Policy Research Working Paper 11006 Is the U.S. Friend-Shoring, Nearshoring, or Reshoring? Evidence from Greenfield Investment Announcements Alen Mulabdic Gaurav Nayyar Prosperity Vertical & A verified reproducibility package for this paper is Prospects Group available at http://reproducibility.worldbank.org, December 2024 click here for direct access. Policy Research Working Paper 11006 Abstract This paper examines the evolution of greenfield investment associated with greenfield investment announcements in announcements—both domestic and international—for the US and its neighboring countries. The paper finds no US multinational companies in response to recent global evidence that US companies are adopting a friend-shoring shocks. The results indicate an intensification of reshoring strategy by investing more in military allies. The paper sug- and nearshoring activities by US companies, especially fol- gests that US supply chains are likely to become less global lowing the Russian Federation’s invasion of Ukraine. This and more regional as these investments become operational. shift is estimated to have doubled the number of direct jobs This paper is a product of the Office of the Chief Economist, Prosperity Vertical and the Prospects Group, Development Economics . It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http:// www.worldbank.org/prwp. The authors may be contacted at amulabdic@worldbank.org and gnayyar@worldbank.org. A verified reproducibility package for this paper is available at http://reproducibility.worldbank.org, click here for direct access. RESEA CY LI R CH PO TRANSPARENT ANALYSIS S W R R E O KI P NG PA The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Is the U.S. Friend-Shoring, Nearshoring, or Reshoring? Evidence from Greenfield Investment Announcements☆ Alen Mulabdic (World Bank) Gaurav Nayyar (World Bank) Keywords: Foreign direct investment, reshoring, friend-shoring, nearshoring, home-bias JEL Codes: F2; F21; F23; F52; F6 ☆ We are grateful to Amat Adarov, Mirco Balatti, Jakob de Haan, Ayhan M. Kose, Aart Kraay, Michele Ruta, Kersten Stamm and seminar participants at the World Bank for comments and conversations. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. The authors may be contacted at amulabdic@worldbank.org and gnayyar@worldbank.org. 1. Introduction Are multinationals likely to reverse 30 years of globalization? Will global value chains that have benefited some developing countries become increasingly regional? How are businesses reacting to friend-shoring requests by policy makers? The reconfiguration of global value chains can have significant economic implications for the world. For instance, Cerdeiro et al. (2021) estimate that technological decoupling can lead to GDP losses of 5 percent for many countries. At the same time, Brenton et al. (2022) predict that reshoring production to high-income countries and China could drive 52 million people in developing countries into extreme poverty. Rising geopolitical tensions between the US and China, the Russian Federation’s invasion of Ukraine, and the Covid-19-related supply chain disruptions have led some policy makers and businesses to revisit the benefits of globalization. These events have also highlighted the risks associated with overdependence on countries that are not considered strategic allies. For instance, US Secretary of the Treasury Janet Yellen called US firms to increase friend-shoring with the aim of having production spread across countries “[…]that have strong adherence to a set of norms and values about how to operate in the global economy and about how to run the global economic system” to reduce trade partners’ geopolitical leverage from market positions in critical raw materials, technologies, or products that could disrupt the US economy. 1 Recent work shows that FDI is increasingly concentrated among countries with similar UN voting patterns (Aiyar et al., 2023; IMF, 2023). 2 In the specific case of the US, Alfaro and Chor (2023) document an increase in greenfield FDI 1 https://home.treasury.gov/news/press- releases/jy0714#:~:text=We%20cannot%20allow%20countries%20to,work%20better%20for%20American %20workers. 2 Aiyar et al. (2023) estimate a gravity model using the information on Ideal Point Distances (IPD) developed by Bailey et al. (2017) to show that “geopolitical distance” is an important determinant of FDI during the 2003-2021 period. 2 by US multinationals in Mexico in recent years, particularly in the automobile and electronics sectors. However, these studies only provide anecdotal evidence of reshoring. The recent Infrastructure Investment and Jobs Act (IIJA) and the Inflation Reduction Act (IRA) have also increased the incentives to relocate production to the US through a combination of tax credits and grants requiring domestic or North American content. For instance, the IRA Section 45X MPTC Advanced Manufacturing Tax Credit applies to components for wind, solar, and battery projects manufactured in the US. To be eligible for the domestic content bonus credit, energy projects and facilities are subject to a domestic content requirement of 40 percent until 2025, which increases to 55 for facilities for which construction begins after 2026. 3 According to Secretary of Energy Jennifer M. Granholm, the rationale for these local content requirements is driven by the aim of “[…] strengthening American manufacturing and enhancing our national security with products stamped ‘Made in the USA.’” 4 To our knowledge, this is the first paper to formally analyze the extent of reshoring by US multinationals, complementing the recent trade literature on the consequences of supply chain disruptions. The paper offers novel insights about the future of US supply chains using information on announcements of greenfield investment projects from the Financial Times fDi Markets database, including projects within the US. The database provides real- time information on cross-border greenfield investments gathered from various sources, including the Financial Times newswire, industry organizations, investment agencies, and company reports. As a result, unlike other data on investments, this information is forward- looking as it captures announcements about new projects and expansions. Furthermore, unlike other studies, the data allow us to leverage information on domestic investment projects between US states to assess reshoring activities. 3 https://www.irs.gov/pub/irs-drop/n-23-38.pdf 4 https://home.treasury.gov/news/press-releases/jy1477 3 First, we analyze the pattern of US investment over time through an event study approach that investigates whether the coefficients on distance, border, NATO, and same country variables have changed. We find limited evidence of nearshoring or friend-shoring as the coefficients of the dummy variables for border and NATO appear relatively stable over time. However, the data suggest an intensification in reshoring activities among US companies. Specifically, the same country dummy, which captures how much companies headquartered in the US are investing domestically, became positive and statistically significantly different in 2020 and further increased in magnitude in 2023. Second, we implement a triple-difference approach comparing US investments to EU investments over time to identify differences in their patterns following the beginning of the US-China trade war, the Covid-19 pandemic, and the Russian invasion of Ukraine. This approach allows us to control for all source country-time, destination country-time, and country-pair unobservable characteristics. The results confirm that US companies intensified their reshoring activities. In addition, we find that, compared to EU firms, US companies also increased nearshoring following the Covid-19 shock. The estimates indicate that US investment in neighboring countries and within the US states doubled during the post Russia’s invasion of Ukraine period, with slightly larger effects for US investment in the US. Finally, we exploit the richness of the fDi Markets database to assess the impact of the shocks on direct job creation associated with greenfield investment announcements. This allows us to determine whether the increase in investment announcements is associated with job creation in the US or if these investments are likely leading to jobless growth in the manufacturing sector with highly automated factories. The results suggest that recent investments are associated with job creation in both the US and neighboring countries. For example, following the invasion of Ukraine, the number of direct jobs in greenfield investment announcements were more than double in the US and its neighbors. These results combined with the ones for investment, may suggest that more labor-intensive projects are being directed to Mexico. 4 This paper contributes to different strands of the literature. First, it contributes to the recent literature on the economic effects of the US-China trade tensions by expanding the analysis to investment, jobs, and more recent shocks. Several studies investigate the impact of tariff increases on trade, prices, and welfare (Alfaro and Chor, 2023; Amiti et al., 2019; Fajgelbaum et al., 2020; Flaaen et al., 2020; Freund et al., 2024; Gopinath et al., 2024). Second, the paper contributes to the literature that analyzes the effects of shocks on the reshaping of global value chains (Antràs, 2020; Freund et al., 2022; Javorcik, 2020). Third, it documents the importance of shocks in influencing FDI decisions (Aiyar et al., 2023; Alfaro and Chen, 2018; Crescenzi et al., 2021; Harding and Javorcik, 2011; Kox and Rojas‐ Romagosa, 2020). In addition, to the best of our knowledge, this is the first paper that quantifies the home bias in greenfield investment. Home bias has been documented in goods trade (e.g., Anderson and Yotov, 2010; McCallum, 1995) and more recently in government procurement (e.g., García-Santana and Santamaría, 2023; Herz and Varela- Irimia, 2020; Mulabdic and Rotunno, 2022; Rickard and Kono, 2014). The paper is organized as follows. The next section describes the empirical strategy and data. Section 3 presents our results. Section 4 presents a series of robustness tests, while section 5 extends the analysis to jobs. Section 6 concludes. 2. Empirical methodology and data This section describes the empirical strategy and data used to analyze changes in greenfield investment patterns following the start of US-China trade tensions, the Covid-19 pandemic, and Russia’s invasion of Ukraine. 2.1. Empirical strategy I: Event study approach As a first exercise, we implement an event study approach and estimate a gravity-like equation and investigate whether the coefficients on distance, border, NATO, and same country variables have changed over time. Specifically, we rely on the following event study specification: 5 2023 = � � ( = ) + + (1) =1 =2015 where is the log of greenfield investment plus one from the US to country in month in a specific year. is a row vector of gravity variables, which include an indicator variable equal to one if country is a member of the North Atlantic Treaty Organization (NATO), the log of the population-weighted distance between the US and country , an indicator variable equal to one if country is the US, a dummy variable equal to one if country is an English speaking country, as well as the log of country j’s GDP and population in the previous year, and time fixed effects . The indicator variable for alliance is used as a proxy for the importance of “friends” or strategic military allies. A positive coefficient that increases over time, especially after the Russian invasion of Ukraine, would suggest an intensification of friend-shoring in nations that provide higher security to US businesses. The coefficients on and capture the sensitivity of investment announcements to physical proximity to the US. Larger negative coefficients for or larger positive coefficients for in recent years would indicate an intensification of nearshoring activity by US companies. Finally, the coefficient on the dummy variable identifies the home bias of the greenfield investment from the US, reflecting the extent to which investment within US national borders differs from investment to other countries, capturing the importance of reshoring. Additionally, we also include controls for population and GDP to account for destination market size and economic conditions. 2.2. Empirical strategy II: Triple-difference approach In a second exercise, we take advantage of the investment data, which allow us to compare US and EU investment flows—the only economies for which we have internal investment. We implement a triple-difference approach and compare the coefficients for distance, 6 border, NATO, and same country variables for the US with respect to the EU over time, while controlling for time-varying shocks specific to destination markets. Specifically, we rely on the following specification: = � + � () + � ( ≥ ℎ) =1 =1 =1 (2) + � () ( ≥ ℎ) + + + + =1 where is the log of greenfield investment from the US or EU in country at time . is a row vector of gravity variables included in specification (1). The vector of coefficients captures the effect of gravity variables for the EU in the baseline period before the shocks. The vector captures the differential effect of the gravity variables for the US during the same control period. The vector identifies changes in the effects of the gravity variables for the EU after the global shocks, while identifies the additional effects for the US. Non-statically significant coefficients in vector would show that US investment reacts similarly to EU investment in the post-shock period. To capture the cumulative effects of various shocks, we restrict the sample to specific periods preceding the next shock. For example, to assess the impact of US-China trade tensions, we limit the sample to the period from January 2015 to February 2020, just before the WHO declared Covid-19 a pandemic on March 11, 2020. To evaluate the effects of the Covid-19 pandemic, we estimate equation (2) using data from January 2015 to January 2022, excluding the post-invasion of Ukraine period. Finally, to analyze the impact of Russia’s invasion of Ukraine, we use the full sample from January 2015 to December 2023. The advantage of the triple-difference approach is that it allows us to include a rich set of fixed effects to control for additional unobservable characteristics. First, we include to control for source country-time fixed effects to control for any US- or EU-specific determinants of outward greenfield investment, such as domestic economic conditions. 7 Second, as we observe investment from both the EU and the US, we can include destination country-time fixed effects to control for any pull factors such as business-friendly reforms in destination markets . Finally, we include country-pair fixed effects to control for any time-invariant characteristics determining bilateral investment between and . 2.3. Data Information on bilateral investment comes from the Financial Times fDi Markets database. The data cover cross-border greenfield investment announcements, including new projects and the expansion of existing projects, at a monthly frequency. In addition to the number of projects, the database provides information on the value of investments. An important feature of the database is that it also tracks US inter-state investments, which allows us to study the impact of recent shocks on reshoring activity by US multinationals. 5 The fDi Markets data have been used extensively in recent research projects and for surveillance purposes by policy institutions, showing that the data tracks well official FDI flows (Aiyar et al., 2023; Alfaro and Chor, 2023; IMF, 2023; Toews and Vézina, 2022; UNCTAD, 2023). The data also cover information on the number of jobs associated with greenfield investment announcements. Toews and Vézina (2022) show that, in the case Mozambique, fDi Markets data is correlated with census data across different sectors, cities, and time but underestimates the total number of jobs created. The greenfield investment data is combined with standard gravity variables from the CEPII’s gravity database. These variables include the bilateral geographical distance and information on whether countries share the same language, border, or colonial history. Finally, we also construct an indicator variable to capture if country-pairs are part of the 5 An important caveat is that the data does not cover intra-US state greenfield investments, which may account for a large share of greenfield investment at the state level. 8 North Atlantic Treaty Organization (NATO). 6 The final sample covers EU and US investment between January 2015 and December 2022 to 156 destination markets. A first inspection of US investment data in Figure 1 shows no indication of permanent shifts in nearshoring or friend-shoring following the trade tension with China (2018-19), the Covid-19 pandemic (2020-21), and the Russian invasion of Ukraine (2022-23). 7 The share of greenfield foreign direct investment in neighboring and NATO countries declined by a few percentage points when comparing the 2015-17 and 2022-23 periods (Figure 1). However, reshoring activity by US multinationals intensified following the Covid-19 pandemic, with the share of greenfield investment in the US increasing from 46 percent to 56 percent, between the 2015-17 and 2020-21 periods. This evidence corroborates the anecdotal evidence of US manufacturers starting to bring back factories to the US. 8 These investments are likely to have an impact on trade flows once they become operational. 6 List of NATO countries: Belgium (1949), Canada (1949), Denmark (1949), France (1949) Iceland (1949), Italy (1949), Luxembourg (1949), Netherlands (1949), Norway (1949), Portugal (1949), United Kingdom (1949), United states (1949), Greece (1952), Türkiye (1952), Germany (1955), Spain (1982), Czechia (1999), Hungary (1999), Poland (1999), Bulgaria (2004), Estonia (2004), Latvia (2004), Lithuania (2004), Romania (2004), Slovak Republic (2004), Slovenia (2004), Albania (2009), Croatia (2009), Montenegro (2017), and North Macedonia (2020). We assume that the EU is a NATO member even if some of the EU countries are not part of the alliance. 7 These results are similar to Aiyar et al. (2023) that show no impact of geopolitical distance for advanced economies. 8 See for instance Bloomberg “US Manufacturers ‘Pumped Up’ About Supply-Chain Reshoring Trend” (November 2, 2022) https://www.bloomberg.com/news/articles/2022-11-02/us-manufacturers-pumped-up- about-supply-chain-reshoring-trend and the Wall Street Journal “U.S. Companies on Pace to Bring Home Record Number of Overseas Jobs” (August 23, 2022) https://www.wsj.com/articles/u-s-companies-on-pace- to-bring-home-record-number-of-overseas-jobs-11660968061. 9 Figure 1: US investment patterns Note: The graph shows average shares based on monthly investment data for the US. The bars for NATO and neighboring countries are presented as a share of foreign investment. In the next section we take a more systematic look at the data by identifying the effects of the various global shocks on the sensitivity of investment announcements with respect to distance, border, and international borders, while controlling for additional determinants of investment flows. 3. Baseline results This section presents the results for the year-by-year and triple-difference specifications. 3.1. Evolution of determinants of US investment flows: Year-by-year To understand if there is an acceleration in nearshoring, friend-shoring, and reshoring trends, we inspect the evolution of the coefficients on the border, NATO, and same country variables. Below are the results obtained from estimating equation (1): 10 Figure 2: Evolution of coefficients for US investment Note: The graph shows point estimates of the same country, NATO, and border dummies on investment obtained by estimating equation (1) and their associated 95% confidence intervals. Figure 2 confirms that there is no systematic evidence of nearshoring. The coefficient on the border variable for US investment flows in recent years is not significantly different from the coefficients in the pre-trade war period (2017), with the coefficient on the border dummy not changing during the entire period. Figure 2 also shows that US investment was not sensitive to military alliances, a proxy for friend-shoring. The importance of these alliances for US greenfield investment announcements has not increased in recent years; if anything, it decreased in 2023. 9 Finally, Figure 2 shows that the importance of the same country variable was decreasing between 2015 and 2018, which is indicative of a decline in the home bias of US greenfield investment announcements. However, this trend was reversed beginning with the Covid- 19 pandemic. The coefficient on the same country variable became statistically significant in 2020, compared to 2017, and increased further in 2023, suggesting that US 9 Results are robust to alternative measures of geopolitical friends using UN voting patterns or formal military alliances from Correlates of War (Gibler, 2009). 11 multinationals increased investment in the US during the Covid-19 pandemic, with further increases following Russia’s invasion of Ukraine. 3.2. Evolution of the determinants of US investment flows: Triple difference One potential issue with the event study approach is that the results could be driven by unobservable characteristics such as global trends or country-specific shocks. For instance, nearshoring of production may be happening, but given that future economic prospects in Asia may be improving, an analysis for the US only would not capture this trend. Also, the reshoring results could be driven by exceptionally favorable investment conditions in the US. Thus, we employ a triple-difference approach where we compare US investment to EU27 investment over time. This specification allows us to control both for source- and destination-time fixed effects as well as county-pair fixed effects. Table 1 reports the estimates from equation (2). Results point to strong evidence of nearshoring for the US FDI compared to the EU. The border dummy increased both in terms of magnitude and statistical significance following the outset of the Covid-19 pandemic (column 2). Estimates from column (3) show that the border coefficient decreased slightly after Russia’s invasion of Ukraine from 0.804 to 0.643. This translates to investment being 90 percent higher in neighboring countries compared to others. 10 This large effect and absence of significance during the trade war period, can be explained by the uncertainty related to the North American Free Trade Agreement (NAFTA) and subsequent ratification of the United States–Mexico–Canada Agreement (USMCA), which came into effect on July 1, 2020, and stimulated US investment in Canada and Mexico. The triple-difference approach confirms the intensification of reshoring activity by US companies which intensified over time. The coefficient of the same country dummy variables for the US in the post-invasion of Ukraine period is 0.795. This implies that, 10 exp(0.643) − 1 = 0.902. 12 everything else being equal, the value of investment between US states in the post-shock period, compared to the EU, is more than double than investment between the US and other countries. 11 Similarly to the event study approach, we do not find evidence of an increase in nearshoring, proxied by distance, or friend-shoring, proxied by NATO. If anything, investment by US multinationals in countries that are military allies of the US decreased over time. 11 (0.795) − 1 = 1.214. 13 Table 1: Triple difference (1) (2) (3) ln(FDI) Trade War Covid-19 Invasion of Ukraine I(t>Event) x I(US) x … … x I(Same country) 0.583* 0.781** 0.795*** (0.337) (0.306) (0.281) … x I(Border) 0.233 0.804** 0.643** (0.417) (0.336) (0.300) … x ln(distance) 0.236** 0.034 0.078 (0.117) (0.090) (0.077) … x I(NATO) -0.067 -0.371* -0.485*** (0.218) (0.200) (0.165) … x I(Common language) -0.099 -0.192** -0.186** (0.116) (0.089) (0.081) I(t>Event) x … … x I(Same country) -0.516* -0.490* -0.486** (0.311) (0.256) (0.237) … x I(Border) -0.251 -0.238 -0.257 (0.233) (0.183) (0.164) ... x ln(distance) -0.311*** -0.105 -0.116* (0.086) (0.070) (0.060) ... x I(Common language) 0.141 0.143* 0.089 (0.110) (0.084) (0.075) Observations 20,088 27,540 34,992 R-squared 0.865 0.863 0.862 Note: The regression includes source country-time, host country-time, and source country-host country fixed effects. The sample covers bilateral US and EU investment. In column (1), the sample covers the period from January 2015 to February 2020 (Trade War); in column (2), from January 2015 to January 2022 (COVID- 19); and in column (3), from January 2015 to December 2023 (Invasion of Ukraine). Robust standard errors, clustered at the source-time level, are reported in parentheses. *** p<0.01, ** p<0.05, * p<0.1 4. Robustness analysis This section presents a series of robustness checks. First, to handle the zero investment flows, we use the Poisson pseudo maximum likelihood (PPML) estimator. Second, to exclude that our results are driven by phantom investment in small countries, we estimate the baseline regressions on the sample of countries with a population greater than 1 million. 14 Table 2 presents the robustness tests for the value of investment. Columns (1) to (3) present the results of estimating equation (2) with a PPML estimator. The results for reshoring and nearshoring are robust to the inclusion of zeros. The results show that investment in the US and neighboring countries has increased over time, particularly following the Covid-19 and Russia’s invasion of Ukraine. The estimated effects are larger than those in Table 2, with investment in the US increasing by a factor of six, while investment in neighboring countries increasing fourfold. Columns (4) to (6) present the results of the baseline specifications estimated using the sub-sample of countries with a population greater than 1 million. This smaller sample excludes small countries such as the Bahamas and Cayman Islands, which may attract investment without having strong economic fundamentals. The results are qualitatively and quantitively similar to those presented in Table 1. The nearshoring and reshoring activities start with Covid-19 and accelerate with the invasion of Ukraine in the case of reshoring. 15 Table 2: Robustness PPML and small countries (1) (2) (3) (4) (5) (6) PPML Small countries excluded FDI ln(FDI) Invasion of Invasion of Trade War Covid-19 Ukraine Trade War Covid-19 Ukraine I(t>Event) x I(US) x … … x I(Same country) 1.172 1.389** 1.832*** 0.613 0.888** 1.087*** (0.785) (0.676) (0.702) (0.489) (0.397) (0.362) … x I(Border) 0.562 1.084* 1.396** 0.264 0.923** 0.904** (0.650) (0.651) (0.651) (0.537) (0.419) (0.372) … x ln(distance) 0.236 0.241 0.377 0.243 0.010 0.135 (0.346) (0.319) (0.342) (0.201) (0.142) (0.125) … x I(NATO) -0.430 -0.370 -0.539 -0.066 -0.517** -0.676*** (0.412) (0.394) (0.341) (0.244) (0.237) (0.200) … x I(Common language) -0.299 -0.377** -0.374** -0.063 -0.159 -0.165* (0.184) (0.174) (0.175) (0.134) (0.101) (0.093) I(t>Event) x … … x I(Same country) -0.453 -0.451 -0.724 -0.531 -0.632** -0.775*** (0.600) (0.524) (0.514) (0.386) (0.305) (0.281) … x I(Border) -0.246 -0.254 -0.540 -0.279 -0.325 -0.418** (0.470) (0.474) (0.457) (0.275) (0.216) (0.194) ... x ln(distance) 0.115 0.138 0.102 -0.318*** -0.108 -0.155** (0.171) (0.197) (0.210) (0.116) (0.087) (0.076) ... x I(Common language) 0.068 0.009 -0.112 0.129 0.133 0.057 (0.239) (0.219) (0.219) (0.118) (0.097) (0.090) Observations 7,622 10,006 12,678 15,996 21,930 27,864 R-squared 0.861 0.860 0.859 Note: The regression includes source country-time, host country-time, and source country-host country fixed effects. The sample covers bilateral US and EU investment. In columns (1) and (4), the sample covers the period from January 2015 to February 2020 (Trade War); in columns (2) and (5), from January 2015 to January 2022 (Covid-19); and in columns (3) and (6), from January 2015 to December 2023 (Invasion of Ukraine). Robust standard errors, clustered at the source-time level, are reported in parentheses. *** p<0.01, ** p<0.05, * p<0.1 5. Extensions to jobs This section expands the analysis to assess the impact of recent shocks on estimated jobs created directly through greenfield investment projects. Given the identified reshoring trend, we test for similar trends using information on the number jobs in the US associated 16 with these investments. If a small number of capital-intensive projects drives the investment results, we would expect a limited impact on direct job creation associated with greenfield investment in the US or neighboring countries. Table 3 reports the results of estimating equation (2), where the dependent variable is the number of jobs created by the announced investment projects instead of the value of projects. All specifications include source-time, destination-time, source-destination fixed effects to control for country-specific time-varying shocks as well as country-pair time invariant characteristics. Results in column (1) show no increase in the number of jobs associated with greenfield announcements in the US, Canada, or Mexico following the beginning of the trade war. This could be the effect of the high uncertainty created by the renegotiation of the North American Free Trade Agreement (NAFTA), which diminished after the ratification of the United States–Mexico–Canada Agreement (USMCA) in March 2020. The border and same country variables become significant in columns (2) and (3) following the Covid-19 and Russia’s invasion of Ukraine shocks. The number of jobs associated with greenfield announcements doubled following Russia’s invasion of Ukraine both in the US and neighboring countries. These effects are similar to those observed for investment value, as presented in Table 1, but slightly stronger for neighboring countries, which could be indicative of more labor intensive projects being directed to Mexico. This evidence suggests that contrary to widespread belief, US multinationals are not merely repatriating factories but are also bringing back some jobs. 17 Table 3: Number of jobs (1) (2) (3) ln(Jobs) Trade War Covid-19 Invasion of Ukraine I(t>Event) x I(US) x … … x I(Same country) 0.469 0.697** 0.723** (0.439) (0.353) (0.322) … x I(Border) 0.424 0.917** 0.723** (0.454) (0.352) (0.317) … x ln(distance) 0.374** 0.151 0.193* (0.151) (0.113) (0.098) … x I(NATO) 0.147 -0.229 -0.367* (0.306) (0.251) (0.217) … x I(Common language) -0.180 -0.267*** -0.262*** (0.130) (0.100) (0.091) I(t>Event) x … … x I(Same country) -0.439 -0.421 -0.589** (0.364) (0.303) (0.282) … x I(Border) -0.438 -0.367* -0.374* (0.276) (0.214) (0.198) ... x ln(distance) -0.457*** -0.187** -0.193** (0.107) (0.086) (0.075) ... x I(Common language) 0.195 0.186* 0.124 (0.135) (0.101) (0.090) Observations 20,088 27,540 34,992 R-squared 0.863 0.863 0.862 Note: The regression includes source country-time, host country-time, and source country-host country fixed effects. The sample covers bilateral US and EU investment. The sample covers bilateral US and EU investment. In column (1), the sample covers the period from January 2015 to February 2020 (Trade War); in column (2), from January 2015 to January 2022 (COVID-19); and in column (3), from January 2015 to December 2023 (Invasion of Ukraine). Robust standard errors, clustered at the source-time level, are reported in parentheses. *** p<0.01, ** p<0.05, * p<0.1 18 6. Conclusion This paper uses data on monthly greenfield investment announcements from 2015 to 2023 to study how US-China trade tensions, the Covid-19 pandemic, and the Russian invasion of Ukraine are reshaping US investment. The paper shows that US multinationals reacted to recent shocks by mainly reshoring their activities. When comparing US multinationals to their EU counterparts, we find that investment patterns differed following the recent shocks. The evidence indicates that the reshoring trend by US multinationals started with the Covid-19 shock and accelerated with the Russia’s invasion of Ukraine. An increase in greenfield investment in neighboring countries has accompanied this reshoring trend. Furthermore, the results suggest that reshoring and nearshoring of investment projects is also associated with the reshoring and nearshoring of jobs. These findings are particularly important for developing countries as a retreat from globalization by US multinational corporations could negatively affect their growth prospects. FDI has played an important in providing capital to, and fostering productivity growth, in developing countries (e.g., Alfaro and Chauvin, 2020; Smarzynska Javorcik, 2004). The linkages between multinational enterprises and domestic firms, especially in global value chains, have been associated with knowledge and technology spillovers (e.g., Dasgupta, 2012; Sinani and Meyer, 2004), which play a fundamental role in income convergence across countries. 19 References Aiyar, S., Malacrino, D., Presbitero, A., 2023. Investing in Friends: The Role of Geopolitical Alignment in FDI Flows (CEPR Discussion Paper No. 18434). Alfaro, L., Chauvin, J., 2020. Foreign direct investment, finance, and economic development. Faculty & Research 1, 231–258. Alfaro, L., Chen, M.X., 2018. Transportation cost and the geography of foreign investment. Chapters 369–406. Alfaro, L., Chor, D., 2023. Global Supply Chains: The Looming “Great Reallocation.” Working Paper Series. https://doi.org/10.3386/w31661 Amiti, M., Redding, S.J., Weinstein, D.E., 2019. The impact of the 2018 tariffs on prices and welfare. Journal of Economic Perspectives 33, 187–210. Anderson, J.E., Yotov, Y.V., 2010. The Changing Incidence of Geography. American Economic Review 100, 2157–86. Antràs, P., 2020. De-Globalisation? Global value chains in the post-COVID-19 age. National Bureau of Economic Research. Bailey, M.A., Strezhnev, A., Voeten, E., 2017. Estimating Dynamic State Preferences from United Nations Voting Data. The Journal of Conflict Resolution 61, 430–456. Brenton, P., Ferrantino, M.J., Maliszewska, M., 2022. Reshaping global value chains in light of covid-19: Implications for trade and poverty reduction in developing countries. World Bank Publications. Cerdeiro, M.D.A., Mano, R., Eugster, J., Muir, M.D.V., Peiris, M.S.J., 2021. Sizing Up the Effects of Technological Decoupling. IMF Working Papers, IMF Working Papers. Crescenzi, R., Di Cataldo, M., Giua, M., 2021. FDI inflows in Europe: Does investment promotion work? Journal of International Economics 132, 103497. https://doi.org/10.1016/j.jinteco.2021.103497 Dasgupta, K., 2012. Learning and knowledge diffusion in a global economy. Journal of International Economics 87, 323–336. https://doi.org/10.1016/j.jinteco.2011.11.012 Fajgelbaum, P.D., Goldberg, P.K., Kennedy, P.J., Khandelwal, A.K., 2020. The return to protectionism. The Quarterly Journal of Economics 135, 1–55. 20 Flaaen, A., Hortaçsu, A., Tintelnot, F., 2020. The production relocation and price effects of US trade policy: the case of washing machines. American Economic Review 110, 2103–27. Freund, C., Mattoo, A., Mulabdic, A., Ruta, M., 2024. Is US trade policy reshaping global supply chains? Journal of International Economics 152, 104011. https://doi.org/10.1016/j.jinteco.2024.104011 Freund, C., Mattoo, A., Mulabdic, A., Ruta, M., 2022. Natural Disasters and the Reshaping of Global Value Chains. IMF Economic Review 70, 590–623. https://doi.org/10.1057/s41308-022-00164-w García-Santana, M., Santamaría, M., 2023. Understanding Home Bias in Procurement: Evidence from National and Subnational Governments. https://doi.org/10.1596/1813-9450-10311 Gibler, D.M., 2009. International Military Alliances, 1648-2008. CQ Press. Gopinath, G., Gourinchas, P.-O., Presbitero, A., Topalova, P.B., 2024. Changing Global Linkages: A New Cold War? Harding, T., Javorcik, B.S., 2011. Roll Out the Red Carpet and They Will Come: Investment Promotion and FDI Inflows*. The Economic Journal 121, 1445–1476. https://doi.org/10.1111/j.1468-0297.2011.02454.x Herz, B., Varela-Irimia, X.-L., 2020. Border effects in European public procurement. Journal of Economic Geography 20, 1359–1405. https://doi.org/10.1093/jeg/lbaa001 IMF, 2023. World economic outlook: A Rocky recovery. IMF. Javorcik, B., 2020. Global supply chains will not be the same in the post-COVID-19 world, in: COVID-19 and Trade Policy: Why Turning Inward Won’t Work. CEPR Press. Kox, H.L., Rojas‐Romagosa, H., 2020. How trade and investment agreements affect bilateral foreign direct investment: Results from a structural gravity model. The World Economy 43, 3203–3242. McCallum, J., 1995. National Borders Matter: Canada-U.S. Regional Trade Patterns. The American Economic Review 85, 615–623. Mulabdic, A., Rotunno, L., 2022. Trade barriers in government procurement. European Economic Review 148, 104204. https://doi.org/10.1016/j.euroecorev.2022.104204 21 Rickard, S.J., Kono, D.Y., 2014. Think globally, buy locally: International agreements and government procurement. The Review of International Organizations 9, 333–352. https://doi.org/10.1007/s11558-013-9177-x Sinani, E., Meyer, K.E., 2004. Spillovers of technology transfer from FDI: the case of Estonia. Journal of Comparative Economics 32, 445–466. https://doi.org/10.1016/j.jce.2004.03.002 Smarzynska Javorcik, B., 2004. Does Foreign Direct Investment Increase the Productivity of Domestic Firms? In Search of Spillovers Through Backward Linkages. American Economic Review 94, 605–627. https://doi.org/10.1257/0002828041464605 Toews, G., Vézina, P.-L., 2022. Resource Discoveries, FDI Bonanzas, and Local Multipliers: Evidence from Mozambique. The Review of Economics and Statistics 104, 1046–1058. https://doi.org/10.1162/rest_a_00999 UNCTAD, 2023. World Investment Report 2023. 22