Internationalisation as a boost for many firms: evidence from Germany

Frey, Rainer; Goldbach, Stefan Article — Published Version Internationalisation as a boost for many firms: evidence from Germany Review of World Economics Provided in Cooperation with: Springer Nature Suggested Citation: Frey, Rainer; Goldbach, Stefan (2024) : Internationalisation as a boost for many firms: evidence from Germany, Review of World Economics, ISSN 1610-2886, Springer, Berlin, Heidelberg, Vol. 161, Iss. 3, pp. 911-963, https://doi.org/10.1007/s10290-024-00567-z This Version is available at: https://hdl.handle.net/10419/330753 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Vol.:(0123456789) Review of World Economics (2025) 161:911–963 https://doi.org/10.1007/s10290-024-00567-z ORIGINAL PAPER Internationalisation asaboost formany firms: evidence fromGermany RainerFrey1· StefanGoldbach1 Accepted: 1 October 2024 / Published online: 2 December 2024 © The Author(s) 2024 Abstract The advantages of globalisation have been increasingly called into question, and protectionist tendencies have entered the stage. So what experiences have firms had after going international in an open economy over the past years? With respect to foreign investor takeovers of initially national firms, we see predominantly positive effects for acquired manufacturing firms in terms of productivity, sales and expenditures on the labour force – likely due to higher employment – in the short and long term. Looking at the results, firm size matters: positive effects are stronger among large firms. In the case of firms starting to invest abroad, positive effects are very rare and limited to short-term sales of acquiring small manufacturing firms. All in all, the largely positive evidence is generally supportive of internationalisation and thus largely contradicts the negative views sometimes present in the public sphere, though even unaffected firms may see themselves as relative losers. Keywords Globalisation· Firm acquisition· M&A· Productivity· Sales· Intangible assets· Knowledge· Technology· Labour costs· Employment· Wages· Firm heterogeneity JEL Classification D22· D24· F23· G34 1 Introduction In the past, globalisation was largely seen as increasing the welfare of all countries involved. However, more recently, some governments have shifted the focus to potential negative effects. In this context, more emphasis is placed on issues * Rainer Frey rainer[email protected] Stefan Goldbach stef[email protected] 1 Directorate General Economics, Deutsche Bundesbank, Mainzer Landstrasse 46, 60325FrankfurtAmMain, Germany 912 R.Frey, S.Goldbach such as cross-border technology transfer to the detriment of the target firm,1 local divestment and labour layoffs, or public security risk through new foreign owners in mergers and acquisitions (M&A) transactions. These considerations played a role in European countries introducing new laws that enable governments to forbid foreign takeovers (see European Union (2019)). Similarly, countries such as the United States have introduced trade restrictions, citing potentially unfair competition while attracting foreign investment through subsidies (Inflation Reduction Act). Conversely, however, limitations imposed on foreign acquirers can lead to foreign governments introducing countermeasures targeting domestic investment abroad. With national restrictions, firms risk the potentially substantial profits from forming multinationals or being part of a multinational group. While in the case of vertical foreign direct investment (FDI), firms distribute the production process across several countries to profit from an international division of labour,2 horizontal FDI is driven by the development of new markets for a firm’s products.3 By engaging in FDI, firms especially intend to increase their efficiency. Addressing the effects of firms’ ownership changes on economic efficiency, as one early example of ample studies in the manufacturing sector, Lichtenberg and Siegel (1987) analyse the strong increase in M&A in the United States back in the 1970s and 1980s, with significant productivity gains in the years after ownership change.4 So does globalisation only have positive effects while restrictions do harm to firms? To answer this question, we need to take a more detailed look at different firm types and various economic indicators. In particular, our study addresses potential internationalisation effects for heterogeneous firms with respect to productivity, sales, andintangible assets. The stock of intangible assets represents different kinds of knowledge and brand reputation and is likely accompanied by positive longerterm productivity and competitiveness implications.5 Finally, we use labour costs 1 Gerstenberger (2018) addresses this discussion with respect to German companies acquired by Chinese firms. More generally, with respect to the performance of firms acquired by Chinese investors, see Fuest etal. (2021). 2 For the distinction between vertical and horizontal FDI, see Helpman (1984) and Markusen (1984). 3 Farrell and Shapiro (1990) show that horizontal mergers can increase welfare – but in combination with significant increases in market concentration, negative effects may arise due to higher sales prices. Caves (1974) sees FDI as a source of allocative and technical efficiency for the host country. 4 In addition, according to Helpman etal. (2004), the most productive firms engage in foreign activities. 5 Intangible assets include components such as software, intellectual property, brands and innovative business processes. Their characteristics facilitate profiting from economies of scale, and legal titles like patents and copyrights make it possible to secure competitive advantages vis-à-vis competitors. This may result in higher productivity, market power and market concentration (see Crouzet and Eberly (2019)). Davies and Markusen (2020) also report positive effects and observe that multinationals provide their foreign affiliates with knowledge-based assets. Thus, after investing abroad, a technology transfer within the newly formed multinational may take place (see Teece (1977)) and there lies an especially strategic motive for expanding intangible assets. 913 Internationalisation asaboost formany firms: evidence from… to proxy employment and wages – knowing that this is a rather crude measure.6,7 The literature documents the general relevance of these variables: following a foreign acquisition, Guadalupe et al. (2012) find a substantial increase in sales and productivity and a greater likelihood of innovating for the acquired firms.8 In Stiebale and Vencappa (2018), the productivity effect seems especially large for small firms. Arnold and Javorcik (2009) find that target firms have higher productivity as a result of restructuring where acquired plants increased investment, employment and wages.9 Aitken etal. (1996) estimate positive wage effects for foreign-owned firms in the United States, Mexico and Venezuela. With respect to knowledge transfer/ innovation, Javorcik and Poelhekke (2017) provide empirical support for continuous injections of headquarter services into foreign plants. We add substantially to this literature by addressing firm heterogeneity. In doing this, the analysis may detect firm types that profit but also those that do not profit or are even negatively affected, such as in the case of labour force layoffs or shifts of intangible assets within the firm conglomerate across borders (“technology theft”) – events that may bring with them negative attitudes towards globalisation. First, to our knowledge, our study is unique as we separately address the manufacturing and the services sectors: previous studies were mostly limited to the manufacturing sector, e.g. Guadalupe etal. (2012) for Spanish firms, Bircan (2019) for Turkish plants and Stiebale and Vencappa (2018) for Indian firms, or they covered all sectors within one sample (e.g. Hijzen etal. (2013) for several countries). We see a separate analysis of the services sector as appropriate, as the services sector has rapidly gained international economic relevance over the last few decades. Second, we investigate the short and long-term effects of internationalisation on firm performance to understand the dynamics after a firm acquisition. Depending on the performance variable, effects are expected in the short term already (e.g. sales may increase rather quickly) or more in the long term (e.g. productivity, intangible assets or employment/wages). We define our short-term horizon as the time span within the year of the acquisition. Meanwhile, our long term comprises the first three years after the acquisition.10 In the case of our variable intangible assets – where an increase may ultimately also result in higher productivity, competitiveness and 6 Our dataset contains far more data on labour costs than, for example, employment. Thus, in our approach, we use total labour costs as a combined measure of employment and wages. The reasons for an increase in labour costs may be higher wages or more employment, ceteris paribus. 7 There may also be some interdependencies between these variables: Conyon etal. (2002) find wage increases due to higher productivity. Egger and Kreickemeier (2013) argue that international companies are more competitive and are thus able to pay higher wages. Koch and Smolka (2019) provide evidence that the acquired firms hire highly skilled workers and provide worker training, which results in higher competitiveness. 8 Complementarily, Bertrand and Zitouna (2008) observe an increase in productivity but not in profits after a merger. They conclude that firms may redistribute efficiency gains within the firm conglomerate. 9 For establishments in the United States, Davis etal. (2014) show that private equity buyouts may yield positive effects for some plants but negative labour effects for other plants, thereby improving the operating margin materially. 10 Egger etal. (2020) find that post-acquisition wage effects take a total of four years to develop and remain constant thereafter. We look only at the first three years after the acquisition, as a further extension of the time horizon comes with a sizeable decrease in observations. 914 R.Frey, S.Goldbach market power – we potentially capture further positive effects that are felt even beyond our three-year time horizon. Third, we allow for potential differences in firm size like, for example, Stiebale and Vencappa (2018), who document positive productivity effects concentrated on small (target) firms.11 With respect to large mergers, Gugler etal. (2003) find that profits may also increase through higher market power over a longer time horizon. However, we cannot consider the structural implications arising from changes in the market structure within our approach.12 Finally, we contribute to the large empirical literature that deals with the effect of outward FDI on domestic economic activity. Desai et al. (2009) find for US manufacturing firms that outward FDI increases domestic investments and wages. Goldbach etal. (2019) confirm the findings of investments for German firms using the Microdatabase Direct investment (MiDi) compiled by the Deutsche Bundesbank. However, they only focus on outward FDI, while we examine the effects of internationalisation in both directions. Other empirical papers also use the MiDi, e.g. Becker and Mündler (2010) examine potential intra-group labour substitution for German manufacturing multinational firms. Our analysis deviates from their approach because we concentrate on German firms engaging in FDI for the first time. By contrast, Becker and Mündler (2010) mostly focus on multinational firms, which have already foreign affiliates. Another example is Conteduca and Kazakova (2018). They even use a variable (“em1” = reason for initial report) from the MiDi that reports the entry mode. However, this information is optional (not mandatory) and often not provided. That is why we decided to use an alternative identification strategy that relies on matching between the Deutsche Bundesbank’s balance sheet database for German firms (JANIS) and MiDi.13 When we address both inward and outward FDI within one setting at the micro level, we use data from Germany, a large open economy. At the end of 2019, the consolidated FDI stocks of German multinationals abroad amounted to €1.4 trillion. The FDI holdings of foreign parent companies in affiliates located in Germany stood at €0.6trillion and are thus also important from an economic perspective. The existing literature mostly investigates the effects on firms stemming from an ownership change to a foreign investor, e.g. Harrison and Lipsey (1996), Guadalupe etal. (2012), while Ashraf etal. (2016) focus on firm productivity and Bertrand et al. (2012) concentrate on research and development. One exception is Szücs (2014), who addresses the involvement of both M&A acquirers and targets and finds that the targets’ R&D decreases and the acquirers’ R&D-to-sales intensity drops due to 11 Deviating from their approach, we use the sum of tangible and intangible assets as a size measure instead of “sales”. We expect this to be more appropriate for our study, which deals with different sectors. 12 If large firms strengthen an already strong position, it may have implications for overall competition in a sector. The prosperity of the economy as a whole does not necessarily increase. If small firms are busy catching up, it initially might not have the same implications for competitors. In addition, in the long term, firms can also profit from technological clusters abroad through local spillovers (see Jaffe etal. (1993)). 13 We conducted a robustness check with an alternative identification strategy that relies on the mode of entry. Since this information is often not provided, we do not have enough observations in the treatment group for a meaningful interpretation of these results. 915 Internationalisation asaboost formany firms: evidence from… strongly increasing sales. However, he does not account for heterogeneity.14 Thus, we add to the extremely small number of studies that also look at domestic firms’ potential opportunities through going global.15 This is important as, to give an example, in light of national protectionist tendencies with respect to cross-border takeovers, reciprocal actions by other countries could, in turn, jeopardise domestic firms’ investment opportunities abroad. For our analysis, we match German micro data for FDI with firm balance sheet data. These datasets are provided by the Deutsche Bundesbank. This allows us to rely on official values provided by firms. In addition, the data also allows for clear identification of purely national firms and of firms that go international. For investments undertaken by German firms abroad, we use total outward FDI, which includes M&A as well as greenfield investment.16 As both types of investment may result in the internationalisation of initially domestic firms, we thus obtain a complete picture of the associated internationalisation effects, although we cannot distinguish between the contributions of the two types.17 To account for a potential selection bias in our empirical analysis, like many authors, e.g. Guadalupe et al. (2012) and Stiebale and Vencappa (2018), we first address the acquisition decision and then conduct propensity score matching to estimate the likelihoods of firms being chosen for acquisition. In doing this, Guadalupe etal. (2012) find for Spanish manufacturing firms that acquisitions are focused on the most productive firms within industries in the acquirers’ interest of increasing innovation and productivity. This two-stage procedure allows for weighted dynamic difference-in-differences (DiD) estimations of post-acquisition effects.18 All in all, our outcome for internationalisation effects shows that in the aftermath of an acquisition in the manufacturing sector, firms experience positive effects on productivity, sales and expenditure on the labour force in the short and long term – with the long-term effects generally being stronger for larger firms. In the services 14 Another strand of literature looks at effects in the host and home countries of FDI, e.g. Harms and Méon (2018). Chen (2011) finds higher productivity gains, sales and employment only when the foreign parents come from industrial countries rather than developing countries. Hijzen etal. (2013) show that wage effects are larger in developing countries. 15 With respect to German investment abroad, we only address German firms investing for the first time in other countries. In addition, the parent companies of already existing multinationals may also profit from entering further foreign markets. In general, if only some countries introduce restrictions, investors may circumvent these policies via other destination economies. 16 Though the Bundesbank data form permits users to differentiate between greenfield investment and M&A, this optional field is usually left empty. Studies relying on these data are very rare, e.g. Bialek and Weichenrieder (2021) with a focus on FDI regulation or Conteduca and Kazakova (2018). 17 Nocke and Yeaple (2007) emphasise the relevance of firms’ choice of investment type. Hennart and Park (1993) find that Japanese investors use M&A in the case of relatively weak competitive advantages; otherwise greenfield investment is a more efficient way to transfer advantages. Davies etal. (2018) find greenfield investment more reliant on firms’ own capacities, on the origin country’s comparative advantages and the destination country’s taxes. 18 Balsvik and Haller (2010) furthermore conclude that foreign firms pick large, high-wage and highproductivity firms, which is also largely corroborated by the literature. In addition, Frey and Hussinger (2011) show that target firms are deliberately chosen to increase the technological competency of the group as a whole. Crouzet and Eberly (2019) even see a shift in relevance from physical to intangible assets in the United States over time. 916 R.Frey, S.Goldbach sector, the major effects broadly disappear. The sales effect is limited to the short term. For intangible assets, the positive/negative effect depends on firm size. In case of firms that start investing abroad, positive effects are very rare and limited to the short-term sales of small manufacturing firms. Again, we find no evidence that firms restructure at the expense of the labour force at home. Thus, our results underline that cross-border investments often strengthen firms and make them more competitive – though there are also numerous firms that are left behind and, in very rare cases, negatively affected. Additionally, participation in international business is not random: larger firms are more likely to be chosen as targets. The remainder of the paper is structured as follows. Section 2 describes our data, followed by the empirical methodology in Sect.3. The results are presented in Sect.4. We conduct several robustness checks in Sect.5. Section6 briefly concludes. 2 Data To analyse the effect of FDI on domestic performance, we use two micro datasets, both provided by the Deutsche Bundesbank.19 Information on foreign investors’ activities in German firms (inward FDI) and on the foreign investment activities of German companies (outward FDI) is obtained from MiDi, a comprehensive annual database of German FDI positions. MiDi provides information on balance sheet items, ownership structure and additional information such as an economic sector classification of each German affiliate owned by a foreign parent company (and foreign affiliates of German parent firms). One particular advantage of MiDi is that reporting by German firms is mandatory under German federal law. Information on domestic performance and on several other parent-level variables is taken from JANIS.20 This dataset is an extension of USTAN, a dataset of corporate balance sheet statistics. The data are primarily extracted from annual accounts (balance sheet, profit and loss accounts) and financial statements. We make use of firmspecific information on total sales, value added (to estimate productivity), intangible assets (to proxy firm relevant knowledge) and labour costs to capture employment and wages.21 Appendix Table6 provides an overview of the definitions of relevant firm variables. According to the balance sheet information from JANIS, intangible assets are the sum of the following positions: “concessions, industrial property and similar rights and assets as well as licences”, “goodwill”, “payments on account for 19 The micro data are confidential and only accessible in anonymised form at the headquarters of the Bundesbank in Frankfurt, Germany. 20 The data report (https:// www. bunde sbank. de/ resou rce/ blob/ 901916/ 4a1c0 d2604 04103 24e8e 306a4 1a0b9 71/ mL/ 202314janisdata. pdf) highlights that “JANIS consists of individual financial statements of non-financial corporations which are provided from several sources: financial statements received by the Deutsche Bundesbank in the context of the credit assessment and from public sources like the Bundesanzeiger.” Although the different data sources are consistent and are subject to quality checks, the raw data exhibit an unbalanced panel structure. We adjust the data to compare our findings with the literature, which distinguishes between short and long-term effects. 21 The firms report their sales of foreign affiliates (from the MiDi dataset) and the sales of their domestic entities (from the JANIS dataset) independently. 917 Internationalisation asaboost formany firms: evidence from… intangible assets” and “internally generated industrial rights and similar rights and assets”. “Labour costs” are the sum of the positions “salaries and wages” and “social security and expenditure for company pension funds and pensions paid”. We match JANIS with MiDi and keep matched observations as well as unmatched observations. Firms that are German parent companies and German affiliates of foreign companies at the same time are excluded from the analysis in order to make a clear distinction between the effects of inward and outward investment.22 In addition, we only keep firm-year observations for five years in a row.23 In the first analysis, we compare German companies that are taken over (“targets”) by foreign firms with firms that remain domestically owned. We only keep those firms in the control group that have no match with any foreign affiliate – no outward FDI – in the period considered. The second analysis compares German companies engaging in FDI for the first time (“acquirers”)with firms that remain purely domestic.24 In turn, we ensure that our comparison group only consists of purely domestic firms that have no match with a foreign investor in MiDi data – no inward FDI. All in all, we end up with an unbalanced panel for the time period from 1999 to 2018, with about 1,900 German firms taken over by foreign companies, 900 domestic firms going global for the first time, 57,000 purely national companies in our control group, and 360,000 (334,000) firm-year observations in total for “target” (“acquirer”) firms.25 Table1 presents data about the German firms involved in takeovers by foreign investors in more detail. In accordance with the literature, we concentrate on possible firm characteristics with a time lag of one year, which we consider relevant for foreign investors. In the table, the panel on the left (“no target”) presents observations for the control group, while the panel on the right (“as target”) refers to only those observations (number of firms) at the time when a foreign investor acquires a German firm. We provide summary statistics for all sectors, the manufacturing 22 Looking at the original MiDi database (1999 to 2018) without any adjustments, we observe about 18,000 reporting units as German parent companies and about 41,000 domestic affiliates of foreign investors. Another approximately 2,000 firms are German parents and German affiliates (of a foreign ultimate investor) in the same year. 23 For our targets and acquirers, we additionally require that they are in our data for at least two years before the takeover and for three years afterwards. This is in line with the related literature. An analysis of firms that closed within this period is not feasible. This is because it is not possible to distinguish between firm closures and missing data. The restrictions we impose do not affect the composition of industries or the regional distribution. We do see some self-selection of firms that survive over time and become larger on average (in terms of total assets, total sales and – when provided – employment). For further details, see Appendix Table8. We also checked whether this imposed restriction affects our baseline results. Since the findings do not change much, we argue that our chosen variables in the first stage of the propensity score matching controls for this potential self-selection bias. 24 “First time” refers to the change in the status within the sample period. Firms investing abroad for the first time do not necessarily need to acquire another foreign company. They can also establish a new foreign affiliate (“greenfield investment”). Our dataset does not allow us to distinguish between the two types of FDI. Nevertheless, we use the term “acquirers” throughout the paper for the sake of simplicity. 25 There is a threshold of €3million for firms’ FDI filings. Thus, we only consider FDI of some significance. However, the relative importance of the foreign investment may differ depending on the parent companies’ size. 918 R.Frey, S.Goldbach sector and the services sector.26 The sectoral distinction is retained for our estimations. We are aware that a clear assignment is rather difficult, especially for large companies as they are often active in both of these areas. Thus, we assume that the core activities reported by the firms are relevant to their foreign investments. As regards German firms going abroad for the first time, we are dealing with relatively small firms for which sectoral assignment is generally clearer. With respect to firm characteristics, Table1 highlights differences between the treatment and control groups – with the exception of prior total factor productivity. The same conclusion can be derived from Table2, in which firms going global for the first time are compared to companies that remain purely domestic. We account for these differences in our estimation strategy. Table3 presents the locations of foreign investors that acquired German companies and the destinations of investments by German firms that are investing abroad for the first time. Again, we also distinguish between the manufacturing and the services sectors. In terms of both the absolute number of firms and the volume of FDI, firms from Western Europe and North America represent the most important investors in German companies, followed by Asian investors. German firms investing abroad concentrate primarily on Western and Eastern European countries. This could relate to cross-border value chains. Asia and North America also play an essential role as destinations. In addition, we observe a much higher volume of “new” FDI in Germany than that of German companies investing abroad. To interpret these differences, we have to consider the likelihood that a large number of foreign firms already form part of a multinational that is adding a unit to its portfolio of firms. In contrast, our German investors are entering the international “playing field” for the first time.27 Thus, these domestic investors are most likely smaller on average than the acquisitions of foreign firms in Germany, which is reflected in the FDI figures. 3 Empirical approach In our empirical setting, we try to identify the causal effect of a change from national to foreign ownership on the target firm’s performance and of a first investment abroad on the parent company’s performance. To keep our description of the econometric procedure short but still comprehensive, we explain our approach below by solely looking at the scenario in which the foreign investor buys a domestic company; the procedure for first-time FDI by a German parent company is the same. For our analysis, we rely on a DiD approach that we combine with propensity score matching techniques. 26 The manufacturing sector comprises the two-digit NACE Rev. 2 codes 10 to 35. The services sector consists of the codes 45 to 63, 68 and 69, 71 to 82, and 85 to 96. Thus, we exclude the finance sector, holdings, households and organisations as they are expected to act differently compared to the rest of the services sector. 27 In our study, we do not consider already existing German multinationals expanding their international investment. 925 Internationalisation asaboost formany firms: evidence from… where yit stands for our performance variables of interest: factor productivity, sales, intangible assets and labour costs of firm i in period t – all in logarithms39; the dummy Fist equals 1 in the case of an acquisition and 0 otherwise; 𝜇i and 𝜌jt are firm and industry-year fixed effects; and 𝛼 is the constant. Due to the inclusion of firm-level fixed effects, the coefficient 𝛽 addresses the effect through the change in ownership – not ownership itself. To account for serial correlation, we cluster at the firm level. The estimated effects may differ not only with respect to sector and time but also with respect to firm size: (4) y it =𝛼+ 2 ∑ k=0 𝛽k∗Fit−k+ 2 ∑ k=0 𝛾k∗smalli∗Fit−k+𝜇i+𝜌jt +𝜀 it Table 4 Probability of being target and being acquirer Probit estimation. The dependent variable is the binary indicator “change” in period t. The unit of observation is firm-year observation. The sample period is 2001 to 2018. Sector fixed effects and time-specific fixed effects are included but not reported. Robust standard errors (clustered by firm) are in parentheses. ***, ** and * denote significance at the 1%, 5% and 10% level, respectively Target Acquirer Manufacturing Services Manufacturing Services (1) (2) (3) (4) Log total assetsit-1 0.129*** 0.195*** 0.279*** 0.245*** (0.028) (0.019) (0.032) (0.025) Log intangible assetsit-1 0.026*** 0.008 0.035*** 0.053*** (0.008) (0.007) (0.009) (0.011) Log TFPit-1 −0.309* 0.064 −0.896*** 0.029 (0.179) (0.061) (0.164) (0.098) Log labour costsit-1 0.186 0.040 0.833*** 0.039 (0.156) (0.054) (0.145) (0.089) Log salesit-1 0.079*** 0.010 0.002 −0.004 (0.030) (0.017) (0.033) (0.020) Return on equityit-1 −0.000 −0.001** −0.000 −0.000*** (0.000) (0.000) (0.000) (0.000) Log fixed assetsit-1 0.028 −0.132*** 0.025 −0.104*** (0.018) (0.010) (0.017) (0.015) TFP growthit-1 0.066 0.077 −0.021 0.143** (0.065) (0.047) (0.090) (0.071) Observations 107,959 199,498 96,459 181,567 39 Propensity score matching is an approach for the cross-section. Since we estimate our effects within a panel dataset, a firm can exhibit a positive weight in one year and a missing value in another year. We carry the propensity score of a firm forward if there is a missing value until the propensity score matching estimates a new weight for this firm. 926 R.Frey, S.Goldbach where smalli equals 1 for small firms below or equal to the median of the sum of tangible and intangible assets, and 0 otherwise. We classify the “treated” firms as small or large based on their status one year before the treatment. Since radius matching allows for multiple control observations, the status of a firm may change over time. To deal with these differences, we count the number of small and large statuses for the firms in the control group (firms with a positive weight from the propensity score matching in a specific year) over the complete time period. If the majority is small (large), we define the firm in the control group as small (large). This definition keeps the status of the treatment and control groups constant over time and we do not have to drop observations. 4 Results Internationalisation is like a shock to a firm and may have an impact on its economic development. Figure 1a presents the event study’s coefficient plots of total factor productivity, sales, intangible assets and labour costs for all firms.40 The upper (lower) row shows the results for targets (acquirers). Furthermore, we distinguish between the manufacturing (left panel) and the services (right panel) sectors. The effects of internationalisation in the event study are plotted for the year of the event and the two subsequent years. To check the validity of the CTA, we additionally plot the two years preceding the event. The insignificant coefficients in periods t-2 and t-1 of Fig. 1a suggest that the CTA is not violated, with only one exception (intangible assets for acquirers in the services sector). We find no significant effect in the periods t, t + 1 and t + 2 for acquirers (lower row), neither in the manufacturing nor the services sector for all four economic variables. Furthermore, the service sector (right panel) is, in general, not affected by the treatment. For “target” firms in manufacturing sector, however, we find some significant positive treatment effects, which are listed in panel (a) of Table5. The incremental effects diminish over time. We estimate a significant treatment effect for productivity of 5.1% in t, 5.0% in t + 1 and 3.2% in t + 2 (sum of all coefficients: 13.3%). The effect on sales is larger: 7.7% in t, 7.2% in t + 1 and 6.5% in t + 2 (sum of all coefficients: 21.4%). Finally, being “targets” leads to higher labour costs of 5.8% in t, 5.9% in t + 1 and 3.4% in t + 2 (sum of all coefficients: 15.1%). For intangible assets, the significant effect is only existent in t (16.6%) and t + 1 (17.4%) but economically larger (sum of both coefficients: 33.9%). Therefore, we see that German affiliates benefit from being a “target” by foreign investors. Fig. 1 a: Event study for all firms (first row: targets; second row: acquirers; left: manufacturing sector; right: services sector). b: Event study for small firms (first row: targets; second row: acquirers; left: manufacturing sector; right: services sector). c: Event study for large firms (first row: targets; second row: acquirers; left: manufacturing sector; right: services sector) ▸ 40 The event study approach estimates Eq.(3) but starts at k = −2 instead of k = 0. Furthermore, instead of using Eq.(4) to distinguish between small and large firms, we use Eq.(3) for small and large firms as subsamples. 927 Internationalisation asaboost formany firms: evidence from… 928 R.Frey, S.Goldbach Figure1b highlights the event study graphs for small firms. Since the results are similar to all firms, we concentrate on the differences between both samples. Again, most coefficients in t-2 and t-1 are insignificant with only one exception (intangible assets for targets in the services sector). Although, we find again only significant results in the manufacturing sector for “targeted” firms, these effects endure shorter and are quantitatively smaller. Only sales are significant in the periods t, t + 1 and t + 2. All other variables become insignificant in the last period compared to all firms. Since almost all coefficients are smaller, with the only exception of intangible assets in period t, we find weaker effects for small firms compared to all firms. Table5 panel b presents the individual short-term and long-term results for all four economic variables of small firms. Finally, Fig.1c provides the event study graphs for large firms. As before, the CTA holds for most specifications. Only for intangible assets in the service sector for acquirers, similarly to the regressions with all firms, we find negative significant differences in pre-trends. The estimated effects for large firms are, on average, stronger and maintain also longer than for small firms. The only exception is sales, where the sum of all three periods t, t + 1 and t + 2 is the same as for small firms. All in all, we can summarize that we find positive significant effects for “targets” in the manufacturing sector on all four economic variables. The CTA seems to hold in most cases. The overall results seems to be mainly driven by large firms. We show the strongly related DiD regression results for the short and long term with the cumulated effects across the three years after internationalisation in Appendix Tables11, 12, 13, 14 and 15 (with the corresponding plots in Figs.2 and 3 for Table 5 Event study graph— estimated treatment effects for “targeted” firms in the manufacturing sector Significant treatment effects of “targeted” firms in event study graphs tt + 1 t + 2 Sum of t, t + 1 and t + 2 All firms Productivity 5.1% 5.0% 3.2% 13.3% Sales 7.7% 7.2% 6.5% 21.4% Intangible assets 16.6% 17.4% 33.9% Labour costs 5.8% 5.9% 3.4% 15.1% Small firms Productivity 3.5% 4.3% 7.7% Sales 6.2% 5.9% 5.9% 17.9% Intangible assets 18.8% 18.8% Labour costs 4.9% 5.8% 10.7% Large firms Productivity 6.2% 5.5% 4.5% 16.5% Sales 6.1% 6.1% 5.6% 17.9% Intangible assets 24.1% 24.1% Labour costs 6.0% 5.5% 4.6% 16.2% 929 Internationalisation asaboost formany firms: evidence from… productivity).41 The interpretation of our results for small and large firms refers to the total effect of the acquisition, including the base effect plus the effect on relevant interaction terms. As before, we examine the manufacturing and the services sectors separately. In our explanations, we concentrate on the output tables as they include the cumulated effects and as they allow better comparisons of the effects across the performance variables and across the sectors through the provision of exact figures for the effects. 4.1 Firms’ post‑acquisition performance 4.1.1 Impact ofownership change oninitially national firms With respect to the productivity of firms in the manufacturing sector, the plots of the event study already show positive effects, especially for large firms. In accordance, our DiD regressions find – at the 1% and 5% level – significantly positive short-term and accumulated long-term effects driven by large firms (for these, the effects are significant at the 1% and 5% level) (see AppendixTable11, columns 2 and 4). For small firms, this significant effect is only observed in the short term. Thus, the new foreign owners are capable of stimulating large firms, in particular, to increase their productivity. There may be stronger overlaps between business activities for larger firms with acquirers that may allow for efficiency gains through the implementation of better technology. We estimate productivity effects for large firms of 3.1% in the short run and 7.7% in the long run.42 Our results are quantitatively rather small compared to the empirical literature. Guadalupe etal. (2012), Bircan (2019) and Arnold and Javorczik (2009) document large productivity effects of 11% for Spanish firms, 13% for Turkish plants and 13.5% for firms in Indonesia. This is reasonable since most other papers concentrate on developing countries (Turkey, India, and Indonesia). Furthermore, they address earlier periods, where globalisation was more pronounced (especially between 1985 and 2011). Furthermore, our outcome contradicts Stiebale and Vencappa (2018). They presume that small manufacturing firms are further away from the technological frontier than large ones and thus can learn more from an acquirer. They estimate productivity effects of 14.8% for small Indian firms. We only estimate effects of 2.7% in the short run and no significant long-run effects for small German firms. While our classifications of small is based on the sum of tangible and intangible assets, they refer to sales. Besides, their sample of foreign acquisitions is rather small. In the case of firms in the services sector, we only see a weakly positive short-term effect and this time only for all firms (significant at the 10% level in column5). In the services sector, productivity growth seems to be less 41 To estimate DiD in an unbiased way, we conducted the described empirical steps in Sect.3: first, we start with probit estimations to estimate propensity scores. The following matching provides us with a weighting scheme that allows us to estimate the effect of a foreign acquisition on target firm performance using DiD in an appropriate way. Our DiD regressions below rely on Eqs.(3) and (4) from Sect.3. 42 The effects slightly differ because the event study approach includes two more treatment variables (starts with k = −2 instead of k = 0). This leads to a drop in observations (and possible self-selection effects). 930 R.Frey, S.Goldbach attainable. Thus, adjustments within a new affiliate are presumably more burdensome in the services sector, which normally prides itself with closer client relations. Fig. 2 Effect of internationalisation on targets’ total factor productivity (left: manufacturing sector; right: services sector) Fig. 3 Effect of internationalisation on acquirers’ total factor productivity (left: manufacturing sector; right: services sector) 931 Internationalisation asaboost formany firms: evidence from… Internationalisation promotes sales, not only for small but also for large firms. One main motive for going global is to benefit from economies of scale. Small and large manufacturing firms profit from significantly positive short-term effects on sales (at the 1 and 5% significance level), asAppendix Table12 shows. These positive effects are, on average, stronger for small firms. In the long term, the effects remain highly significant for small firms, while significance vanishes to some extent for large firms. It is probably the case that investing companies choose firms to create new distribution channels for their previous products – thus indicating a horizontal merger (see Markusen (1984)). In the case of large-scale mergers, market concentration effects may be intended and the sales effect on the various units of the conglomerate may be initially unclear. Again, the economic effects are smaller than in the empirical literature. Guadalupe etal. (2012) document strong effects of 18% for Spanish firms. We find 5.5% in the short run and 15.6% in the long run. For the services sector, we see positive effects in the short term for small and large firms (at the 5% level). Here, the impact is stronger for large firms. The pattern largely changes when we turn to the effects on intangible assets (see Appendix Table 13). Here, we see no effects within the manufacturing sector. Thus, transfers of knowledge reported in the balance sheet do not seem to be relevant within the time span we address. However, the picture is different for the services sector where we find mixed results: large services firms can be seen as knowledge profiteers, experiencing positive short and long-term effects (significant at the 1% level). Thus, it appears that investors strengthen new affiliates and that they may establish new concepts (e.g. franchising). In contrast, we find significantly negative short and long-term effects for smaller firms (at the 5% level). Intangible assets could be transferred to the new parent company abroad or intangibles assets could become useless and written off. The latter could be the case for trademarks, for example. Finally, AppendixTable14 shows positive short and long-term effects on labour costs applying to large and small firms in the manufacturing sector (at the 1 and 5% significance level in the short term and the 5 and 10% level in the long term). Either employment or wages increased –the latter perhaps going hand in hand with the employment of more highly qualified staff. This result would be in line with Hijzen etal. (2013) who find positive wage effects through employment growth in highly skilled jobs, with no evidence of greater job insecurity. Egger et al. (2020) also provide reasons for higher wages: the application of new technologies may require worker training, which is likely to be accompanied by higher wages. Furthermore, a wage premium may be paid to protect the technological advantage of the multinational.43 Thus, the fear of job losses seems to be unfounded, and the opposite may instead be true with respect to small manufacturing firms (though it is also possible that, actually, wages increased). Conversely, in the services sector, we do not find any significant results. As highlighted previously, there are many missing values for employment in the dataset. Nevertheless, we conducted a robustness check using 43 According to Koch and Smolka (2019), output gains are highest when firms engage in both technology and skill upgrading at the same time. Thus, new foreign owners share profits with the highly skilled workers who are already in the plant before the acquisition (see Balsvik and Haller (2010)). 932 R.Frey, S.Goldbach employment instead of labour costs (seeAppendix Table12). We find significantly positive results in the manufacturing sector for large firms only. Thus, the positive labour cost effects could result from higher employment. We would interpret these results with caution since we cannot rule out a potential selection bias. Thus, our strong results with respect to the development of the performance variables of our German manufacturing target firms show that foreign investments make their new German affiliates profit for the most part. However, in the case of the services sector, the evidence is much weaker and less significant. An explicit empirical comparison of the coefficients of the two sectors shows only minor differences (seeAppendix Table11). 4.1.2 Impact onfirms investing abroad AppendixTable15 presents the coefficients of short and long-term effects of firsttime FDI on productivity for the manufacturing (upper panel) and the services (lower panel) sectors.44 An impact on the performance of the parent company would not come as a surprise as international investments are likely to trigger firm restructuring. On the other hand, the acquiring firm is, on average, larger than the acquired firm and its firm structure more complex. Thus, effects may be relatively smaller or may take more time to unfold. Our results in Table15 point to the second strand of interpretation, as most effects are insignificant. With respect to productivity, intangible assets and labour costs, we do not find any significant effects. Thus, it may be rather burdensome or at least time-consuming for an initially national enterprise to restructure in such a way as to profit either through the exploitation of economies of scale by focusing on production processes or through process innovation in the newly established cross-border value or distribution chain. Internationalisation may render the firm structure more complex. We find at least some positive effects in the aftermath of internationalisation for sales: small firms increase their sales in the short term in the manufacturing sector (at the 5% significance level). In the services sector, we generally see no positive sales effects, indicating that the local market continues to be served by the newly acquired affiliate. The stock of intangible assets is generally not affected by expanding abroad. The small dynamic firms participating in internationalisation probably already have some know-how in a particular area that they may intend to also exploit abroad and, at least at the starting stage of the multinational firm’s creation, no further efforts 44 The three different coefficients for all, small and large firms base on two different regressions. In the following, we explain how we compute them. The coefficient Changeit (0.031) in AppendixTable11 column (1) is equivalent to “All” (0.007) in AppendixTable15. This corresponds to β0 in Eq.(3).Appendix Table11 column (2) distinguishes then between small and large firms. The coefficient Changeit + Changeit x Smalli (0.027) is the total effect for small “target” firms. The corresponding effect is “Small” (0.003) in AppendixTable 15. This is equal to β0 + γ0 in Eq.(4). The coefficient Changeit (0.035) is the total effect for large “target” firms. The corresponding effect is “Large” (0.012) in AppendixTable15. This is equal to β0 in Eq.(4). All explanations also hold for the long-term effect, where we use instead of β0 and γ0 only, the sum of β0 + β1 + β2 and γ0 + γ1 + γ2. 933 Internationalisation asaboost formany firms: evidence from… leading to fast benefits, e.g. with respect to the parent’s technological stance, are observed. In addition, we could not detect any effect on labour costs for firms starting to go abroad. Thus, at least there is no evidence of negative labour outsourcing effects that may be detrimental to the domestic labour force. Goldbach etal. (2019) find with respect to German multinationals that new foreign affiliates bring with them even more investment at home and employment increases as well. However, Lichtenberg and Siegel (1987) find that firms sell plants that are less productive. In addition, Maksimovic etal. (2011) find extensive restructuring in the aftermath of acquisitions – which is likely to increase the competitiveness of a company. For German firms engaging in FDI for the first time, we even see significantly positive employment effects though only for large companies in the manufacturing sector in the short term, asAppendix Table17 shows. 4.2 Characteristics offirms taking part ininternationalisation The probit estimations, which rely on Eq.(2) from Sect.3, are an econometric prerequisite for the DiD estimations in the preceding Sect.4.1. We briefly look at the outcome tables from these estimations as they reveal the specific characteristics of firms participating in internationalisation. First, the coefficients of targets’ and acquirers’ firm size (measured as log of total assets) in the period preceding the acquisition are significantly positive at the 1% level (see Table4, columns (1)–(4) for both the manufacturing and the services sectors). Foreign investors may acquire German firms to benefit from their infrastructure and reputation; conversely, national firms need economic resources to enter foreign markets through FDI. Thus, large enterprises may be generally better equipped for such an undertaking. Furthermore, with respect to potential long-term effects, Gugler etal. (2003) see large firm mergers as a means to increase market power. Turning to specific assets, we find that intangible assets are of strong relevance for targets in the manufacturing sector and for acquirer firms even in both sectors – again with a significantly positive coefficient at the 1% level. The reason for this is that knowledge or brand-based advantages can also be applied abroad. They would allow for scale effects and ultimately result in higher competitiveness. Aitken etal. (1996) observe that multinational firms possess intangible productive assets such as technological know-how, marketing and management skills, export contacts, coordinated relationships with suppliers and customers, and reputation. Additionally, the intention behind a cross-border firm buy-off may also be to harness additional knowledge to boost productivity in the future, meaning that the competitiveness of the conglomerate as a whole could increase (see also Frey and Hussinger (2011)). By contrast, we find a significantly negative coefficient for tangible assets in the case of the services sector. Thus, equipment and buildings seem to be more of a burden than an advantage for firms with international ambitions. In a rather surprising deviation from the literature, in the case of the manufacturing sector, firms participating in the internationalisation process are less productive. Also surprisingly, firms investing abroad for the first time also seem to have other 934 R.Frey, S.Goldbach comparative advantages. Accordingly, acquirers in this sector have higher labour costs; this may indicate higher qualification levels (see Egger et al. (2020)) or a larger labour force.45 Empirical evidence of the positive relevance of sales is limited to target firms in the manufacturing sector. Additionally, we find significantly negative results for profitability, defined as return on equity in the case of the services sector. Finally, in our sample, the number of firms in the services sector that invest abroad is considerably lower than in the manufacturing sector (see Table2). On the one hand, this underlines the high importance of the German manufacturing sector; on the other hand, it raises questions about the international strength of German firms in the services sector. This is even more striking given that foreign acquirers invest fairly equally in manufacturing and services firms in Germany (see Table1). 5 Robustness 5.1 Firms’ investment abroadthroughM&A orhorizontal investment Our empirical setup first examines acquisitions of German firms by foreign companies, followed by the other investment direction (first-time investment abroad by German firms), which encompasses M&A and greenfield investments. This raises the question as to which type of outward investment drives our baseline results. To overcome this data limitation, we provide several robustness checks with different approximations of M&A transactions. We start with German outward transactions with other industrialised countries (see AppendixTable18). The literature argues that firms are more likely to conduct horizontal FDI in high-income countries to serve foreign markets, while firms engage in vertical FDI in low-income countries with lower labour costs. We identify horizontal FDI in countries within the upper quartile of national GDP per capita within our total country sample. Descriptive statistics show that, within this quartile, 68% of FDI is horizontal, conducted by German firms going abroad, probably predominately by means of M&A. We find only insignificant results, which confirms our baseline findings. Beside the distinctions greenfield vs. M&A and horizontal vs. vertical FDI, the empirical literature proposes another way to distinguish between horizontal and vertical FDI. If the parent company and its foreign affiliate are active in the same sector, then it is more likely that the firm will engage in horizontal FDI. Engaging in different sectors is defined as vertical FDI, since firms engage in cross-border supply chains. We define horizontal FDI as a German parent company and its foreign affiliate declaring the same two-digit sector classification. According to this definition, we find results differing from our baseline estimates, especially in the manufacturing sector (seeAppendix Table19). Manufacturing firms engaging in FDI for the first time also benefit from going international. The results for this investment direction 45 Additionally, Balsvik and Haller (2010) conclude that foreign firms pick large, high-wage and highproductivity firms, which is also largely corroborated by the literature. 941 Internationalisation asaboost formany firms: evidence from… Table 8 Sample composition of five year restriction (“targets”) Manufacturing Services No restrictions Restrictions No restrictions Restrictions Important 2-digit economic sectors Manufacture of machinery and equipment n.e.c 203 (19.5%) 191 (19.4%) Manufacture of computer, electronic and optical products 119 (11.4%) 104 (10.6%) Manufacture of chemicals and chemical products 102 (9.8%) 86 (8.7%) Wholesale trade, except of motor vehicles and motorcycles 328 (35.5%) 304 (33.8%) Computer programming, consultancy and related activities 96 (10.4%) 93 (10.3%) Firm size Mean total assets (in thsd. €) 125,911 159,167 70,184 88,043 Mean total sales (in thsd. €) 149,767 183,559 112,431 115,544 Mean employment 459 531 304 373 Geographical region Asia 62 (6.0%) 57 (5.8%) 70 (7.6%) 47 (5.2%) North America 123 (11.8%) 112 (11.4%) 90 (9.7%) 78 (8.7%) Western Europe 812 (78.1%) 779 (79.0%) 737 (79.8%) 736 (81.8%) Number of firms 1,040 986 924 900 942 R.Frey, S.Goldbach Table 9 Balancing property test of “targets” for total factor productivity The standardised percentage bias is the percentage difference of the sample means in the treated and non-treated (full or matched) sub-samples as a percentage of the square root of the average of the sample variances in the treated and non-treated groups (see Rosenbaum and Rubin (1985)) Sample Mean t test Treated Control % Bias p value Manufacturing Log total assetsit-1 Unmatched 10.16 8.62 90.9 0.000 Matched 10.20 10.18 1.4 0.688 Log intangible assetsit-1 Unmatched 4.27 2.34 70.8 0.000 Matched 4.66 4.73 −2.5 0.551 Log TFPit-1 Unmatched 4.08 2.92 65.5 0.000 Matched 4.34 4.34 −0.5 0.875 Log labour costsit-1 Unmatched 8.88 7.47 84.6 0.000 Matched 9.08 9.09 −0.4 0.898 Log salesit-1 Unmatched 10.54 8.94 87.4 0.000 Matched 10.66 10.64 0.8 0.805 Return on equityit-1 Unmatched 4.84 2.11 3.4 0.047 Matched 0.87 1.18 −0.4 0.791 Log fixed assetsit-1 Unmatched 7.71 6.61 41.5 0.000 Matched 8.39 8.44 −1.6 0.591 TFP growthit-1 Unmatched 0.05 0.03 6.2 0.004 Matched 0.04 0.04 −1.2 0.730 943 Internationalisation asaboost formany firms: evidence from… Table 10 Balancing property test of “targets” for total factor productivity The standardised percentage bias is the percentage difference of the sample means in the treated and non-treated (full or matched) sub-samples as a percentage of the square root of the average of the sample variances in the treated and non-treated groups (formulae from Rosenbaum and Rubin (1985)) Sample Mean t test Treated Control % Bias p value Services Log total assetsit-1 Unmatched 10.16 8.62 90.9 0.000 Matched 9.75 9.64 6.3 0.170 Log intangible assetsit-1 Unmatched 4.27 2.34 70.8 0.000 Matched 3.75 3.66 3.2 0.553 Log TFPit-1 Unmatched 4.08 2.92 65.5 0.000 Matched 3.54 3.45 4.7 0.284 Log labour costsit-1 Unmatched 8.88 7.47 84.6 0.000 Matched 8.47 8.36 6.1 0.209 Log salesit-1 Unmatched 10.54 8.94 87.4 0.000 Matched 10.30 10.23 3.6 0.422 Return on equityit-1 Unmatched 4.84 2.11 3.4 0.047 Matched 1.67 2.75 −1.3 0.425 Log fixed assetsit-1 Unmatched 7.71 6.61 41.5 0.000 Matched 6.92 6.90 0.7 0.889 TFP growthit-1 Unmatched 0.05 0.03 6.2 0.031 Matched 0.05 0.05 2.4 0.654 944 R.Frey, S.Goldbach Table 11 Effect of internationalisation on targets’ total factor productivity OLS panel estimation. The dependent variable is total factor productivity (in logs). The unit of observation is firm-year observation. The sample period is 1999 to 2018. Firm fixed effects and sector-time-specific fixed effects are included but not reported. Robust standard errors (clustered by firm) are in parentheses. ***, ** and * denote significance at the 1%, 5% and 10% level, respectively Manufacturing Services (1) (2) (3) (4) (5) (6) (7) (8) Changeit 0.031*** 0.035*** 0.030*** 0.034** 0.022* 0.029 −0.005 0.008 (0.008) (0.012) (0.011) (0.014) (0.013) (0.019) (0.018) (0.025) Changeit-1 0.030*** 0.025** 0.013 0.026 (0.010) (0.012) (0.018) (0.024) Changeit-2 0.014 0.016 0.012 0.032 (0.010) (0.011) (0.014) (0.020) Changeit x Smalli−0.008 −0.008 −0.014 −0.027 (0.017) (0.023) (0.025) (0.032) Changeit-1 × Smalli0.008 −0.028 (0.022) (0.031) Changeit-2 × Smalli−0.003 −0.039 (0.020) (0.026) Changeit + Changeit x Smalli0.027** 0.015 p value: 0.019 p value: 0.373 2 ∑ j =0 Changeit− j 0.074** 0.075** 0.020 0.067 p value: 0.011 p value: 0.022 p value: 0.664 p value: 0.278 2 ∑ j =0 Changeit−j+ 2 ∑ j =0 Changeit−j x Smalli 0.072 −0.027 p value: 0.146 p value: 0.654 Observations 124,182 124,182 103,563 103,563 290,577 290,577 236,148 236,148 Adj. R2 0.943 0.943 0.953 0.953 0.953 0.953 0.966 0.966 945 Internationalisation asaboost formany firms: evidence from… Table 12 Effect of internationalisation on targets’ sales OLS panel estimation. The dependent variable is total sales (in logs). The unit of observation is firm-year observation. The sample period is 1999 to 2018. Firm fixed effects and sector-time-specific fixed effects are included but not reported. Robust standard errors (clustered by firm) are in parentheses. ***, ** and * denote significance at the 1%, 5% and 10% level, respectively Manufacturing Services (1) (2) (3) (4) (5) (6) (7) (8) Changeit 0.054*** 0.035** 0.053*** 0.029* 0.056*** 0.065** 0.023 0.023 (0.012) (0.014) (0.016) (0.017) (0.018) (0.027) (0.027) (0.037) Changeit-1 0.049*** 0.028* 0.037 0.025 (0.015) (0.017) (0.028) (0.041) Changeit-2 0.043*** 0.026 0.041** 0.050* (0.015) (0.018) (0.019) (0.029) Changeit x Smalli0.037 0.051 −0.017 0.001 (0.024) (0.033) (0.035) (0.046) Changeit-1 × Smalli0.043 0.022 (0.032) (0.049) Changeit-2 × Smalli0.037 −0.017 (0.029) (0.040) Changeit + Changeit x Smalli0.073*** 0.048** p value: 0.000 p value: 0.041 ∑2 j =0Changeit− j 0.145*** 0.082* 0.101 0.098 p value: 0.000 p value: 0.069 p value: 0.143 p value: 0.321 ∑2 j =0Changeit−j + ∑2 j =0Changeit− j x Smalli 0.214*** 0.104 p value: 0.004 p value: 0.226 Observations 126,443 126,443 105,366 105,366 301,720 301,720 244,756 244,756 Adj. R2 0.866 0.866 0.915 0.915 0.870 0.870 0.908 0.908 946 R.Frey, S.Goldbach Table 13 Effect of internationalisation on targets’ intangible assets OLS panel estimation. The dependent variable is intangible assets (in logs). The unit of observation is firm-year observation. The sample period is 1999 to 2018. Firm fixed effects and sector-time-specific fixed effects are included but not reported. Robust standard errors (clustered by firm) are in parentheses. ***, ** and * denote significance at the 1%, 5% and 10% level, respectively Manufacturing Services (1) (2) (3) (4) (5) (6) (7) (8) Changeit 0.005 −0.046 0.023 −0.009 0.042 0.244*** 0.030 0.313*** (0.042) (0.061) (0.055) (0.077) (0.049) (0.073) (0.070) (0.098) Changeit-1 0.038 0.057 0.049 0.291*** (0.055) (0.077) (0.065) (0.088) Changeit-2 −0.056 −0.046 0.056 0.251*** (0.053) (0.075) (0.060) (0.082) Changeit x Smalli0.099 0.061 −0.394*** −0.586*** (0.088) (0.112) (0.100) (0.136) Changeit-1 × Smalli−0.035 −0.506*** (0.110) (0.132) Changeit-2 × Smalli−0.019 −0.416*** (0.106) (0.127) Changeit + Changeit x Smalli0.054 −0.150** p value: 0.377 p value: 0.025 ∑2 j =0Changeit− j 0.005 0.002 0.136 0.856*** p value: 0.973 p value:0.992 p value: 0.446 p value: 0.001 ∑2 j =0Changeit−j + ∑2 j =0Changeit− j x Smalli 0.009 −0.654** p value: 0.966 p value: 0.011 Observations 126,310 126,310 105,270 105,270 301,266 301,266 244,350 244,350 Adj. R2 0.745 0.745 0.763 0.763 0.807 0.808 0.831 0.832 947 Internationalisation asaboost formany firms: evidence from… Table 14 Effect of internationalisation on targets’ labour costs OLS panel estimation. The dependent variable is labour costs (in logs). The unit of observation is firm-year observation. The sample period is 1999 to 2018. Firm fixed effects and sector-time-specific fixed effects are included but not reported. Robust standard errors (clustered by firm) are in parentheses. ***, ** and * denote significance at the 1%, 5% and 10% level, respectively Manufacturing Services (1) (2) (3) (4) (5) (6) (7) (8) Changeit 0.035*** 0.037*** 0.033*** 0.032** 0.020 0.017 −0.019 −0.024 (0.009) (0.013) (0.012) (0.015) (0.017) (0.027) (0.022) (0.032) Changeit-1 0.035*** 0.025** −0.008 −0.027 (0.011) (0.013) (0.022) (0.035) Changeit-2 0.012 0.016 0.009 0.008 (0.012) (0.012) (0.016) (0.022) Changeit x Smalli−0.004 0.002 0.006 0.011 (0.020) (0.026) (0.034) (0.043) Changeit-1 × Smalli0.019 0.037 (0.023) (0.044) Changeit-2 × Smalli−0.008 0.004 (0.023) (0.033) Changeit + Changeit x Smalli0.033** 0.023 p value: 0.016 p value: 0.251 ∑2 j =0Changeit− j 0.079*** 0.073** −0.018 −0.043 p value: 0.009 p value: 0.037 p value: 0.744 p value: 0.577 ∑2 j =0Changeit−j + ∑2 j =0Changeit− j x Smalli 0.086* 0.008 p value: 0.099 p value: 0.912 Observations 126,698 126,698 105,622 105,622 294,650 294,650 239,250 239,250 Adj. R2 0.936 0.936 0.947 0.947 0.929 0.929 0.944 0.944 948 R.Frey, S.Goldbach Table 15 Effect of internationalisation on acquirers’ productivity, sales, intangible assets and labour costs Coefficients of interest with clustered standard errors (by firms) in parentheses Short-term effects Long-term effect Log TFP Log sales Log intangible assets Log labour costs Log TFP Log sales Log intangible assets Log labour costs Manufacturing All 0.007 0.026** −0.058 0.006 −0.009 0.020 −0.136 −0.014 (0.008) (0.010) (0.046) (0.008) (0.029) (0.035) (0.159) (0.027) Small 0.003 0.033** −0.073 −0.003 −0.001 0.038 −0.209 −0.023 (0.011) (0.016) (0.055) (0.011) (0.039) (0.050) (0.186) (0.041) Large 0.012 0.018 −0.042 0.017 −0.018 0.003 −0.064 −0.005 (0.011) (0.014) (0.074) (0.012) (0.040) (0.048) (0.258) (0.042) Observations 67,275 68,290 68,244 68,861 56,716 57,520 57,457 57,978 Adj. R20.934 0.876 0.667 0.931 0.949 0.925 0.693 0.948 Services All 0.013 0.003 −0.133* 0.020 0.067 0.020 −0.146 0.079 (0.014) (0.025) (0.072) (0.017) (0.061) (0.100) (0.245) (0.067) Small −0.001 0.002 −0.103 −0.007 0.062 0.030 −0.319 0.022 (0.016) (0.028) (0.109) (0.018) (0.077) (0.101) (0.392) (0.084) Large 0.026 0.004 −0.163* 0.046 0.072 0.010 −0.001 0.128 (0.025) (0.041) (0.094) (0.032) (0.095) (0.163) (0.304) (0.104) Observations 228,207 236,931 241,579 234,086 187,024 193,856 197,378 191,550 Adj. R20.967 0.897 0.824 0.957 0.976 0.922 0.846 0.967 949 Internationalisation asaboost formany firms: evidence from… Table 16 Differences between manufacturing and services Coefficients of interest with clustered standard errors (by firms) in parentheses Short-term effects Long-term effects Log TFP Log sales Log intangible assets Log labour costs Log TFP Log sales Log intangible assets Log labour costs Targets All 0.005 0.002 −0.060 0.008 0.053 0.070 −0.175 0.073 (0.016) (0.022) (0.066) (0.019) (0.056) (0.080) (0.235) (0.063) Small 0.008 0.060 0.407*** 0.015 0.112 0.187* 0.279 0.138 (0.034) (0.047) (0.136) (0.045) (0.080) (0.107) (0.325) (0.093) Large −0.006 −0.033 −0.323*** −0.004 −0.064 −0.073 −0.954*** −0.034 (0.027) (0.037) (0.097) (0.038) (0.083) (0.129) (0.348) (0.096) Observations 410,273 421,507 420,963 415,903 336,748 345,557 345,184 341,155 Adj. R20.952 0.878 0.777 0.932 0.963 0.916 0.799 0.946 Acquirers All −0.008 −0.037 0.095 −0.022 −0.079 −0.216* −0.004 −0.108 (0.017) (0.035) (0.084) (0.018) (0.067) (0.130) (0.288) (0.075) Small −0.039 0.004 −0.142 −0.021 −0.143 −0.225 −0.105 −0.167* (0.037) (0.065) (0.170) (0.045) (0.089) (0.159) (0.388) (0.096) Large 0.012 −0.038 0.177 −0.013 −0.015 −0.212 0.104 −0.054 (0.029) (0.058) (0.127) (0.038) (0.099) (0.186) (0.414) (0.115) Observations 301,468 312,836 307,183 306,547 248,539 257,330 253,024 252,439 Adj. R20.957 0.857 0.730 0.934 0.968 0.913 0.755 0.949 950 R.Frey, S.Goldbach Table 17 Employment as dependent variable Coefficients of interest with clustered standard errors (by firms) in parentheses Targets Acquirers Manufacturing Services Manufacturing Services Short-term Long-term Short-term Long-term Short-term Long-term Short-term Long-term All 0.031*** 0.072* 0.015 0.013 0.017 −0.020 0.010 0.009 (0.011) (0.038) (0.027) (0.089) (0.020) (0.058) (0.035) (0.132) Small 0.024 0.027 −0.021 −0.035 −0.025 −0.089 −0.008 −0.131 (0.018) (0.062) (0.033) (0.137) (0.024) (0.056) (0.043) (0.164) Large 0.037*** 0.108** 0.048 0.048 0.065** 0.053 0.034 0.174 (0.013) (0.045) (0.036) (0.092) (0.030) (0.101) (0.052) (0.200) Observations 76,901 66,066 133,892 113,652 40,202 34,665 118,854 100,643 Adj. R20.964 0.969 0.972 0.975 0.856 0.870 0.928 0.933 957 Internationalisation asaboost formany firms: evidence from… Table 24 German parent companies with first-time FDI (coarsened exact matching) Coefficients of interest with clustered standard errors (by firms) in parentheses Short-term effects Long-term effects Log TFP Log sales Log intangible assets Log labour costs Log TFP Log sales Log intangible assets Log labour costs Manufacturing All 0.002 0.015 0.027 0.003 −0.015 0.026 0.011 −0.014 (0.008) (0.014) (0.050) (0.011) (0.025) (0.037) (0.179) (0.033) Small 0.004 0.032 0.015 0.013 −0.020 0.030 −0.048 −0.018 (0.011) (0.020) (0.071) (0.013) (0.041) (0.054) (0.283) (0.048) Large 0.001 0.003 0.036 −0.005 −0.012 0.023 0.048 −0.012 (0.010) (0.020) (0.070) (0.017) (0.032) (0.049) (0.233) (0.044) Observations 12,462 13,179 13,182 12,365 10,472 11,154 11,094 10,461 Adj. R20.962 0.903 0.786 0.954 0.970 0.943 0.782 0.965 Services All −0.007 −0.002 −0.135 0.015 −0.025 0.081 −0.127 0.005 (0.017) (0.035) (0.086) (0.019) (0.049) (0.113) (0.306) (0.061) Small −0.008 0.036 −0.051 −0.002 0.011 0.090 0.055 0.030 (0.016) (0.038) (0.115) (0.021) (0.061) (0.142) (0.412) (0.084) Large −0.006 −0.038 −0.213* 0.030 −0.056 0.073 −0.294 −0.018 (0.022) (0.063) (0.118) (0.030) (0.071) (0.176) (0.422) (0.087) Observations 13,855 12,846 15,191 11,721 11,245 10,062 12,291 9,403 Adj. R20.984 0.884 0.881 0.976 0.986 0.889 0.880 0.978 958 R.Frey, S.Goldbach Table 25 Foreign takeovers of German companies (“target” is never in control group) Coefficients of interest with clustered standard errors (by firms) in parentheses Short-term effects Long-term effects Log TFP Log sales Log intangible assets Log labour costs Log TFP Log sales Log intangible assets Log labour costs Manufacturing All 0.031** 0.060*** 0.068 0.028*** 0.076* 0.158*** 0.168 0.079* (0.011) (0.016) (0.057) (0.011) (0.041) (0.061) (0.199) (0.041) Small 0.031* 0.078*** 0.098 0.022 0.094 0.265** 0.175 0.092 (0.018) (0.026) (0.087) (0.020) (0.078) (0.110) (0.291) (0.077) Large 0.030* 0.044** 0.041 0.034*** 0.061 0.074 0.163 0.068 (0.014) (0.019) (0.077) (0.013) (0.043) (0.062) (0.271) (0.045) Observations 118,553 120,506 120,294 120,960 98,740 100,294 100,131 100,708 Adj. R20.943 0.868 0.742 0.941 0.953 0.916 0.762 0.951 Services All 0.026 0.083*** 0.093 0.021 −0.011 0.070 0.349 −0.016 (0.017) (0.029) (0.067) (0.024) (0.063) (0.102) (0.250) (0.075) Small 0.028 0.072** −0.155* 0.040 −0.018 0.136 −0.420 0.061 (0.023) (0.036) (0.093) (0.033) (0.084) (0.123) (0.353) (0.113) Large 0.021 0.094** 0.359*** −0.001 −0.003 0.005 1.091*** −0.103 (0.025) (0.042) (0.099) (0.036) (0.083) (0.150) (0.345) (0.103) Observations 285,346 296,271 295,662 289,244 231,714 240,146 239,608 234,670 Adj. R20.953 0.860 0.802 0.926 0.966 0.902 0.832 0.940 959 Internationalisation asaboost formany firms: evidence from… Table 26 German parent companies with first-time FDI (“acquirer” is never in control group) Coefficients of interest with clustered standard errors (by firms) in parentheses Short-term effects Long-term effects Log TFP Log sales Log intangible assets Log labour costs Log TFP Log sales Log intangible assets Log labour costs Manufacturing All −0.005 0.016 −0.048 −0.007 −0.047 −0.022 −0.052 −0.060 (0.012) (0.013) (0.063) (0.013) (0.041) (0.046) (0.205) (0.042) Small −0.008 0.019 0.022 −0.020 −0.036 −0.033 0.103 −0.078 (0.018) (0.019) (0.075) (0.019) (0.059) (0.074) (0.238) (0.061) Large −0.003 0.015 −0.105 0.005 −0.057 −0.013 −0.170 −0.045 (0.016) (0.018) (0.100) (0.017) (0.056) (0.057) (0.321) (0.057) Observations 63,328 64,259 64,198 64,806 53,264 53,997 53,925 54,435 Adj. R20.932 0.887 0.678 0.930 0.948 0.932 0.702 0.949 Services All 0.040 0.033 −0.092 0.055 0.041 0.131 −0.015 0.037 (0.032) (0.060) (0.107) (0.034) (0.115) (0.197) (0.297) (0.120) Small −0.029 −0.072 −0.086 −0.060 −0.041 −0.031 −0.448 −0.200 (0.043) (0.086) (0.210) (0.045) (0.184) (0.259) (0.431) (0.175) Large 0.078* 0.087 −0.095 0.116** 0.080 0.207 0.179 0.147 (0.044) (0.079) (0.122) (0.046) (0.145) (0.265) (0.385) (0.151) Observations 225,798 234,443 239,052 231,661 184,936 191,702 195,191 189,448 Adj. R20.965 0.878 0.820 0.949 0.975 0.906 0.848 0.962 960 R.Frey, S.Goldbach Table 27 Staggered DiD with Callaway and Sant’Anna (2021) estimator Coefficients of interest with clustered standard errors (by firms) in parentheses Manufacturing Services Log TFP Log sales Log intangible assets Log labour costs Log TFP Log sales Log intangible assets Log labour costs Targets All −0.003 0.011 −0.080 −0.008 0.022 0.111*** 0.073 0.023 (0.009) (0.013) (0.057) (0.011) (0.014) (0.038) (0.061) (0.016) Small 0.005 0.033* −0.071 −0.003 0.023 0.158** 0.012 0.018 (0.013) (0.018) (0.089) (0.014) (0.017) (0.077) (0.087) (0.020) Large −0.010 −0.011 −0.075 −0.012 0.022 0.072*** 0.131 0.028 (0.012) (0.017) (0.070) (0.015) (0.021) (0.025) (0.084) (0.023) Acquirers All 0.035*** 0.028** 0.208*** 0.033*** 0.048** 0.062 0.142 0.055** (0.011) (0.014) (0.069) (0.012) (0.019) (0.044) (0.098) (0.026) Small 0.053*** 0.050*** 0.225*** 0.047** 0.079** 0.135 0.308** 0.071** (0.014) (0.015) (0.078) (0.015) (0.035) (0.084) (0.138) (0.028) Large 0.008 −0.006 0.157 0.010 0.018 −0.003 −0.001 0.035 (0.015) (0.022) (0.107) (0.017) (0.029) (0.048) (0.140) (0.041) 961 Internationalisation asaboost formany firms: evidence from… Funding Open Access funding enabled and organized by Projekt DEAL. Declarations Conflict of interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/ licenses/by/4.0/. References Ackerberg, D. A., Caves, K., & Frazer, G. (2015). Identification properties of recent production function estimators. Econometrica, 83(6), 2411–2451. Aitken, B., Harrison, A., & Lipsey, R. E. (1996). Wages and foreign ownership: a comparative study of Mexico, Venezuela, and the United States. Journal of International Economics, 40, 345–371. Arnold, J. M., & Javorcik, B. S. (2009). Gifted kids or pushy parents? Foreign direct investment and plant productivity in Indonesia. Journal of International Economics, 79, 42–53. Ashraf, D., Herzer, D., & Nunnenkamp, P. (2016). The effects of greenfield FDI and cross-border M&As on total factor productivity. World Economy, 39(11), 1728–1755. Baker, A. C., Larcker, D. F., & Wang, C. (2022). How much should we trust staggered difference-indifferences estimates? Journal of Financial Economics, 144(2), 370–395. Balsvik, R., & Haller, S. (2010). Picking “Lemons” or Picking “Cherries”? Domestic and foreign acquisitions in Norwegian manufacturing. Scandinavian Journal of Economics, 112(2), 361–387. Balsvik, R., & Haller, S. (2020). Worker-plant matching and ownership change. Scandinavian Journal of Economics, 122(4), 1286–1314. Becker, S. O., & Muendler, M.-A. (2010). Margins of multinational labor substitution. American Economic Review, 100(5), 1999–2030. Bertrand, O., Hakkala, K. N., Norbäck, P.-J., & Persson, L. (2012). Should countries block foreign takeovers of R&D champions and promote greenfield entry? Canadian Journal of Economics, 45(3), 1083–1124. Bertrand, O., & Zitouna, H. (2008). Domestic versus cross-border acquisitions: Which impact on the target firms’ performance? Applied Economics, 40(17), 2221–2238. Bialek, S., & Weichenrieder, A. J. (2021). Do stringent environmental policies deter FDI? M&A versus greenfield. Environmental & Resource Economics, 80, 603–636. Bircan, C. (2019). Ownership structure and productivity of multinationals. Journal of International Economics, 116, 125–143. Caliendo, M., Kopeinig, S. (2005). Some practical guidance for the implementation of propensity score matching. IZA Discussion Paper Series No 1588, May 2005. Callaway, B., & SantAnna, P. H. C. (2021). Difference-in-differences with multiple time periods. Journal of Econometrics, 225(2), 200–230. Cassiman, B., Colombo, M. G., Garrone, P., & Veugelers, R. (2005). The impact of M&A on the R&D process: an empirical analysis of the role of technologicaland market-relatedness. Research Policy, 34(2), 195–220. Caves, R. E. (1974). Multinational firms, competition, and productivity in host-country markets. Economica, 41(162), 176–193. 962 R.Frey, S.Goldbach Chen, W. (2011). The effect of investor origin on firm performance: domestic and foreign direct investment in the United States. Journal of International Economics, 83(2), 219–228. Conteduca, F., Kazakova, E. (2018). Serving abroad: Export, M&A, and greenfield investment. CRC TR224 Discussion Paper Series No. 008, March 2018. Conyon, M. J., Girma, S., Thompson, S., & Wright, P. W. (2002). The productivity and wage effects of foreign acquisition in the United Kingdom. The Journal of Industrial Economics, 50(1), 85–102. Crass, D., & Peters, B. (2014). Intangible assets and firm-level productivity. SSRN Electronic Journal. https:// doi. org/ 10. 2139/ ssrn. 25623 02 Crouzet, N., Eberly, J. (2019). Understanding weak capital investment: The role of market concentration and intangibles. NBER Working Paper No 25869. Davies, R., Markusen, J. R. (2020). The structure of multinational firms’ international activities. NBER Working Paper No. 26827, March 2020. Davies, R. B., Desbordes, R., & Ray, A. (2018). Greenfield versus merger and acquisition FDI: Same wine, different bottles? Canadian Journal of Economics/revue Canadienne D’économique, 51(4), 1151–1190. https:// doi. org/ 10. 1111/ caje. 12353 Davis, S., Haltiwanger, J., Handley, K., Jarmin, R., Lerner, J., & Miranda, J. (2014). Private equity, jobs, and productivity. American Economic Review, 104(12), 3956–3990. Dehejia, R., & Wahba, S. (2002). Propensity score-matching methods for nonexperimental causal studies. The Review of Economics and Statistics, 84(1), 151–161. Desai Hines, M. A., Foley, C. F., & James, R. (2009). Domestic effects of the foreign activities of US. American Economic Journal: Economic Policy, 1(1), 181–203. Egger, H., Jahn, E., & Kornitzky, S. (2020). Reassessing the foreign ownership wage premium in Germany. The World Economy, 43(2), 302–325. Egger, H., & Kreickemeier, U. (2013). Why foreign ownership may be good for you. International Economic Review, 54(2), 693–716. Farrell, J., & Shapiro, C. (1990). Horizontal mergers: an equilibrium analysis. American Economic Review, 80(1), 107–126. Frey, R., & Hussinger, K. (2011). European market integration through technology-driven M&As. Applied Economics, 43(17), 2143–2153. Fuest, C., Hugger, F., Sultan, S., & Xing, J. (2021). What drives Chinese overseas M&A investment? Evidence from micro data. Review of International Economics, Online: Gandhi, A., Navarro, S., & Rivers, D. (2020). On the identification of gross output production functions. Journal of Political Economy, 128(8), 2973–3016. Gerstenberger, J. (2018). More Chinese M&A deals in the German SME sector—but the share remains negligible. KfW Research Paper 229. Girma, S., & Görg, H. (2007). Evaluating the foreign ownership wage premium using a difference-indifferences matching approach. Journal of International Economics, 72(1), 97–112. Goldbach, S., Nagengast, A., Steinmüller, E., & Wamser, G. (2019). The effect of investing abroad on investment at home: on the role of technology, tax savings, and internal capital markets. Journal of International Economics, 116, 58–73. Guadalupe, M., Kuzmina, O., & Thomas, C. (2012). Innovation and foreign ownership. American Economic Review, 102(7), 3594–3627. Gugler, K., Mueller, D., Yurtoglu, B., & Zulehner, C. (2003). The effects of mergers: an international comparison. International Journal of Industrial Organization, 21(5), 625–653. Harms, P., & Méon, P.-G. (2018). Good and useless FDI: the growth effects of greenfield investment and mergers and acquisitions. Review of International Econonomics, 26, 37–59. Helpman, E. (1984). A simple theory of international trade with multinational corporations. Journal of Political Economy, 92(3), 451–471. Helpman, E., Melitz, M., & Yeaple, S. (2004). Export versus FDI with heterogeneous firms. American Economic Review, 91(4), 300–316. Hennart, J.-F., & Park, Y.-R. (1993). Greenfield versus acquisition: the strategy of Japanese investors in the United States. Management Science, 39(9), 1054–1070. Hijzen, A., Martins, P. S., Schank, T., & Upward, R. (2013). Foreign-owned firms around the world: a comparative analysis of wages and employment at the micro-level. European Economic Review, 60, 170–188. Iacus, S. M., King, G., & Porro, G. (2011). Multivariate matching methods that are monotonic imbalance bounding. Journal of the American Statistical Association, 106(493), 345–361. 963 Internationalisation asaboost formany firms: evidence from… Iacus, S. M., King, G., & Porro, G. (2012). Causal inference without balance checking: coarsened exact matching. Political Analysis, 20(1), 1–24. Jaffe, A. B., Trajtenberg, M., & Henderson, R. (1993). Geographic localization of knowledge spillovers as evidenced by patent citations. The Quarterly Journal of Economics, 108(3), 577–598. Javorcik, B., & Poelhekke, S. (2017). Former foreign affiliates: Cast out and outperformed? Journal of the European Economic Association, 15(3), 501–539. Koch, M., & Smolka, M. (2019). Foreign ownership and skill-biased technological change. Journal of International Economics, 118, 84–104. Levinsohn, J., & Petrin, A. (2003). Estimating production functions using inputs to control for unobservables. Review of Economic Studies, 70(2), 317–341. Lichtenberg, F. R., Siegel, D., Jorgenson, D., & Mansfield, E. (1987). Productivity and changes in ownership of manufacturing plants. Brookings Papers on Economic Activity, 1987(3), 643–683. Maksimovic, V., Phillips, G., & Prabhala, N. R. (2011). Post-merger restructuring and the boundaries of the firm. Journal of Financial Economics, 102(2), 317–343. Markusen, J. (1984). Multinationals, multi-plant economies, and the gains from trade. Journal of International Economics, 16(3), 205–226. Nocke, V., & Yeaple, S. (2007). Cross-border mergers and acquisitions vs. greenfield foreign direct investment. The role of firm heterogeneity. Journal of International Economics, 72(2), 336–365. https:// doi. org/ 10. 1016/j. jinte co. 2006. 09. 003 Regulation (EU) (2019). Establishing a framework for the screening of foreign direct investments into the Union. Official Journal of the European Union L, 79, 1–14. Roth, J., Sant’Anna, P. H. C., Bilinski, A., & Poe, J. (2023). What’s trending in difference-in-differences? A synthesis of the recent econometrics literature. Journal of Econometrics, 235(2), 2218–2244. Stiebale, J., & Reize, F. (2011). The impact of FDI through mergers and acquisitions on innovation in target firms. International Journal of Industrial Organization, 29(2), 155–167. Stiebale, J., & Vencappa, D. (2018). Acquisitions, markups, efficiency, and product quality: evidence from India. Journal of International Economics, 112, 70–87. Szücs, F. (2014). M&A and R&D: asymmetric effects on acquirers and targets? Research Policy, 43(7), 1264–1273. Teece, D. J. (1977). Technology transfer by multinational firms: the resource cost of transferring technological know-how. Economic Journal, 87(346), 242–261. Van Biesebroeck, J. (2008). The sensitivity of productivity estimates. Journal of Business & Economic Statistics, 26, 311–328. Wang, J., & Wang, X. (2015). Benefits of foreign ownership: evidence from foreign direct investment in China. Journal of International Economics, 97(2), 325–338. Wooldridge, J. M. (2009). On estimating firm-level production functions using proxy variables to control for unobservables. Economics Letters, 104(3), 112–114. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.