The Great Recession and its Aftermath: Evidence from Micro-Data: Themenheft 6/Bd. 234(2014) Jahrbücher für Nationalökonomie und Statistik 9783110511161, 9783828206120

Citation preview

The Great Recession and its Aftermath: Evidence from Micro-Data Edited by Holger Görg and Joachim Wagner

With Contributions by Bellmann, Lutz, University of ErlangenNuremberg Driffield, Nigel, Aston University, Birmingham, UK Gustavsson Tingvall, Patrik, The Ratio Institute, Stockholm Hübler, Olaf, Leibniz Universität Hannover Hundt, Christian, Ruhr Universität Bochum Kaitila, Ville, The Research Institute of the Finnish Economy (ETLA), Helsinki Kokko, Ari, Copenhagen Business School and The Ratio Institut, Stockholm M c Quinn, John, The Economic and Social Research Institute (ESRI), Dublin

Lucius &c Lucius • Stuttgart 2 0 1 4

Siedschlag, Iulia, The Economic and Social Research Institute (ESRI), Dublin Söderlund, Bengt, Stockholm School of Economics and The Ratio Institut, Stockholm Sternberg, Rolf, Leibniz Universität Hannover Temouri, Yama, Aston University, Birmingham, UK Wagner, Joachim, Leuphana University Lueneburg Weche Gelübcke, John P., Leuphana University Lueneburg Zhang, Xiaoheng, The Economic and Social Research Institute (ESRI), Dublin

Guest Editors Prof. Dr. Holger Görg Christian-Albrechts-Universität zu Kiel Wirtschafts- und Sozialwissenschaftliche Fakultät Institut für V W L Olshausenstraße 4 0 2 4 1 1 8 Kiel [email protected] Prof. Dr. Joachim Wagner Leuphana University Lueneburg Institute of Economics Scharnhorststr. 1, C 4 . 2 0 4 2 1 3 3 5 Lüneburg [email protected]

Bibliografische Information der Deutschen Nationalbibliothek Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http://dnb.d-nb.de abrufbar ISBN 9 7 8 - 3 - 8 2 8 2 - 0 6 1 2 - 0

© Lucius 8c Lucius Verlagsgesellschaft mbH • Stuttgart • 2 0 1 4 Gerokstraße 5 1 , D - 7 0 1 8 4 Stuttgart Das Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt. Jede Verwertung außerhalb der engen Grenzen des Urheberrechtsgesetzes ist ohne Zustimmung des Verlags unzulässig und strafbar. Das gilt insbesondere für Vervielfältigungen, Übersetzungen und Mikroverfilmungen und die Einspeicherung und Verarbeitung in elektronischen Systemen.

Satz: inMEDIAlo, Digital- und Printmedien UG, Plankstadt Druck und Bindung: Neumann Druck, Heidelberg Printed in Germany

Jahrbücher f. Nationalökonomie u. Statistik (Lucius & Lucius, Stuttgart 2014) Bd. (Vol.) 234/6

Inhalt / Contents Guest Editorial

660-661

Abhandlungen/Original Papers Siedschlag, lulia, Ville Kaitila, John Mc Quinn, Xiaoheng Zhang, International Investment and Firm Performance: Empirical Evidence from Small Open Economies Kokko, Ari, Bengt Söderlund, Patrik Gustavsson Tingvall, Redirecting International Trade: Contracts, Conflicts, and Institutions Hundt, Christian, Rolf Sternberg, How Did the Economic Crisis Influence New Firm Creation? Wagner, Joachim, John P. Weche Geliibcke, Risk or Resilience? The Role of Trade Integration and Foreign Ownership for the Survival of German Enterprises During the Crisis 2008-2010 Driffield, Nigel, Yama Temouri, Inward Investment and the Drivers of Post Recession Recovery in Germany Bellmann, Lutz, Olaf Hübler, The Skill Shortage in German Establishments Before, During and After the Great Recession

662-687 688-721 722-756 757-774 775-799 800-828

Buchbesprechungen / Book Reviews Haucap, Justus, Carolin Klein, Jürgen Kühling, Die Marktintegration der Stromerzeugung aus erneuerbaren Energien Nell, Edward /., Karim Errouaki, Rational Econometric Man: Transforming Structural Econometrics Rumpf, Dominik, Zinsbereinigung bei der Dualen Einkommensteuer Sylvers, Malcolm, Brigitte Domurath-Sylvers, Mythen und Kritik der Ideengeschichte der USA: 25 Portraits Bandinhalt des 234. Jahrgangs der Zeitschrift für Nationalökonomie und Statistik Contents of Volume 234 of the Journal of Economics and Statistics

829 831 834 837

Jahrbücher f. Nationalökonomie u. Statistik (Lucius & Lucius, Stuttgart 2014) Bd. (Vol.) 234/6

Guest Editorial In 2008/2009 the world economy was hit by the hardest economic crisis for decades. This period has been labeled The Great Recession, and it was (at least, in many countries, including Germany) followed by a recovery starting in 2010. Numerous empirical studies using aggregate data at the level of countries, regions or industries have been published that contribute to our understanding of the economic developments during these turbulent times. What is lacking, however, are micro-economic studies using data for individuals and firms that help to understand the micro-structure of the crisis and the recovery. This gap in the literature was (at least, in part) caused by a lack of suitable longitudinal data that allow following individuals and firms over the years from 2008 to 2010 and after. The growing availability of longitudinal micro-level data of this type changes this situation gradually. With the papers in this special issue we intend to contribute to this emerging literature. The paper by Siedschlag et al. starts with the observation that international mergers and acquisition (M&A) activity declined sharply during the Great Recession. However, this created opportunities for such multinationals that had ample access to finance. Based on this observation they use firm level data from Amadeus and Zephyr to look at the impact of international M&As on the performance of targets in six European countries, Austria, Belgium, Denmark, Finland, the Netherlands and Sweden. The analysis suggests evidence for cherry picking - foreign investors tend to acquire larger firms in all countries. The impact of M&As on the targets is highly country specific, however. For example, in Belgium the estimated effect on productivity growth after M&A is negative, while it is positive in Finland. This shows the importance of taking into account country specificities when drawing up policy responses to the crisis. Kokko et al. take the collapse of international trade during the crisis as their starting point. They use Swedish firm level data from Statistics Sweden covering the period 1997 to 2009. Based on this they analyse how Swedish firms have adjusted their trade structure in the wake of the crisis. They find that Swedish exports and imports of intermediates have shifted over time from the OECD region toward rapidly growing non-OECD economies that generally exhibit lower institutional quality than the traditional OECD markets. This shift is driven strongly by differences in economic growth. However, weak institutions in the destination markets are an obstacle for Swedish exports in particular from industries characterized by high conflict-intensity. Four papers in this special issue focus on the German economy. Hundt and Sternberg look at new firm formation before, during, and after the great economic crisis in German regions. Their analysis is based on micro data from the Adult Population Survey of the Global Entrepreneurship Monitor (GEM), and it explicitly distinguishes between the individual level, the time level, and the spatial level within a multi-level system. They find, among others, that both time-related and time-invariant regional determinants contribute significantly to explaining new firm formation; that personal attributes of an individual, time and space are interrelated when it comes to individual entrepreneurial behavior; that high-capacity urban regions were not better off in terms of entrepreneurship during the crisis compared to high-capacity non-urban regions; that differences between opportunity-driven and necessity-driven entrepreneurial activities matter; and that regions with a high GDP per capita profit from a global crisis due to a relative increase in

Guest Editorial • 6 6 1

the number of opportunity-driven start-ups. Based on their empirical findings, the authors discuss implications for entrepreneurial support policies. Wagner and Weche Geliibcke use a newly available tailor-made data set for enterprises from manufacturing industries in Germany to investigate the link between different forms of internationalization (importing, exporting, and foreign ownership) and firm survival during the 2 0 0 8 / 2 0 0 9 crisis. They find a disadvantage of exporting for survival chances of a firm, while importing is positively related with survival. Furthermore, their results do not support the hypothesis that foreign multinationals are more volatile during a crisis. Driffield and Temouri use enterprise level data from the ORBIS data base to look at productivity differences between various groups of firms in Germany. They demonstrate that foreign owned firms are more productive than German multinational enterprises and purely domestic firms, with the gap narrowing in the manufacturing sector, but growing in the service sector after the start of the crisis. Bellmann and Hiibler investigate the development of skill shortages in Germany before, during and after the crises. Their econometric analysis is based on data from the German Establishment Panel of the Institute for Employment Research (IAB). They report differences over time in relative skill shortage. Furthermore, they point out that apprenticeship and further training serve to reduce the number of unfilled, high-skill jobs. The papers in this special issue demonstrate that micro-econometric studies which use longitudinal data for individuals and firms can help to understand the micro-structure of an economic crisis and recovery. Therefore, we are looking forward to seeing more studies of this kind. Holger Joachim

Görg

Wagner

Jahrbücher f. Nationalökonomie u. Statistik (Lucius & Lucius, Stuttgart 2014) Bd. (Vol.) 234/6

Abhandlungen / Original Papers International Investment and Firm Performance: Empirical Evidence from Small Open Economies lulia Siedschlag*bc, Ville Kaitiladr John Mc Quinn ab , and Xiaoheng Zhang"* * The Economic and Social Research Institute (ESRI), Dublin b Department of Economics, Trinity College Dublin c Institute for International Integration Studies, Trinity College Dublin d The Research Institute of the Finnish Economy (ETLA), Helsinki a

JEL F16; F23; J24 Multinational firms; productivity; employment; propensity score matching.

Summary The global financial and economic crisis has severely affected foreign direct investment, particularly the cross-border mergers and acquisitions in advanced economies. This paper examines the effects of foreign mergers and acquisitions on labour productivity and employment growth over the period 2001-2009 in six small open economies in the European Union: Austria, Belgium, Denmark, Finland, the Netherlands and Sweden. We show that the severity of the crisis has been uneven across these six economies. Taken together, our estimates suggest that foreign direct investment had stronger effects on firm performance in services than in manufacturing.

1

Introduction

The global financial and economic crisis has severely affected foreign direct investment (FDI), particularly in developed economies. In 2008, FDI inflows to developed countries contracted by 2 9 % , mainly due to a sharp decline in cross-border mergers and acquisitions (M&A) sales that fell by 3 9 % in value in comparison to 2007 (UNCTAD 2009). 1 In 2009, FDI inflows to developed economies declined further, by 4 4 % , again mainly due to a severe contraction of 6 5 % in the value of cross-border M & A sales, particularly in manufacturing (UNCTAD 2010). 2 The sharp decline in cross-border M & A sales is linked to their higher sensitivity to financial conditions, given their shorter investment cycles, than those of greenfield investments. In addition, the turmoil in stock markets distorted the price signals upon which M & A sales rely. However, while depressed stock prices reduced the value of transactions, in combination with global restructuring, they also

* This paper was part of the SERVICEGAP project funded by the European Commission under the 7th Framework Programme, Grant Agreement No. 244522. We thank Holger Görg, Mika Maliranta, participants in research presentations at a workshop and an international conference held at the Economic and Social Research Institute in Dublin, the Editor, and two anonymous referees for their useful comments and suggestions on earlier versions. 1 Cross-border M8cA peaked in 2007 after a five-year worldwide boom. 1 In comparison to 2008, in 2009 cross-border M & A sales in manufacturing were down by 77%, while in services, they declined by 57%.

International Investment and Firm Performance • 663

generated opportunities for m u l t i n a t i o n a l s t h a t w e r e still able t o access finance ( U N C T A D 2010). Given these developments in international investment activity during the recent global financial a n d e c o n o m i c crisis, it appears pertinent t o analyse the impacts o f c r o s s - b o r d e r M & A on firm p e r f o r m a n c e . Foreign mergers a n d acquisitions imply a c h a n g e o f o w n ership a n d they thus provide a n a t u r a l e x p e r i m e n t w h i c h c a n help to identify the effects o f foreign o w n e r s h i p on firm p e r f o r m a n c e . W h i l e m o s t existing analyses have focused o n firms in m a n u f a c t u r i n g , the evidence for firms in services is scarce. T o fill this evidence gap, w e use t w o rich m i c r o data sets 3 available f o r the period 2 0 0 1 - 2 0 0 9 , a n d analyse the effects o f foreign mergers a n d acquisitions on productivity a n d e m p l o y m e n t g r o w t h in service firms in six small open E u r o p e a n U n i o n (EU) countries. 4 E c o n o m i c g r o w t h in small open e c o n o m i e s is m o r e dependent on F D I inflows a n d it is therefore m o r e vulnerable to changes in i n t e r n a t i o n a l investment flows. As d o c u m e n t e d in previous studies (Johne/Storey 1 9 9 8 ; M i l e s 2 0 0 5 ) , services have a n u m ber o f distinct characteristics, such as: (i) their intangibility; (ii) simultaneity o f their p r o duction a n d c o n s u m p t i o n ; a n d (iii) perishability. T h e s e specific characteristics together w i t h the fact t h a t services a c c o u n t for a g r o w i n g share o f e c o n o m i c activity in advanced e c o n o m i e s , m o t i v a t e the focus o f this analysis o n services. T o identify service specific effects o f foreign acquisition o n productivity a n d e m p l o y m e n t g r o w t h , w e c o m p a r e these results with the c o r r e s p o n d i n g evidence o b t a i n e d f o r m a n u f a c t u r i n g firms. O u r d a t a , described in detail b e l o w , spans the period 2 0 0 1 to 2 0 0 9 , i.e., it includes the recent g l o b a l financial and e c o n o m i c crisis. W e s h o w t h a t the severity o f the crisis has been uneven across these six e c o n o m i e s . In p a r t i c u l a r , the decline in real G D P a n d e m p l o y m e n t g r o w t h in 2 0 0 8 a n d 2 0 0 9 w a s m o r e severe in D e n m a r k , F i n l a n d and Sweden than in the o t h e r three analysed small open e c o n o m i e s , Austria, Belgium and the N e t h e r l a n d s . Productivity g r o w t h declined in all six e c o n o m i e s , with the biggest decrease in Finland. Against this m a c r o e c o n o m i c b a c k g r o u n d , the a n n u a l average over 2 0 0 8 - 2 0 0 9 f o r crossb o r d e r M & A sales declined in Austria, F i n l a n d , the N e t h e r l a n d s and D e n m a r k , while it w a s higher than their value in 2 0 0 7 in Belgium a n d Sweden. O u r evidence indicates t h a t , in b o t h m a n u f a c t u r i n g and services sectors, foreign investors tend t o acquire larger firms. O t h e r characteristics o f acquired firms differ across c o u n tries a n d between m a n u f a c t u r i n g a n d services. T a k e n together, o u r estimates suggest t h a t foreign direct investment h a d stronger effects o n firm p e r f o r m a n c e in services in c o m parison t o m a n u f a c t u r i n g . O v e r a l l , w e find t h a t the effects o f foreign direct investment o n l a b o u r productivity and e m p l o y m e n t g r o w t h w e r e c o u n t r y specific. T h e rest o f this p a p e r is organised as f o l l o w s . Section 2 reviews the relevant theoretical a n d empirical literature. Section 3 presents o u r d a t a a n d s u m m a r y statistics while o u r empirical m e t h o d o l o g y is outlined in Section 4 . Section 5 discusses o u r empirical results. Finally, Section 5 concludes.

3 4

Amadeus and Zephyr, provided by Bureau van Dijk, http://www.bvdinfo.com. Austria, Belgium, Denmark, Finland, the Netherlands, and Sweden.

664 • lulia Siedschlag, Ville Kaitila, John Mc Quinn, and Xiaoheng Zhang

2

Theoretical and empirical framework

Firms' post-acquisition performance depends on the pre-acquisition performance and characteristics of both the acquired and acquirer firms. In relation to the pre-acquisition performance of the acquired firms, two hypotheses have been put forward in the existing literature. O n the one hand, the synergy-effects hypothesis argues that 'cherries' (i.e. good performers) are more likely to be acquired. On the other hand, the management's comparative advantage (or managerial-discipline) hypothesis suggests that 'lemons' (i.e. bad performers) are more likely to be acquired. In both cases, the performance of the acquired firms is expected to improve after acquisition. These two hypotheses maintain that the aim of the acquisition is to maximise profits. Productivity is expected to rise when foreign investors transfer their superior firm-specific advantages to their foreign affiliates. 5 Table 1 summarises the main empirical findings of these studies with respect to preacquisition performance. Out of 42 studies reviewed in Table 1, 22 conclude that foreign firms "cherry picked" high productivity firms. On the other hand, two studies find evidence that foreign firms acquired local firms with below-average productivity, 6 while six studies do not analyse this question. 7 Overall, the bulk of the existing evidence suggests that foreign investors tend to acquire high productivity firms. Cross-border M & A may involve either the most or the least efficient foreign investors. Nocke and Yeaple (2007) show that, in industries in which the source of firm heterogeneity is linked to internationally mobile capabilities, such as R&D-intensive industries, foreign investors are the most productive, while foreign acquirers in industries with low or non-mobile capabilities are the least productive. This evidence implies that potential productivity spillovers are expected to be the highest when the foreign acquirer is in a R&D-intensive industry and the lowest, or even negative, when the foreign investor operates in industries with low R & D intensity. The productivity impact of foreign investment on the acquired firm may depend on its absorptive capacity, i.e. the level of education of its employees (see for example, Nelson/Phelps 1966). Thus, it may be that only a firm with a higher productivity when acquired will be able to absorb the more advanced technology of the foreign-owned firm (Lapan/Bardhan 1973). Consequently, an acquired exporting firm may receive greater benefits than an acquired domestic firm, as found by Bandick and Gorg (2010). On the other hand, it has also been suggested that a large technological gap between the foreignowned firm and the acquired firm may lead to a larger productivity boost in the latter (Findlay 1978). This situation has been analysed, for example, by Girma (2005a).

5 6

7

See, for example, Gugler et al. (2003), Fukao et al. (2008) and Balsvik and Haller (2010). According to Gioia and Thomsen (2004), foreign buyers tend to buy poor performers in Denmark as measured by return on assets and factor productivity. They argue that this is because of information disadvantages leading to a double "lemons problem". Bertrand and Zitouna (2008) find evidence of lemons picking in French manufacturing industries. Similar evidence is found by Girma and Gorg (2007) for the UK electronics and food industries, and Harris (2009) for UK service industries. 12 of the studies referred to in Table 1 use data from the UK. However, even these find different answers to the question about cherry-picking.

International Investment and Firm Performance • 665

Table 1 Empirical Evidence on Pre-Acquisition Performance and Foreign Acquisitions in Manufacturing and Services Authors

Country

Manu-

Services

facturing Aitken and Harrison (1999) Conyon et al. (2002a) Harris and Robinson (2002) Girma and Gorg (2004) Gioia and Thomsen (2004) Gugler and Yurtoglu (2004) llmakunnas and Maliranta (2004) Fukao, Ito, and Kwon (2005) Fukao and Murakami (2005) Girma (2005b) Piscitello and Rabbiosi (2005) Bellak, Pfaffermayr, and Wild (2006) Benfratello and Sembenelli (2006) Girma and Gorg (2007) Girma, Kneller, and Pisu (2007) Almeida (2007) Hanley and Zervos (2007) Gong, Gorg, and Maioli (2007) Huttunen (2007) Karpaty (2007) Bertrand and Zitouna (2008) Csengódi, Jungnickel, and Urban (2008) Fukao et al. (2008) Lehto and Bockerman (2008) Girma, Gorg, and Pisu (2008) Salis (2008) Arnold and Javorcik (2009) Criscuolo and Martin (2009) Bandick and Hansson (2009) Chari, Chen, and Domínguez (2009) Harris (2009) Schiffbauer, Siedschlag, and Ruane (2009) Arndt and Mattes (2010) Balsvick and Haller (2010) Lipsey, Sjóholm, and Sun (2010) Bandick and Gorg (2010) Mattes (2010) Bandick (2011) Bandick and Karpaty (2011) Chen (2011) Vahter (2011) Greenaway, Guaraglia, and Yu (2012)

Venezuela UK UK UK Denmark US, UK, Europe Finland Japan Japan UK Italy Austria Italy UK UK Portugal UK China Finland Sweden France Hungary Japan Finland UK Slovenia Indonesia UK Sweden USA UK UK Germany Norway Indonesia Sweden Germany Sweden Sweden USA Estonia China

Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

Cherry Picking

No No No No No No No No No No No No No No No No No No No No No No Yes Yes No No No No No No Yes Yes No No No No No No No No No No

Yes No Yes Yes No Yes Yes Yes No Yes Yes No Yes Yes Yes No No No Yes Yes Yes Yes Yes No No No No Yes Yes Yes Yes Yes Yes No

Existing evidence suggests that while in the short term productivity and employment growth may be negatively correlated, in the long term productivity growth correlates positively with employment g r o w t h (Pissarides/Vallanti 2 0 0 4 ; L a n d m a n n 2004). In the short term, restructuring and performance-boosting measures following acquisition may

666 • lulia Siedschlag, Ville Kaitila, John Mc Quinn, and Xiaoheng Zhang lead to lower employment and higher labour productivity.8 Existing studies have typically analysed total factor productivity which also accounts for the efficiency of the capital stock use. Foreign-owned firms are often more capital-intensive than domestic firms.9 With respect to empirical methodologies, early analyses of effects of foreign acquisitions on firm performance have used Ordinary Least Square (OLS) estimators (Aitken/Harrison 1999; Conyon et al. 2002a; Gioia/Thomsen 2 0 0 4 ; Piscitello/Rabbiosi 2005; Fukao/Murakami 2 0 0 5 ; Hanley/Zervos 2 0 0 7 ; Balsvik/Haller 2010) or system G M M (Harris/Robinson 2 0 0 2 ; Gugler/Yurtoglu 2 0 0 4 ; Harris 2009). To capture the causal link between foreign ownership and firm performance, more recent studies use propensity score matching combined with difference-in-differences estimators. These studies include Girma (2005a, 2005b); Bellak et al. (2006); Girma et al. (2007); Karpaty (2007); Huttunen (2007); Salis (2008); Bertrand and Zitouna (2008); Arnold and Javorcik (2009); Bandick and Hansson (2009); Schiffbauer et al. (2009); Lipsey et al. (2010). In most cases, firms were analysed for at least two years before and after the acquisition. However, in some studies (for example, Chen 2011), acquisition effects are found only five years after the event, which suggests the need to extend the analysed period. Nevertheless, it is difficult to assess a priori how many years it takes for the possible effects of an ownership change to fully sink in. Evidence from the reviewed literature indicates that foreign acquisitions tend to result in higher productivity growth and that the productivity level remains higher relative to the pre-acquisition period. 10 This productivity boost can be linked to restructuring of inefficient plants, which may involve labour shedding and new capital investments. Most existing studies use data on manufacturing firms, with only a few including also service firms. Using data from the UK, Harris (2009) found TFP gains in the acquired service sector plants. However, it appears that these productivity gains decline over time. In contrast, Schiffbauer et al. (2009) found no TFP effects of foreign acquisitions on service firms in the UK. In comparison to the evidence on effects of foreign acquisitions on productivity, the evidence with respect to the employment effects of foreign acquisitions is less conclusive Faster employment growth after acquisition is found by Piscitello and Rabbiosi (2005) for Italy; Gong et al. (2007) for privatised Chinese firms; Almeida (2007) for Portugal; Arnold and Javorcik (2009), and Lipsey et al. (2010) for Indonesia; Balsvik and Haller (2010) for Norway; and Bandick and Gorg (2010), as well as Bandick and Karpaty (2011) for Sweden. Negative employment effects have been found by, among others, Conyon et al. (2002b) for the UK; Csengodi et al. (2008) for Hungary; and Chari et al. (2009) for the United States. In a number of studies, the employment effects have been found to depend on the sector, the size of the acquired firms, or the skill-level of the labour force. Girma (2005b)

8

9

10

For example G i r m a ( 2 0 0 5 b ) found that foreign acquisitions in the UK led to an increase in labouruse efficiency. O n the other hand, Piscitello and R a b b i o s i ( 2 0 0 5 ) as well as Arnold and J a v o r c i k ( 2 0 0 9 ) find that there has been a rise in both labour productivity and employment in foreignacquired Italian and Indonesian firms, respectively. For UK manufacturing, Schiffbauer et al. ( 2 0 0 9 ) found no effect o f foreign M & A on total factor productivity (TFP). T h e y found that labour productivity rose due to capital deepening. Furthermore, they found positive T F P effects when the acquirer was in R8cD-intensive industries and negative effects when the acquirer was in marketing-intensive industries. Also negative productivity effects have been found, e.g. Hanley and Zervos ( 2 0 0 7 ) for UK manufacturing.

International Investment and Firm Performance • 667

found negative employment effects in larger acquired British firms and positive effects in smaller ones. Using Swedish data, Bandick and Gorg (2010) found that the increase in employment was larger in exporters and smaller in acquired multinational firms, but both effects occurred only if the takeover was vertical. There were no effects if the target was a purely domestic firm or if the acquisition was horizontal. Huttunen (2007) as well as Lehto and Bockerman (2008) found negative employment effects of foreign acquisitions in Finland albeit with some variation depending on the skill groups and sectors. Huttunen's results indicate that the share of highly-skilled workers declined in the post-acquisition period. On the other hand, Bandick and Hansson (2009) found that in Sweden, the relative demand for skilled labour rose in foreign-acquired nonmultinational firms (but not in acquired multinational firms). Also Bandick and Karpaty (2011) found an increase in skilled employment in Sweden following foreign acquisitions. Girma and Gorg (2004) found slower employment growth in the electronics industry in the UK, in particular for unskilled labour, but no significant effects in the food sector. Lipsey et al. (2010) found that in foreign-acquired firms in Indonesia, blue-collar employment grew faster than white-collar employment. Only few studies report evidence of employment effects of foreign acquisitions of service firms. Fukao et al. (2008) report a fall, albeit temporary, in non-manufacturing employment following a foreign takeover. Lehto and Bockerman (2008) found some evidence of negative employment effects in construction and other services in Finland, but no effects in trade, hotels and restaurants. Harris (2009) found that post-acquisition employment changed very little in service sectors in the UK. The review of the literature discussed above suggests that foreign acquisition leads most frequently to productivity increases, while the employment performance of firms in the post-acquisition period appears to be more mixed, depending on firm and sector characteristics. While most of the previous evidence has been obtained for manufacturing firms, the evidence on the effects of foreign acquisitions on service firms is very limited. 3

Data and descriptive statistics

We use firm level data from the Bureau van Dijk's Amadeus and Zephyr datasets for firms in the six small and open economies we analyse in this paper. The period under examination covers the years 2001 to 2 0 0 9 . Amadeus is a large micro dataset including information on firm characteristics, financial performance and legal structure while Zephyr has detailed information on mergers and acquisitions, notably cross-border transactions. Using common identifiers we combine these two datasets for this study. Considering their legal form, companies are grouped into three broad categories: limited companies, limited liability companies, and other forms. We use data on unconsolidated accounts for only the first two categories to allow comparability across countries as these two categories correspond to public and private limited companies, respectively. Firms are classified according to their two-digit NACE code (Rev.1.1), which enables us to separate service firms (NACE 5 0 - NACE 74) from manufacturing firms (NACE 15 - NACE 37) and explore heterogeneity between the two sectors. We define a foreign acquisition as any change of ownership from domestic to foreign passing over a threshold of 10 per cent of total shareholding in line with officially recognised definitions of foreign direct investment.11 Given data availability, we only consider

11

For a definition of FDI see International Monetary Fund ( 1 9 9 3 ) .

668 • lulia Siedschlag, Ville Kaitila, John Mc Quinn, and Xiaoheng Zhang

medium-sized and large firms. 12 These are defined in the Amadeus data set as firms fulfilling at least one of the following conditions: the n u m b e r of employees is greater t h a n 15, operating revenue is greater t h a n one million euros and/or total assets are greater t h a n t w o million euros. Financial institutions and insurance companies are excluded f r o m the Amadeus database due to compatibility issues with respect to the f o r m a t of financial accounts. Using available data, we construct the following firm variables: the age of the firm; employment; the ratio of debt to fixed assets; tangible fixed assets per employee (capitallabour ratio); turnover per employee 13 (labour productivity); employment growth rate; turnover per employee growth rate (labour productivity growth); a foreign acquisition d u m m y (binary variable equal to one in the year w h e n the acquirer's stake passes 10 per cent); and industry, region and year dummies. We use industry producer price indices at the t w o digit level to deflate manufacturing firm monetary variables with 2 0 0 5 as the base year and a G D P deflator with 2 0 0 6 as the base year for service firms. Finally, our sample is restricted to non-negative observations for tangible fixed assets and the number of employees while debt is restricted to values equal to or greater than zero. 14 T h e available data are limited by missing values. Assuming that missing data are randomly missing, w e generate these data using a weighted hotdeck imputing methodology. This is a multiple imputation process whereby five datasets are generated using a stochastic process and combined using the Rubin's Rule. 15 A detailed description of the imputation method is given in Appendix A. Summary statistics are presented in Tables B1-B6 in Appendix B. Relative to manufacturing, firms in services are smaller, younger, more productive, more capital-intensive a n d have a higher debt burden. Further, relative to manufacturing, foreign-acquired firms in services are smaller, older, more productive (with the exception of Belgium and the Netherlands), more capital-intensive (with the exception of the Netherlands), and with higher debt burden. Comparing foreign acquired firms in services across countries, the average size is the largest in Austria and the smallest in Sweden; the average age is the highest in Belgium and the lowest in Finland; average productivity - the highest in the Netherlands and the lowest in Sweden and Finland; average capital intensity - the highest in Austria and the lowest in Finland; the average debt burden - the highest in the Netherlands and the lowest in Austria. T o put the results of our analysis in the context of the global financial and economic crisis, we provide a brief overview of descriptive statistics of macroeconomic performance and cross border M & A activity in 2 0 0 8 and 2 0 0 9 . The severity of the global economic and financial crisis has been uneven across the six small open economies we analyse in this paper: Austria, Belgium, D e n m a r k , Finland, the Netherlands and Sweden.

12 13

14 15

Data is more frequently missing in the case of small firms. The choice of labour productivity measure based on turnover is motivated by concerns over measurement errors given the lack of prices for intermediates if value added were chosen as output measure. These choices are motivated by using in the analysis logarithmic transformations of these variables. See Andridge and Little (2010).

International Investment and Firm Performance • 669

• 2007 • 2008 • 2009 Source-.

Eurostat

Figure 1 Real GDP growth, percentage change on previous year 4

• 2007 Source:

2008

• 2009

Eurostat

Figure 2 Employment growth, percentage change on previous year

Figures 1 and 2 indicate that the decline in real G D P and employment growth in 2 0 0 8 and 2 0 0 9 was more severe in the three Nordic countries, D e n m a r k , Finland, and Sweden than in the other three small open economies, Austria, Belgium and the Netherlands. Figure 3 shows that the economic and financial crisis also resulted in a decline in productivity growth in all six economies. A m o n g these countries, Finland experienced the biggest decrease.

670 • lulia Siedschlag, Ville Kaitila, John M c Quinn, and Xiaoheng Zhang

• 2007 «2008 • 2009 Source-. Eurostat

Figure 3 Real labour productivity growth, percentage change on previous year

• 2001-2006

• 2007

• 2008

• 2009

Source: Eurostat

Figure 4 Cross-border M & A sales, mill. US dollars

Against this macroeconomic background, the annual average over 2 0 0 8 - 2 0 0 9 for crossborder M & A sales declined in Austria, Finland, the Netherlands and Denmark, while in Belgium and Sweden it was higher than their value in 2007.

International Investment and Firm Performance • 671

4

Empirical methodology

We use propensity score matching combined with difference-in-differences estimators (Heckman et al. 1997) to examine the causal effect of foreign acquisition on firm productivity and employment growth. T o this purpose, we first estimate the propensity of foreign acquisition (the treatment, D) conditioned by the observed firm characteristics, X. We then use the propensity score to match foreign-acquired firms and domestic nonacquired firms assuming conditional independence, i.e. that foreign acquisitions are only determined by observables X and not by any unobservable characteristics. In combination with this assumption, a substantial overlap between the propensity score of the treated and untreated firms, also referred to as the common support assumption, allows matching non-acquired (control) firms to acquired (treated) firms such that: (Y1,Y0).LD|p(X)andO 5 M- > « « ° S g rt O = u _q S £ O

£ «j a>

Redirecting International Trade: Contracts, Conflicts, and Institutions • 707

In summary, the results from Table 2 show that institutions, relationship-specificity, and conflict risk have a systematic effect on the extensive margins of exports and offshoring. A weak institutional environment hampers exports and offshoring in general, and in particular exports in conflict-intensive sectors and offshoring in relationship-specific sectors. Translating these findings into a context where the role of non-OECD economies is expected to increase further suggests that many firms may find it difficult to restructure their export relationships. The institutional environment in many non-OECD countries is weaker, which suggest that there is a higher threshold for engaging in trade. In particular, there is reason to expect that industries recording high values for the Cl-index (RS-index) may find it hard to shift their exports (offshoring) from OECD to non-OECD countries. Some changes in the structure of trade can therefore be expected, at least during a transition period while firms learn to navigate new institutional environments. Tables 3 - 6 explore some extensions and test the robustness of the findings stated above. In Table 3, we explore some of the asymmetries between OECD and non-OECD countries in closer detail. More specifically, we analyze how the effect of institutions and related variables vary across countries with different levels of institutional quality and economic growth. Columns 1 - 2 (for offshoring) and 5 - 6 (for exports) compare countries with different levels of institutional quality. Countries are classified in the "good institutions" group if their value for the variable I n s t c j is above average and in the "weak institutions" group otherwise. Columns 3 - 4 (for offshoring) and 7 - 8 (for exports) examine the link between economic growth, trade, and institutions. The question here is whether differences in institutional quality have any influence on the expected trade-promoting impact of high growth rates in the partner country. A first point to note from columns 1 - 2 and 6 - 7 in Table 3 is that the marginal impact of institutional quality is larger in countries with lower-than-average institutional quality than in countries with well-developed institutions. This asymmetry between country groups is statistically significant for exports, but not for offshoring. A second notable point is that the inclusion of per capita income growth in the destination market has a strongly significant and positive impact on exports (column 7), but a weaker and less significant effect on offshoring (column 3). Columns 4 and 8 explore the impact of growth further by appending an interaction term between income growth and institutional quality ( p c i _ g r o w t h c t - I n s t c t ) • It can be seen that there are no systematic asymmetric effects of income growth on offshoring, while exports to high growth economies are boosted by strong institutions. This means that exports are drawn to growing economies in general but that weak institutions hamper this reallocation. For offshoring, it should be noted that income growth has two opposite effects that may cancel each other. On the one hand, income growth and development are expected to raise the variety of products available for offshoring, suggesting a positive effect. On the other hand, high growth is associated with increasing wages and prices, which could have a negative impact on offshoring.

5.2

Robustness: varying the multilateral trade resistance term

To capture the effects of relative prices on trade patterns and to overcome the omitted variable bias, the specification of the multilateral trade resistance term plays a central role in the estimation of gravity models (Anderson/van Wincoop 2003; Feenstra 2004). To analyze the robustness of the results with respect to various specifications of the M T R term, Table 4 presents results with different dummy variable set-ups. In columns 1 (offshoring) and 5 (exports), we introduce industry dummies at the two-digit level. These

7 0 8

• Ari

Kokko, Bengt

Söderlund,

and

Patrik

(N i

o o

-

o

8 P

J 8

CI íT\

o

9

o

o

fs

co in o o o o o o

l

ÛO 00

o

2

S

ö

I

t

S g ® o 9 S

»

Gustavsson

c

t

Tingvall

J)

3 2 E E -a o 3 ~ c O -a I^ oí) • =

Q.

S Q. •e

o

¥ « m a E ^ O e M E J 3 O-

fM O o

o o

rv m VC o o o o o

t ! c

in

j t ' Ê o r-;

i

oo r^ m o o o o

; .s !

i 1 3 ; -o

,

9 o

O

KDra

rv o o o o o d

-t-» . t c I3 _ ~ O

i: "K c 3 3 -o O c U

VO o d

00

c¡ E t00

n c

•

8

g S

W M •o E S £

- I Üß rt c

Institutions

o i

s< 15 «

o o o o

'co "D £ Ol — ( £3 «2

ü |

d' SjìOc -0a0

a» 5 o -

gl

£ =

t

D; £

°

*

4) "O E > 1C 3T C3

®

4> 0_

c

3 S Jß : '5. Q -S3 • 8 c oc C Z. o c4 nL» 3 «^ £4>: £ Ä Q R O. > Os O Q. Q. •

£ 'S

£

:

5i w U U

ä

5 & ^ rt « O ~o -O t "g £ E ti f >, 'C CL 3 O — — £ I § 5 ^ 5 fi I § > 6 0 b C >N 0) o o "o > oo E .£ .E E l l 4 M) H C < •••

; 5 LO

H o w Did t h e Economic Crisis Influence N e w Firm Creation? • 7 4 7

1 II 111

o d

I 11 I

ill

¡ I ! I s S ± ±5

1 1 1 i a ±

±a j-s

¡ 1 1

SIS lilt

:| 1 II

a á a

SIS

5 3

IS

Si 13 3! II 11 if 11

1

g

*A

¡a

aS ai

II.. II H 'llll

I

II

1 I 1

áá s s ; § S 5 3

?

!H ^

s s 8?

SI

:IS

SS2 ° a a

748 • Christian Hundt and Rolf Sternberg

Appendix A

X 2 : Household Income

Control Variables Dependent Variable

Individual level

X 3 : Knowing Entrepreneurs

Dummy: 1 = upper third

Dummy: 1 = yes

Gender

Dummy: 1 = male

Age

Dummy: 1 = 25-44

Entrepreneurial Experience

Dummy: 1 = yes

Y: Involved in Entrepreneurial Activity

Dummy: 1 = yes

Min 0.000 Max 1.000 Mean 0.372 SD 0.482 Min 0.000 Max 1.000 Mean 0.282 SD 0.405 Min 0.000 Max 1.000 Mean 0.286 SD 0.453 Min 0.000 Max 1.000 Mean 0.457 SD 0.498 Min 0.000 Max 1.000 Mean 0.421 SD 0.494 Min 0.000 Max 1.000 Mean 0.015 SD 0.122 Min 0.000 Max 1.000 Mean 0.049 SD 0.217 N 64,053

Model 3: Necessitydriven early-stage entrepreneurship

X 1 : Educational Attainment

Dummy: 1 = university of applied sciences entrance diploma or higher

Model 2: Opportunitydriven early-stage entrepreneurship

Model 1: Early-stage entrepreneurship

Descriptive statistics for variables at the individual level

0.000 1.000 0.371 0.482 0.000 1.000 0.282 0.404 0.000 1.000 0.283 0.451 0.000 1.000 0.455 0.498 0.000 1.000 0.420 0.494 0.000 1.000 0.014 0.119 0.000 1.000 0.036 0.187 63,160

0.000 1.000 0.364 0.480 0.000 1.000 0.275 0.402 0.000 1.000 0.272 0.445 0.000 1.000 0.450 0.500 0.000 1.000 0.416 0.493 0.000 1.000 0.014 0.116 0.000 1.000 0.014 0.119 61,771

Please note: Missing values were estimated by interpolating the available data at fixed rates. Data source: GEM Adult Populations Surveys Germany 2000-2012

How Did the Economic Crisis Influence New Firm Creation? • 749 Appendix B

Regional level (n=39)

R 1 : GDP per capita (In)

Logarithmic value of regional GDP per capita

R 2 : Population Density

Total population / size of NUTS2-region in square km

Time level (n=105)

Descriptive statistics for variables at the context levels

T 1 : Shock

Dummy: 1 = occurence of the economic recession in 2008/2009

T 2 : Change in Umemployment Rates

Percentaged change in regional unemployment rates from period to period

Data source: Cambridge Econometrics 2000-2012

Min Max Mean SD Min Max Mean SD Min Max Mean SD Min Max Mean SD

4.340 5.300 4.713 0.220 0.700 3.820 0.433 0.702 0.000 1.000 0.200 0.401 (-)6.900 6.500 (-)0.605 2.158

750 • Christian Hundt and Rolf Sternberg

Appendix C Multicollinearity test Variable X,

Tolerance

Variance Inflation Factor

R 1 : GDP per capita (In)

0.890

1.124

R 2 : Population Density

0.890

1.124

T 1 : Shock

0.633

1.579

T 2 : Change in Umemployment Rates

0.633

1.579

X 1 : Educational Attainment

0.943

1.106

X 2 : Household Income

0.951

1.051

X 3 : Knowing Entrepreneurs

0.962

1.039

Gender

0.980

1.020

Age

0.982

1.018

Entrepreneurial Experience

0.993

1.007

CD >

.Si "S o

tuo OJ

"Ö) > CJ a;

E i-

"Ö) >

n)

-a >

i/) aj -Q

nS > O c

o (J

Tolerance and Variance Inflation Factor are definded as '1 - R 2 , and '1/1 - R?' respectively, where R? is the multiple R 2 for the regression of X, on the other covariates at the same level (without the response variable Y). Data source: GEM Adult Populations Surveys Germany 2000-2012; Cambridge Econometrics 2000-2012

How Did the Economic Crisis Influence New Firm Creation? • 751

Appendix D Source and description of the independent variables Independent variable

Description

Source

R1: Regional GDP per capita

GDP is the sum of gross value added of the various institutional sectors or the various industries plus taxes and less subsidies on products (logarithmic value of GDP per capita)

CAMBRIDGE ECONOMETRICS > Sub-National > Sub-National: Europe > Regional Database > NUTS2 Other > GDP http://www.camecon.com/SubNational/ SubNationalEurope/RegionalDatabase.aspx

R2: Population density

Total Population / size of NUTS2-region in square km

CAMBRIDGE ECONOMETRICS > Sub-National > Sub-National: Europe > Regional Database > NUTS2 Other > 'Population', 'Area' http://www.camecon.com/SubNational/ SubNationalEurope/RegionalDatabase.aspx

T1: Economic Occurrence of the economic shock recession in 2008/2009 (Dummy: 1 = yes, 0 = no)

Own calculation

T2: Change in unemployment rates

Regional (NUTS2 level) unemployment rate represents unemployed persons as a percentage of the economically active population (percentaged change from period to period)

CAMBRIDGE ECONOMETRICS > Sub-National > Sub-National: Europe > Regional Database > NUTS2 Other > Unemployment rate http://www.camecon.com/SubNational/ SubNationalEurope/RegionalDatabase.aspx

X1: Educational attainment

Highest educational attainment GEM adult population surveys Germany 2000-2012: GLOBAL ENTREPRENEURSHIP MONITOR > DATA completed? http://www.gemconsortium.org/Data

X2: Household income

Total annual income of all the members of your household, including your income, as one combined figure?

GEM adult population surveys Germany 2000-2012: GLOBAL ENTREPRENEURSHIP MONITOR > DATA http://www.gemconsortium.org/Data

Knowing other entrepreneurs? X3: Knowing other (Dummy: 1 = yes, 0 = no) entrepreneurs

GEM adult population surveys Germany 2000-2012: GLOBAL ENTREPRENEURSHIP MONITOR > DATA http://www.gemconsortium.org/Data

Gender (control variable)

Respondent's gender

GEM adult population surveys Germany 2000-2012: GLOBAL ENTREPRENEURSHIP MONITOR > DATA http://www.gemconsortium.org/Data

Age (control variable)

Respondent's age

GEM adult population surveys Germany 2000-2012: GLOBAL ENTREPRENEURSHIP MONITOR > DATA http://www.gemconsortium.org/Data

Entrepreneurial experience (control variable)

Shut down, discontinued, or quit a business in the past 12 months? (Dummy: 1 = yes, 0 = no)

GEM adult population surveys Germany 2000-2012: GLOBAL ENTREPRENEURSHIP MONITOR > DATA http://www.gemconsortium.org/Data

752 • Christian Hundt and Rolf Sternberg

Appendix E Source and description of the dependent variables Dependent variable

Description

Source

Nascent entrepreneur

You are, alone or with others, currently trying to start a new business, including any self-employment or selling any goods or services to others. (Yes/No) If 'Yes': Over the past twelve months have you done anything to help start a new business, such as looking for equipment or a location, organising a start-up team, working on a business plan, beginning to save money, or any other activity that would help launch a business? (Yes/No)

GEM adult population surveys Germany 2000-2012: GLOBAL ENTREPRENEURSHIP MONITOR > DATA http://www.gemconsortium.org/Data

Young entrepreneur

You are, alone or with others, currently the owner of a company you help manage, self-employed, or selling any goods or services to others. (Yes/No) If 'Yes': What was the first year the owners received wages, profits, or payments in kind?

GEM adult population surveys Germany 2000-2012: GLOBAL ENTREPRENEURSHIP MONITOR > DATA http://www.gemconsortium.org/Data

(year) 'Payment' must not date back more than 3.5 years. Early-stage entrepreneur (model 1)

18 to 64 years old individual who is either a nascent or a young entrepreneur (as defined above)

GEM adult population surveys Germany 2000-2012: GLOBAL ENTREPRENEURSHIP MONITOR > DATA http://www.gemconsortium.org/Data

Necessitydriven early-stage entrepreneur (model 3)

18 to 64 years old early-stage entrepreneur (as defined above) who has no better choice for work alternative classification for the year 2000: no employee now or in five years

GEM adult population surveys Germany 2000-2012: GLOBAL ENTREPRENEURSHIP MONITOR > DATA http://www.gemconsortium.org/Data

Opportunitydriven early-stage entrepreneur (model 2)

18 to 64 years old early-stage entrepreneur (as defined above) who a) takes advantage of a business opportunity or b) takes advantage of a business opportunity and, at the same time, has no better choice for work alternative classification for the year 2000: at least 75 employees now or in five years

GEM adult population surveys Germany 2000-2012: GLOBAL ENTREPRENEURSHIP MONITOR > DATA http://www.gemconsortium.org/Data

How Did the Economic Crisis Influence New Firm Creation? • 753

References Acs, Z.J., N. Bosma, R. Sternberg (2011), Entrepreneurship in world cities. Pp. 125-152 in: M. Minniti (ed.), The Dynamics of Entrepreneurial Activity. Oxford: Oxford University Press. Amoros, E., N. Bosma (2014), Global Entrepreneurship Monitor 2013 Global Report. GERA. http://www.gemcons0rtium.0rg/d0cs/3106/gem-2013-global-report. Audretsch, D., O. Falck, M. Heblich (2011), Who's got the aces up his sleeve? Functional specialization of cities and entrepreneurship. Annuals of Regional Science 46(3): 621-636. Beaudry, C., A. Schiffauerova (2009), Who's right, Marshall or Jacobs? The localization versus urbanization debate. Research Policy 38: 318-337. Bilsborrow, R. (1998), The state of the art and overview of the chapters. Pp. 1 - 5 6 in: R. Bilsborrow (ed.), Migration, urbanization, and development: New directions and issues. Kluwer Academic Publishers, Boston et al. Bosma, N. (2009), The geography of entrepreneurial activity and regional economic development: Multilevel analyses for Dutch and European regions. Utrecht: University of Utrecht. Bosma, N. (2012), The Global Entrepreneurship Monitor (GEM) and its impact on entrepreneurship research. GEM Working Paper Series, No. 12-01 (version 9 august 2012); http://ideas.repec.Org/p/gem/wpaper/1201.html Bosma, N., J. Hessels, V. Schutjens, M. van Praag, I. Verheul (2012), Entrepreneurship and role models. Journal of Economic Psychology 33: 410^424. Bosma, N., J. Levie (2010), Global Entrepreneurship Monitor: 2009 Global Report. Babson Park, MA, Santiago de Chile, Reykjavik, London: Babson College, Universidad del Desarrollo, Reykjavik University, London Business School. Bosma, N., V. Schutjens (2009), Mapping entrepreneurial activity and entrepreneurial attitudes in European regions. International Journal of Entrepreneurship and Small Business 9(2): 191-213. Bosma, N., R. Sternberg (2014), Entrepreneurship as an urban event? Empirical evidence from European cities. Regional Studies 48(6): 1016-1033. DOI: 10.1080/00343404.2014.904041 Brixy, U., R. Sternberg, H. Stiiber (2012), The selectiveness of the entrepreneurial process. Journal of Small Business Management 50(1): 105-131. Brixy, U., C. Hundt, R. Sternberg, A. Vorderwülbecke (2011), Global Entrepreneurship Monitor - Länderbericht Deutschland 2010. Hannover, Nürnberg: Leibniz Universität Hannover; IAB. Burda, M., J. Hunt (2011), What explains the German labor market miracle in the great recession? Brookings Papers on Economic Activity 42(1): 273-335. Caliendo, M., A.S. Kritikos (2010), Start-ups by the unemployed: Characteristics, survival and direct employment effects. Small Business Economics 35(1): 71-92. Caliendo, M., S. Kiinn (2011), Start-up subsidies for the unemployed: Long-term evidence and effect heterogeneity. Journal of Public Economics 95(3-4): 311-331. Caliendo, M., S. Kiinn (2014), Regional effect heterogeneity of start-up subsidies for the unemployed. Regional Studies 48(6): 1108-1134. DOI: 10.1080/00343404.2013.851784 Caliendo, M., J. Hogenacker, S. Künn, F. Wießner (2012), Gründungszuschuss für Arbeitslose. Bislang solider Nachfolger der früheren Programme. IAB-Kurzbericht 2/2012. Capello, R. (2002), Entrepreneurship and spatial externalities: Theory and measurement. The Annals of Regional Science 36: 386-402. Cheng, S. (2011), Business cycle, industrial composition, or regional advantage? A decomposition analysis of new firm formation in the United States. Annals of Regional Science 47: 147-161. Chlosta, S., H. Patzelt, S.B. Klein, C. Dormann (2012), Parental role models and the decision to become self-employed: The moderating effect of personality. Small Business Economics 38: 121-138. Coleman, J.S. (1991), Grundlagen der Sozialtheorie 1, Handlungen und Handlungssysteme. München: Oldenbourg Verlag (Scientia Nova).

754 • Christian Hundt and Rolf Sternberg

Congregado, E., A.A. Golpe, S. Parker (2009), The dynamics of entrepreneurship: Hysteresis, business cycles and government policy. IZA discussion paper N o 4093; ftp://repec.iza.org/ RePEc/Discussionpaper/dp4093.pdf. Congregado, E., A. Golpe, A. van Stel (2012), The 'recession-push' hypothesis reconsidered. International Entrepreneurship Management Journal 8: 325-342. Cowling, M., W. Liu, A. Ledger (2013), Small business financing in the UK before and during the current financial crisis. International Small Business Journal 30(7): 778-800. Davidsson, P., S.R. Gordon (2011), Panel studies of new venture creation: a methods-focused review and suggestions for future research. Small Business Economics 39: 853-876. Etzioni, A. (1987), Entrepreneurship, adaptation and legitimation: a macro-behavioral perspective. Journal of Economic Behavior & Organization 8(2): 175-189. Fairlie, R.W. (2013), Entrepreneurship, economic conditions, and the Great Recession. Journal of Economics & Management Strategy 22(2): 207-231. Feldman, M.P. (2001), The entrepreneurial event revisited: Firm formation in a regional context. Industrial and Corporate Change 10: 861-891. Foster, K.A. (2007), A case study approach to understanding regional resilience. Working Paper 2007-08, Institute of Urban and Regional Development, University of California, Berkeley. Franke, N., M . Gruber, D. Harhoff, J. Henkel (2006), W h a t You Are Is What You Like: Similarity Biases in Venture Capitalists' Evaluations of Start-Up Teams. Journal of Business Venturing 21: 802-826. Fritsch, M., O. Falck (2007), N e w business formation by industry over space and time: a multidimensional analysis. Regional Studies 41(2): 157-172. Fritsch, M., P. Mueller (2007), The persistence of regional new business formation-activity over time - Assessing the potential of policy promotion programs. Journal of Evolutionary Economics 17: 299-315. Fritsch, M., M. Wyrwich (2014), The long persistence of regional entrepreneurship: Germany 1925 to 2005. Regional Studies 48(6): 955-973. Glaeser, E. (2007), Entrepreneurship and the city. NBER Working Paper 13551. Glaeser, G.L., W.R. Kerr, G.A.M. Ponzetto (2010), Clusters of entrepreneurship. Journal of Urban Economics 67(1): 150-168. Hamilton, R., D. Harper (1994), The entrepreneur in theory and practice. Journal of Economic Studies 21(6): 3 - 1 8 . Helsley, R.W., W.C. Strange (2011), Entrepreneurs and cities: complexity, thickness and balance. Regional Science and Urban Economics 41: 550-559. Hessels, J., M. van Gelderen, R. Thurik (2008), Entrepreneurial aspirations, motivations, and their drivers. Small Business Economics 31(3): 323-339. Hindle, K. (2010), H o w community context affects entrepreneurial process: A diagnostic framework. Entrepreneurship & Regional Development 9(7-8): 599-647. Holling, C.S., L.H. Gunderson, (2002), Resilience and adaptive cycles. Pp. 2 5 - 6 2 in: L. Gunderson, C.S. Holling (eds.), Panarchy: Understanding transformations in human and natural systems. Washington, DC: Island Press. Hox, J. (2002), Multilevel analysis: Techniques and applications. Mawah, N.J.: Lawrence Erlbaum Associates. Hundt, C. (2012), Zur Erklärung von Gründungsaktivitäten - Eine Mehrebenenanalyse aus individueller, regionaler und nationaler Perspektive [Explaining new firm creation: A multilevel analysis based upon data of individuals, regions, and countries] (Wirtschaftsgeographie, No. 54). LIT-Verlag, Münster. Hundt, C., R. Sternberg (2014), Explaining new firm creation in Europe from a spatial and time perspective - A multilevel analysis based upon data of individuals, regions and countries. Papers in Regional Science. Online first: doi:10.1111/pirs.12133. Kacperczyk, A. (2013), Social influence and entrepreneurship: The effect of university peers on entrepreneurial entry. Organization Science 24: 664-683. Kelley, D., N . Bosma, J.E. Amoros (2011), Global Entrepreneurship Monitor 2010 Global Report. GERA.

How Did the Economic Crisis Influence New Firm Creation? • 755

Kim, G., J. Cho (2009), Entry dynamics of self-employment in South Korea. Entrepreneurship 8c Regional Development 21(3): 3 0 3 - 3 2 3 . Klapper, L., I. Love (2011), The impact of the financial crisis on new firm registration. Economics Letters 113(1): 1 - 4 . Koellinger, P.D., A.R. Thurik (2012), Entrepreneurship and the business cycle. The Review of Economics and Statistics 94(4): 1 1 4 3 - 1 1 5 6 . L0we Nielsen, S.L., K. Klyver, M . R . Evald, T. Bager (2012), Entrepreneurship in theory and practice: Paradoxes in play. Cheltenham, N o r t h a m p t o n : Edward Elgar. Luke, D.A. (2004), Multilevel modeling. Sage University Paper Series on Quantitative Applications in Social Sciences, 0 7 - 1 4 3 . N e w b u r y Park, CA: Sage. Malecki, E.J. (2009), Geographical environments for entrepreneurship. International Journal of Entrepreneurship and Small Business 7(2): 1 7 5 - 1 9 0 . Martinez, J., I. M i r a , J . M . Gómez (2013), Influence of the economic cycle on the determinants of nascent entrepreneurial activity. An empirical analysis of the Spanish case. Investigaciones Regionales 26, special issue: 1 9 - 4 6 . McMullen, J.S., R.D. Bagby, L.E. Palich (2008), Economic freedom and the motivation to engage in entrepreneurial action. Entrepreneurship Theory and Practice 32(5): 8 7 5 - 8 9 5 . Minniti, M . (2005), Entrepreneurship and network externalities. Journal of Economic Behavior and Organization 57(1): 1 - 2 7 . Möller, J. (2010), The german labor market response in the world recession: De-mystifying a miracle. Journal for Labour M a r k e t Research 42(4): 3 2 5 - 3 3 6 . N a n d a , R., J.B. S0renson (2010), Workplace peer effects and entrepreneurship. M a n a g e m e n t Science 56, 1 1 6 - 1 1 2 6 . Neffke, F.H.M., R. Boschma, K.J. Lundquist, L.O. Olander (2011), T h e dynamics of agglomeration externalities along the life cycle of industries. Regional Studies 45(1): 4 9 - 6 5 . Parker, S.C. (2009), The economics of entrepreneurship. Cambridge: Cambridge University Press. Parker, S.C. (2012), Theories of entrepreneurship, innovation and the business cycle. Journal of Economic Surveys 26(3): 3 7 7 - 3 3 9 . Rabe-Hesketh, S., A. Skrondal (2008), Multilevel and longitudinal modeling using Stata. 2 n d ed. College Station: Stata Press. Raspe, O . (2009), The regional knowledge economy: A multilevel perspective on firm performance and localized knowledge externalities. Utrecht: University of Utrecht. Raudenbush, S.W., A. Bryk (2002), Hierarchical linear models: Applications and data analyses methods. 2 n d ed. N e w b u r y Park, CA: Sage Publications. Raudenbush, S.W., A. Bryk, Y.F. Cheong, R. C o n g d o n , M . du Toit (2011), H L M 7: Hierarchical linear and nonlinear modelling. Lincolnwood, IL: SSI Scientific Software International. Reynolds, P.D., N . Bosma, E. Autio, S. H u n t , N . de Bono, I. Serváis, P. Lopez-Garcia, N . Chin (2005), Global Entrepreneurship M o n i t o r : Data collection design and implementation 1 9 9 8 - 2 0 0 3 . Small Business Economics 24(3): 2 0 5 - 2 3 1 . Reynolds, P.D., R.T. Curtin (2011), N e w business creation: An international overview. N e w York, Dordrecht, Heidelberg, London: Springer. Reynolds, P.D., D. Storey, P. Westhead (1994), Cross-national comparisons of the variation in new firm formation rates - an editorial overview. Regional Studies 28: 3 4 3 - 3 4 6 . Rinne, U., K. Z i m m e r m a n n (2013), Is Germany the N o r t h star of labor market policy? IMF Economic Review 61(4): 7 0 2 - 7 2 9 . Robinson, W. (1950), Ecological correlations and behaviour of individuals. American Sociological Review 15: 3 5 1 - 3 5 7 . Rosenthal, S.S., W.C. Strange (2004), Evidence on the nature and sources of agglomeration economies. Pp. 2 1 1 9 - 2 1 7 2 in: J.V. Henderson, J.F. Thisse (eds.), H a n d b o o k of Urban and Regional Economics, Vol. 4. Amsterdam: Elsevier. Saxenian, A. (1994), Regional advantage: Culture and competition in Silicon Valley and Route 128. Cambridge, M A : H a r v a r d University Press.

756 • Christian Hundt and Rolf Sternberg

Schultz, T.W. (1961), Investment in human capital. The American Economic Review 51(1): 1-17. Schultz, T.W. (1980), Investment in entrepreneurial ability. The Scandinavian Journal of Economics 82(4): 437-348. Schumpeter, J.A. (1939), Business cycles: A theoretical, historical and statistical analysis of the capitalist process. New York/London: McGraw-Hill. Shane, S. (2003), A general theory of entrepreneurship: The individual-opportunity-nexus. Cheltenham, UK, Northampton, USA: Edward Elgar Publishing. Shane, S. (2009), Why encouraging more people to become entrepreneurs is bad public policy. Small Business Economics 33(2): 141-149. Simmie, J., R. Martin (2010), The economic resilience of regions: Towards an evolutionary approach. Cambridge Journal of Regions, Economy and Society 3: 27—43. Snijders, T., R. Bosker (2004), Multilevel analysis: An introduction to basic and advanced multilevel modelling. ed. Sage, London et al. Sorenson, O., P.G. Audia (2000), The social structure of entrepreneurial activity: Geographic concentration of footwear production in the United States, 1940-1989. American Journal of Sociology 106: 424^162. Stam, E. (2007), Why butterflies don't leave: Locational behavior of entrepreneurial firms. Economic Geography 83(1): 27-50. Stam, E. (2010), Entrepreneurship, evolution and geography. Pp. 139-161 in: R. Boschma, R. Martin (eds.), The handbook of evolutionary economic geography. Cheltenham: Edward Elgar. Sternberg, R. (2005), Gründungsförderung in Deutschland und seinen Regionen - eine Bewertung auf Basis des Regional Entrepreneurship Monitor (REM). Pp. 175-204 in: F. Welter (ed.), Dynamik im Unternehmenssektor: Theorie, Empirie und Politik. Berlin: Duncker & Humblot. Sternberg, R. (2009), Regional dimensions of entrepreneurship. Boston, Delft: Now Publishers (Series Foundations and Trends in Entrepreneurship 5(4)). Sternberg, R., H.O. Rocha (2007), Why entrepreneurship is a regional event: Theoretical arguments, empirical evidence, and policy consequences. Pp. 215-238 in: M.P. Rice, T.G. Habbershon (eds.), Entrepreneurship: The engine of growth, Volume 3. Westport/CT, London: Praeger. Sternberg, R., A. Vorderwülbecke, U. Brixy (2014), Global Entrepreneurship Monitor (GEM). Länderbericht Deutschland 2012. Hannover, Nürnberg: Institut für Wirtschafts- und Kulturgeographie, Universität Hannover, Institut für Arbeitsmarkt- und Berufsforschung der Bundesagentur für Arbeit (IAB). Wagner, J., R. Sternberg (2005), Personal and regional determinants of entrepreneurial activities: Empirical evidence from the Regional Entrepreneurship Monitor (REM). Jahrbuch für Regionalwissenschaft 25(1): 91-105. Wong, P.K., Y.P. Ho, E. Autio (2005), Entrepreneurship, innovation and economic growth: Evidence from GEM data. Small Business Economics 24(3): 335-350. Zarth, M. (2011), Zur Entwicklung der deutschen Regionen in den langfristigen Konjunkturzyklen. Informationen zur Raumentwicklung (2): 99-112. Dr. Christian Hundt, Institute of Geography, Ruhr Universität Bochum, Universitätsstraße 150, 44780 Bochum, Germany. [email protected] Prof. Dr. Rolf Sternberg, Institute of Economic and Cultural Geography, Leibniz Universität Hannover, Schneiderberg 50, 30167 Hannover, Germany. [email protected]

Jahrbücher f. Nationalökonomie u. Statistik (Lucius & Lucius, Stuttgart 2014) Bd. (Vol.) 234/6

Risk or Resilience? The Role of Trade Integration and Foreign Ownership for the Survival of German Enterprises During the Crisis 2008-2010 Joachim Wagner Leuphana University Lueneburg, and CESIS, Stockholm John P. Weche Gelübcke Leuphana University Lueneburg JEL F23; L60 Exports, imports; foreign ownership; firm survival; economic crisis, Germany.

Summary This is the first study of the link between internationalization and firm survival during the 2008/2009 crisis in Germany, a country which was hit relatively lightly compared to other countries. Moreover, it is the first study which looks at the role of importing, exporting and FDI simultaneously in the context of a global economic recession. We use a tailor-made representative dataset that covers all enterprises from the manufacturing sector with at least 20 employees. Our most striking result is to demonstrate the disadvantage of exporting for the chances of survival of a firm during the crisis in western Germany. Importing instead reveals a positive correlation with survival and firms that both export and import do not show a different exit risk relative to non-traders. A plausible explanation is that in a global recession, deteriorating markets abroad cause demand losses for exporters and improved conditions on factor markets which result in an advantage for firms sourcing from factor markets abroad. Two-way traders do not show a link with exit risk, supporting the idea that they were able to outweigh their losses from exporting with their gains from importing, in what could be called an exportimport hedge. Furthermore, we cannot support the hypothesis that foreign multinationals are more volatile during times of economic crisis.

1

Introduction

Germany is one of the economies most integrated into the international division of labor. Only a few years ago Germany has been considered the world's leading export nation in terms of the total value of its exported goods. In the year 2012, Germany was the third largest exporter, and, at the same time, the third largest importer, demonstrating a considerable level of trade integration (World Trade Organization 2013, Appendix Table 1.2). In the dimension of foreign direct investment (FDI), only the US, Hong Kong, the UK, and France had a larger inward FDI stock than Germany and direct investments of German multinationals abroad made Germany ranked fourth in terms of outward FDI stocks in 2010 (UNCTAD 2011, Annex Table 1.2).

758 • Joachim W a g n e r and John P. Weche Gelubcke

Although trade integration generally enables welfare-increasing efficiency gains and the exploitation or generation of firms' competitive advantages, the recent global economic crisis, which started in 2008/09, shed light also on the negative aspects of economic internationalization. For example, Kleinert et al. (2012) investigate the transmission of economic shocks within multinational firms' affiliate networks and Wagner (2013a) finds that idiosyncratic shocks to only a few internationalized large firms characterized the export collapse in German manufacturing. The recent economic crisis is an appropriate event for evaluating the general hypothesis of whether or not negative idiosyncratic shocks, such as a shrinking world demand or decreasing availability of investment funds, affect internationalized firms to a greater extent than nationally oriented competitors. In other words, whether internationalization increases an individual firm's vulnerability in times of economic deterioration. The link between firm performance and international activities has been the subject of a huge strand of the empirical literature since the seminal work by Bernard and Jensen (1995). 1 Apart from other dimensions of firm performance, such as productivity and profitability, survival chances are a key aspect of firm performance which is of high importance to all stakeholders. Developments in trade theory have given an additional incentive to study firm survival in the context of international trade, since the so-called new new trade theory predicts a considerable impact of trade on industry structure, meaning in particular that less productive firms leave the market (see Melitz 2003 and Helpman 2013). Previous work on internationalization and firm closure can be separated into that referring to trade activities, including exporting and importing, and that referring to FDI and foreign ownership. 2 To start with the trade criterion, why should exporting activities be linked to firm survival? Exporting can be considered a form of risk diversification through the spread of sales over different markets with different business cycle conditions or being in different phases of the product cycle (see Hirsch/Lev 1971). Therefore, exports might provide a chance to substitute sales abroad for sales at home when a negative demand shock to the home market would have otherwise forced a firm to close down. Furthermore, Baldwin and Yan (2011) argue that non-exporters are in general less efficient than exporters (younger, smaller, and less productive) and that, as a result, one expects that non-exporters are more likely to fail than exporters. Regarding imports, imported intermediate inputs or capital goods might be cheaper and/or technically more advanced than inputs bought on the national market. Gibson and Graciano (2011) argue that the benefit of using imported inputs lies in a combination of the relative price and the technology embodied in the inputs. Imports, therefore, lead to an increase in price competitiveness and non-price competitiveness of importers compared to firms that do not import. Furthermore, there is empirical evidence of a positive link between imports and productivity (discussed in Vogel/Wagner 2010), documented by a significant productivity differential between firms that import and firms that do not trade internationally. Therefore, the probability of surviving can be expected to be higher for importers than for non-importers, ceteris paribus.

1 2

See Wagner (2012) for a survey of recent studies. For a more detailed survey of these two parts of the literature, see Wagner (2011 and 2013b) and Wagner and Weche Gelubcke (2012). The following summary is based on these publications.

Risk or Resilience? • 759

Firms that both export and import can be expected to benefit from the positive effects of both forms of international trade on firm survival. Furthermore, two-way traders tend to be more productive than firms that either only import, or only export, or do not trade at all (see Vogel/Wagner 2010). Therefore, we expect the probability of firm exit to be smaller for two-way traders than for firms that only export or only import. Surveying the empirical literature on the link between the international trade activities of a firm and its survival chances, Wagner (2011 and 2013b) concludes that the survival chances of exporters are generally higher even after controlling for individual firm characteristics, such as size, age, and productivity. His results for Germany show that importers and two-way traders have lower probabilities of closure, but that exporting does not seem to lower the exit risk. Regarding foreign ownership ties, Baldwin and Yan (2011) argue that from a theoretical point of view the relation between foreign ownership and firm exit is not clear. On the one hand, foreign owned firms may have access to superior technologies belonging to their foreign owners that might increase their efficiency and lower the risk of exit. The greater propensity to invest in R & D found in foreign owned firms in Germany might lead to more innovation, improved competitiveness in Germany and in foreign markets, and might therefore increase the chances of survival. On the other hand, Baldwin and Yan (2011) point out that foreign owned firms are less rooted in the host country economy and that they can shift their activities to another country when the local economy deteriorates. This should increase the probability of shutdown compared to nationally owned firms. Empirical studies reveal that the evidence is ambiguous and highly country dependent: foreign owned firms turn out to be more likely to exit in some countries and less likely in other countries. The only studies for Germany, by Andrews et al. (2012) and by Wagner and Weche Geliibcke (2012), conclude that there is a higher exit risk for foreign firms than for domestically owned firms when only dependent subsidiaries are considered, but that foreign owned firms do not differ when also domestically owned independent firms and group heads are part of the control group. Firm-level studies on the link between firms' international status and their exit risk in the context of an economic crisis are rare. Narjoko and Hill (2007) investigate firm survival during the 1997/1998 Indonesian crisis and find export orientation and foreign ownership to be highly significant determinants of both survival and recovery. Although a local crisis is never without an impact on other countries in an integrated economic world and the distinction between a local and a global economic crisis cannot be straightforward, it is a major feature of the recent economic crisis 2008/2009 that it can be characterized as a global phenomenon rather than a local one. 3 This has a significant impact on theoretical considerations regarding the link between international exposure and performance. For example, in a local crisis framework, exporting would be clearly seen as risk diversifying and, therefore, as a measure for overcoming economic slow-downs. In contrast, if a crisis ramifies mainly abroad, export orientation would be regarded as much more risky than a focus on domestic sales. The underlying problem is twofold: i) firms are usually not able to perfectly substitute sales abroad with sales at home and ii) firms are normally not able

3

When we talk a b o u t the impact of the recent economic crisis in the context at hand, we refer mainly to the aspect of a shrinking demand for manufactured products. It is k n o w n that such a demand shock happened in many of the German export markets and that the general demand in Germany was relatively stable, for example, due to stable employment levels.

760 • Joachim Wagner and John P. Weche Gelübcke

to shrink rapidly in reaction of a deteriorating d e m a n d as they m a y expand in prospering times. Therefore, a shift to the h o m e market c a n n o t be regarded as a perfect alternative for most exporting firms, although it m a y serve as an easing factor. Alfaro a n d Chen (2012) find multinational subsidiaries worldwide to have been more resilient during the 2 0 0 8 / 2 0 0 9 global crisis. T h e key determinants t u r n out to be a vertical production link with the parent c o m p a n y and being more closely linked in financial terms. G o d a r t et al. (2012) focus on Ireland and conclude that foreign firms were not more likely to exit than domestically owned firms during the crisis. Amendola et al. (2012) bring together the aspects of trade relatedness, foreign ownership, and o u t w a r d FDI as determinants for firm survival over the crisis with Italian data. Their findings point to more volatile multinational subsidiaries and more resilient Italian exporters. It is self evident that there are other factors shaping the survival chances in times of economic d o w n t u r n . An important feature of economic distress is an increase in interest rates and a potential "credit c r u n c h " . Consequently, Abildgren et al. (2013) stress the importance of bank links in shaping survival chances. They use data on Danish non-financial firms and find a higher default probability of firms with links to " w e a k " banks during the crisis, but they d o not focus on the role of internationalization. 4 T h e contribution of our paper is to provide the first empirical study of the link between internationalization and firm survival during the 2 0 0 8 / 2 0 0 9 crisis in Germany. Hence, it is the first study for a country which was hit relatively lightly compared to other countries, and whose domestic d e m a n d experienced a relatively lesser decline. Furthermore, it is the first analysis adding the role of input sourcing and importing activities to FDI and export status. We use a tailor-made representative dataset that merges information f r o m surveys performed by the G e r m a n statistical offices, f r o m administrative data collected by the T a x Authorities, and f r o m a commercial data provider. The data covers all enterprises f r o m the manufacturing sector with at least 2 0 employees. W e have been able to study the link between international integration and firm exits during non-crisis years previously with the same data (Wagner 2 0 1 3 b ; Wagner/Weche Geliibcke 2012) and are therefore able to c o m p a r e the results f r o m the crisis period with those f r o m a non-crisis period. O u r previous studies focus on different aspects of international integration (Wagner 2 0 1 3 b on trade integration; Wagner/Weche Geliibcke 2 0 1 2 on foreign ownership) which are combined in the study at h a n d . Both studies were performed w h e n data on the crisis period h a d not been available. T o anticipate the most important results, exporting appears to negatively affect survival chances in western Germany during the crisis. Importing instead reveals a positive correlation with survival and firms that both export and import do not show a different exit risk relative to non-traders. A plausible explanation is that in a global recession, deteriorating markets a b r o a d cause d e m a n d losses for exporters and improved conditions in factor markets, something which results in an advantage for firms sourcing a b r o a d . Furthermore, foreign subsidiaries d o not s h o w any different exit probabilities t h a n Germ a n firms, therefore not supporting the hypothesis of more volatile multinationals. In eastern Germany, all our indicators for different aspects of internationalization remain statistically insignificant.

4

Naidoo (2010) also looks at firms' behavior during the recent crisis and specifically at the role played by marketing. His analysis of Chinese export oriented SMEs reveals a positive correlation between marketing innovations and the perceived likelihood of survival.

Risk or Resilience? • 761

The rest of the paper is organized definitions of the variables. Section probit estimates of survival premia a n d other firm-level characteristics.

2

as follows. Section 2 describes the dataset and the 3 presents the descriptive results. Section 4 gives the dependent on several aspects of internationalization Section 5 concludes.

Data and variables

This study uses a tailor-made enterprise level dataset that contains information f r o m surveys performed by the German statistical offices, f r o m data collected by the T a x Authorities, and f r o m a commercial data provider. The first source of data is the monthly and annual reports for establishments in mining, quarrying, and manufacturing industries described in Konold (2007). These surveys cover all establishments f r o m the mining, quarrying, and manufacturing industries that employ at least 2 0 people in the local production unit or in the c o m p a n y that o w n s the unit as a whole. The participation of firms in the survey is m a n d a t e d by law. Participation in this survey is used to identify surviving and exiting firms (discussed in more detail below). This survey is also the source for information on the location of the firm in western Germany or eastern Germany, industry affiliation, whether a firm exports or not, labor productivity (measured as sales per employee), and the n u m b e r of employees (used to measure firm size). Furthermore, given that the data start with the year 1995, this survey is used to distinguish between old firms (that were already covered by the survey in 1995) and new firms (that entered the survey in 1996 or later). N o t e that in this dataset, export refers to the a m o u n t of sales to a customer in a foreign country plus sales to a German export trading company; indirect exports (for example, tires produced in a plant in Germany that are delivered to a German m a n u f a c t u r e r of cars w h o exports some of its products) are not covered by this definition. For this project, the information collected at the establishment level has been aggregated at the enterprise level to match the unit of observation f r o m the other sources of data used here. The second source of data is the G e r m a n Turnover T a x Statistics Panel (described in detail in Vogel/Dittrich 2008). This dataset is based on the yearly turnover tax: all enterprises with a turnover that exceeds a rather low threshold (17,500 EUR since 2003) are covered. This dataset informs us whether a firm imports or not. N o t e , however, that imports are not directly recorded therein completely. Imports f r o m EU member countries are reported under the item of "intra-Community acquisitions". The a m o u n t of imports f r o m countries outside the EU is not included in the turnover tax statistics. In this case an import turnover tax is charged by the customs authorities. Nonetheless, this import turnover tax is deductible as an input tax, and is therefore reported in the dataset. From this information we k n o w whether the enterprise imports f r o m non-EU countries or not. The third source of data is the survey of products (Produktionsstatistik). This survey is used to distinguish between firms that produce only one product and multi-product firms. Information on the foreign ownership status of a firm is based on data f r o m the commercial database MARKUS, a joint product of the commercial data providers Bureau van Dijk and Creditreform. This database reports whether an enterprise is an affiliate, g r o u p head, or independent entity, and whether the g r o u p head of an affiliate is located a b r o a d . Starting with the reporting year 2 0 0 7 , this information was linked to the German enterprise register system (Unternehmensregistersystem) by the German Federal Statistical Office (see Weche Geliibcke 2 0 1 1 for details). A firm is regarded as foreign owned if

762 • Joachim Wagner and John P. Weche Geliibcke

it is an affiliate with a group head located in a foreign country and if more than 5 0 % of the voting rights of the owners or more than 5 0 % of the shares are controlled (directly or indirectly) by a firm or a person/institution located outside Germany. The data from these sources were linked by using the enterprise register system that includes, among other things, information about the unique enterprise identifier used in the surveys conducted by the statistical offices and the unique turnover tax identifier used by the Tax Authorities. Our data covers the years from 2007 to 2010. The global financial crisis started in 2007 and unfolded its impact on the real economy in the last quarter of 2008. Since we are interested in the economic crisis that followed the financial crisis (although these two aspects are of course not easy to separate) we try to start our analysis immediately before the crisis unfolded its real economic impact. Since we use yearly data, to start our analysis in 2007 seems too far from the major breakout at the end of 2008 and 2008 may already be affected by the economic crisis. We decided to start the analysis in 2008 as we assume firm exits to appear with a certain lag after the economic impact and we therefore call 2008 the pre-crisis period. A firm is identified as an exit if it has reported to either the monthly report or the annual report for establishments in mining, quarrying, and manufacturing industries in 2008 but not in the recovery year 2010. Consequently, we assume the identified firms to have exited the market at some point in 2009 or 2010. It should be noted that the definition of firm exit used here is not without problems. First of all, if a firm relocates outside Germany or changes its activities from mining, quarrying, or manufacturing to services or agriculture, it no longer reports to the monthly report or the annual report for establishments in mining, quarrying, and manufacturing industries and, therefore, is considered as an exit. To the best of our knowledge and according to information from the employees in the official statistical office who are in charge of preparing the data used here, this is only rarely the case. Second, the industry classification of the monthly reports and the Turnover Tax Statistics was subject to changes in the year 2009. This means that some identified exits may not be real exits but rather firms which became classified as being outside the scope of the surveys used here in 2009. Therefore we had to exclude the whole of those industries which experienced such re-classifications, namely the publishing sector and the recycling sector.s Third, firms that shrink below the threshold of 20 employees in the local production unit or in the company that owns the unit are no longer obliged to report to the survey (but often do so at least for some years anyway), and if they did not report in 2009 or 2010 they are considered as exits here but are in fact survivors. To reduce the uncertainty in the classification of a firm as an exit related to the threshold of 20 employees, we excluded all firms below a threshold of 30 employees in 2 0 0 8 / Note that neither a change in the legal form of the firm nor a change in the ownership (due to a merger or an acquisition) nor a relocation of the firm inside Germany leads to an erroneous classification of a firm as an exit, because the identification number of the firm used in official statistics will not change. Unfortunately, it is not possible to investigate further the data for firms identified

5

6

Within the publishing sector, there were 2 , 3 7 1 observations dropped from the sample, including 3 3 exits. In the recycling sector, there were 1 9 1 observations and no exits. For the descriptive statistics of the final sample, see Section 3. All estimates are also reported without this additional threshold in the Appendix.

Risk or Resilience? • 763

as exits according to the definition used here due to the strict confidentiality of the firm level data. A certain degree of fuzziness, therefore, remains, and this should be kept in mind when putting the results from the empirical investigation into perspective.

3

Descriptive results

The final sample contains information about 36,183 enterprises, of which 288 left the market in the years 2 0 0 9 or 2 0 1 0 . The overall exit rate in our sample is hence below one percent. This seems very low compared to other results for the pre-crisis period from the same database by Wagner and Weche Geliibcke (2012), who report an exit rate of 2 . 7 7 % for the 2 0 0 7 exit cohort. This huge difference is mainly due to the exclusion of firms with less than 30 employees and points to higher exit rates among small firms, which is in line with theoretical considerations that assume a "liability of smallness" due to, for instance, disadvantages of scale, more restrictive access to capital markets, and a lower level of management skills (Audretsch 1995: 149 and Strotmann 2007). Even after excluding firms with below 30 employees from our sample, still 7 6 % of exiting firms had between 30 and 50 employees (see Table 1). Surprisingly, the descriptive statistics in Table 1 also reveal that only 2 5 % of exits were firms without any trade activities and 6 7 % of exiting firms reported either solely export activities or export and import activities in 2 0 0 8 . Only 8 % of exits happened within the group of only importing firms. This somehow contradicts the general expectation of only domestically oriented firms' being more likely to exit than internationally oriented firms due to risk diversification and generally higher productivity levels. In this respect, the picture appears to be different from the pre-crisis evidence, which shows an exit rate among non-trading firms that is almost twice the exit rate among firms that are involved in exporting activities (Wagner/Weche Geliibcke 2012). However, if we consider the risk diversification reasoning in the light of a global crisis it may not be too surprising to see exporting firms more vulnerable. The figures of exits by ownership categories in Table 1 do not surprise. The major share of exiting firms ( 5 7 % ) was labeled independent entity and was therefore not part of any domestic or foreign enterprise group. Only 7 % of exits were foreign subsidiaries, supporting the assumption that affiliates of multinationals have higher survival chances due to a network effect and access to internal resources. Furthermore, younger and less productive firms show higher exit rates, in line with the assumption of a "liability of newness" due to a lack of experience in the particular market (Audretsch 1995: 149) and the predictions by theoretical models of industrial dynamics, such as Jovanovic (1982), for the role of firms' productivity levels for entry, exit, and growth. The number of products seem to have little influence on the exit rate as both categories - one-product firms and two or more product firms - show similar exit rates. Multi-product firms are generally regarded as the more resilient firms because they are diversified across product markets and thus are able to reduce their sales default risk (e.g., Jovanovic/Gilbert 1993 and Lipczynski/Wilson 2001: 324f.). 8 4 % of exiting firms were located in western Germany, which reflects almost exactly the share of observations in the total sample located in western Germany, which is 8 2 % .

7 6 4 • Joachim W a g n e r a n d J o h n P. W e c h e C e l ü b c k e

Table 1 Descriptive statistics f o r 2 0 0 8 Survivors N u m b e r of firms

Share ( % )

Exits N u m b e r of firms

Share ( % )

All firms

35,895

99.20

288

0.80

Number of employees >

1/1

rsi o «- o r* o vo O T- o ^ O o I d d d d d d o ^

VO VO o o O 00 ^ r— >3- — Ól—V ói 1 • O (N O ro O VO VO VO O T^ O VO (N O O Ó (N Ó Ó O Ó Ó I I I i;

^ ? oí —* oo\o«oino E 01 VI 01 E ÍN ¿ o •o in

VO a\ Oí oí

E >

_o DE OI V

o in rM

•D 01

•o o

•o o o -O

>

o 3 "O o

3 -o o

-e 0 CL X 01

•e o CL E

O

O

3 -o o

>N

O c 1)

m.

omooomomo'-s oso^oroifioot d d d d d d d d o E | |— l o (N m O rv Oo m O O1S ® ^ O

V— ^ in ^ O oo o in o t- o rv | io o o o

i

O io —. ¿ _ p t o yo (Ti ir> o 00 IX i i i i i i j) O O ¡N VDO *

vo o ^ ^ Q * ^m ^TJ^ ^ r- a\ o m oin T-in vo (N q o o q u i q ^l- o00 v-G\ i i i i i i J> O ¡N o o og. o rN 00 1o

VO o •tf ¿ m vd ^ oroinooinovo O N m o o N o o dv^rsidodd-^ I ~ — | |—

lililí

*

* * (N ÍN o (N * * * O O O . NOVOO*i)OCOOmO»

c « B 2 ^ai o O -C

Ó 'a o S

vo a\ "D nJ

-C '

in

CO 10 rs IV Ol in ii) vo IV T— 00 00 CM in T— T— To\ IV IV in rn

IV IV 00 in 00 IV m 00 in IV cri o in O 00 T- Oi O in rn T— m T— fN

4-» i/j

5

§ «

§
o o o CM o o o o

—, O r^. in To in uj t m vc o oo O To o I

•s- oo ^ in