Back in 2013, China initiated a new framework called the Belt and Road Initiative (BRI) to promote regional economic development. The BRI originally aimed to develop transport infrastructure from China through West Asia to Europe. However, it soon expanded to South Asia, Southeast Asia, Africa, and the rest of the world and became engaged in the development of other types of infrastructure, such as information and communication technology (ICT), energy, and mining. For example, in 2013, or the initial year of the BRI, the share of energy projects in total BRI engagement was 52 percent, whereas the share of transport projects was 17 percent. In 2023, the share of energy projects had declined to 31 percent, and the shares of transport and mining projects were 16 and 21 percent, respectively. By 2023, 150 countries had signed memorandums of understanding (MOUs) for the BRI, and the cumulative engagement for BRI-related investments exceeded 1 trillion US dollars. The BRI countries include most low- and middle-income countries, with several exceptions such as India, Bhutan, Brazil, and Mexico, and some developed countries, most of which are East European countries, except Portugal. Other developed countries, such as the United States (US), most West European countries, Japan, and Australia, have not yet signed an MOU for the BRI.
Although the development of energy and transport infrastructure is crucial, the BRI is also supposed to strengthen economic ties between China and participating countries, including trade and foreign direct investment (FDI). By applying difference-in-differences (DID) estimations to gravity models, several studies have indeed shown that participating in the BRI has a positive effect on the level of Chinese FDI in the partner country. These studies find that Chinese FDI in BRI countries has increased more significantly than that in other developing countries. However, the effect of the BRI is heterogeneous. For example, participating in the BRI is negatively correlated with Chinese investment in advanced economies.
In addition to FDI from China, the BRI may influence FDI from other countries through changes in the characteristics of the host country and bilateral relationships between the host and source countries. For example, infrastructure development due to the BRI could encourage FDI inflows regardless of the source country. Productivity growth in the host country due to BRI projects may also attract FDI from any country. In addition, participation in the BRI signals closer political alignment with China because China intends to assert greater international influence through the BRI. Therefore, bilateral relationships between BRI countries and non-China source countries may change and affect FDI, particularly when the source countries compete or, conversely, cooperate with China. However, whether participation in the BRI influences FDI from countries other than China has not been studied in the literature. Moreover, the literature does not fully examine the mechanisms behind the effect of the BRI, particularly failing to distinguish between effects through changes in host country characteristics, such as infrastructure, and changes in bilateral relationships due to strategic competition and investment cooperation with China.
Furthermore, the current literature relies on DID estimations that assume a single timing of the treatment. All countries that signed an MOU for the BRI participated in the BRI in 2013 when President Xi Jinping first announced the BRI. In other words, these studies assume that all BRI member countries participated in the BRI in a particular period and thus compare FDI from China between the pre- and post-participation periods. However, in practice, countries began to participate in the BRI in different periods, as we will show in detail later. Therefore, estimations assuming the single timing of the treatment may lead to biased results. In addition, the recently growing literature on staggered DID argues that two-way fixed effect DID estimations can be biased when the treatment effect is heterogeneous across treated periods.
This paper makes several contributions to the literature by filling these gaps. First, in addition to examining the effect of the BRI on FDI from China to BRI countries, we also estimate the effect of the BRI on FDI from major source countries, such as the US, the United Kingdom (UK), France, Germany, and Japan, some of which compete with China in economic and political relationships with BRI countries, while others are more cooperative with China. Second, we distinguish between the BRI effect through changes in host country characteristics and bilateral relationships by comparing the results from estimations with and without country-year fixed effects. The effect estimated in the specifications using country-year fixed effects that control for any time-varying host country attribute, such as the level of infrastructure and productivity, can be interpreted as the effect through changes in bilateral relationships.
The results indicate that FDI from China, Hong Kong SAR, the US, Switzerland, Japan, and France to BRI countries has increased in the post-BRI period, whereas FDI from the UK, the Netherlands, and Luxembourg has decreased. After controlling for host country-year fixed effects, the post-BRI upward trend in FDI from the US, Switzerland, and France and the downward trend in FDI from the UK, the Netherlands, and Luxembourg remain, whereas the positive effect on FDI from Japan disappears. These Findings suggest that changes in bilateral relationships are an important determinant of FDI to BRI countries. We presume that the US invests more in BRI countries to strategically compete with China, whereas French and Swiss firms cooperate with Chinese firms in investment projects in BRI countries supported by policies. In contrast, the UK has possibly reduced its FDI in BRI countries to mitigate the risks of supply chains, as BRI countries are strongly linked with China. FDI from Japan to BRI countries is most likely affected by host country characteristics rather than bilateral relationships.
the positive effect on FDI from the US, which has signed no MOU with China for third-party market cooperation, implies that US firms increase investment in BRI countries by following the policy support of the US government to strategically compete with China in BRI countries, such as BDN, BUILD, and B3 W. Although the US government also encourages friendshoring to like-minded countries and thus may discourage investment in countries closely linked with China, the positive effect through strategic competition is more likely to surpass the negative effect through risk mitigation.
Second, FDI from France and Switzerland to BRI countries has increased because the governments and private firms of both countries have signed several MOUs with China for third-party market cooperation and conducted investment jointly with China in BRI countries, particularly in Africa. Notably, France is engaged in cooperation with China most extensively among European countries. For example, Alstom France and China Water Conservancy and Hydro Power Group cooperate for hydropower projects in Uganda, Ghana, and Cote d’Ivoire. Additionally, the CMA CGM of France and the MSC Cruises of Switzerland collaborate with Hunan Road & Bridge for an infrastructure project in Benin). These examples support the positive effect of the BRI on FDI from the two countries.
Third, FDI from three other European countries, namely, the UK, the Netherlands, and Luxembourg, shows a downward trend in FDI in BRI countries. Although the UK and the Netherlands have signed an MOU with China for third-party market cooperation, they take a cautious approach to address China. For example, the UK government’s view of China and its BRI shifted from positive to substantially negative in 2019, formally stating “systemic challenges” from China as the reason for this shift. The Dutch government published “The Netherlands-China: A New Balance” in 2019 to emphasize cautious engagement with China, particularly in critical economic sectors. Therefore, these countries may account for the geopolitical risks of supply chains with unfriendly BRI countries more seriously than others do.
Finally, FDI from Japan or Germany does not show an upward or downward trend after controlling for country-year fixed effects, implying that the positive effect through third-party market cooperation and the negative effect through minimizing supply chains with unfriendly BRI countries cancel each other out. Although Japan has signed an MOU with China for third-party cooperation, practical cooperation between the two is reported to have faced challenges because of political interference and potential economic risks in BRI countries. Although many German companies, including Siemens and GAUFF, have collaborated with Chinese companies, the German government has not signed an MOU with China for third-party market cooperation. These observations are in line with the lack of a positive effect of the BRI on FDI. We observe a positive and significant effect of the BRI on FRI from Japan when we control for attributes of the host country. The different results between estimations using country attributes and country-year fixed effects suggest that FDI from Japan to BRI countries is seen as increasing because of changes in attributes of the host country that are not explicitly controlled for by our measures of productivity, ICT infrastructure, and governance, such as improvements in transport infrastructure and systems for investment.
This paper investigates the impact of the Belt and Road Initiative (BRI) on foreign direct investment (FDI) from China and other top 10 source countries, including the United States, the United Kingdom, Germany, and Japan, to BRI countries. We apply staggered difference-in-differences (DID) event study estimations to a gravity model. Our contributions to the literature are threefold. First, we estimate the effect of the BRI on FDI from countries other than China. Second, to highlight the mechanisms of the BRI effect on FDI from non-China countries, we distinguish between the effects of the BRI through changes in the characteristics of the host country and changes in the bilateral relationships between the source and host countries by utilizing country attributes and fixed effects at various levels. Changes in host country characteristics include improvements in productivity and infrastructure, whereas bilateral relationships are affected by strategic competition and investment cooperation with China and the mitigation of risks in supply chains with BRI countries closely linked with China. Finally, we examine the heterogeneous effects of the BRI on FDI in democratic and autocratic countries.
Our results using country-pair and year fixed effects show that signing an MOU for the BRI with China significantly affects FDI from China and Hong Kong SAR to BRI countries, which is consistent with results from the literature. We also find a positive and significant effect of the BRI on FDI from the US and Switzerland and a negative and significant effect on FDI from the UK several years after participating in the BRI. We observe an upward trend of FDI from Japan and France and a downward trend of FDI from the Netherlands and Luxembourg in the post-BRI period, although these effects are not statistically significant. We further control for source and host country fixed effects and still observe an upward post-BRI trend in FDI from the US, Switzerland, and France and a downward trend in FDI from the UK, the Netherlands, and Luxembourg. Country-year fixed effects represent effects through changes in any characteristic of the host country due to the BRI, including market size, productivity, infrastructure, and governance. Therefore, these results from estimations with country-year fixed effects imply that changes in bilateral relationships due to the BRI have affected FDI from some non-China countries.
From anecdotal evidence of policies and business practices, we conclude that the US has increased its FDI in BRI countries to strategically compete with China over the economic and political presence there. France and Switzerland have done so because of their active cooperation with China in private investment in BRI countries, particularly in Africa. By contrast, the UK and the Netherlands, which have taken a cautious approach to relationships with China, have decreased their FDI in BRI countries, possibly to reduce the risk of supply chains with countries close to China. While FDI from Japan to the BRI has increased, this is possibly because of changes in host country characteristics rather than changes in bilateral relationships.
Furthermore, when we distinguish between subsamples depending on whether the host country was initially democratic or autocratic, we find that FDI from the US to autocratic BRI countries has increased more than FDI to democratic countries has not. This finding confirms our conclusion that the US has increased its FDI in BRI countries because of strategic competition with China.
Finally, we note several caveats of this paper. First, although our estimations using source and host country-year fixed effects show an upward or downward trend in FDI from several source countries after participating in the BRI, the effect is often insignificant at the 5-percent level because of large standard errors, suggesting substantial heterogeneity in the BRI effect. Although we examine heterogeneity depending on the level of democracy, there could be other sources of heterogeneity. Examining heterogeneity more deeply may reveal more significant effects of the BRI. Second, our analysis reveals that changes in bilateral relationships have affected FDI from non-China countries to BRI countries and suggests different reasons for such changes, i.e., strategic competition, investment cooperation, and the mitigation of supply chain risk, on the basis of anecdotal evidence. However, quantitative analysis could provide clearer evidence for the mechanism for each country. Finally, this study does not consider spillover effects, i.e., a possible negative effect of a host country’s participation in the BRI on FDI in non-BRI countries due to diversion from non-BRI countries to BRI countries. Therefore, the estimated effect of the BRI in this study should be interpreted as the net effect of the BRI, i.e., the pure effect of the BRI on FDI from a source country to a BRI country minus the average spillover effect on FDI from the source country to non-BRI countries.
By Yasuyuki Todo, Shuhei Nishitateno and Sean Brown for Science Direct.
