Short and medium-term effects of foreign acquisitions on manufacturing firms: Evidence from Germany

Görg, Holger; Lehr, Jakob Article — Published Version Short and medium-term effects of foreign acquisitions on manufacturing firms: Evidence from Germany Economics Letters Provided in Cooperation with: Kiel Institute for the World Economy – Leibniz Center for Research on Global Economic Challenges Suggested Citation: Görg, Holger; Lehr, Jakob (2024) : Short and medium-term effects of foreign acquisitions on manufacturing firms: Evidence from Germany, Economics Letters, ISSN 1873-7374, Elsevier BV, Amsterdam, Vol. 241, pp. 1-6, https://doi.org/10.1016/j.econlet.2024.111826 This Version is available at: https://hdl.handle.net/10419/302104 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. http://creativecommons.org/licenses/by-nc/4.0/ Economics Letters 241 (2024) 111826 Available online 21 June 2024 0165-1765/© 2024 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/bync/4.0/). Contents lists available at ScienceDirect Economics Letters journal homepage: www.elsevier.com/locate/ecolet Short and medium-term effects of foreign acquisitions on manufacturing firms: Evidence from Germany✩ Holger Görg a, Jakob Lehr b,∗ aKiel Institute for the World Economy and University of Kiel, Germany bUniversity of Mannheim, Germany ARTICLE INFO JEL classification: F23 F61 Keywords: Foreign direct investment Foreign acquisitions Firm behavior Ex-post evaluation ABSTRACT This paper, for the first time, investigates the impact of foreign acquisitions on German manufacturing firms using, newly available unique administrative micro data spanning 25 years. Based on an event study design combined with propensity score matching techniques, we find that foreign acquisitions significantly increase labor productivity and average wages in acquired firms. A reduction in employment drives both effects. 1. Introduction Germany’s economy, particularly its manufacturing sector, is deeply integrated into the global economy. This shows itself through high export and import activity but also through foreign direct investments (FDI). For instance, the FDI stock in Germany reached approximately 1000 billion USD in 2021 (cf. United Nations Conference on Trade and Development (2023)). Despite the significance of foreign investments, little is known about their specific impacts on acquired companies. How does their competitive position evolve? What happens to employment levels, and how do wages develop? Against the backdrop of a potential disintegration of the global economy, e.g., due to political instabilities or security considerations, these already critical questions gain even greater significance. This paper contributes to answering these questions. ✩The project generating the dataset used in this paper has been contracted by and received funding from the German Federal Ministry for Economic Affairs and Climate Action. We thank the Research Data Center (RDC) the Federal Statistical Office and the Research Data and Service Centre (RDSC) of the Bundesbank for granting us access to the micro-data. The specific project title in the data usage agreement with the Bundesbank is ‘‘Evaluating the impact of foreign acquisitions in Germany (Project ID 2023/0047)’’. All results in this paper were generated via Controlled Remote Execution. Jakob Lehr acknowledges additional funding through the German Research Foundation (DFG) through CRC TR 224 (Project B07). ∗Corresponding author. E-mail address: [email protected] (J. Lehr). 1Literature also looks at the impact of inward FDI on aggregate productivity and performance more generally (cf. Görg and Greenaway (2004), Greenaway and Kneller (2007), Keller (2024)). 2The most notable exceptions are Wagner and Weche-Gelübcke (2015) and Weche-Gelübcke (2015) who look at the impact of foreign acquisitions on access to finance, and employment and productivity of target firms respectively. They use data for 2007 to 2010. Their definition of foreign ownership is based on Foreign Affiliate Statistics (FATS) for Germany, which only includes majority foreign owned firms. Our definition is based on foreign investments equivalent to at least a 10 percent ownership share (see Kruse et al. (2023) for a discussion). More importantly, our data cover a much longer time period allowing us to consider also medium to long run effects. Moreover the data we use are available for approved researchers at the Research Data and Service Centre at the Bundesbank and the Research Data Centre of the Federal Statistical Office. We investigate the response of firm-level measures of economic performance – productivity, output, employment and average wages – to an inflow of foreign investment through an acquisition of a domestic firm by a foreign owner. The analysis is based on comprehensive, new and unique firm-level panel data for Germany, covering 25 years from 1995 until 2019. We implement an event study design, which we combine with propensity score matching techniques, in order to identify the effects of foreign acquisitions at the firm level. Our paper makes three contributions: First, we estimate the effect of foreign investment on German manufacturing firms. There are related papers looking at the effects of cross-border M&As or FDI more broadly on firm performance in the context of other countries, also using similar methods (e.g. Arnold and Javorcik (2009), Criscuolo and Martin (2009), Guadalupe et al. (2012), Girma et al. (2015), Arnold et al. (2023)).1However, comparable evidence for Germany is scarce.2 https://doi.org/10.1016/j.econlet.2024.111826 Received 15 April 2024; Received in revised form 18 June 2024; Accepted 18 June 2024 Economics Letters 241 (2024) 111826 2 H. Görg and J. Lehr Filling this gap is particularly important given the degree of openness of the German economy, its size, level of technology, and its role as Europe’s industrial powerhouse. Second, we conduct our analysis linking, for the first time, two large administrative micro datasets: the German census of the manufacturing industry provided by the German Federal Statistical Office and the Microdatabase Direct Investment collected by the German Bundesbank. The link between the two datasets has been established only recently, and we are the first to exploit this new and unique database. With this data, spanning 25 years, we are able to estimate short and mediumterm effects of foreign acquisitions. Generally, studies as the ones cited above only look at short run effects. Third, we employ an empirical specification that recognizes recent advancements in the difference-in-differences literature, which might be particularly relevant in our setting given the staggered treatment adoption. We implement the event study estimator proposed by Sun and Abraham (2021), combining it with propensity score matching techniques to estimate the causal effects of acquisitions on firms. Our results show that foreign acquisitions lead to a strong and lasting increase in labor productivity and average wages. For instance, over an eight-year period following the takeover, average labor productivity and wages increase by 5% and 2%, respectively, in acquired firms compared to the control group. Both effects are driven by firms reducing employment while maintaining a constant output level and only moderately decreasing total labor costs. Effects are robust to a variety of alternative specifications. 2. Data We can, uniquely, combine the German census of the manufacturing industry with the Microdatabase Direct Investment – MiDi – to investigate the effect of foreign acquisitions on German firms.3The census data spans the period 1995–2019 and covers the population of manufacturing firms with at least 20 employees. It contains basic information on firms’ characteristics, like the economic sector, and measures of economic performance, such as gross output, the number of employees and the total wage bill. The MiDi database contains comprehensive information on foreign investments in Germany at the investmentfirm-year level, starting in 1999.4The data cover the universe of FDI, subject to two reporting thresholds: firstly, the balance sheet total of the German firm receiving the investment has to exceed 3 million Euros; secondly, the investor must own at least 10% of the voting rights in the German company. We combine the two datasets using common firm identifiers and define a foreign-ownership dummy equal to one if the firm appears in the MiDi data. To ensure that a switch of the foreign-owned dummy from 0 to 1 captures a real change in the firm’s ownership status, rather than an already foreign-owned firm trespassing the 3 million Euro threshold, we exclude all firms with a balance sheet total of less than four million euro in the first year they appear in MiDi.5We also exclude all firms in foreign ownership in 1999, the first year of MiDi. 3For further information on the Microdatabase Direct Investment, see Friederich et al. (2021) and Blank et al. (2020) (DOI: 10.12757/ Bbk.MiDi.9920.09.09). The specific dataset for the manufacturing industry has been compiled for this project and is not a standard data product offered by the Research Data Centers of the Federal Statistical Office. This project-based dataset combined NACE Rev. 2 sections B-F: including mining and quarrying, manufacturing, energy and water supply, as well as construction. It is derived from the RDC products AFiD-Panel Industrieunternehmen (DOI: 10.21242/42221.2021.00.01.1.1.0), AFiD-Panel Unternehmenstrukturstatistiken (DOI: 10.21242/42231.2020.00.05.1.1.0) and AFiD-Panel Energieunternehmen (DOI: 10.21242/43221.2021.00.01.1.1.0). For the combined data set cf. also Boddin et al. (2024). 4A firm can thus appear multiple times annually if it has received more than one foreign investment. In this case, we aggregate to the firm level to have one firm observation per year. 5A change in the foreign-owned dummy can indicate an FDI shock along the extensive or the intensive margin, given that it might reflect a crossing 3. Empirical approach We employ an event study specification to identify the effect of foreign acquisitions on German firms. The central identifying assumption requires treated firms’ counterfactual trends in outcomes to be parallel to trends in the control group. While parallel trends prior to treatment yield strong support for this assumption, we aim to further increase confidence in the parallel trends assumption by choosing a control group similar to the treatment group concerning the distribution of covariates. To achieve this, we use propensity score matching techniques.6 Specifically, each firm acquired by a foreign investor in 𝑡+ 1 gets paired with a domestic firm based on the firm’s propensity score (1:1 matching), which we predict from a logistic regression of a dummy indicating a foreign acquisition on 3-digit-sector dummies, year dummies, the logs of the number of employees, the average wage, labor productivity (measured as output per worker) as well as output and wage growth. All variables relate to two years prior to acquisition. We then estimate the following event study equation on the matched sample: 𝑦𝑖𝑡𝑐 =𝛽0+ 8 ∑ 𝑗=−4 𝛽𝑗𝐷𝑗 𝑖,𝑡 +𝜏𝑡𝑐 +𝜙𝑖+𝜉𝑖𝑡𝑐 (1) The dependent variable, 𝑦𝑖𝑡𝑐 , can be any outcome of firm 𝑖at time 𝑡. The subscript 𝑐indicates the cohort, i.e., the year the firm was acquired or matched as a control. We are interested in the vector of coefficients 𝛽on the event indicators 𝐷𝑗 𝑖,𝑡. The coefficient for the last year before the acquisition (𝑗= −1) is normalized to zero, and therefore, all estimates are to be interpreted relative to that year. The coefficients for years prior to takeover (𝑗 < 0) test pre-trends while the coefficients (𝑗≥0) capture the post-acquisition dynamics. The effects are estimated conditional on the time-cohort fixed effect 𝜏𝑡𝑐 and the firm fixed effect 𝜙𝑖. While the latter absorbs all time-invariant firm characteristics, 𝜏𝑡𝑐 controls for time effects by cohort of treated and control firms. Finally, 𝜉𝑖𝑡𝑐 is a random disturbance term. We follow Abadie and Spiess (2022) and cluster standard errors at the level of the matched pairs. We also recognize recent advancements in the difference-in-differences and event study literature (de Chaisemartin and D’Haultfœuille,2020;Goodman-Bacon,2021;Sun and Abraham,2021;Callaway and Sant’Anna,2021) and employ the event study estimator suggested by Sun and Abraham (2021).7 4. Descriptive statistics Panel A of Table 1 provides summary statistics of main variables for the full sample. This contains over 900,000 firm-year observations from 1995 to 2019. Panel B shows summary statistics for firms acquired by a foreign investor, and Panel C for the matched controls (figures for both groups come from the year firms were taken over/matched). Comparing Panel A and Panel B, it is notable that foreign investors pick larger and more productive firms. Panel C shows that the matching successfully homogenizes the estimation sample: acquired firms and their matched controls are of similar size – in particular medians are closely aligned – and also productivity proxies are of comparable magnitude. Fig. A.1 in the appendix provides background information on the origin of investors and the number of takeovers in each year from the post-matching sample. of the voting-right threshold. It is standard practice in this literature, though, to define a firm as foreign if the share of foreign ownership exceeds 10% (cf. Bajgar and Javorcik (2020)). In practice, most ownership shares are much higher, suggesting that changes happen mainly along the extensive margin. 6PSM techniques have been used extensively in this literature, see Girma and Görg (2007), Arnold and Javorcik (2009), Guadalupe et al. (2012), Brucal et al. (2019). 7We use the Stata command eventstudyinteract which implements the estimator suggested by Sun and Abraham (2021). Economics Letters 241 (2024) 111826 3 H. Görg and J. Lehr Table 1 Summary statistics. Mean Sd p50 p10 p90 N A. Full Sample Gross Output (in 1000e) 48 237.74 722 988.78 7789.09 1994.44 63 602.36 924 391 # Employees 170 1438 57 26 283 924391 Average Wage (in 1000e) 35.58 15.43 33.25 17.00 56.42 924 391 Investment (in 1000e) 1671.10 25 359.55 180.46 8.62 2195.14 924 391 Gross Output/# Employees 181.16 399.89 130.84 53.58 322.52 924391 B. Firms Taken Over Gross Output (in 1000e) 57 502.29 154 182.98 25 476.11 8562.27 122 605.23 1354 # Employees 218 306 134 50 434 1354 Average Wage (in 1000e) 39.31 11.57 37.83 26.11 53.91 1354 Investment (in 1000e) 1975912.6 5253759.9 654 821.31 85 696 4 069 002 1354 Gross Output/# Employees 261.66 289.80 186.89 96.93 466.30 1354 C. Control Firms Gross Output (in 1000e) 63 628.8 137 233.58 23 364.35 6081.98 135 089.03 1354 # Employees 249 479 128 43 506 1354 Average Wage (in 1000e) 38.86 11.33 37.97 25.82 53.29 1354 Investment (in 1000e) 2135069.3 6180518.3 521 790.33 47 520.67 4575142.5 1354 Gross Output/# Employees 244.48 282.13 175.37 92.56 445.12 1354 Notes: Part A of the Table shows summary statistics for the main variables from all observations pooled across years. Parts B and C of the Table show summary statistics from the matched sample, separately for treated firms (Part B) and the matched control firms (Part C) from the year firms were matched. Source: Research Data Centre of the Federal Statistical Office and Research Data and Service Centre (RDSC) of the Deutsche Bundesbank: Project-specific Data based on the Census of Manufacturing Industry (1995–2019) and Microdatabase Direct Investment (1999–2019), own calculations. Fig. 1. Main results: Event Study Estimates. Notes: Figures a–e show event study estimates from the matched sample following the approach by Sun and Abraham (2021). . (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Source: Research Data Centre of the Federal Statistical Office and Research Data and Service Centre (RDSC) of the Deutsche Bundesbank:Project-specific Data based on the Census of Manufacturing Industry (1995–2019) and Microdatabase Direct Investment (2000–2019), own calculations. 5. Results Subfigures a–e in Fig. 1 plot the main results from the baseline event study specification shown in Eq. (1). The number of observations in the estimation sample is 54,055. All estimates show the differential evolution in outcomes in the treatment group relative to the control group and the year prior to the acquisition. In the plots, the vertical gray line indicates the acquisition time and the shaded blue areas demarcate the 95th confidence intervals. Fig. 1(a) plots the results with the log of gross output on the lefthand side. Reassuringly, all pre-treatment coefficients are close to zero and statistically insignificant, both individually and jointly. The point estimates for the post-period indicate an increase in output during the first two to three years after the ownership change. These positive effects fade away quickly, as indicated by the statistically insignificant coefficients in later years. In fact, we cannot reject the null hypothesis that all post-coefficients are jointly statistically insignificant. Subfigure b proceeds with the effect on employment: again, pretrends are flat and indistinguishable from zero. By contrast, all posttreatment coefficients are negative, increasing over time and jointly significant. For example, two years after a firm was acquired, employment is 2% lower among treated firms relative to the matched control Economics Letters 241 (2024) 111826 4 H. Görg and J. Lehr Table 2 Effect of acquisitions by region of investment origin. Post treatment Log of Go Log of #Worker Log of Wagebill Log of Wage Log of GO/Worker EU −0.0056 −0.0706*** −0.0552*** 0.0154** 0.0650*** (0.0221) (0.0173) (0.0184) (0.0079) (0.0162) ROW −0.0061 −0.0152 −0.0000 0.0152*0.0091 (0.0248) (0.0177) (0.0200) (0.0084) (0.0183) Pre treatment EU 0.0092 0.0094 0.0083 −0.0012 −0.0002 (0.0158) (0.0120) (0.0124) (0.0065) (0.0123) ROW 0.0048 0.0007 0.0004 −0.0003 0.0041 (0.0180) (0.0129) (0.0133) (0.0067) (0.0131) Notes: The Table reports the joint preand post-takeover coefficients from the event study estimates on a matched sample. The number of observations in the EU-takeover sample is 24,968 and 21,074 in the ROW sample. Source: Research Data Centre of the Federal Statistical Office and Research Data and Service Centre (RDSC) of the Deutsche Bundesbank: Project-specific Data based on the Census of Manufacturing Industry (1995–2019) and Microdatabase Direct Investment (2000–2019), own calculations. * Significance levels is indicated as p <0.1. ** Significance levels is indicated as p <0.05. *** Significance levels is indicated as p <0.01. Fig. A.1. Takeovers in the post-matching sample. Notes: Subfigure (a) reports the number of takeovers per year in the matched sample, separately for takeovers from an investor based in an EU country (blue) and the rest of the world (red). Subfigure (b) shows the number of takeovers in the estimation sample for the ten countries of investment origin. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Source: Research Data Centre of the Federal Statistical Office and Research Data and Service Centre (RDSC) of the Deutsche Bundesbank: Project-specific Data based on the Census of Manufacturing Industry (1995–2019) and Microdatabase Direct Investment (2000–2019), own calculations. firms; and eight years after the takeover, employment has been reduced relative to control firms by 8%. The post-treatment effects are jointly highly significant and indicate an overall reduction in employment as a consequence of the takeover by 4%. A negative effect on the wage bill, i.e., the sum of all wages, as shown in Subfigure c, reflects the decline in employment. In the years immediately after treatment, the wage bill remained stable and only dipped four years after the takeover. The joint effect of the postcoefficients implies a 2% reduction in labor costs, albeit statistically insignificant. We find an immediate increase in gross output per worker and the average wage (Subfigures d and e). In the first year after the takeover, gross output per worker increased by almost 3.5% and wages are about 1% higher relative to the control group. These effects grow larger over time, with all post-coefficients being individually significant. Jointly, the post-coefficients indicate a 2% increase in the average wage and a 5% increase in gross output per worker. We check the robustness of our results towards alternative specifications. Fig. A.2 in the appendix plots the estimates for all outcomes from four alternative specifications: a conventional event study ignoring recent refinements to the difference-in-differences framework, our baseline specification with year-cohort-sector fixed effects instead of year-cohort fixed effects only, the baseline specification estimated on a balanced panel and finally an exact matching within economic sectors.8 Results are robust towards the choice of the empirical specification. We split the sample of treated firms into those whose investor is from an EU country and those with an investor from the rest of the world. Table 2 reports the subsample results; the joint coefficients from the post-treatment years are in the upper part of the Table, and the lower part of the Table reports the joint pre-treatment coefficients. Reassuringly, all pre-treatment coefficients are small and statistically insignificant and thus support the parallel trends assumption. It is evident from the Table that EU countries drive the main results, as shown in Eq. (1). The average reduction in employment after an investment from an EU country over the 8-year post-treatment event window is 7%. This estimate is 3% larger than in the main results. Moreover, we also find a substantial and highly significant reduction in the wage bill. Nonetheless, average wages increase by about 1.5% after an acquisition, and we also find a positive effect on labor productivity. 8We use the same propensity score as in the baseline specification and then match within two-digit industries, enforcing common support within each industry. Economics Letters 241 (2024) 111826 5 H. Görg and J. Lehr Fig. A.2. Robustness: Alternative specifications. Notes: Each subfigure reports results for one outcome from four different empirical specifications: a conventional event study as opposed to the approach by Sun and Abraham (2021), a specification with year-cohort-sector fixed effect, a balanced panel specification and an alternative matching, i.e., matching within economic sectors. The number of observations varies by specification and is, in the above order, 54055, 51978, 27098, 53365. The dots in the figures mark the point estimates, and the vertical lines show the 95th confidence intervals, with standard errors clustered at the level of the matched pairs. Source: Research Data Centre of the Federal Statistical Office and Research Data and Service Centre (RDSC) of the Deutsche Bundesbank: Project-specific Data based on the Census of Manufacturing Industry (1995–2019) and Microdatabase Direct Investment (2000–2019), own calculations. The effects of acquisitions from the rest of the world are all signconsistent with the main effects but smaller and insignificant except for the effect on average wages. 6. Conclusion This paper presents first evidence on the effect of foreign acquisitions on German manufacturing firms based on newly available unique administrative micro data covering two and a half decades. We find that acquired firms experience a positive and lasting boost to productivity: Output per worker increases by about 5%, and average wages by roughly 2% compared to the control group. The effects are mainly due to investments from other EU countries. The increase in productivity results from a reduction in employment and overall labor costs while maintaining the same output level. This mechanism might be most relevant in countries close to the technological frontier, such as Germany, whereas upgrading technology after a takeover presumably plays a larger role in developing countries for any positive effects on productivity. From a policy point of view, our results suggest that foreign acquisitions can benefit the economy by increasing aggregate productivity and improving competitiveness. This, however, may come at the expense of somewhat lower employment. Declaration of competing 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. Data availability The data that has been used is confidential. Appendix See Figs. A.1 and A.2 References Abadie, Alberto, Spiess, Jann, 2022. Robust post-matching inference. J. Amer. Statist. Assoc. 117 (538), 983–995. http://dx.doi.org/10.1080/01621459.2020.1840383. Arnold, Jens Matthias, Javorcik, Beata S., 2009. Gifted kids or pushy parents? Foreign direct investment and plant productivity in Indonesia. J. Int. Econ. 79 (1), 42–53. http://dx.doi.org/10.1016/j.jinteco.2009.05.004. Arnold, David, Milligan, Kevin S., Moon, Terry, Tavakoli, Amirhossein, 2023. Job transitions and employee earnings after acquisitions: Linking corporate and worker outcomes. Working Paper 31866, National Bureau of Economic Research, http: //dx.doi.org/10.3386/w31866. Bajgar, Matej, Javorcik, Beata, 2020. Climbing the rungs of the quality ladder: FDI and domestic exporters in Romania. Econ. J. 130 (628), 937–955. http://dx.doi.org/ 10.1093/ej/ueaa003. Blank, S., Lipponer, A., Schild, C.-J., Scholz, D., 2020. Microdatabase Direct Investment (MiDi) – A full survey of German inward and outward investment. Ger. Econ. Rev. 21 (3), 273–311. Boddin, Dominik, Kruse, Hendrik, Merkle, Hariolf, Walter, Susanne, Zapf, Benedikt, 2024. Combined international trade and investment data ‘‘CITID’’ – A new innovative and comprehensive data landscape for Germany. Jahrbüucher für Nationalökonomie und Statistik http://dx.doi.org/10.1515/jbnst-2024-0024. Brucal, Arlan, Javorcik, Beata, Love, Inessa, 2019. Good for the environment, good for business: Foreign acquisitions and energy intensity. J. Int. Econ. 121, 103247. http://dx.doi.org/10.1016/j.jinteco.2019.07.002. Callaway, Brantly, Sant’Anna, Pedro H.C., 2021. Difference-in-differences with multiple time periods. J. Econometrics 225 (2), 200–230. http://dx.doi.org/10.1016/j. jeconom.2020.12.001, Themed Issue: Treatment Effect 1. Criscuolo, Chiara, Martin, Ralf, 2009. Multinationals and U.S. productivity leadership: Evidence from Great Britain. Rev. Econ. Stat. 91 (2), 263–281. de Chaisemartin, Clément, D’Haultfœuille, Xavier, 2020. Two-way fixed effects estimators with heterogeneous treatment effects. Amer. Econ. Rev. 110 (9), 2964–2996. http://dx.doi.org/10.1257/aer.20181169. Friederich, K., Pham-Dao, L., Schild, C., Scholz, D., Schumacher, J., 2021. Microdatabase Direct Investment: Data Report 2023-22. Deutsche Bundesbank, Research Data and Service Centre. Economics Letters 241 (2024) 111826 6 H. Görg and J. Lehr Girma, Sourafel, Gong, Yundan, Görg, Holger, Lancheros, Sandra, 2015. Estimating direct and indirect effects of foreign direct investment on firm productivity in the presence of interactions between firms. J. Int. Econ. 95 (1), 157–169. http: //dx.doi.org/10.1016/j.jinteco.2014.11.007. Girma, Sourafel, Görg, Holger, 2007. Evaluating the foreign ownership wage premium using a difference-in-differences matching approach. J. Int. Econ. 72 (1), 97–112. Goodman-Bacon, Andrew, 2021. Difference-in-differences with variation in treatment timing. J. Econometrics 225 (2), 254–277. http://dx.doi.org/10.1016/j.jeconom. 2021.03.014, Themed Issue: Treatment Effect 1. Görg, Holger, Greenaway, David, 2004. Much ado about nothing? Do domestic firms really benefit from foreign direct investment? World Bank Res. Observer 19 (2), 171–198. http://dx.doi.org/10.1093/wbro/lkh019. Greenaway, David, Kneller, Richard, 2007. Firm heterogeneity, exporting and foreign direct investment. Econom. J. 117 (517), F134–161. Guadalupe, Maria, Kuzmina, Olga, Thomas, Catherine, 2012. Innovation and foreign ownership. Amer. Econ. Rev. 102 (7), 3594–3627. Keller, Wolfgang, 2024. Knowledge spillovers, international trade, and foreign direct investment. In: Oxford Research Encyclopedia of Economics and Finance. http: //dx.doi.org/10.1093/acrefore/9780190625979.013.705. Kruse, Hendrik W., Hieber, Florian, Limberg, Heiko, Zapf, Benedikt, Boddin, Dominik, 2023. Außenhandelsaktive Unternehmen: Neue Analysemöglichkeiten durch Mikrodatenverknüpfung. WISTA (5), 13–26. Sun, Liyang, Abraham, Sarah, 2021. Estimating dynamic treatment effects in event studies with heterogeneous treatment effects. J. Econometrics 225 (2), 175–199. http://dx.doi.org/10.1016/j.jeconom.2020.09.006, Themed Issue: Treatment Effect 1. United Nations Conference on Trade and Development, 2023. World investment report, 2023. https://unctad.org/publication/world-investment-report-2023. (Accessed: 01 February 2024). Wagner, Joachim, Weche-Gelübcke, John P., 2015. Access to finance, foreign ownership and foreign takeovers in Germany. Appl. Econ. 47 (29), 3092–3112. Weche-Gelübcke, John P., 2015. The impact of foreign takeovers: Comparative evidence from foreign and domestic acquisitions in Germany. Appl. Econ. 47 (8), 739–755.