to 2019M12,
total assets, which reflects the extent of investments in machinery and
just before the start of COVID–19. We focus on FDI in 19 EMs for which equipment that potentially embed technology (Samaniego & Sun, 2015).
GPR data are available (see Table 1).
To check the importance of skilled labor, who possess knowledge about
production and management, we also consider SkilledLabori, the total
wage bill divided by the number of employees. This measure is higher
6
when a larger proportion of employees are highly skilled. All these We conduct most of our analysis using country-level data because the
industry–level measures of technological characteristics are from GPR index is only available at country level. If firm-level theories are
Samaniego and Sun (2015). general enough to capture the investment patterns of all (or at least the
majority) of MNCs, aggregate FDI should respond to GPR in patterns
3.2. Summary statistics similar to firm-level FDI. By connecting firm-level theory with country-
level analysis, we test the generality of existing firm-level theories.
Table 1 provides summary statistics of key variables for the full Nonetheless, we also check the robustness of our results using firm-level,
sample and each of the 19 EMs. During our sample period, in terms of industry-level, and bilateral FDI.
the average growth rate, the GPR increased the most in Ukraine (0.30%
per month or 3.54% per year) and decreased the most in Indonesia 3.3.2. The role of governance
(0.60% per month or 7.22% per year). The GPR is the most volatile in To explore how good governance impacts the effect of GPR on FDI,
South Africa (standard deviation = 39.3%) and the most stable in China we expand Eq. (1) to include the interaction between GPR and indicators
(standard deviation = 17.84%). Further examination of the time series of good governance:
reveals that GPR in China is relatively stable because of the sample yd,t = βlog(GPRd,t ) + γ 0 log(GPRd,t ) × Governanced,y + cd,y + ct + εd,t . (3)
period before the 2018 US–China trade war. Table A.4 displays the
distribution of FDI from the five most important source countries for Here, the good governance indicator Governanced,y is a dummy variable
each destination country. The US is an important investor in most that equals 1 for relatively good governance (top quartile) in destination
destination markets; other top investors appear to be related market d in year y and 0 otherwise. Note that since our measure of
geographically or historically to the destination country. For example, governance is at annual frequency, its direct impact on FDI is absorbed
Spain is an important investor in several Latin American countries. The by destination-year fixed effects cd,y. The estimated coefficient on the
top five source countries account for 60% of all FDI flows on average interaction term, γ 0, captures the differential impact of GPR on FDI
across destinations, which suggests that the most important investing between destination markets with good and bad governance. If good
countries have considerable experience in dealing with investments in governance safeguards FDI from GPR (Hypothesis 2), the estimated
the destination market. γ 0 should be positive and statistically significant.
3.3. Methodology 3.3.3. The role of information

To understand how information affects MNCs’ response to GPR, we
3.3.1. Baseline model delve into FDIs,d,t, the bilateral FDI from source market s to destination
To evaluate how FDI responds to GPR in the destination market, we market d at period t, and estimate the following model:
estimate the following baseline model:
log(1 + FDIs,d,t ) = β1 log(GPRd,t ) + γ1 log(GPRd,t ) × Informations,d,t
yd,t = βlog(GPRd,t ) + cd,y + ct + εd,t . (1)
+ δ1 Informations,d,t + cs,y + cd,y + cs,d + ct + εs,d,t . (4)
The dependent variable yd,t refers to different measures of FDI flows to
The variable Information measures information flows between destina
destination market d in period t. It takes the form log(1 + FDId,t) in the
tion market d and source market s, gained through geographical, cul
main regressions, where FDId,t is the dollar amount of total FDI flows to
tural, and commercial ties. Stronger ties correspond to better
destination market d in period t. It also refers to log(FDId,t) and FDId,t, the
information about GPR. If better access to information drives MNCs to
average size and total number of FDI projects in various estimations,
cut their investment more aggressively in response to GPR, the coeffi
respectively. The key explanatory variable, log(GPRd,t), is the logarith
cient on the interaction term, γ 1, should be negative and statistically
mic geopolitical risk and uncertainty in destination market d in period t.
significant. Otherwise, γ1 should be positive. Note that on top of desti
In light of Hypothesis 1, we expect the coefficient on log(GPRd,t), β, to be
nation–year fixed effects that capture the time–varying factors that
negative and statistically significant, which indicates that rising GPR
attract FDI to destination markets, we also include source–year fixed
discourages FDI.
effects cs,y to absorb time–varying source–specific factors that drive FDI,
To control for economic and political factors that could potentially
such as low funding costs and weak business growth prospects in the
affect FDI that mostly vary at yearly frequency, we include destina
source market. Again, time fixed effects are included to control for the
tion–year fixed effects cd,y. For example, if MNCs in the destination
global forces that drive FDI. We also control for source–destination fixed
market purchase or renew political risk insurance annually, its effects on
effects cs,d to incorporate the impact of various source–destination
FDI would be absorbed by cd,y. Such fixed effects also absorb constant
linkages such as geographic contiguity, common language, common
destination–specific characteristics such as natural resource endow
colonial history, cultural linkages, etc. Thus, when information is
ment, culture, and institutions, which may affect FDI (Buckley et al.,
proxied for by geographic and cultural ties, its direct impact on FDI is
2007; Dunning, 2004). The allocation of FDI flows may also be driven by
absorbed by cs,d. The coefficient on Information (δ1) only needs to be
global factors such as international funding cost, global economic policy
reported when information is proxied for by bilateral investment treaties
uncertainty, liquidity, and risk. We control for the impact of time
and commercial ties calculated by cumulative investment experience.
–varying global factors on FDI allocation by time fixed effects ct. Finally,
εd,t is the error term.
3.3.4. The role of technology
In an attempt to check for the importance of geopolitical uncertainty
To evaluate how technological characteristics impact the effect of
relative to geopolitical risk, we further explore whether the effect of GPR
GPR, we dig further into bilateral industry–level FDI from source market
on FDI differs across destination markets with high versus low growth:
s to destination market d in industry i, FDIs,d,i,t, and estimate the
yd,t = βlog(GPRd,t ) + γlog(GPRd,t ) × Growthd,y + cd,y + ct + εd,t . (2) following:
Here, Growthd,y is a dummy that equals 1 if the economic growth in log(1 + FDIs,d,i,t ) = β2 log(GPRd,t ) + γ2 log(GPRd,t ) × Technologyi + cd,y
destination market d is in the top quartile, and 0 otherwise. If geopo + cs,y + cs,d,i + ci,t + εs,d,i,t .
litical risk dominates geopolitical uncertainty in driving the impact of (5)
GPR on FDI (Hypothesis 1), the coefficient of the interaction term, γ,
should be positive and statistically significant. Note that the direct Here, Technologyi is the technology intensity in industry i measured by
impact of Growthd,y on FDI is absorbed by cd,y.
7
Table 2
Baseline results on the impact of GPR on FDI.
(1) (2) (3) (4) (5) (6) (7)
Estimator OLS PPML Tobit OLS
Dependent variable log (1 + FDI) log (1 + N) FDI log (FDI) FDI log (1 + FDI) log (1 + FDI)
log(1 + )
N
log(GPR) − 0.258 − 0.057 − 0.192 − 0.219 − 0.222 − 0.260 − 0.190

(0.000) (0.044) (0.000) (0.004) (0.000) (0.000) (0.001)
log(GPR) × Growth − 0.014
(0.937)
Constant 7.122 2.684 4.553 7.150 8.803 6.096 4.607
(0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
Adjusted R2 0.603 0.858 0.277 0.550 0.214
Observations 3876 3876 3876 3737 3876 3876 3816
FDI
FDI is the total dollar amount of foreign direct investment, GPR is the perception of geopolitical risk and uncertainty, N is the number of investment projects, and is
N
the average project size in destination market d at period t (subscript abstracted). Growth is a dummy that equals 1 if the destination’s 5-year moving average of GDP
growth is in the top quartile and 0 otherwise. Columns 1–4 and 7 use ordinary least square (OLS), column 5 uses Poisson pseudo maximum likelihood (PPML), and
column 6 uses Tobit with a lower bound of 0. All regressions include destination, time, and destination × year fixed effects. p–values associated with standard errors
clustered by destination are shown in parentheses.
R&D and various technological characteristics introduced in Section affect FDI, such as political risk and macroeconomic risk. Thus our
3.1.5. We follow the common practice in the literature to adopt such finding implies that the text-based GPR index contains new information
industry-specific measures, because the relative technology intensity is that is not incorporated by traditional risk measures but relevant for
stable across different countries and over time (see, for example, Ilyina MNCs’ FDI decisions.
& Samaniego, 2011; Nunn, 2007; Rajan & Zingales, 1998; Samaniego,
2010).9 The variation in technology intensity across industries facilitates 4.1.2. Extensive and intensive margins of FDI
identification of technology’s impact: We can compare Let Nd,t be the number of FDI projects in destination market d at
technology-intensive FDI (akin to the treatment group) with period t, and
FDId,t
the average project size. We can decompose FDI into
Nd,t
technology-light FDI (akin to the control group) during periods of
the number of FDI projects and the average project size such that
relatively high versus low GPR. FDId,t
If technology intensity increases the resilience of FDI to GPR (Hy FDId,t = Nd,t × Nd,t , which can be transformed into log(FDId,t ) = log(Nd,t )
FDI FDId,t
pothesis 4), the coefficient on the interaction term, γ2, should be positive +log( Nd,td,t ) and implies that log(1 + FDId,t ) ≈ log(1 + Nd,t ) + log(1 + Nd,t ).
and statistically significant. Other than destination–year and source– We now turn to examining whether the extensive margin of FDI
year fixed effects, we further control for source–destination–industry
measured by log(1 + Nd,t) or the intensive margin measured by log(1 +
fixed effects, cs,d,i, to account for the idiosyncratic comparative advan
FDId,t
tage in each industry for either the source market, destination market, or Nd,t ) is driving the negative response of FDI to GPR.
the source–destination pair. The industry–time fixed effects, ci,t, capture The results in columns 2 and 3 of Table 2 show that both the
the role of industrial cycles and also the direct effect of technology in extensive and intensive margins of FDI decline significantly in response
tensity on FDI. The inclusion of industry × time, source × year, desti to GPR (β = − 0.057, p = 0.044 for the extensive margin; β = − 0.192,
nation × year, and source × destination × industry fixed effects rules out p < 0.01 for the intensive margin). FDI falls more aggressively at the
most potentially confounding factors. intensive than at the extensive margin in response to GPR: For a one-
standard-deviation shock (0.363) to log(GPR), the intensive margin de
4. Empirical results clines by 7.0%—more than triple that of the extensive margin (2.1%).
The sum of the coefficients of log(GPR) in columns 2 and 3 is 0.249,
4.1. Response of FDI to GPR which is very similar to that in column 1. A simple accounting exercise
suggests that the extensive and intensive margins contribute 23.1% ( =
4.1.1. Baseline result 2.1% 7%
2.1%+7%) and 76.9% ( = 2.1%+7%) of the FDI response to GPR, respectively.
Table 2 reports the baseline result that summarizes the impact of These results provide additional support to Hypothesis 1.
rising GPR on FDI. Based on ordinary least square (OLS) estimation of
Eq. (1), we show in column 1 that the coefficient on log(GPR) is negative 4.1.3. Dealing with zero observations
and statistically significant (β = − 0.258, p = 0.000). This suggests that Using log(1 + FDI) as the dependent variable allows us to tackle
FDI to a destination market falls as the GPR there increases, which scenarios in which there is no FDI to a destination market at a given
supports Hypothesis 1. In particular, for a one-standard-deviation shock point in time, which is common in the literature (Head & Mayer, 2014).
(0.363) to log(GPR) (GPR increases by 36.3%), FDI drops by 9.4%.10 However, the zeros may simply be missing observations that are
Note that we have controlled for destination × year fixed effects, which wrongly recorded as 0, which leads to inconsistent estimation. By
absorb annually varying destination-specific factors that could possibly replacing the dependent variable with log(FDI) to treat all zero obser
vations as missing, we find similar evidence that rising GPR discourages
FDI, as indicated by the negative and statistically significant coefficient
9
This practice differs from that of Castellani, Jimenez, and Zanfei (2013), on log(GPR) (β = − 0.219, p = 0.004) in column 4 of Table 2.
Castellani and Lavoratori (2020), who focus on FDI projects whose main The presence of heteroskedasticity in log–linearized models and zero
business activities are labeled by the data vendor as R&D. In these papers, all
R&D FDI are treated as if their R&D intensity were similar across
projects/MNCs.
10
This is calculated as the multiplier of the coefficient on log(GPR) in column
1 of Table 2, − 0.258, and the shock to log(GPR), 0.363.
8
Table 3
The safeguarding role of governance.
(1) (2) (3) (4) (5) (6) (7)
Dependent variable log (1 + FDI)
log(GPR) − 0.287 − 0.293 − 0.315 − 0.302 − 0.288 − 0.242 − 0.302

(0.000) (0.000) (0.000) (0.000) (0.000) (0.001) (0.000)
log(GPR) × GovEffectiveness 0.124
(0.031)
log(GPR) × RegulatoryQuality 0.157
(0.043)
log(GPR) × CorruptionControl 0.273
(0.066)
log(GPR) × RuleLaw 0.184
(0.061)
log(GPR) × PoliticalStability 0.090
(0.320)
log(GPR) × Accountability − 0.059
(0.522)
log(GPR) × GovernanceIndex 0.186
(0.019)
2
Adjusted R 0.603 0.603 0.603 0.603 0.603 0.603 0.603
Observations 3876 3876 3876 3876 3876 3876 3876
This table reports how different measures of governance shape the impact of geopolitical risk on FDI. FDI is the total dollar amount of FDI and GPR is the perception of
geopolitical risk and uncertainty in destination market d in period t (subscript abstracted). The dummy variables GovEffectiveness, PoliticalStability, RegulatoryQuality,
CorruptionControl, RuleLaw, Accountability, and GovernanceIndex equal 1 if their corresponding values are in the top quartile and 0 otherwise. GovernanceIndex is an
index calculated as the average of the six governance indices in columns 1–6. All regressions include destination, time, and destination × year fixed effects. p–values
associated with standard errors clustered by destination are shown in parentheses.
Table 4 Table 5
The role of information channeled through close ties. The role of technology.
(1) (2) (3) (4) (5) (1) (2) (3) (4)
log (1 + FDI) Dependent variable log (1 + FDI)
log(GPR) − 0.020 − 0.016 − 0.013 0.016 − 0.000 log(GPR) − 0.005 − 0.007 − 0.003 − 0.006
(0.020) (0.078) (0.148) (0.163) (0.985) (0.029) (0.003) (0.217) (0.020)
log(GPR) × Contiguity − 0.082 log(GPR) × DR&D 0.011
(0.030) (0.006)
log(GPR) − 0.061 log(GPR) × DAssetFixity − 0.010
× CommonLanguage (0.016)
(0.030) log(GPR) × DSkilledLabor − 0.005
log(GPR) × Experience − 0.009 (0.380)
2
(0.007) Adjusted R 0.159 0.159 0.159 0.159
log(GPR) × Treaty − 0.050 Observations 1148,928 1148,928 1148,928 1148,928
(0.004)
Experience 0.129 This table summarizes how the effect of GPR on FDI varies across industries with
(0.000) different technology characteristics. FDI is the dollar amount of foreign direct
Treaty 0.242 investment from source market s to destination market d in industry i at period t
(0.004) (subscript abstracted), and GPR is the perception of geopolitical risk and un
Constant 0.539 0.557 0.556 0.015 0.443 certainty in destination market d in period t. The dummy variables DR&D, DAs
(0.000) (0.000) (0.000) (0.780) (0.000) setFixity and DSkilledLabor equal 1, respectively, if asset fixity and skilled labor are in
2
Adjusted R 0.455 0.454 0.454 0.459 0.455 the top quartile and 0 otherwise. Due to the availability of technology data, only
Observations 245,004 239,088 239,088 245,004 245,004
manufacturing industries are included. All regressions include destina
This table reports how information gained through close ties between source and tion–source pair × industry, industry × time, destination × year, and source
destination markets affects the impact of GPR on FDI. Here FDI is the dollar × year fixed effects. p–values associated with standard errors clustered by des
amount of foreign direct investment from source market s to destination market tination–source pair are shown in parentheses.
d in period t (subscript abstracted) and GPR is the perception of geopolitical risk
and uncertainty in destination market d in period t. The dummy variables discourages FDI (β = − 0.222, p = 0.000). We also apply a Tobit esti
Contiguity and CommonLanguage equal 1 if source and destination markets share mator and show in column 6 of Table 2 that our baseline result remains
common borders and common languages, respectively. Experience is the log of 1
robust.11
plus cumulative FDI from source market s to destination market d up to period t.
Treaty is a dummy that equals to 1 if source and destination markets have a
bilateral investment treaty. All regressions include destination × source, time,
source × year, and destination × year fixed effects. The coefficients of static
variables—Contiguity and CommonLanguage—are absorbed by fixed effects.
P–values associated with standard errors clustered by destination–source pair 11
In addition, we conduct a battery of other robustness checks using different
are shown in parentheses.
model specifications and estimation techniques, controlling for additional
variables including various risk measures and macroeconomic shocks and
observations could lead to a biased estimation of the coefficients of in applying FDI aggregated at different levels. Our finding that GPR has a negative
terest. To simultaneously deal with these two issues, we employ the effect on FDI remains robust. We further show that the effect of GPR on FDI is
Poisson pseudo maximum likelihood estimator from Silva and Tenreyro more pronounced when GPR is high and volatile; this is consistent with Pástor
(2006) and repeat our analysis. The result in column 5 of Table 2 yields a and Veronesi (2013), who show that investors demand higher risk premiums
similar finding, whereby rising GPR in the destination market when they face greater uncertainty. These results (not reported) are available
upon request.
9
4.1.4. Geopolitical risk versus geopolitical uncertainty p = 0.066). Similar weak evidence is found for a stronger rule of law (see
GPR incorporates both geopolitical risk and uncertainty but does not column 4), which reflects public confidence in and abidance by the rules
differentiate the two. The risk-return tradeoff implies that geopolitical of society (β = 0.184, p = 0.061).
risk, which acts like an additional cost of doing business (Click & There is no evidence that either political stability (which measures
Weiner, 2010), is of less concern as long as returns to investment are the reverse likelihood of politically motivated instability and violence)
sufficiently large. As a result, the impact of geopolitical risk on FDI or accountability (which quantifies the extent of freedom of speech, the
should be less pronounced when economic growth is faster, which im media, association, and election) affect the impact of GPR on FDI (see
plies higher investment returns on average. However, growth prospects columns 5 and 6 of Table 3).13
are unlikely to affect the impact of geopolitical uncertainty, since in These different measures of governance are highly correlated (see
vestors typically delay irreversible investments to prioritize safety over Table A.3), which makes it difficult to evaluate simultaneously their
profitability. If geopolitical risk dominates geopolitical uncertainty, mitigating effects. To understand the overall impact of governance, we
then the negative effect of GPR on FDI should be less pronounced. combine these measures to form a composite governance index. The
Column 7 of Table 2 shows that the coefficient on log(GPR) × Growth is results in column 7 of Table 3 show that good governance mitigates the
not statistically significant (β = − 0.014, p = 0.937). Thus there is no negative impact of GPR on FDI, which is consistent with our previous
credible evidence to support Hypothesis 1, that the effect of GPR on FDI finding and supports Hypothesis 2.
varies with economic growth. This result implies that geopolitical un
certainty is likely to be the key driver of the negative effect of GPR on 4.3. The information channels
FDI.
After examining the role of governance in safeguarding FDI against
GPR, we turn to examining how MNCs may protect their investments by
4.2. The safeguarding roles of good governance
leveraging information on destination markets. So far, we have focused
on destination–aggregated FDI to understand its response to GPR. In this
If MNCs expect that good governance will safeguard their investment
section, we differentiate FDI from different source markets to under
against GPR, FDI should be less responsive to rising GPR in destination
stand how the close ties between source and destination markets that
markets with better governance. We measure the perception of gover
reduce information asymmetry may reshape the response of bilateral
nance quality, based on WGI data, in terms of overall government
FDI to GPR.
effectiveness, regulatory quality, control of corruption, the rule of law,
We start by evaluating how bilateral FDI responds to GPR in the
political stability, and accountability. We first create a dummy variable
destination market by estimating Eq. (4) without the interaction term.
that equals 1 if destination market d in period t falls in the top quartile of
We show in column 1 of Table 4 that rising GPR reduces bilateral FDI,
the given measure of governance and 0 otherwise. We then estimate Eq.
which is consistent with the baseline result. However, the coefficient on
(3), which includes the interaction between log(GPR) and each of these
log(GPR) is much smaller than that based on destination–aggregated FDI
dummy variables to identify whether the impact of GPR on FDI varies
due to a larger number of zero observations in bilateral FDI. The nega
with governance characteristics. Table 3 reports the estimation results.
tive impact of GPR on FDI remains robust when we deal with the zeros
The coefficient on the interaction between GPR and government
using alternative estimators (not reported).
effectiveness in column 1 of Table 3 is positive and statistically signifi
To understand how information gained from geographic, cultural,
cant (β = 0.124, p = 0.031). This suggests that effective governments,
and commercial ties between source and destination markets affect the
which make high–quality and independent public policy as well as
impact of GPR on bilateral FDI, we replace information in Eq. (4) with (i)
credible commitments even during geopolitical turbulence, weaken the
dummy variables Contiguity and CommonLanguage, which equal 1
negative impact of GPR on FDI. The sum of the coefficients of log(GPR)
respectively, if the source and destination markets share common bor
and log(GPR) × GovEffectiness is − 0.163, which is statistically signifi
ders or have common languages; (ii) Experience, the logarithm of 1 plus
cant (p = 0.003). This means that rising GPR still deters FDI in desti
the cumulative FDI from source market s to destination market d from
nation markets with good governance, though to a smaller extent. In
the first sample period (January 2003) to period t; and (iii) Treaty, a
particular, for every one-standard-deviation (0.363) increase in log
dummy that equals 1 if there is a bilateral investment treaty between the
(GPR), FDI declines by 5.9% in destinations with effective government,
source and destination markets.
which is 4.5% points less than that in destinations with relatively inef
The coefficient on the interaction between log(GPR) and Contiguity in
fective governments (10.4%).12
column 2 of Table 4 is negative and statistically significant (β = −
The coefficient on the interaction between GPR and regulatory
0.082, p = 0.030). This suggests that better information channeled
quality shown in column 2 of Table 3 is also positive and statistically
through closer geographic ties strengthens the negative impact of GPR
significant (β = 0.157, p = 0.043), which suggests that high regulatory
on FDI, which supports Hypothesis 3. The finding also implies that
quality attenuates the impact of GPR on FDI. This provides evidence that
geopolitical uncertainty plays a more important role than geopolitical
governments that are capable of formulating and implementing regu
risk in driving the negative impact of GPR on FDI. Intuitively, close
lations to promote private sector development safeguard FDI against
geographic ties mitigate information asymmetry, which enables MNCs
GPR.
to gain better access to information about geopolitical uncertainty and
The result in column 3 of Table 3 provides weak evidence that tighter
increases their confidence in acting on the information assertively and
control of corruption, which curbs the practice of trading public power
avoiding irreversible investments.
for private gain, mitigates the impact of GPR on FDI (β = 0.273,
We find similar evidence in column 3 of Table 4 that information
gained from cultural ties, measured by common language, reinforces the
12 negative effect of GPR on FDI. The coefficient on the interaction between
The change in FDI in destination markets with high government effective
ness is calculated as ( − 0.290 + 0.155) × 0.363 = 4.9%, the coefficient sum of
log(GPR) and CommonLanguage in column 3 is negative and statistically
log(GPR) and log(GPR) × GovEffectiness, multiplied by the standard deviation of significant (β = − 0.061, p = 0.030). This lends further support to
log(GPR). The net impact of FDI in destination markets with low government
effectiveness is calculated as − 0.290 × 0.363 = 10.5%, the coefficient on log
13
(GPR) multiplied by the standard deviation of log(GPR). The differential FDI It is difficult to establish whether accountability moderates the negative
response to the same change in geopolitical risk is calculated as effect of GPR on FDI because the public does not condemn geopolitical con
0.155 × 0.363 = 5.6%, the coefficient on log(GPR) × GovEffectiness multiplied frontations, or because the disciplinary effects of public opinion on government
by the standard deviation of log(GPR). actions are small or nonexistent.
10
Hypothesis 3—that information that originates from close cultural ties characteristics summarized by Samaniego and Sun (2015) and find no
increases the sensitivity of FDI to GPR. evidence that they impact the effects of GPR on FDI (not reported).
Other than geographic and cultural ties, we can also measure in
formation from the perspective of commercial ties. This is because MNCs 5. Conclusion
could accumulate knowledge through commercial activities, which en
ables them to better assess geopolitical risk in destination markets. We document robust evidence that rising GPR deters FDI even after
Column 4 of Table 4 shows that the coefficient on the interaction be controlling for various risks. This suggests that GPR, which reflects
tween log(GPR) and Experience is negative and statistically significant perceptions of geopolitical risk and uncertainty, has incremental effects
(β = − 0.009, p = 0.007). We document a similar finding in column 5 of on MNCs’ FDI decisions. We further show that good governance miti
Table 4: The coefficient on the interaction term between log(GPR) and gates the negative effect of GPR on FDI by ensuring consistent, inde
Treaty is negative and statistically significant (β = − 0.05, p = 0.004). pendent, and effective policymaking, while good information
These results mean that MNCs from source markets that have closer strengthens the negative effect by motivating MNCs to avoid rather than
commercial ties with destination markets cut FDI more aggressively in to deal with GPR. Moreover, the technology embeddedness of FDI pro
response to rising GPR in those markets, which supports Hypothesis 3. vides a shield against GPR because MNCs can leverage their internal
linkages and organization to transfer intangible technology-intensive
4.4. The resilience of technology-intensive FDI assets across borders.
Our findings further understanding of how MNCs incorporate
We further delve into the industry composition of bilateral FDI to geopolitical risk and uncertainty in their investment decisions and point
understand how industry–specific technological characteristics affect to potential channels for navigating the negative effects of GPR on FDI,
the impact of GPR on FDI. We are interested in whether technology which connect international theories with public policy. This study’s
embedded in FDI, especially in the form of R&D, may increase the contributions to the literature are threefold. First, it adds to the inter
resilience of FDI to GPR. Whereas our main focus is on R&D, we also national business literature by highlighting the incremental effects of
explore other technological characteristics used in the trade and mac perceived geopolitical risk and uncertainty on FDI, which have largely
roeconomic literature. been neglected by prior studies that focus on traditional political risks or
Before analyzing the role of technology, we first verify the baseline realized adverse events. Like political risk, geopolitical risk and uncer
results using bilateral–industry level FDI by estimating Eq. (5) without tainty perceived by market participants undermine FDI, which suggests
the interaction term. Column 1 of Table 5 shows that the negative that existing theories about political risk and international business are
impact of GPR on FDI remains robust. general enough to incorporate new sources of risk in dynamic geopo
We first measure technology by R&D and examine whether it pro litical environments. Unlike political risk of a domestic origin, which is
vides a shield against GPR. We estimate Eq. (5) and present the result in at the center of previous studies, geopolitical risk and uncertainty
column 2 of Table 5. The coefficient on the interaction between log(GPR) involve multiple nations and may require different response strategies.
and DR&D, a dummy that equals 1 for industries with R&D intensity Our study contributes to current theories that emphasize MNCs’ pro
ranked in the top quartile, is positive and statistically significant active management of political risk (e.g., through strengthening con
(β = 0.011, p = 0.006). This means that the negative impact of GPR on nections with local governments) by uncovering new evidence that
FDI is less pronounced in R&D-intensive industries, which supports informed MNCs choose to avoid rather than to manage geopolitical risk
Hypothesis 4—that FDI in more technology-intensive industries is more and uncertainty, which implies that different navigation strategies in
resilient to GPR. Intuitively, R&D generates intangible assets that are response to different risks. Finally, this study adds to the literature on
precious to the destination market and can be transferred across borders governance quality by revealing its role in safeguarding FDI from
within the firm with relative ease, and is therefore less constrained by geopolitical risk and uncertainty.
GPR in a particular destination. The sum of the coefficients of log(GPR) Our findings have important implications for policymakers and MNC
and log(GPR) × DR&D is 0.004, which is economically small and statis managers. First, geopolitical risk and uncertainty are highly relevant for
tically insignificant (p = 0.328). This suggests that the negative impact the investment decisions of MNCs. Prior experience gained with other
of GPR on FDI is concentrated in less R&D-intensive industries. types of political risks and hedging strategies developed to address
Besides R&D, technology can be embedded in factors of production them—which mainly rely on developing ties with domestic actors in the
such as machinery, equipment, and skilled labor (see, for, example destination market—may not necessarily be useful in the context of
Samaniego, 2010). Advances in machinery can be identified by the geopolitical risk and uncertainty, which involves tensions among mul
capital expenditure on fixed assets, and knowledge of the production tiple nations. Second, to attract FDI, governments may consider
process is often embedded in skilled labor. Following this strand of the investing in good governance as a way to signal to investors a stable
literature, we follow Samaniego and Sun (2015) to measure these policy environment that is less likely to change unpredictably in
technological characteristics using (i) asset fixity, the fraction of assets response to geopolitical tensions. This is especially important given that
invested in machine and equipment and other fixed assets, and (ii) better-informed investors—who also tend to be those who invest the
skilled labor, the average wage bill. We then create dummy variables most—tend to curb FDI more aggressively in anticipation of geopolitical
DAssetFixity and DSkilledLabor that equal 1 if the value of asset fixity and uncertainty. Third, technology-intensive FDI may be especially valuable
skilled labor, respectively, is in the top quartile and 0 otherwise. to EMs if they are prepared to protect it in the face of geopolitical risk
We first estimate Eq. (5) including the interaction between log(GPR) and uncertainty, since firms react less strongly in the face of geopolitical
and DAssetFixity and report the result in column 3 of Table 5. The coeffi turmoil when the fruits of their investment are intangible and can easily
cient on the interaction term is negative and statistically significant be moved across borders within their internal organization.
(β = − 0.010, p = 0.016), which means that FDI to industries with larger There are several caveats in this study. First, we only document how
asset fixity is more responsive to GPR. Unlike R&D investments, in MNCs respond to GPR through FDI allocation but not their implications
vestments in tangible assets such as machinery and equipment cannot be for corporate performance. Second, we uncover how good governance
easily reversed or transferred to other markets. Thus, MNCs need to be and high technology safeguard FDI from GPR without considering the
more careful while investing in industries that require substantial fixed resource constraints of doing so. Third, we mostly rely on country-level
assets whenever they face rising GPR. data for our analysis, which can be much improved if firm-level data on
By repeating the analysis for skilled labor, we show in column 4 of GPR, technology, and information were available. More interdisci
Table 5 that there is no evidence that skilled labor matters for the impact plinary studies are needed to inform corporate investment decisions and
of GPR on FDI. We also repeat the analysis using various technological public policymaking in an environment of growing geopolitical risk and
11
uncertainty. It would be interesting to explore how MNCs’ heteroge Data availability

neous response to geopolitical risk affect their long–term growth and
whether avoiding or managing GPR yields better corporate perfor Data will be made available on request.
mance. Another promising research direction would be evaluating the
costs and benefits of devoting public resources to safeguarding FDI. Appendix A. Additional Tables
Tables A.1, A.2, A.3, A.4.
Table A.1
Variable definitions.
Variable Definition Source
GPRd,t Geopolitical risk index for destination market d at period t Caldara and Iacoviello (2022)
FDI
FDId,t Aggregate FDI to destination market d at period t fDi Intelligence
FDIs,d,t Bilateral FDI from source market s to destination market d at period t fDi Intelligence
FDIs,d,i,t Bilateral industry–level FDI from source market s to destination market d in industry i at period t fDi Intelligence
Governance
GovEffectiveness A dummy that equals 1 if the government effectiveness index is in the top quartile WGI
RegulatoryQuality A dummy that equals 1 if the regulatory quality index is in the top quartile WGI
CorruptionControl A dummy that equals 1 if the corruption control index is in the top quartile WGI
RuleLaw A dummy that equals 1 if the rule of law index is in the top quartile WGI
PoliticalStability A dummy that equals 1 if the political stability index is in the top quartile WGI
Accountability A dummy that equals 1 if the government accountability index is in the top quartile WGI
Information
Contiguitys,d A dummy that equals 1 if the source and destination markets share common borders CEPII
CommonLanguages, A dummy that equals 1 if the source and destination markets share common language CEPII
d
ColonialRelations,d A dummy that equals 1 if the source and destination markets were in colonial relations historically CEPII
Experiences,d,t Cumulative FDI from source market s to destination market d from January 2003 to period t fDi Intelligence
Treatys,d,t A dummy that equals 1 if the source and destinations markets have a bilateral investment treaty. UNCTAD
Technology
R&Di Industry–level R&D intensity measured by the average share of R&D expenditures in total capital expenditures for industry i Samaniego and Sun (2015)
AssetFixityi The average ratio of fixed assets to total assets for industry i. Samaniego and Sun (2015)
SkilledLabori The total wage bill divided by the number of employees for industry i Samaniego and Sun (2015)
Other measures
Growth A dummy that equals 1 if GDP growth is in the top quartile
HighRisk A dummy that equals 1 for a high–mean and high–variance state of GPR authors’ calculations
Crisis A dummy that equals 1 if the economy is in an economic crisis Laeven and Valencia (2012),
Recession A dummy that equals 1 if the economy is in a recession Braun and Larrain (2005)
PoliticalRisk Risk score for the political stability of a country ICRG
EconomicRisk Risk score for a country’s current economic strengths and weaknesses ICRG
FinancialRisk Risk score for a country’s ability to pay its way by financing its official, commercial, and trade debt obligations ICRG
GovernmentStability A country’s ability to reach its declared objectives and to stay in power ICRG
Abbreviated data sources: Worldwide Governance Indicators (WGI); Centre d′ etudes prospectives et d′ informations internationales (CEPII); United Nations Conference
on Trade and Development (UNCTAD); The International Country Risk Guide (ICRG).
Table A.2
Summary of the hypothesized roles of governance, information, and technology.
Channel Hypothesis
Geopolitical risk Geopolitical uncertainty GPR
Better Mitigates negative effect (+) Effective and swift policymaking reduces the impact of Mitigates negative effect (+) Reduces policy uncertainty +
governance adverse geopolitical events on business, facilitates conflict resolution, and prevents following geopolitical events.
escalation of the conflict.
More Mitigates negative effect (+) Allows investors to reduce risk exposure through hedging or Intensifies negative effect ( − ) Investors have better +∕
information management and act more decisively on information about risk. information about uncertainty and delay investment to −
avoid it.
Higher Mitigates negative effect (+) Protection from governments due to their strategic Mitigates negative effect (+) Flexible transfer across +
technology importance. borders to bypass uncertainty.
This table sketches how better governance, more information, and higher technology reshape the negative effects of GPR and its two components, geopolitical risk and
geopolitical uncertainty, on FDI.
Table A.3
Correlation between different measures of governance.
GovEffectiveness RegulatoryQuality CorruptionControl RuleLaw PoliticalStability Accountability
GovEffectiveness 1.000
RegulatoryQuality 0.926 1.000
(0.000)
CorruptionControl 0.914 0.900 1.000
(0.000) (0.000)
(continued on next page)
12
Table A.3 (continued )

GovEffectiveness RegulatoryQuality CorruptionControl RuleLaw PoliticalStability Accountability
RuleLaw 0.940 0.928 0.932 1.000

(0.000) (0.000) (0.000)
PoliticalStability 0.537 0.473 0.565 0.546 1.000
(0.000) (0.000) (0.000) (0.000)
Accountability 0.410 0.434 0.441 0.442 0.233 1.000
(0.000) (0.000) (0.000) (0.000) (0.000)
This table reports the correlation coefficients between different (continuous) indices of governance quality, as defined in Table A.1. p–values associated with cor
relation coefficients are in parentheses.
Table A.4
Fraction of FDI by source country.
Destination Source Fraction of total FDI (%) Destination Source Fraction of total FDI (%)
Argentina United States 18.7 Mexico United States 33.4

Spain 16.1 Spain 12.8
Italy 9.0 Germany 9.3
Brazil 7.0 Japan 8.2
Canada 6.9 Canada 4.6
Brazil United States 20.2 Philippines Japan 17.9
Germany 9.4 United States 16.2
Spain 7.7 China 14.3
Italy 7.2 Thailand 8.4
United Kingdom 6.3 Korea, Rep. 5.3
China United States 20.7 Russian Federation Germany 13.0
Germany 11.0 China 11.3
Japan 10.7 United States 11.0
Taiwan 8.6 France 6.1
Hong Kong, China 7.0 United Kingdom 5.5
Colombia United States 16.5 Saudi Arabia United States 24.6
Spain 15.8 France 15.3
Canada 14.2 United Arab Emirates 10.8
Brazil 8.8 Japan 8.7
Switzerland 5.8 China 5.6
Hong Kong, China United States 28.1 South Africa United Kingdom 17.7
United Kingdom 14.9 United States 14.3
China 13.1 Germany 7.3
Germany 5.3 Australia 6.3
Japan 5.0 China 5.8
India United States 24.4 Thailand Japan 31.3
United Kingdom 11.0 United States 19.9
Japan 8.8 Hong Kong, China 8.0
China 7.2 China 5.0
Germany 6.7 Singapore 4.3
Indonesia China 17.1 Turkey Germany 14.2
Japan 15.2 United States 12.0
United States 11.8 Azerbaijan 11.6
Korea, Rep. 7.3 Netherlands 6.9
Singapore 4.8 France 5.3
Israel United States 76.1 Ukraine Russian Federation 13.9
China 3.3 United States 10.0
France 2.7 Germany 8.6
Spain 2.6 Austria 8.6
United Kingdom 2.3 France 6.9
Korea, Rep. United States 29.9 Venezuela United States 27.3
Japan 14.5 Brazil 14.0
Netherlands 10.4 Spain 12.8
France 9.5 Russian Federation 8.3
Germany 6.8 China 4.8
Malaysia United States 14.8
Japan 13.9
China 13.3
Germany 6.8
Netherlands 5.3
This table reports the fraction of total FDI over the sample period in the destination market originating from the 5 most important source markets.
Appendix B. Supporting information References
Supplementary data associated with this article can be found in the Arikan, I., & Shenkar, O. (2013). National animosity and cross-border alliances. ISSN:
00014273 Academy of Management Journal, 56(6). https://doi.org/10.5465/
online version at doi:10.1016/j.ibusrev.2023.102136. amj.2011.0210.
Baker, S. R., Bloom, N., & Davis, S. J. (2016). Measuring economic policy uncertainty.
The Quarterly Journal of Economics, 131(4), 1593–1636.
13
Beazer, Q. H., & Blake, D. J. (2018). The conditional nature of political risk: How home Head, K., & Mayer, T. (2014). Gravity equations: Workhorse, toolkit, and cookbook. Vol.
institutions influence the location of foreign direct investment. American Journal of 4. Handbook of international economics (pp. 131–195). Elsevier. Vol. 4.
Political Science, 62(2), 470–485. Hegre, H., & Nygård, H. M. (2015). Governance and conflict relapse. Journal of Conflict
Bernanke, B. S. (1983). Irreversibility, uncertainty, and cyclical investment. The Resolution, 59(6), 984–1016.
Quarterly Journal of Economics, 98(1), 85–106. Henisz, W. J. (2000). The institutional environment for multinational investment. The
Björkman, A., & Piekkari, R. (2009). Language and foreign subsidiary control: An Journal of Law, Economics, and Organization, 16(2), 334–364.
empirical test. Journal of International Management, 15(1), 105–117. Henisz, W. J., & Zelner, B. A. (2005). Legitimacy, interest group pressures, and change in
Braun, M., & Larrain, B. (2005). Finance and the business cycle: International, inter- emergent institutions: The case of foreign investors and host country governments.
industry evidence. ISSN: 00221082 The Journal of Finance, 60(3), 1097–1128. Academy of Management Review, 30(2), 361–382.
https://doi.org/10.1111/j.1540-6261.2005.00757.x. Huang, T., et al. (2015). Political risk and dividend policy: Evidence from international
Buckley, P. J., et al. (2007). The determinants of Chinese outward foreign direct political crises. ISSN: 14786990 Journal of International Business Studies, 46(5),
investment. Journal of International Business Studies, 38(4), 499–518. 574–595. https://doi.org/10.1057/jibs.2015.2. ISSN: 14786990
Busse, M., & Hefeker, C. (2007). Political risk, institutions and foreign direct investment. 〈https://www-jstor-org.libproxy1.nus.edu.sg/stable/43653868?pq-origsite=360lin
ISSN: 01762680 European Journal of Political Economy, 23(2). https://doi.org/ k&seq=1#metadata_info_tab_contents〉.
10.1016/j.ejpoleco.2006.02.003. Husted, L., Rogers, J., & Sun, B. (2020). Monetary policy uncertainty. Journal of Monetary
Butler, K. C., & Joaquin, D. C. (1998). A note on political risk and the required return on Economics, 115, 20–36.
foreign direct investment. Journal of International Business Studies, 29(3), 599–607. Ilyina, A., & Samaniego, R. M. (2011). Technology and financial development. ISSN:
Caldara, D., & Iacoviello, M. (2022). Measuring geopolitical risk. American Economic 00222879 Journal of Money, Credit and Banking, 43(5), 899–921. https://doi.org/
Review, 112(4), 1194–1225. 10.1111/j.1538-4616.2011.00401.x.
Castellani, D., Jimenez, A., & Zanfei, A. (2013). How remote are R&D labs? Distance Javorcik, B. S. (2004). The composition of foreign direct investment and protection of
factors and international innovative activities. ISSN: 00472506 Journal of intellectual property rights: Evidence from transition economies. European Economic
International Business Studies, 44(7), 649–675. https://doi.org/10.1057/jibs.2013.30. Review, 48, 39–62. 〈www.sciencedirect.comwww.elsevier.com/locate/econbase〉.
Castellani, D., & Lavoratori, K. (2020). The lab and the plant: Offshore R&D and co- Jens, C. E. (2017). Political uncertainty and investment: Causal evidence from US
location with production activities. Journal of International Business Studies, 51(1), gubernatorial elections. Journal of Financial Economics, 124(3), 563–579.
121–137. Jensen, N. (2008). Political risk, democratic institutions, and foreign direct investment.
Chan, C. M., & Makino, S. (2007). Legitimacy and multi-level institutional environments: The Journal of Politics, 70(4), 1040–1052.
Implications for foreign subsidiary ownership structure. Journal of International Julio, B., & Yook, Y. (2016). Policy uncertainty, irreversibility, and cross-border flows of
Business Studies, 38(4), 621–638. 〈www.jibs.net〉. capital. Journal of International Economics, 103, 13–26.
Chan, J.M. L. and H. Zheng (2019). FDI on the Move: Cross-border M&A and Migrant Kastner, S. L. (2007). When do conflicting political relations affect international trade?.
Networks. Tech. rep. Tech. rep. working paper. ISSN: 00220027 Journal of Conflict Resolution, 51(4). https://doi.org/10.1177/
Click, R. W., & Weiner, R. J. (2010). Resource nationalism meets the market: Political 0022002707302804.
risk and the value of petroleum reserves. ISSN: 00472506 Journal of International Keller, W., & Yeaple, S. R. (2013). The gravity of knowledge. American Economic Review,
Business Studies, 41(5). https://doi.org/10.1057/jibs.2009.90. 103(4), 1414–1444.
Col, B., Durnev, A., & Molchanov, A. (2018). Foreign risk, domestic problem: Capital King, G., & Zeng, L. (2001). Explaining rare events in international relations. International
allocation and firm performance under political instability. Management Science, 64 Organization, 55(3), 693–715.
(5), 2102–2125. Knight, F. H. (1921). Risk, uncertainty and profit. Vol. 31. Houghton Mifflin.
Colino, A. (2013). Conflict resolution processes, uncertainty and investment dynamics: Kobrin, S. J. (1979). Political risk: A review and reconsideration. Journal of International
Evidence for the Basque Country. Defence and Peace Economics, 24(3), 229–245. Business Studies, 10(1), 67–80.
Cordero, A. M., & Miller, S. R. (2019). Political party tenure and MNE location choices. Laeven, L. and F. Valencia (2012). Systemic Banking Crises Database: An Update. IMF
Journal of International Business Studies, 50, 973–997. https://doi.org/10.1057/ Working Paper, 12/163.
s41267-019-00220-0 Li, Q. (2008). Foreign direct investment and interstate military conflict. Journal of
Cuervo-Cazurra, A. (2006). Who cares about corruption? Journal of International Business International Affairs, 62(1). ISSN: 0022-197X.
Studies, 37(6), 807–822. Li, Q., & Vashchilko, T. (2010). Dyadic military conflict, security alliances, and bilateral
Cuypers, I. R. P., et al. (2021). Transaction cost theory: Past progress, current challenges, FDI flows. Journal of International Business Studies, 41(5), 765–782. 〈www.jibs.net〉.
and suggestions for the future. Academy of Management Annals, 15(1), 111–150. Liu, X., & White, S. (2001). Comparing innovation systems: A framework and application
Czinkota, M. R., et al. (2010). Terrorism and international business: A research agenda. to China’s transitional context. Research Policy, 30(7), 1091–1114.
ISSN: 00472506 Journal of International Business Studies, 41(5), 826–843. https://doi. Makino, S., & Tsang, E. W. K. (2011). Historical ties and foreign direct investment: An
org/10.1057/jibs.2010.12. ISSN: 00472506 〈https://www-jstor-org.libproxy1.nus. exploratory study. ISSN: 00472506 Journal of International Business Studies, 42(4).
edu.sg/stable/40731534?pq-origsite=360link&seq=1#metadata_info_tab_conten https://doi.org/10.1057/jibs.2010.53.
ts〉. Moyo, D. (2019). Are businesses ready for deglobalization. Harvard Business Review, 6.
Darendeli, I. S., & Hill, T. L. (2016). Uncovering the complex relationships between Narula, R., & Dunning, J. H. (1998). Explaining international R&D alliances and the role
political risk and MNE firm legitimacy: Insights from Libya. Journal of International of governments. International Business Review, 7(4), 377–397.
Business Studies, 47(1), 68–92. Nunn, N. (2007). Relationship-specificity, incomplete contracts, and the pattern of trade.
Daude, C., & Stein, E. (2007). The quality of institutions and foreign direct investment. ISSN: 0033-5533 The Quarterly Journal of Economics, 122(2), 569–600. https://doi.
Economics & Politics, 19(3), 317–344. org/10.1162/qjec.122.2.569.
Demirbag, M., & Glaister, K. W. (2010). Factors determining offshore location choice for Oh, C. H., & Oetzel, J. (2011). Multinationals’ response to major disasters: how does
R&D projects: A comparative study of developed and emerging regions. Journal of subsidiary investment vary in response to the type of disaster and the quality of
Management Studies, 47(8), 1534–1560. country governance? Strategic Management Journal, 32(6), 658–681.
Dunning, J. H. (2004). Determinants of foreign direct investment: globalization-induced Papanastassiou, M., Pearce, R., & Zanfei, A. (2020). Changing perspectives on the
changes and the role of policies. Towards pro poor policies, 279–290. internationalization of R&D and innovation by multinational enterprises: A review
Eden, L., & Miller, S. R. (2004). Distance matters: Liability of foreignness, institutional of the literature. Journal of International Business Studies, 51(4), 623–664.
distance and ownership strategy. Theories of the Multinational Enterprise: Diversity, Pástor, L., & Veronesi, P. (2013). Political uncertainty and risk premia. ISSN: 0304405X
Complexity and Relevance. Emerald Group Publishing Limited. Journal of Financial Economics, 110(3). https://doi.org/10.1016/j.
EY (2020). Geostrategy in Practice 2020. Tech. rep. EY. jfineco.2013.08.007.
Filippaios, F., et al. (2019). Political governance, civil liberties, and human capital: Peng, G. Z., & Beamish, P. W. (2014). MNC subsidiary size and expatriate control:
Evaluating their effect on foreign direct investment in emerging and developing Resource-dependence and learning perspectives. Journal of World Business, 49(1),
economies. Journal of International Business Studies, 50, 1103–1129. https://doi.org/ 51–62.
10.1057/s41267-019-00239-3 Rajan, R. G., & Zingales, L. (1998). Financial dependence and growth. ISSN: 00028282
Friberg, R. (2015). Managing risk and uncertainty: A strategic approach. MIT Press. American Economic Review, 88(3), 559–586. https://doi.org/10.2307/116849. ISSN:
Gao, G. Y., et al. (2018). Impact of historical conflict on FDI location and performance: 00028282 〈https://www.jstor.org/stable/116849〉.
Japanese investment in China. Journal of International Business Studies, 49, Rauch, J. E., & Trindade, V. (2002). Ethnic Chinese networks in international trade.
1060–1080. https://doi.org/10.1057/s41267-016-0048-6 Review of Economics and Statistics, 84(1), 116–130.
Giambona, E., Graham, J. R., & Harvey, C. R. (2017). The management of political risk. Samaniego, R. M. (2010). Entry, exit, and investment-specific technical change. ISSN:
Journal of International Business Studies, 48(4), 523–533. 00028282 American Economic Review, 100(1), 164–192. https://doi.org/10.1257/
Gray, C. S., & Sloan, G. (2014). Geopolitics, geography and strategy. Routledge. aer.100.1.164.
Guimón, J., et al. (2018). Policies to attract R&D-related FDI in small emerging countries: Samaniego, R. M., & Sun, J. Y. (2015). Technology and contractions: Evidence from
Aligning incentives with local linkages and absorptive capacities in Chile. Journal of manufacturing. European Economic Review, 79, 172–195.
International Management, 24(2), 165–178. Shi, W. S., et al. (2017). Institutional fragility and outward foreign direct investment
Guiso, L., Sapienza, P., & Zingales, L. (2009). Cultural biases in economic exchange. from China. ISSN:14786990 Journal of International Business Studies, 48(4), 452–476.
ISSN: 00335533 Quarterly Journal of Economics, 124(3). https://doi.org/10.1162/ https://doi.org/10.1057/s41267-016-0050-z. ISSN:14786990
qjec.2009.124.3.1095. 〈https://www-jstor-org.libproxy1.nus.edu.sg/stable/26170008?pq-origsite=360lin
Hassan, T. A., et al. (2019). Firm-level political risk: Measurement and effects. ISSN: k&seq=1#metadata_info_tab_contents〉.
15314650 Quarterly Journal of Economics, 134(4), 2135–2202. https://doi.org/ Silva, J. M. C. S., & Tenreyro, S. (2006). The log of gravity. The Review of Economics and
10.1093/qje/qjz021. ISSN: 15314650 〈https://academic.oup.com/qje/article/134/ Statistics, 88(4), 641–658.
4/2135/5531768〉. Simonin, B. L. (1999). Ambiguity and the process of knowledge transfer in strategic
alliances. Strategic Management Journal, 20(7), 595–623.
14
Sriram, C., J. Large, and S. Brown (2009). Governance in Conflict Prevention and Witte, C. T., et al. (2017). Dodging bullets: The heterogeneous effect of political violence
Recovery: A Guidance Note. United Nations Development Programme. on greenfield FDI. ISSN: 14786990 Journal of International Business Studies, 48(7),
Stephan, M., Silvia, M., & Arie Y, L. (2008). A dynamic perspective on next-generation 862–892. https://doi.org/10.1057/s41267-017-0079-7. ISSN: 14786990
offshoring: The global sourcing of science and engineering talent. Academy of 〈https://www-jstor-org.libproxy1.nus.edu.sg/stable/45149356?pq-origsite=360lin
Management Perspectives, 22(3), 35–54. k&seq=1#metadata_info_tab_contents〉.
Sun, P., Mellahi, K., & Wright, M. (2012). The contingent value of corporate political ties. Zaheer, S. (1995). Overcoming the liability of foreignness. Academy of Management
Academy of Management Perspectives, 26(3), 68–82. Journal, 38(2), 341–363.
Von Zedtwitz, M., et al. (2015). A typology of reverse innovation. Journal of Product Zhao, M. (2006). Conducting R&D in countries with weak intellectual property rights
Innovation Management, 32(1), 12–28. protection. Management Science, 52(8), 1185–1199.
Williamson, O. E. (1998). The institutions of governance. The American Economic Review, Zheng, H. (2020). Coordinated bubbles and crashes. Journal of Economic Dynamics and
88(2), 75–79. Control, 120, Article 103974.
Witt, M. A. (2019). De-globalization: Theories, predictions, and opportunities for
international business research. Journal of International Business Studies, 50(7),
1053–1077.
15

1 s2.0 S0969593123000367 Main

Uploaded by

Copyright:

Available Formats

1 s2.0 S0969593123000367 Main

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S0969593123000367 Main

Uploaded by

Copyright:

Available Formats

International Business Review xxx (xxxx) xxx

Contents lists available at ScienceDirect

International Business Review

Responses of FDI to geopolitical risks: The role of governance, information,

1. Introduction In response to the changing world order, do MNCs incorporate their

Table 1 3.1.2. Geopolitical risk and uncertainty

3.3. Methodology 3.3.3. The role of information

log(GPR) − 0.258 − 0.057 − 0.192 − 0.219 − 0.222 − 0.260 − 0.190

log(GPR) − 0.287 − 0.293 − 0.315 − 0.302 − 0.288 − 0.242 − 0.302

uncertainty. It would be interesting to explore how MNCs’ heteroge Data availability

Tables A.1, A.2, A.3, A.4.

Geopolitical risk Geopolitical uncertainty GPR

Table A.3 (continued )

RuleLaw 0.940 0.928 0.932 1.000

Argentina United States 18.7 Mexico United States 33.4

Appendix B. Supporting information References

You might also like

1 s2.0 S0969593123000367 Main

Uploaded by

Copyright:

Available Formats

1 s2.0 S0969593123000367 Main

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S0969593123000367 Main

Uploaded by

Copyright:

Available Formats

International Business Review xxx (xxxx) xxx

Contents lists available at ScienceDirect

International Business Review

Responses of FDI to geopolitical risks: The role of governance, information,

1. Introduction In response to the changing world order, do MNCs incorporate their

Table 1 3.1.2. Geopolitical risk and uncertainty

3.3. Methodology 3.3.3. The role of information

log(GPR) − 0.258 − 0.057 − 0.192 − 0.219 − 0.222 − 0.260 − 0.190

log(GPR) − 0.287 − 0.293 − 0.315 − 0.302 − 0.288 − 0.242 − 0.302

uncertainty. It would be interesting to explore how MNCs’ heteroge­ Data availability

Tables A.1, A.2, A.3, A.4.

Geopolitical risk Geopolitical uncertainty GPR

Table A.3 (continued )

RuleLaw 0.940 0.928 0.932 1.000

Argentina United States 18.7 Mexico United States 33.4

Appendix B. Supporting information References

You might also like

uncertainty. It would be interesting to explore how MNCs’ heteroge Data availability