Employment performance in times of crisis: A multilevel analysis of economic resilience in the German biotechnology industry

Abstract

The purpose of this study is to elucidate the determinants of economic resilience at various levels of analysis. While the economic benefits of regional clustering are well-documented, the impact of external shocks on regional clusters has only recently gained attention. Combining multilevel regression and egocentric network analyses, the study explores the antecedents of economic resilience, defined as sustained employment growth, prior to and during the global financial crisis within the German biotechnology industry. The findings indicate that while specialization at the network and context-level is conducive to firm growth prior to the crisis, these configurations seem to be particularly susceptible to external shocks. Conversely, diversity (diversified regional agglomerations and diverse networks) seems to be associated with economic resilience during the crisis.

Full text

Employment performance in
times of crisis
A multilevel analysis of economic resilience
in the German biotechnology industry
Julian Kahl and Christian Hundt
Department for Urban and Regional Economics,
Ruhr University of Bochum, Bochum, Germany
Abstract
Purpose – The purpose of this study is to elucidate the determinants of economic resilience at
various levels of analysis. While the economic benets of regional clustering are well-documented,
the impact of external shocks on regional clusters has only recently gained attention. This study
explores the antecedents of economic resilience, dened as sustained employment growth, prior to
and during the global nancial crisis within the German biotechnology industry.
Design/methodology/approach – This study combines multilevel linear regression analysis
with egocentric network analysis. This allows us to distinguish micro- and context-level effects in
the analysis of economic resilience.
Findings – The ndings of this study indicate that while specialization at the network and
context-level is conducive to rm growth prior to the crisis, these congurations seem to be
particularly susceptible to external shocks. Conversely, diversity (diversied regional
agglomerations and diverse networks) seems to be associated with economic resilience during the
crisis. Moreover, we nd that economic resilience is connected to adaptive capability at the
micro-level, that is, the ability to expand and diversify a rms’ portfolio of network ties in the face
of an external shock. Finally, we show that these adaptive processes are facilitated by geographical
proximity among collaborating organizations.
Originality/value – This study contributes to the existing literature by showing that the
antecedents of economic resilience are located at multiple levels of analysis. An important
implication of this study is that the examination of the resilience of regional clusters may thus be
signicantly enhanced by disentangling effects at the rm, network and regional (i.e. context) level.
Keywords Resilience, Networks, Employment performance, External shocks, Multilevel analysis,
Regional clusters
Paper type Research paper
1. Introduction
While the recent global nancial crisis caused a severe downturn of economic
activity across a wide range of economic systems, the impact of this external shock
differed widely across sectors and regions. Uncertainties over market conditions
and reduced liquidity in the nancial system dramatically altered the competitive
environment within high-technology industries, where access to venture capital is
crucial (Gompers and Lerner, 2004;OECD, 2012). Moreover, it is well documented
that high-technology industries – such as the biotechnology industry – tend to
cluster in particular locations (Porter, 1998;Swann et al., 1998). Geographically
bounded knowledge spillovers, specialized labor pools as well as intensive local
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/1059-5422.htm
Employment
performance
in times of
crisis
371
Received 17 December 2014
Revised 20 February 2015
Accepted 1 March 2015
Competitiveness Review
Vol. 25 No. 4, 2015
pp. 371-391
© Emerald Group Publishing Limited
1059-5422
DOI 10.1108/CR-12-2014-0038

competition have been identied as important drivers of regional clustering giving
rise to innovation, employment as well as productivity growth (Baptista and Swann,
1998;Delgado et al., 2010;Jaffe et al., 1993;Porter, 1998). However, while the benets
of the spatial concentration of economic activity are well-documented, the ways in
which regional clusters as well as the rms representing a cluster’s main constituent
components respond to external shocks has only recently gained attention under the
umbrella concept of “economic resilience” (Martin and Sunley, 2011;Martin, 2012;
Pendall et al., 2010;Wrobel, 2013). More generally, the question whether regional
clusters promote or inhibit economic resilience has not been addressed sufciently.
It has been argued that economic resilience is associated with a cluster’s capacity to
adapt in the face of external shocks (Pendall et al., 2010;Simmie and Martin, 2010).
Moreover, recent studies suggest that a cluster’s adaptive capacity is determined by
its constituent components, that is, its rms (Menzel and Fornahl, 2010). A large
body of literature has shown that rm performance, in turn, depends on the ability
to adapt to and exploit changes in the business environment. In this context, rms
face a fundamental trade-off between explorative and exploitative adaptation
(March, 1991). The ways in which rms balance exploration and exploitation may
thus have an important inuence on economic resilience at the rm and regional
levels.
The recent global nancial crisis represents an external shock which allows us to
study the relationship between economic resilience and regional clusters. At the same
time, this provides us with the rare opportunity to extend the examination of regional
clusters to its constituent components, which is of particular importance given that a
cluster’s response to an external shock is dened by rms (Holm and Ostergaard, 2013;
Martin and Sunley, 2011;Menzel and Fornahl, 2010;Wrobel, 2013). An important
starting point of this paper is that the antecedents of economic resilience are located at
multiple levels of analysis. While extant research has tended to examine economic
resilience at the regional level, we posit that this may considerably underestimate
heterogeneity at lower-order levels of analysis. At the rm-level, external shocks may
erode competitive positions, which is why in the face of changing environmental
circumstances rms must seek to adapt accordingly. Moreover, rms are not isolated
from their regional environment. That is, economic resilience may be supported (or
inhibited) by different context-level features concerning, for instance, location within a
cluster as well as the economic structure of regional clusters.
Our empirical multilevel framework allows us to disentangle micro-level effects
from context-level features (pertaining to the regional environment in which rms
are embedded). Moreover, we examine the impact of changing environmental
conditions in two temporal brackets that delineate two phases of homogeneous
environmental conditions (i.e. pre-crisis and crisis).
The remainder of this paper is structured as follows. Section 2 summarizes the
theoretical background. Section 3 presents the data and methods. The results are
reported in Section 4. Section 5 concludes this paper.
2. Literature background and hypotheses
2.1 Economic resilience and regional clusters
In the wake of the recent global nancial crisis, scholarly interest has increasingly
focused on the resilience of regional economies and regional clusters. While it has
CR
25,4
372

been shown that rms in regional clusters, that is, “geographical concentrations of
interconnected companies […] in a particular eld” (Porter, 1998, p. 197), outperform
non-clustered rms in terms of growth (Beaudry and Schiffauerova, 2009), as well as
innovation and productivity growth (Baptista and Swann, 1998;Porter, 1998), the
performance implications of regional clusters during economic crises remain
unclear. Under the umbrella concept of “economic resilience”, recent studies have
addressed the role of changing environmental conditions and regional clustering
(Wrobel, 2013). An important starting point was the observation that while a large
range of clusters was adversely affected by external shocks, some clusters
successfully navigated discontinuities in the external environment, thus proving to
be resilient. Economic resilience may be dened as the “ability […] to recover
successfully from shocks” (Hill et al., 2008, p. 4). More specically, economic
resilience relates to the capacity of a regional economy or regional cluster to adapt
its structure – a process that is referred to as “industrial mutation” by Simmie and
Martin (2010) –soastosustain growth in output (Martin, 2012). Other conceptions
of resilience have placed a slightly different focus on the persistence of social and
ecological systems (pertaining to the magnitude a system can tolerate and still
persist) (Adger, 2000)[1]. Empirical studies show that a cluster’s vulnerability to
external shocks and its ability to adapt depends on the capabilities as well as the
network interactions of individual rms (Holm and Ostergaard, 2013;Wrobel, 2013).
This study draws on related conceptual and empirical work that uses employment
as a central variable of cluster evolution (Menzel and Fornahl, 2010) and the
economic resilience of regional clusters (Wrobel, 2013). Employment performance
gives an indication of economic resilience at the rm and cluster levels. On the one
hand, external shocks may adversely affect a rms’ competitive advantage and
economic situation which is reected in the level of employment of individual rms.
On the other hand, employment losses may cause regional economies to enter into
recessions. We dene economic resilience as a rms’ ability to sustain or augment
employment performance during crisis as compared to a previous level of
employment prior to the crisis. More specically, in this paper, we examine
economic resilience (dened as employment performance) in the context of an
external shock in the form of the recent global nancial crisis which led to the
dramatic deterioration of venture capital availability in the years 2009-2010 (Ernst
& Young, 2011) compared to the pre-crisis situation in 2007-2008.
2.2 Micro-level effects
It is widely acknowledged that sustained competitive advantage requires the
deployment of resources and capabilities that are appropriate to a rms’ external
environment (Tushman and O’Reily, 2004). Although rm-internal considerations
may prompt rms to change their strategy, changes may also be imposed upon rms
by variations in the external environment. External shocks including the emergence
of new technological paradigms, changes in regulation, market demand – or, in our
case, capital market conditions – may give rise to discontinuities that trigger high
degrees of uncertainty (Dess and Beard, 1984;Pfeffer and Salancik, 1978), as well as
competence destruction (Tushman and Anderson, 1986), thus rendering former
core-capabilities obsolete. This, in turn, has an important bearing on the economic
situation of rms as well as the ways in which rms allocate resources.
373
Employment
performance
in times of
crisis

In high-technology industries, inter-organizational networks represent one of the
most widely dispersed forms of coordination providing access to various
complementary resources such as nancial and human capital, as well as
technological capabilities and marketing skills (Ahuja, 2000;Burt, 1992). When
engaging in inter-organizational networks, rms are confronted with a fundamental
trade-off in the decision whether resources are allocated to the renement of existing
capabilities (exploitation) or to the discovery of new knowledge and capabilities
(exploration) (Gilsing, 2005;Lavie and Rosenkopf, 2006;March, 1991). While
exploitation relating to the efcient use and renement of existing assets and
capabilities is needed for survival in the short term, exploration, that is, the
development of new capabilities is needed for long-term survival (Tushman and
O’Reily, 2004). Balancing these activities poses a considerable trade-off for rms. In
essence, this trade-off arises from the role of diversity and specialization in
inter-organizational learning. While diversity, referring to the heterogeneity of
partners or knowledge and cognitive distance within the network (Baum et al., 2000;
Koka and Prescott, 2002;Lavie and Rosenkopf, 2006;Rodan and Galunic, 2004), is a
primary factor for the generation of Schumpeterian novel combinations (Nelson and
Winter, 1982), it is also associated with important drawbacks. Although diversity
increases the potential for innovative outcomes (Koka and Prescott, 2002),
organizations are limited in their capacity to assimilate and make use of novel
knowledge as rms need to learn how to bridge cognitive distances (Cohen and
Levinthal, 1990;Lane and Lubatkin, 1998). By contrast, the renement of existing
capabilities and resources (i.e. exploitation) is associated with short-term efciency,
as absorptive capacity and cognitive proximity facilitate inter-organizational
learning. Exploitation has been observed in industry environments characterized by
low uncertainty and competence-enhancing technological change in which
efciency considerations are crucial. These regimes have been shown to promote
specialization, that is, specic knowledge in a narrow range of issues (Gilsing, 2005).
However, exploitation is also associated with risks, as specialization on a narrow
range of issues may leave rms particularly exposed to environmental change
(Levitt and March, 1988), whereas, in environments characterized by
competency-destroying discontinuities, survival and growth may be associated
with the need to gain access to new and diverse resources that are crucial for
adaptation (explorative adaptation) (Tushman and O’Reily, 2004).
It has been shown that cognitive proximity between collaborating rms, which,
for the purpose of this study, is dened as low partner diversity, is conducive to the
efcient transfer of knowledge in inter-organizational networks (Boschma and
Frenken, 2010;Sampson, 2007;Simonin, 1999). That is, common skills and related
technological platforms facilitate knowledge transfer and cumulative specialization
by reducing the costs and time needed for the assimilation and economic use of
external knowledge (Cohen and Levinthal, 1990;Kahl, 2014;Lane and Lubatkin,
1998). We thus assume that rms engaged in networks with low partner diversity
will focus on exploitation (Rothaermel and Deeds, 2004), which we assume to be
positively associated with rm growth in the competency-enhancing regime prior to
the crisis. By contrast, in inter-organizational networks involving a high degree of
diverse agents with heterogeneous knowledge bases, rms need to develop an
in-depth understanding of a broad variety of technological elds which may be
CR
25,4
374

associated with inefciency and higher costs (Cohen and Levinthal, 1989;Gilsing,
2005;Lane and Lubatkin, 1998;Sampson, 2007;Simonin, 1999):
H1. Network diversity will be negatively associated with employment
performance prior to the crisis.
However, while cognitive proximity among network ties enables efcient knowledge
transfer, it may not be sufcient to stimulate innovative resource combinations (Ahuja,
2000;Noteboom, 2000). Empirical studies suggest that exploration is particularly
important in changing environments (Gilsing, 2005) because gaining access to new and
diverse resources provides the exibility needed to respond to external shocks
(Boschma and Frenken, 2010;Fleming, 2001). We thus assume that network diversity is
positively associated with rm growth during the crisis:
H2. Network diversity will be positively associated with employment
performance during the crisis.
Inter-organizational networks are characterized by a specic geography. Recent studies
suggest that geographical proximity among collaborating rms is neither a necessary
nor a sufcient condition for innovation (Boschma and Frenken, 2010;Boschma, 2005).
Rather than geographical proximity, it has been argued that various forms of proximity
among network partners matter for performance and innovation (Boschma, 2005;
Capaldo and Petruzzelli, 2014). Therefore, we do not expect the geography of network
ties per se to inuence rm growth prior to the crisis:
H3. The geography of network ties will have no inuence on employment
performance prior to the crisis.
By contrast, in environments characterized by competence-destroying change and
uncertainty, rm growth may be associated with regional interactions. In contrast to
extra-regional linkages, the transactions costs for the identication of collaboration
partners, as well as the costs arising from collaboration and monitoring activities,
are substantially lower within geographical proximity (Cooke, 2001;Marzucchi
et al., 2013). Moreover, face-to-face interactions may facilitate the development of
shared trust among collaboration partners. Uncertainty reduction and efciency
gains derived from regional networking may thus be particularly important to
resolve the conundrum arising from the need for exploration in times of crisis – as
both exploration and economic crises are characterized by high degrees of
uncertainty:
H4. Regional networks will be positively associated with employment
performance during the crisis.
2.3 Context-level effects
Firms are embedded in a specic regional economic context beyond their immediate
portfolio of network ties (Cooke, 2001). Importantly, the clustering of economic
activity has been shown to exert context-effects which are crucial for rm growth,
productivity and innovation (Beaudry and Schiffauerova, 2009;Porter, 1990,1998).
However, less scholarly attention has been on the relationship between changing
external conditions and regional clustering.
375
Employment
performance
in times of
crisis

Two broad categories of cluster externalities associated with the process of
knowledge creation and diffusion may be distinguished. One the one hand, this
relates to localization economies, which are stimulated by regional industry
specialization or the strength of the industry within a particular region (Baptista
and Swann, 1999). The benets derived from regional clustering of industries is
attributed to knowledge spillovers (Jaffe et al., 1993;Glaeser et al., 1992), reduced
transport costs for inputs and outputs, as well as specialized labor markets. More
specically, the concentration of an industry within a region is associated with
Marshall–Arrow–Romer externalities derived from intra-industry knowledge
spillovers between rms. Specialization is thus assumed to promote the
transmission and exchange of tacit and codied knowledge through imitation
(Pouder and St. John, 1996) and inter-rm circulation of skilled workers as well as
local competition among rms (Porter, 1990). On the other hand, Jacobs (1969)
argues that the most important source of knowledge spillovers and innovation is
diversity. Urbanization economies arise from large functional agglomerations as
well as industry variety within a region. In this line of argumentation, diverse
regional economic structure is associated with higher opportunities for search and
experimentation as well as for the recombination of resources and capabilities
across industries (Beaudry and Schiffauerova, 2009).
While empirical studies report on a positive relationship between rm growth
and localization (Baptista and Swann, 1998), as well as urbanization economies
(Beaudry and Swann, 2009), Neffke et al. (2011) nd that these effects vary across the
industry life cycle. Drawing on these studies, we expect the benets derived from
regional economic structure to vary across different environmental conditions. In
stable and competence-enhancing environments, we expect specialized regional
clusters of biotechnology rms to be associated with rm growth. In these
environments, we posit that cumulative regional specialization will foster
absorptive capacity and intra-industry knowledge spillovers which, in turn,
facilitate rm growth:
H5. Localization economies will be positively associated with employment
performance prior to the crisis.
By contrast, competence-destroying discontinuities may place pronounced
constraints on rms located in narrowly specialized clusters due to established
routines and high restructuring costs. These types of clusters may be particularly
vulnerable to external shocks (Grabher, 1993). Narrow specialization may thus
impede a cluster’s adaptability in the face of changing environmental conditions
thereby undermining the resilience of the rms in the cluster:
H6. Localization economies will be negatively associated with employment
performance during the crisis.
Recent empirical studies suggest that the diversity of the economic structure of a
cluster promotes economic resilience (Holm and Ostergaard, 2013;Wrobel, 2013).
We thus assume that urbanization economies are associated with employment
growth in competence-destroying environments. In essence, diversied regional
clusters may succeed at spreading risk associated with external shocks across
various industries. Moreover, the varied stock of resources available in diversied
CR
25,4
376

regional agglomerations may enable rms to exibly recombine a broad set of
resources which, in turn, may promote adaptation in the face of an external shock:
H7. Urbanization economies will be positively associated with employment
performance during the crisis.
3. Methods and data
3.1 Structure of the empirical model
In this section, we present the structure of our empirical model. This paper combines
egocentric network (Wasserman and Faust, 1994) and multilevel analysis
(Raudenbush and Bryk, 2002) in two separate sets of models (pre-crisis and crisis).
To capture the effect of the external shock represented by the global nancial crisis,
we distinguish two temporal brackets of homogeneous environmental conditions,
that is, pre-crisis including the years 2007-2008 characterized by relative stability
and capital market municence, on the one hand, and crisis marked by
competency-destroying change, uncertainty and the deterioration of capital market
conditions in the years 2009-2010 on the other.
Choosing a multilevel approach allows us to reect an important part of
(economic) reality inasmuch as rms are not isolated from their environment but
nested within a specic regional economic and institutional context (Luke, 2004;
Snijders and Bosker, 2004). In addition, high-technology rms are commonly
embedded in a broad set of inter-organizational relationships. Using egocentric
network analysis allows one to capture the effects of a focal rms’ portfolio of
inter-organizational ties on employment performance.
We gauge the impact of ve rm- (X
1
-X
5
), four network-related (X
6
-X
9
) and two
context-level predictors (R
1
-R
2
) on employment performance (Y) in the German
biotechnology industry. We use the statistics program “HLM 7” from Raudenbush
et al. (2011) for the model estimates. A formal notation of this two-level linear
regression model is provided below:
Yij ⫽[
␦
00 ⫹
␦
10Xij ⫹
␦
01Rj⫹
关
uj⫹rij]
with,
Y
ij
⫽Response variable at the rm level.
Employment growth (in per cent):
␦
00
⫽Regression intercept;
X
ij
a
⫽Explanatory variables at the rm level;
X
ij
1
⫽Product;X
ij
2
:Product orientation;X
ij
3
:Service;X
ij
4
: Servitization;X
ij
5
:
Firm size;X
ij
6
:Network diversity;X
ij
7
:Network regional;X
ij
8
:Network
size; and X
ij
9
:Network churn;
R
j
b
⫽explanatory variables at the context level;
R
j
1
⫽Concentration;
R
j
2
⫽Population density;
R
ij;
u
j
⫽Error terms of the rm (r) and the regional level (u).
The structure of the model is strictly hierarchical in the sense that i biotech rms are
nested in j regions. As suggested by Hox (2010), we chose a bottom-up strategy
377
Employment
performance
in times of
crisis

consisting of six consecutive analytical steps to construct our models. We rst
estimated the impact of our ve rm- (Column 1) and four network-related
predictors (Column 2), separately. We then excluded the highly insignicant
variables (p-value ⱖ0.20) and merged the remaining micro-level predictors in a joint
model (Column 3). We then controlled for the two context-level predictors (Column 4)
and nally, to test the robustness of our results, we added four control variables
covering the regional economic capacity (Column 5) and the regional labor market
(Column 6), respectively. As expected, we observe an improving model t
(documented by declining deviance values) as more predictor variables are included.
3.2 Sample
The data for the response variable as well as for the rm and network predictors are
derived from the German Biotechnology Year and Address Book (BIOCOM) for the
years 2007 to 2010. For the purpose of this paper, BIOCOM data provide at least
three advantages. First, it represents the most comprehensive independent
repository of the German biotechnology industry with a response rate of
approximately 80 per cent. This enabled us to identify 266 biotechnology rms in
both brackets representing approximately 50 per cent of the population of German
biotechnology rms. Second, the database reports micro-data including location,
year of establishment, business models and employment as well as network
partners of individual rms. The database is thus amenable to multilevel analysis
allowing us to capture the hierarchical relationship between the rm and its regional
context. Third, the database reports longitudinal data which allows us to construct
a panel of two-year temporal brackets (pre-crisis and crisis) comprising 266
biotechnology rms. Moreover, even though employment losses were signicant in
the German biotechnology industry, a negligible number of exits was observed
(Ernst & Young, 2012), thus reducing sample bias. Finally, for the predictors at the
context-level, we draw on EUROSTAT data.
3.3 Variables
For our response variable at the micro-level Employment Performance (Y
ij
), used as
a proxy for economic resilience (Wrobel, 2013;Baptista and Swann, 1999), we rst
compute the within-bracket mean employment of individual biotechnology rms
and subsequently gauge the cross-bracket employment growth (in per cent). We
specify ve network-related explanatory variables. Network Size (X
ij
2
), Network
Churn (X
ij
3
), Network Diversity (X
ij
4
) and Network Regional (X
ij
5
). Network Size (X
ij
2
)
captures a focal rm’s number of direct ties. Linked to that, we dene Network
Churn (X
ij
3
) as the growth rate of Network Size across the brackets identied above.
The fourth predictor, i.e. Network Diversity X
ij
4
, is based on the Blau Index (Blau,
1977) that captures the diversity of a rms’ portfolio of network ties. To compute
network diversity, in line with Powell et al. (2005), we rst identify ve categories of
partner organizations including biotechnology rms, public research organizations,
pharmaceutical corporations and government institutes, as well as biomedical
companies. We then compute the Blau Index, a measure commonly used to gauge
network diversity, for individual biotechnology rms:
Blau’s index d⫽1⫺兺i⫽1
npi2
CR
25,4
378

Nrepresents the total number of categories of network partners and pcomputes the
percentage of objects in a certain category. This measures ranges from 0 to 1, where
0 represents completely homogeneous networks and 1 completely diversied
networks. If a rm has a wide range of network partners, the measure will be high (1
or close to 1), whereas if most collaboration partners are concentrated in a specic
category, it will be low. Finally, Network Regional (X
ij
5
) captures the share of
regional network ties within a focal rms’ portfolio of network ties. To construct this
measure, we coded the geographical localization of all network ties in different
geographical categories (regional, national, international and global).
Collaborations within the same federal state were classied as “regional”. All
collaborations with German network partners were dened as “national”, whereas
collaborations with European partners and non-European partners were assigned
“international” and “global”, respectively.
At the context-level, we distinguish two predictors, that is, Concentration (R
j
1
)
and Population Density (R
j
2
). Concentration (R
j
1
) computes the degree of
disproportionality of the spatial distribution of the German biotechnology industry
in relation to a national reference. We assess the share of biotechnology rms within
NUTS-2 regions in relation to the number of biotechnology rms at the national
level. This variable is used as a proxy for localization economies. Population Density
(R
j
2
) computes population density in NUTS-2 regions (number of inhabitants per
square kilometer) and serves as a proxy for urbanization economies.
Moreover, several control variables were used. We controlled for Firm Size
relating to the absolute number of employees (in t–1) and for Network Size, several
business models as well as business model adaptation. At the regional-level we
controlled for Log Regional GDP per Capita and the share of Regional
Unemployment as well as for the change in Regional GDP and Regional
Unemployment at the NUTS-2 level across the brackets identied above.
4. Results
4.1 Descriptive results
Tables II-Vpresent the descriptive results of our empirical analysis including
means, standard deviations and correlations. None of the correlations between the
predictor variables indicates multicollinearity. All micro-level predictors were
estimated for the years 2007-2010 for 266 rms. During the crisis, we observe a
sharp decline in employment growth, as well as a lower standard deviation.
Although total employment growth remains positive, it is the result of fewer rms.
During the crisis, we observe increasing Firm Size (in t⫺1). This is attributable to
growth in the Pre-Crisis model. While average Network Size is slightly higher in the
Crisis-model, Network Churn is reduced during crisis. Moreover, the means for
Network Diversity and Network Regional are relatively stable in both models. At the
context-level, all predictors were estimated for 34 NUTS-2 regions. We observe the
impact of the global nancial crisis as evidenced by the decline of Regional GDP Per
Capita in the Crisis-model. Interestingly enough, prior to the impact on regional
GDP, as well as employment in the biotechnology industry, we observe an increase
in Unemployment in the Pre-Crisis model. High standard deviation indicates that
this varied considerably across NUTS-2 region (Tables I-IV).
379
Employment
performance
in times of
crisis

Table I.
Variable means,
standard deviations
and correlations
(Pre-Crisis;
micro-level)
a
Variable Mean SD Y X
1
X
2
X
3
X
4
X
5
X
6
X
7
X
8
X
9
Y: Empl_Performance 27.34 65.03 1 ⫺0.01 ⫺0.05 ⫺0.06 0.17** 0.09 ⫺0.06 ⫺0.08 0.09 0.12*
X
1
: Product 0.23 0.42 ⫺0.01 1 0.04 ⫺0.26** 0.28** 0.10 ⫺0.13* ⫺0.11 ⫺0.08 ⫺0.06
X
2
: Product orientation 0.20 0.40 ⫺0.05 0.04 1 ⫺0.23** ⫺0.27** 0.02 0.02 0.14* 0.07 ⫺0.04
X
3
: Service 0.18 0.39 ⫺0.06 ⫺0.26** ⫺0.23** 1 ⫺0.12 ⫺0.11 ⫺0.04 0.00 ⫺0.04 0.08
X
4
: Servitization 0.23 0.42 0.17** ⫺0.28** ⫺0.27** ⫺0.12 1 0.01 ⫺0.01 ⫺0.03 0.06 0.05
X
5
: Firm size (in t⫺1) 44.23 133.74 0.09 0.10 0.02 ⫺0.11 0.01 1 ⫺0.08 ⫺0.16** 0.04 ⫺0.10
X
6
: Network diversity 0.37 0.25 ⫺0.06 ⫺0.13* 0.02 ⫺0.04 ⫺0.01 ⫺0.08 1 ⫺0.15* 0.43** 0.20**
X
7
: Network regional 35.41 34.26 ⫺0.08 ⫺0.11 0.14* 0.00 ⫺0.03 ⫺0.16** ⫺0.15* 1 ⫺0.22** 0.02
X
8
: Network size 5.64 4.97 0.09 ⫺0.08 0.07 ⫺0.04 0.06 0.04 0.43** ⫺0.22** 1 0.11
X
9
: Network churn 25.28 69.08 0.12* ⫺0.06 ⫺0.04 0.08 0.05 ⫺0.10 0.20** 0.02 0.11 1
Notes:
a
n⫽266; *pⱕ0.05; **pⱕ0.01
CR
25,4
380

4.2 Regression results
The results of the linear multilevel regressions are presented in Tables V (pre-crisis)
and VI (crisis). The results obtained here provide support for H1 and H2 relating to
the effects of network structure on Employment Performance. Prior to the crisis, we
observe a negative effect of Network Diversity and, conversely, a positive impact of
specialization. By contrast, during the crisis, we report on a positive effect of
Network Diversity. These observations may be explained by the
exploitation– exploration framework. While specialized network structures with
high degrees of cognitive proximity have been shown to be conducive to the efcient
use and exploitation of external knowledge (Boschma and Frenken, 2010), these
network structures may not be sufcient to stimulate new resource combinations
that facilitate adaptation during crisis. Economic resilience thus seems to be
connected to diverse inter-organizational networks.
Moreover, we provide support for H3 and H4 relating to the role of regional
network ties during the crisis. While regional network ties had no effect on
Employment Performance prior to the crisis, they gain importance in the face of an
external shock. More specically, H3 provides support to the view that geographical
proximity among network partners is neither a necessary nor a sufcient condition
for positive network outcomes (Boschma, 2005). However, in line with H4, this only
seems to hold for relatively stable environmental conditions. By contrast, during
crisis regional network ties gain signicance in facilitating Employment
Performance. This seems to indicate the importance of face-to-face interactions and
shared trust among organizations in the biotechnology industry which may be
particularly relevant for the effective reconguration of inter-organizational
networks as well as for dealing with high levels of uncertainty induced by changing
environmental conditions.
Consistent with H5, we report positive localization effects at the context-level
before the crisis. Moreover, in line with H7, urbanization economies are positively
associated with Employment Performance during the crisis. Finally, our results
provide support for H6, i.e. specialized regional clusters are negatively related to
Employment Performance during the crisis. The results reported here thus seem to
indicate that the benets derived from regional clusters are moderated by the state
of the external environment. Although regional specialization seems to be conducive
to employment growth in relatively stable and competence-enhancing environments
(H5), regional specialization is associated with risks. That is, as clustered rms
follow existing routines and technological capabilities in a process of cumulative
Table II.
Variable means,
standard deviations
and correlations
(Pre-Crisis;
context-level)
a
Variable Mean SD R
1
R
2
C
1
C
2
C
3
C
4
R
1
: Concentration 2.87 2.82 1 0.39* 0.33 ⫺0.16 0.03 0.12
R
2
: Population density 0.47 0.75 0.39* 1 0.39* ⫺0.14 0.27 0.26
C
1
: Log GDP per capita 10.25 0.23 0.33 0.39* 1 ⫺0.45** – –
C
2
:⌬GDP per capita 8.27 1.66 ⫺0.16 ⫺0.14 ⫺0.45** 1 – –
C
3
: Unemployment 8.41 3.65 0.03 0.27 – – 1 0.51**
C4: ⌬Unemployment ⫺2.52 5.41 0.12 0.26 – – 0.51** 1
Notes:
a
n⫽34; *pⱕ0.05; **pⱕ0.01
381
Employment
performance
in times of
crisis

Table III.
Variable means,
standard deviations
and correlations
(Crisis; micro-level)
a
Variable Mean SD Y X
1
X
2
X
3
X
4
X
5
X
6
X
7
X
8
X
9
Y: Empl_Performance 10.96 45.24 1 ⫺0.03 0.01 0.07 ⫺0.04 ⫺0.05 0.11 0.09 0.04 0.11
X
1
: Product 0.28 0.45 ⫺0.03 1 ⫺0.01 ⫺0.36** ⫺0.27** 0.03 ⫺0.13* ⫺0.14* ⫺0.13* ⫺0.08
X
2
: Product orientation 0.22 0.42 0.01 ⫺0.01 1 ⫺0.22** ⫺0.23** ⫺0.02 0.08 ⫺0.07 0.09 0.13*
X
3
: Service 0.24 0.43 0.07 ⫺0.36** ⫺0.22** 1 0.06 ⫺0.09 ⫺0.01 0.09 ⫺0.13* ⫺0.03
X
4
: Servitization 0.16 0.37 ⫺0.04 ⫺0.27** ⫺0.23** 0.06 1 0.06 0.02 0.01 0.02 ⫺0.03
X
5
: Firm size (in t⫺1) 63.66 251.07 ⫺0.05 0.03 ⫺0.02 ⫺0.09 0.06 1 ⫺0.08 ⫺0.14* 0.09 0.02
X
6
: Network diversity 0.38 0.26 0.11 ⫺0.13* 0.08 ⫺0.01 0.02 ⫺0.08 1 ⫺0.09 0.42** 0.11
X
7
: Network regional 34.80 34.15 0.09 ⫺0.14* ⫺0.07 0.09 0.01 ⫺0.14* ⫺0.09 1 ⫺0.18** ⫺0.10
X
8
: Network size 6.13 5.48 0.04 ⫺0.13* 0.09 ⫺0.13* 0.02 0.09 0.42** ⫺0.18** 1 0.24**
X
9
: Network churn 18.79 69.48 0.11 ⫺0.08 0.13* ⫺0.03 ⫺0.03 0.02 0.11 ⫺0.10 0.24** 1
Notes:
a
n⫽266; *pⱕ0.05; **pⱕ0.01
CR
25,4
382

specialization, the barriers to new technologies, procedures and strategies become
more pronounced. This in turn seems to render specialized regional clusters less
equipped to react to changes in the external environment (H6). By contrast,
diversied regional agglomerations seem to be particularly well equipped to
withstand external shocks (H7). These types of clusters provide access to a diverse
set of assets which may be critical for the exible recombination of resources and
adaptation in times of crisis.
Considering the lack of signicance of the control variables pertaining to the type
of business models and (nearly all variables related to) business model adaptation,
rather than changing the architecture of the value creation, economic resilience
seems to be associated with the restructuring of inter-organizational networks. Two
control variables warrant further discussion. First, the results for the control
variable Servitization which refers to business model changes from product-based
business models to service-based business models indicate that while these types of
business model changes were benecial to Employment Performance prior to the
crisis, during the crisis they had no effect. This is somewhat surprising given that
service-based business models are associated with more short-term revenue
generating potential as well as considerably lower capital requirements and risk –
all of which are seemingly important during crisis. Second, the results for Network
Churn relating to the growth rate of a focal rms’ portfolio of network ties indicate
a positive effect prior to and during the crisis. The expansion and renewal of
networks and thus access to a larger set of resources seems to be benecial to
Employment Performance, irrespective of the nature of the business environment.
However, the results reported above relating to the varying effects of network
structure provide a more nuanced picture. It may thus be inferred that while prior
to the crisis the expansion of networks in terms of specialization supports
Employment Performance, during the crisis diversication fosters economic
resilience (Tables V and VI).
5. Conclusion
The aim of this paper was to explore the economic resilience of regional clusters and
rms in times of crisis. This study contributes to the existing literature by showing
that the antecedents of economic resilience are located at multiple levels of analysis.
An important implication of this study is that the examination of the resilience of
regional clusters may thus be signicantly enhanced by disentangling effects at the
rm, network and regional (i.e. context) levels.
Table IV.
Variable means,
standard deviations
and correlations
(Crisis;
context-level)
a
Variable Mean SD R
1
R
2
C
1
C
2
C
3
C
4
R
1
: Concentration 2.87 2.82 1 0.39* 0.35* 0.05 0.04 0.07
R
2
: Population density 0.47 0.75 0.39* 1 0.42* 0.00 0.29 ⫺0.17
C
1
: Log GDP per capita 10.25 0.22 0.35* 0.42* 1 ⫺0.52** – –
C
2
:⌬GDP per capita ⫺0.02 2.46 0.05 0.00 ⫺0.52** 1 – –
C
3
: Unemployment 7.60 2.70 0.04 0.29 – – 1 ⫺0.67**
C
4
:⌬Unemployment ⫺0.65 9.59 0.07 ⫺0.17 – – ⫺0.67** 1
Notes:
a
n⫽34; *pⱕ0.05, **pⱕ0.01
383
Employment
performance
in times of
crisis

Table V.
Regression results
(Pre-Crisis) (with
robust standard
errors)
ab
Response variable: rm
growth (%)
Firm-related variables
(X1-X5)
Network-related variables
(X6-X9)
Firm- and network-related
variables
Regional context variables
(R1-R2)
Regional control variables
(C1-C2)
Regional control variables
(C3-C4)
Coefcient (standard errors) Coefcient (Standard errors) Coefcient (standard errors) Coefcient (standard errors) Coefcient (standard errors) Coefcient (standard errors)
Regression intercept (⫹) 27.31*** (3.35) (⫹) 27.31*** (3.25) (⫹) 27.29*** (3.43) (⫹) 23.58*** (3.81) (⫹) 21.08*** (2.99) (⫹) 23.73*** (3.81)
X1: Product (⫹) 3.18 (6.25)
X2: Product orientation (⫺) 2.36 (7.57)
X3: Service (⫺) 5.44 (6.08)
X4: Servitization (⫹) 25.75** (10.33) (⫹) 23.45** (10.19) (⫹) 24.44** (10.60) (⫹) 21.47** (10.90) (⫹) 24.77** (10.68)
X5: Firm size (in t⫺1) (⫹) 0.04 (0.04)
X6: Network diversity (⫺) 39.44** (16.71) (⫺) 37.40** (17.13) (⫺) 37.93** (16.61) (⫺) 37.09** (16.42) (⫺) 37.99** (16.72)
X7: Network regional (⫺) 0.16**** (0.12) (⫺) 0.15 (0.12)
X8: Network size (⫹) 1.63* (0.85) (⫹) 1.49* (0.85) (⫹) 1.81** (0.76) (⫹) 1.66** (0.71) (⫹) 1.76** (0.83)
X9: Network churn (⫹) 0.14**** (0.09) (⫹) 0.13**** (0.08) (⫹) 0.12**** (0.08) (⫹) 0.12**** (0.08) (⫹) 0.12**** (0.08)
R1: Concentration (⫹) 1.23** (0.46) (⫹) 1.48*** (0.52) (⫹) 1.29*** (0.44)
R2: Population density (⫹) 0.39 (1.20)
C1: GDP pc (ln) (⫹) 42.38** (15.79)
C2:⌬GDP pc (%) (⫹) 7.33*** (1.77)
C3: Unemployment (⫹) 0.28 (0.87)
C4:⌬Unemployment (%) (⫺) 0.59 (0.68)
Level-1 –Variance 4,054.46 4,021.80 3,924.98 3,920.28 3,826.84 3,913.05
Level-2 –Variance 1.81 1.58 2.22 1.00 0.75 0.98
Deviance (null) 2,974.89 2,974.89 2,974.89 2,974.89 2,974.89 2,974.89
Deviance (model) 2,964.78** 2,962.61** 2,956.17*** 2,955.77*** 2,949.34*** 2,955.28***
Notes: an⫽266; ball predictors are grand mean centered; spatial autocorrelation does not occur; the table displays unstandardized coefcients and standard errors; *** pⱕ0.01; **pⱕ0.05; * pⱕ
0.10; **** pⱕ0.20
Sources: BIOCOM AG (N⫽266); EUROSTAT (N⫽34), own calculations
CR
25,4
384

Table VI.
Regression results
(Crisis) (with robust
standard errors)
ab
Response variable: rm
growth (%)
Firm-related variables
(X1-X5)
Network-related variables
(X6-X9)
Firm- and network-related
variables
Regional context variables
(R1-R2)
Regional control variables
(C1-C2)
Regional control variables
(C3-C4)
Coefcient (Standard errors) Coefcient (Standard errors) Coefcient (Standard errors) Coefcient (Standard errors) Coefcient (Standard errors) Coefcient (Standard errors)
Regression intercept (⫹) 10.96*** (2.48) (⫹) 10.95*** (2.68) (⫹) 10.96*** (2.66) (⫹) 10.97*** (2.12) (⫹) 11.14*** (2.13) (⫹) 11.05*** (1.84)
X1: Product (⫺) 1.71 (6.67)
X2: Product orientation (⫹) 1.04 (9.34)
X3: Service (⫹) 7.23 (7.20)
X4: Servitization (⫺) 5.06 (7.55)
X5: Firm size (in t⫺1) (⫺) 0.01** (0.01) (⫺) 0.01* (0.003) (⫺) 0.01**** (0.01) (⫺) 0.01**** (0.01) (⫺) 0.01**** (0.01)
X6: Network diversity (⫹) 19.31**** (13.94) (⫹) 17.90* (10.73) (⫹) 19.25* (11.13) (⫹) 18.94* (11.19) (⫹) 19.01* (11.22)
X7: Network regional (⫹) 0.14** (0.07) (⫹) 0.14** (0.07) (⫹) 0.13** (0.07) (⫹) 0.13** (0.06) (⫹) 0.13** (0.07)
X8: Network size (⫺) 0.11 (0.57)
X9: Network churn (⫹) 0.08** (0.03) (⫹) 0.08** (0.03) (⫹) 0.08** (0.03) (⫹) 0.08** (0.03) (⫹) 0.08** (0.03)
R1: Concentration (⫺)0.76*** (0.21) (⫺) 1.14*** (0.40) (⫺) 0.99*** (0.22)
R2: Population density (⫹)5.89*** (0.43) (⫹) 5.23*** (1.19) (⫹) 7.69*** (1.13)
C1: GDP pc (ln) (⫹) 13.32 (11.27)
C2:⌬GDP pc (%) (⫹) 0.95 (1.36)
C3: Unemployment (⫺) 0.88 (0.92)
C4:⌬Unemployment (%) (⫹) 0.03 (0.36)
Level-1 –Variance 2,019.12 1,965.63 1,964.31 1,920.27 1,917.60 1,915.12
Level-2 –Variance 1.19 3.94 3.35 0.32 0.31 0.30
Deviance (null) 2,781.91 2,781.91 2,781.91 2,781.91 2,781.91 2,781.91
Deviance (model) 2,779.37 2,772.60* 2,772.34** 2,765.91** 2,765.54** 2,765.19**
Notes: an⫽266; ball predictors are grand mean centered. Spatial autocorrelation does not occur; the table displays unstandardized coefcients and standard errors; *** pⱕ0.01; **pⱕ0.05; * pⱕ
0.10; **** pⱕ0.20
Sources: BIOCOM AG (N⫽266); EUROSTAT (N⫽34), own calculation
385
Employment
performance
in times of
crisis

The following three main results are identied. First, while being located in a
narrowly specialized cluster was shown to facilitate Employment Performance prior
to the crisis, regional economic specialization seemed to undermine economic
resilience in the face of an external shock. This may be explained by the process of
cumulative specialization rendering these regional clusters particularly exposed to
competency-destroying change, thus turning core competencies into core rigidities.
Second, the results show that diversied regional agglomerations are associated
with economic resilience. These regional clusters seem to provide a greater potential
for the innovative recombination of resources and capabilities thus enabling local
adaptive processes. Third, economic resilience is associated with micro-level
adaptive capability, that is, a rms’ ability to renew and recongure its portfolio of
networks ties so as to increase the size and diversity in the face of changing
environmental conditions. Moreover, these adaptive processes seem to be fostered
by geographical proximity. Geographical proximity thus seems to resolve an
important conundrum relating to the need for exploration in times of crisis. That is,
both exploration and rapidly changing environments are characterized by
uncertainty, which is why these activities may be particularly difcult in the face of
an external shock. However, geographical proximity seems to reduce uncertainty
among collaborating organizations and facilitate explorative adaptation.
The results obtained in this study have several implications for policymakers.
The results pertaining to the economic resilience of regional clusters seem to reect
the exploration– exploitation trade-off identied at the micro-level relating to the
conicting nature of specialization and diversity. In reference to the notion of
organizational ambidexterity (Tushman and O’Reily, 2004), sustainable cluster
growth may thus entail balancing explorative and exploitative activities.
Policymakers may thus choose to foster diversity by supporting new rm formation
as well as launching cluster-oriented policies that promote local discovery and
exploration processes in related, yet new elds to increase the resilience of regional
clusters in anticipation of external shocks. Moreover, cluster management agencies
may promote economic resilience by increasing diversity at the network-level
through the provision of collaboration platforms and R&D-support schemes.
The results documented in this study also relate to a parallel discussion on
organizational resilience which offers important insights in the proactive
management of resilience. Two of our main results at the micro-level indicate that
economic resilience is supported by adaptive capability and by the diversication of
the portfolio of network ties. The organizational resilience literature (Lee et al., 2013)
shows that to engage in these restructuring processes, rms must develop planning
strategies relating to the management of vulnerabilities within a rms’ business
environment. More specically, economic resilience necessitates a proactive
posture. Firms need to display a strategic and behavioral readiness reacting to early
warning signals in an organizations’ external environment. Firms also need to build
an understanding of the resources and relationships they may need to build and
access during crisis (Lee et al., 2013;McManus et al., 2008). Moreover, adaptive
capability presupposes a range of factors and processes within the rm including
the minimization of social, cultural and behavioral silos, the mobilization of extra
capacity or resources as well strong leadership during crisis. Moreover, it has been
CR
25,4
386

shown that environments in which staff members are encouraged and rewarded for
designing mechanisms for solving existing and new problems support economic
resilience (Lee et al., 2013;McManus et al., 2008).
Future research may build on the ndings presented here along various
dimensions. Future research may focus on the role external shocks play in the
evolution of regional clusters (Zettinig and Vincze, 2012). In a longitudinal research
design, these studies may assess how external shocks relate to the development of
regional clusters examining the size and heterogeneity of the composition of
regional clusters (Menzel and Fornahl, 2010). These longitudinal case studies may
elucidate how the observed changes or persistence in the composition of the regional
clusters relate to various outcomes including employment growth or innovation.
Finally, economic resilience may be studied at the interface of the network and
cluster levels. Recent empirical ndings suggest that knowledge ows and network
positions differ signicantly within clusters (Alberti and Pizzurno, 2015). Future
research may therefore shed light on the ways in which varied network positions
and knowledge ows within regional clusters relate to economic resilience. For
instance, future research may examine whether central network nodes or brokers
are more readily able to withstand external shocks. Finally, a further promising
avenue for research lies in the investigation of the degree of the persistence of
knowledge ows in regional clusters in the face of an external shock.
Note
1. The authors wish to acknowledge the anonymous reviewer #2 for raising this point.
References
Adger, W.N. (2000), “Social and ecological resilience: are they related?”, Progress in Human
Geography, Vol. 24 No. 3, pp. 347-364.
Ahuja, G. (2000), “Collaboration networks, structural holes, and innovation: a longitudinal study”,
Administrative Science Quarterly, Vol. 45 No. 3, pp. 425-455.
Alberti, F.G. and Pizzurno, E. (2015), “Knowledge exchanges in innovation networks: evidences
from an Italian aerospace cluster”, Competitiveness Review, Vol. 25 No. 3.
Baptista, R. and Swann, P. (1998), “Do rms in clusters innovate more?”, International Journal of
Industrial Organization, Vol. 18, pp. 515-535.
Baptista, R. and Swann, P. (1999), “A comparison of clustering dynamics in the US and UK
computer industries”, Journal of Evolutionary Economics, Vol. 9, pp. 373-399.
Baum, J.A.C., Calabrese, T. and Silverman, B.S. (2000), “Don’t go it alone’: alliance network
composition and startups’ performance in Canadian biotechnology”, Strategic
Management Journal, Vol. 21, pp. 267-294.
Beaudry, C. and Schiffauerova, A. (2009), “Who’s right, Marshall or Jacobs? The localization
versus urbanization debate”, Research Policy, Vol. 38 No. 2, pp. 318-337.
Beaudry, C. and Swann, M.P. (2009), “Firm growth in industrial clusters of the United Kingdom”,
Small Business Economics, Vol. 32 No. 2, pp. 409-424.
Blau, P.M. (1977), Inequality and Heterogeneity, Free Press, New York, NY.
387
Employment
performance
in times of
crisis

Boschma, R.A. (2005), “Proximity and innovation. A critical assessment”, Regional Studies, Vol. 39
No. 1, pp. 61-74.
Boschma, R.A. and Frenken, K. (2010), “The spatial evolution of innovation networks: a proximity
perspective”, in Boschma, R.A. and Martin, R. (Eds), The Handbook of Evolutionary
Economic Geography, Edward Elgar, Cheltenham, pp. 120-135.
Burt, R.S. (1992), Structural Holes: The Social Structure of Competition, Harvard University Press,
Cambridge, MA.
Capaldo, A. and Petruzzelli, A.M. (2014), “Partner geographic and organizational proximity and
the innovative performance of knowledge-creating alliances”, European Management
Review, Vol. 11 No. 1, pp. 63-84.
Cohen, W.M. and Levinthal, D. (1989), “Innovation and learning: the two faces of R&D”, Economic
Journal, Vol. 99 No. 397, pp. 569-596.
Cohen, W.M. and Levinthal, D. (1990), “Absorptive capacity: a new perspective on learning and
innovation”, Administrative Science Quarterly, Vol. 35 No. 1, pp. 128-152.
Cooke, P. (2001), “Regional innovation systems, clusters, and the knowledge economy”, Industrial
and Corporate Change, Vol. 10 No. 4, pp. 945-974.
Delgado, M., Porter, M.E. and Stern, S. (2010), “Clusters, convergence and economic performance”,
Working Paper No. 10-34, CES, Cambridge, MA.
Dess, G.G. and Beard, D.W. (1984), “Dimensions of organizational task environments”,
Administrative Science Quarterly, Vol. 29, pp. 52-73.
Ernst & Young (2011), Deutscher Biotechnologie Report 2012, Ernst & Young, Mannheim.
Ernst & Young (2012), Deutscher Biotechnologie Report 2012, Ernst & Young, Mannheim.
Fleming, S. (2001), “Recombinant uncertainty in technological search”, Management Science,
Vol. 47, pp. 117-132.
Gilsing, V. (2005), The Dynamics of Innovation and Interrm Networks: Exploration, Exploitation
and Co-evolution, Edward Elgar, Cheltenham.
Glaeser, E.L., Kallal, H.D., Schinkmann, J.A. and Shleifer, A. (1992), “Growth in cities”, Journal of
Political Economy, Vol. 100, pp. 1126-1152.
Gompers, P. and Lerner, J. (2004), “The venture capital revolution”, Journal of Economic
Perspectives, Vol. 15 No. 29, pp. 145-168.
Grabher, G. (1993), “The weakness of strong ties: the lock-in of regional development in the Ruhr
area”, in Grabher, G. (Ed.), The Embedded Firm: On the Socioeconomics of Industrial
Networks, Routledge, London, pp. 255-277.
Hill, E.W., Wial, H. and Wolman, H. (2008), “Exploring regional economic resilience”, Working
Paper 2008-04, University of California, Berkeley.
Holm, J.R. and Ostergaard, C.R. (2013), “Regional employment growth, shocks and regional
industrial resilience: a quantitative analysis of the Danish ICT sector”, Regional Studies,
Vol. 49 No. 1.
Hox, J.J. (2010), Multilevel Analysis. Techniques and Applications, Routledge, New York, NY.
Jacobs, J. (1969), The Economy of Cities, Vintage Books, New York, NY.
Jaffe, A.B., Trajtenberg, M. and Henderson, R. (1993), “Geographic localization of knowledge
spillovers as evidenced by patent citations”, The Quarterly Journal of Economics, Vol. 108
No. 3, pp. 577-598.
CR
25,4
388

Kahl, J. (2014), “The impact of mergers and acquisitions on performance and innovation.
empirical evidence from the North American pharmaceutical industry”, Dr Kovac,
Hamburg.
Koka, B. and Prescott, J.E. (2002), “Strategic alliances as social capital: a multidimensional view”,
Strategic Management Journal, Vol. 23, pp. 795-816.
Lane, P.J. and Lubatkin, M. (1998), “Relative absorptive capacity and interorganizational
learning”, Strategic Management Journal, Vol. 19 No. 5, pp. 461-477.
Lavie, D. and Rosenkopf, L. (2006), “Balancing exploration and exploitation in alliance formation”,
Academy of Management Journal, Vol. 49 No. 4, pp. 797-818.
Lee, A.V., Vargo, J. and Seville, E. (2013), “Developing a tool to measure and compare
organizations’ resilience”, Natural Hazards Review, Vol. 14 No. 1, pp. 29-41.
Levitt, B. and March, J. (1988), “Organizational learning”, Annual Review of Sociology, Vol. 14,
pp. 319-340.
Luke, D. (2004), Multilevel Modeling, Sage, Thousand Oaks, CA.
McManus, S., Seville, E., Vargo, J. and Brundson, D. (2008), “A facilitated process for improving
organizational resilience”, Natural Hazards Review, Vol. 9 No. 2, pp. 81-90.
March, J. (1991), “Exploration and exploitation in organizational learning”, Organization Science,
Vol. 2, pp. 71-87.
Martin, R. (2012), “Regional economic resilience, hysteresis and recessionary shocks”, Journal of
Economic Geography, Vol. 12 No. 1, pp. 1-32.
Martin, R. and Sunley, P. (2011), “Conceptualizing cluster evolution: beyond the life cycle model?”,
Regional Studies, Vol. 45 No. 10, pp. 1299-1319.
Marzucchi, A., Antonioli, D. and Montresor, S. (2013), “Industry-research cooperation within and
across regional boundaries. What does innovation policy add?”, Papers in Regional Science,
forthcoming.
Menzel, M. and Fornahl, D. (2010), “Cluster life cycles: dimensions and rationales of cluster
evolution”, Industrial and Corporate Change, Vol. 19 No. 1, pp. 205-238.
Neffke, F., Henning, S.M. Boschma, R.A. Lundequist, K.J. and Olander, L.O. (2011), “The dynamics
of agglomeration externalities along the life cycle of industries”, Regional Studies, Vol. 45
No. 1, pp. 49-65.
Nelson, R.R. and Winter, S.G. (1982), An Evolutionary Theory of Economic Change, Harvard
University Press, Cambridge, MA.
Noteboom, B. (2000), Learning and Innovation in Organizations and Economies, Oxford
University Press, Oxford.
OECD (2012), Science, Technology and Industry Outlook 2012, OECD, Paris.
Pendall, R., Foster, K.A. and Cowell, M. (2010), “Resilience and regions: building
understanding of the metaphor”, Cambridge Journal of Regions: Economy and Society,
Vol. 3 No. 1, pp. 71-84.
Pfeffer, J. and Salancik, G. (1978), The External Control of Organizations, The Maple Press
Company, New York, NY.
Porter, M.E. (1990), The Competitive Advantage of Nations, The Free Press, New York, NY.
Porter, M.E. (1998), On Competition, Harvard Business School Press, Boston, MA.
389
Employment
performance
in times of
crisis

Pouder, R. and St. John, C.H. (1996), “Hot spots and blind spots: geographical clusters of rms and
innovation”, Academy of Management Review, Vol. 21 No. 4, pp. 1192-1225.
Powell, W.W., Koput, K.W., White, D.R. and Owen-Smith, J. (2005), “Network dynamics and eld
evolution. The growth of interorganizational collaboration in the life sciences”, American
Journal of Sociology, Vol. 110, pp. 1132-1205.
Raudenbush, S.W. and Bryk, A. (2002), Hierarchical Linear Models: Applications and Data
Analyses Methods, Sage Publications, Newbury Park, CA.
Raudenbush, S.W., Bryk, A., Cheong, Y.F., Congdon, R. and du Toit, M. (2011), HLM 7:
Hierarchical Linear and Nonlinear Modelling, SSI Scientic Software International,
Lincolnwood, IL.
Rodan, S. and Galunic, C. (2004), “More than network structure: how knowledge heterogeneity
inuences managerial performance and innovation”, Strategic Management Journal,
Vol. 25 No. 6, pp. 541-562.
Rothaermel, F.T. and Deeds, D.L. (2004), “Exploration and exploitation alliances in biotechnology:
a new system of product development”, Strategic Management Journal, Vol. 25 No. 3,
pp. 201-221.
Sampson, R.C. (2007), “R&D alliances and rm performance: the impact of technological diversity
and alliance organization on innovation”, Academy of Management Journal, Vol. 50 No. 2,
pp. 364-386.
Simmie, J. and Martin, R. (2010), “The economic resilience of regions: towards an evolutionary
approach”, Cambridge Journal of Regions, Economy and Society, Vol. 3 No. 1, pp. 27-43.
Simonin, B.L. (1999), “Ambiguity and the process of knowledge transfer in strategic alliances”,
Strategic Management Journal, Vol. 27 No. 7, pp. 595-623.
Snijders, T. and Bosker, R. (2004), Multilevel Analysis: An Introduction to Basic and Advanced
Multilevel Modeling, Sage, London.
Swann, G.P., Prevezer, J.J. and Stout, D. (1998), The Dynamics of Industrial Clustering:
International Comparisons in Computing and Biotechnology, Oxford University Press,
Oxford.
Tushman, M.L. and Anderson, P. (1986), “Technological discontinuities and organizational
environments”, Administrative Science Quarterly, Vol. 31 No. 3, pp. 439-465.
Tushman, M.L. and O’Reily, A.C. (2004), “The ambidextrous organization”, Harvard Business
Review, Vol. 82 No. 4, pp. 74-81.
Wasserman, S. and Faust, K. (1994), Social Network Analysis: Methods and Applications,
Cambridge University Press, Cambridge.
Wrobel, M. (2013), “‘One for all and all for one’: cluster, employment, and the global economic
crisis. Evidence from the German mechanical engineering industry”, Papers in Regional
Science, Vol. 94 No. 2, pp. 273-294.
Zettinig, P. and Vincze, Z. (2012), “How cluster evolve”, Competitiveness Review, Vol. 2 No. 2,
pp. 110-132.
Further reading
Ernst & Young (2004), Deutscher Biotechnologie Report 2004, Ernst & Young, Mannheim.
Ernst & Young (2008), Deutscher Biotechnologie Report 2008, Ernst & Young, Mannheim.
Martin, R. and Sunley, P. (2006), “Path dependence and regional economic evolution”, Journal of
Economic Geography, Vol. 6 No. 4, pp. 395-437.
CR
25,4
390

Park, E. and Ostergaard, C.R. (2012), “Cluster decline and resilience – the case of the wireless
communication cluster in North Jutland”, paper presented at the DRUID 2012, Barcelona, 4
January, available at: http://papers.ssrn.com/sol3/papers.cfm?abstract_id⫽2196445
Prevezer, M. (1997), “The dynamics of industrial clustering in biotechnology”, Small Business
Economics, Vol. 9 No. 3, pp. 255-271.
Van der Panne, G. (2004), “Agglomeration externalities: Marshall versus Jacobs”, Journal of
Evolutionary Economics, SSI Scientic Software International, Vol. 14 No. 5, pp. 593-604.
About the authors
Julian Kahl is a Research Associate at the Department for Urban and Regional Economics at
the Institute of Geography at the Ruhr University of Bochum. Julian Kahl is the corresponding
author and can be contacted at: julian.kahl@rub.de
Christian Hundt is a Research Associate at the Department for Urban and Regional
Economics at the Institute of Geography at the Ruhr University of Bochum.
For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com
391
Employment
performance
in times of
crisis