The geography of startups in Amsterdam : insights on the locational behaviour of venture capital backed startups

(1)

THE GEOGRAPHY OF STARTUPS IN

AMSTERDAM:

Insights on the locational behaviour of venture capital

backed startups

Wai Yin Liu - 10787720 Supervised by Prof. Dennis Arnold Second reader: Prof. Niels Beerepoot 12 August 2019 Master Economic Geography Graduate School of Social Sciences

(2)

1 ABSTRACT

In much of the Global North, a turn to a knowledge economy can be witnessed in the past few decades. Subsequently, entrepreneurship has become important in the policy of Amsterdam in recent years. Startup formation, creating a startup ecosystem and attracting startups has become one of the main ambitions of the municipality of Amsterdam. This research aims to add to the gap of literature on the geography of startups. By using a multi-scalar approach, this research aims to understand the locational behaviour of venture capital (VC) backed startups in Amsterdam. By looking at why Amsterdam is chosen and why specifically an area or neighbourhood in the city is chosen by VC-backed startups as office location. The research has found that within Amsterdam, a micro level geography of VC-backed startups is visible in the form of clusters around incubator, co-working and university spaces. In addition, interviews with high ranking startup employees have indicated that perceived factors influencing locational behaviour can be distributed under macro (city) and micro (neighbourhood) level characteristics.

(3)

2

1. Introduction

In the past few decades, a shift in dominance from a manufacturing economy to a knowledge economy in the Global North can be witnessed (Baldwin, 2009; Kloosterman, 2015), this also manifests itself in policy. In Europe for instance, a number of leaders have pushed to speed up the transition towards a knowledge-based economy since the late 1990s, in order to match the economic growth levels of the United States and the Asian tigers (Van Winden et al., 2007). Consequently, the Lisbon Agenda became the action plan of the European Union which committed European governments to increase their investments in R&D, education, information and communication technology and the promotion of commercialisation of research (ibid.). The main aim of this action plan was to transform the European Union into the most competitive knowledge-based economy and dynamic knowledge-based economy in the world, capable of sustainable economic growth with more and better jobs and greater social cohesion.1 This is not to denote the importance of manufacturing in the west, but now more than ever, attracting the presence of knowledge intensive businesses and activities is high on agendas of cities. Having financial companies, large multinationals and research institutions is often desired by municipalities.

However, in the knowledge economy it also means that in the past 30 years, the innovative advantage of large firms has transferred over to small and new firms due to new technologies and the reduced importance of scale economies in many sectors (van Stel & Suddle, 2008). In debates concerning economic development, it is the relevance of exactly this new form of entrepreneurship that have increasingly been emphasized (Stam, 2007). The necessity for the creation of new firms, or startups in general is underlined by the OECD (2017, p.72):

“The birth of new enterprises is a key indicator of business dynamism. It reflects an important

dimension of entrepreneurship in a country, namely the capacity to start up an entirely new business. New enterprises are considered as key drivers of growth due to their contribution to aggregate job creation and because of the productivity-enhancing effect associated with a pace of firm entry and exit.”

Consequently, entrepreneurship in the form of startup creation is increasingly being promoted by municipalities. For instance, the municipality of Amsterdam is always keen on strengthening the economic position of the Amsterdam Metropolitan Area (AMA) for economic wealth and welfare for its citizens (Gemente Amsterdam, 2011). But in recent years, there has been a particular focus from the municipality to focus on creating a ‘city of knowledge’ through the promotion of innovation and startups (Gemeente Amsterdam, n.d. a). Amsterdam has a rich history when it comes to hosting

1

“The Lisbon Strategy in short”

(5)

4 successful start-ups. Companies like Booking.com, Adyen, WeTransfer and Travelbird are some of the most well-known success stories (StartupAmsterdam, 2015).

Thus, entrepreneurship is important for modern day economies and Joseph Schumpeter underlines that it drives innovation as well (see Sledzik, 2013). However, the spatial distribution of innovation is not evenly spread across regions. For one, cities and urban areas tend to attract and harness innovation, entrepreneurship and new economic activities (Glaeser et al., 1992; Florida & Mellander, 2016). Nevertheless, the relationship between urban areas and entrepreneurship is nuanced. The geographical relationship between the city and entrepreneurial activities has undergone changes since the turn to a knowledge economy. Florida and Mellander (2016) argue that entrepreneurial activities have increasingly been taking place within the city as opposed to suburban peripheral areas. During the 1970’s, 80’s and 90’s, innovation took place in suburban outposts, in particular within the high-tech sector. One example is California’s Silicon Valley, which houses Intel, Apple, Google and Facebook. Researchers documented this rise of so-called edge cities of industries and technology at the suburban periphery, and identified clusters of high technology enterprises and venture capital in peripheral areas such as Silicon Valley (ibid.). The success story of Silicon Valley has (in)directly led to many regions in industrialized countries to set up their own science parks and venture capital and financial innovation support schemes (Breschi & Malerba, 2001). Still, recent studies have shown that the movement of talent and jobs have shifted back into denser urban locations and centres (Florida & Mellander, 2016). One reason for this process is most likely related to agglomeration effects. Denser and more diverse areas can create more opportunities in terms of a higher variety in demand for products and services (Bosma, van Stel & Suddle, 2008). Innovation therefore cannot be seen as a solitary activity that can happen everywhere, it is to a certain extent dependent on its environment. Hence, exploring where entrepreneurship and innovation concentrate can allow for a better understanding of the environmental characteristics that foster them. Again, there are variances in the geography of entrepreneurship in cities based on the aforementioned. Innovation and entrepreneurship, as argued by Adler et al. (2019), is not dispersed homogeneously across the city, as it is inherently heterogeneous in nature when looking at social, economic and political characteristics. Instead a micro-level geography of entrepreneurship within the city is visible in the U.S. based on their study (ibid.). It is the neighbourhood level that constitutes this micro-level geography.

Following these interesting findings in the U.S., this research focuses primarily on the spatial pattern of entrepreneurship in Amsterdam in particular, in order to provide more insight on the geography of start-ups within cities. Current policy directions suggest that the municipality of Amsterdam is keen on promoting a startup ecosystem in the city through attracting startups and having incubator and accelerator spaces (StartupAmsterdam, 2015), which makes this study fitting.

(6)

5 In general, the research aim is twofold: first, the research provides a visual representation of the geography of startups in Amsterdam. The geography of start-ups within the city is mapped through the use of the geographic information system (GIS) ArcMap, showing where clusters of startups are situated. Secondly, the research aims to understand what influences the locational behaviour of startups using a multi-scalar approach. The research tries to unveil why Amsterdam and why a specific location or neighbourhood within Amsterdam is chosen by startups. Therefore, the following research question is answered in this paper:

- To what extent do perceived local socio-economic and physical environmental factors influence the geography of start-ups on the city and neighbourhood scale in Amsterdam?

2. Conceptual framework

In order to fulfil the research aims defined by the main question, firstly an understanding the (societal) relevance of the topic is necessary. Consequently, this section frames the discussion regarding the increasing economic significance of start-ups in cities, and why cities tend to be the places where start-ups and entrepreneurship can be found. What are some of the (in)direct causes that have led to this turn to entrepreneurship and innovation in the policy and developmental vision of municipalities such as Amsterdam? Categorizing in broad terms, insights are drawn from globalisation literature and the (re)structuring of economies, economic and social geographical literature on the competitiveness of regions, literature related to the theories of Joseph Schumpeter, and economic geography literature on the spatial distribution of entrepreneurship. Based on the literature reviewed, the following concepts are promising for this research:

- The knowledge economy - Urban/regional competitiveness - Entrepreneurship

- Start-ups and the locational behaviour of startups - Agglomeration effects

2.1. The knowledge economy

The societal relevance of the shift to a knowledge economy cannot be underestimated. Explaining the knowledge economy requires a description of the processes that paved the way for a knowledge turn as discussed briefly in the introduction. Globalization has played a huge role in changing how economies are structured, especially what is called the second unbundling of globalization. The first unbundling of globalization, which occurred from the steam revolution to the mid or late 1980s, was much about falling trade costs. The second unbundling was much more about the diminishing communication and transmission costs due to the ICT revolution (Baldwin, 2011). In other words, it

(7)

6 became much easier for one section of a firm to communicate with other sections. This had huge impacts on economies and in particular on the way industrialization was structured globally. Manufacturing and labour intensive jobs of advanced-nation firms were outsourced in large numbers from the global North to the Global South (Baldwin, 2011; Audretsch & Thurik, 2000).

Kloosterman et al. (2015) argue that the service sector has become increasingly digitized, commodified and tradable due to advances in ICT. Moreover, they argue that the advances in ICT, in combination with neoliberal deregulation policies, have created a new phase in global capitalism. First, they underpin a fundamental shift in the production system of advanced economies towards cognitive-cultural activities that depend on the input of highly skilled labour. Likewise, there is a growth of employment in what is called transactional activities, to a point where it has outgrown that in manufacturing in advanced economies (ibid.). Secondly, there is a specific type of global division of labour visible. The distribution of high-end cognitive-cultural activities is not only limited to advanced economies in the global north, but they are also present in China, India, Latin America and Africa. The traditional view of core-periphery (i.e. global north and global south) has now developed into a more complex ‘global mosaic’ of large metropolitan areas (see Krugman, 1991; Dieleman & Faludi, 1998; Kloosterman & Lambregts, 2001; Brenner, 2003; Pain, 2008).

This socio-economic restructuring process coincided with the global economic crisis of the 1970s and highlights the shift from relatively low skilled labour to a ‘knowledge’ turn in the global north. The notion of a knowledge economy has been used often for the past 20 years, appearing in OECD and World Banks reports (see World Bank, 2007; Unger, n.d.). The knowledge economy refers to the increased economic significance of knowledge production. According to the OECD (1996, as cited in Van Winden, 2010), knowledge economies are “…economies which are directly based on the

production, distribution and the use of knowledge information”. The knowledge economy remains,

however, a broad concept. Van Winden et al. (2007) provides us two perspectives that complement each other. The first perspective sees the knowledge economy as a separate entity in the economy. It is the sector in which actors such as universities and corporate research establishments conduct fundamental or applied research. Through research they ultimately produce knowledge that leads to new, products, production methods and productivity growth. Using this perspective, the performance of nations and regions is often measured in terms of patents, R&D spending and innovations. The second perspective is more inclusive, it sees knowledge as a dominant factor throughout the economy and does not only view scientific or technological knowledge as drivers of growth. The knowledge economy should have an economic and institutional regime that provides incentives for the efficient use of existing knowledge, the creation of knowledge and entrepreneurship. Moreover, it should have an educated and skilled population that can create and use knowledge. Furthermore, it needs an information infrastructure that can facilitate the effective

(8)

7 communication, dissemination and processing of knowledge. Using this perspective, cities specialised in advanced services, creative industries and innovative manufacturing sectors, besides those having a ‘formal’ knowledge base, can also become successful ‘knowledge cities’.

It is documented that knowledge, created through innovation and technological progress, as well as creativity, are considered engines for economic growth (Van Winden, 2010; Asongu, 2015). As such, human capital and R&D have obtained a more prominent place in economic theory and policy (see OECD, 2007). While it might be clear that innovation and entrepreneurship power economic growth, most theories dating back to the theories of classical economists Marx and Schumpeter, as well as more modern theories associated with Robert Solow, view innovation and entrepreneurship as something operating within the confines of the firm or the individual level (Florida et al, 2017). The exact role of cities and geography in a knowledge economy was not focused upon until the last decades. Van Winden (2010) argues that that the ‘needs’ of the knowledge economy has reinforced the role of cities in the current economic structure. Firstly, cities are inherently diverse when it comes to the supply of people, firms and cultures. As such, they are incubators for new ideas and innovations. Secondly, knowledge spill-overs is geographically concentrated in urban areas due to the close proximity and density of knowledge workers and knowledge based firms (Audretsch & Feldman, 1996; Van Winden, 2010). Thirdly, related to the diversity of cities, urban areas have large and specialized labour markets. Both for firms and workers this is attractive due to the fact that firms can find specialised staff and workers have more career opportunities (Van Winden, 2010). Lastly, cities are often specialised in creative industries and knowledge-intensive services, which are two sectors that show high rates of growth in the knowledge economy. Therefore, urban agglomeration economies have become more important in the current economic structure. Hence, planning policies are aimed often at the regional development and governance (see Albrechts et al., 2003), which is also related to the competitive turn of the urban agenda. However, there are variances in the success of urban agglomerations. London, Amsterdam and Paris are examples of successful agglomerations in northwest Europe, but not all cities benefit equally from a knowledge economy (Van Winden et al., 2007).

2.2. Urban competitiveness

The origins of competitiveness is in line with the economic restructuring of the global economy from the 1970s onwards, which involved an increased growth of multinationals, the emergence of rapid information exchanges, and the liberalisation of trade and capital flows. Cities have been at the forefront of this restructuring process (Peck et al, 2009). With the role of urban areas in the knowledge economy being increasingly more significant, (international) competitiveness between cities has increased as well (Gordon, 1999; Chorianopoulos et al. 2014). Turok (2004) underline the

(9)

8 two main reasons for increased competitiveness as the increasing international mobility of capital and more open national markets. Economies have become more interconnected through rising exports and increased foreign direct investments (FDI). Competition is usually between places similar in size and scope, competing for things such as but not limited to investment, skilled workers and tourism (Malecki, 2007). However, competitiveness itself is a tricky concept due to the fact that it is used at different economic levels (Martin & Simmi, 2008), from the individual firm (the micro-economic level) to the national economy (macro-micro-economic level). Paul Krugman (1996, as cited in Martin and Simmie, 2008) even argues that the obsession with competitiveness could lead to bad economy policies. This can be related to the fact that there is no one magic formula to make all economies competitive, one cannot simply disregard the historically determined economic track on which a city (or nation) finds itself (Musterd & Deurloo, 2006). However, the significance of urban competitiveness is addressed by Sáez and Periáñez (2015), who describe a “global/local duality”. It is the notion that global opportunities are dependent and created by local capabilities and initiatives. Underlining that cities and regions are incubators for global socioeconomic growth, which in turn offers those cities and regions political and economic power. Hence, competition between regions also involves enhancing a city’s attributes that makes it attractive in the first place, so that it remains a ‘sticky place’ (Malecki, 2007). The aforementioned is partly incorporated in the definition of urban competitiveness provided by Martin and Simmie (2008, p.4.), they consider it:

“…the ability of cities to continually upgrade their business environment, skill base, and

physical, social and cultural infrastructures, so as to attract and retain high-growth, innovative and profitable firms, and an educated, creative and entrepreneurial workforce, to thereby enable it achieve a high rate of productivity, high employment rate, high wages, high GDP per capita, and low levels of income inequality and social exclusion.”

D’Arcy and Keogh (1999, p.917) claim that “urban competitiveness is taken to mean the ability of a

city to exploit or create comparative advantage, and thereby to generate high and sustainable economic growth relative to its competitors”.

Based on the above, urban competitiveness is mainly used as a measure of economic growth or development, by having an edge over other peer cities. In Europe, this competitive turn is also witnessed in urban policy over the recent years. At the subnational level, economic development strategy documents often talk about having local comparative advantages and attempts to improve the regions attractiveness for the business environment, as well as catering the knowledge economy (Chorianopoulos et al, 2014) (see also Musterd, 2004; Musterd & Deurloo, 2005; gemeente Amsterdam, 2011; StartupAmsterdam, 2015).

(10)

9

2.3. Entrepreneurship

As has become apparent in the previous sections, entrepreneurship can be considered part of enhancing urban competitiveness. Entrepreneurship, as well as innovation, are arguably two of Joseph Schumpeter’s most distinctive contributions to economics. Generally formulated, the entrepreneur is someone who is concerned only with functions and activities related with innovation in the first entrepreneurship theory of Schumpeter (Sledzik, 2013). This differed from other visions of the entrepreneur presented in the literature of his time since he also defined the entrepreneur not simply as an employer and owner of capital. The entrepreneur is someone who breaks with tradition and would revolutionize the pattern of production, creating new tradition. Or as Schumpeter would call, an innovator (ibid.). In his second theory of entrepreneurship, Schumpeter refined his definition of an entrepreneur to something less individualistic. Here, the entrepreneur does not have to be one person, it could also be the country itself or even its agenda (ibid.). Nowadays, when this second definition is utilized, firms can also be seen as entrepreneurs. As Sledzik (2013) argues: “Today’s

knowledge-based economies are dependent by a dynamic technological progress. The generation of innovation no longer depends on individual personalities but involves the cooperation of many different actors.” Audretsch (2018) views the role of entrepreneurship as a conduit. It enables

spill-over of knowledge from the source or organization creating that knowledge, to the new organization where it is ultimately commercialized and transformed into innovative activity.

In any case, it is well documented that entrepreneurship in the current knowledge economy plays a very significant role. It is a crucial determinant of (regional) economic development and the formation of new firms is a fundamental process in economic geography (Stam, 2005; Ferreira et al., 2018; Malecki, 2018). Economists and policymakers often claim that urban success depends on a city’s level of entrepreneurship (Glaesser et al., 2015). Consequently, empirical studies have indeed shown a correlation between small establishment size (e.g. start-ups), and urban employment growth across sectors in the U.S. (ibid.).

Nonetheless, one should not leave the exogenic forces of the environment out of the equation when looking at the level of entrepreneurship. Many other factors such as local public policies and the historic path of a city influence entrepreneurship levels as well (ibid.), which can be called the entrepreneurial ecosystem. This highlights the geographic nature of entrepreneurship. In short, the entrepreneurial ecosystem describes how entrepreneurship is enabled by a comprehensive set of resources and actors, which are often present locally. Forms of governance that enable investment and innovation to take place, and the availability of formal and informal institutions that enable these forms of governance, are all part of the entrepreneurial ecosystem (Stam, 2015). Furthermore, institutional and physical characteristics of the entrepreneurial environment play a role as well (ibid.). These characteristics can often be found in urban areas. As

(11)

10 aforementioned, urban areas have characteristics that ‘feed’ the knowledge economy. As such, there is a specific geographical distribution of Schumpeterian entrepreneurship, or (innovative) startups, visible in the form of clustering in cities (see Kloosterman & Lambregts, 2001; Florida et al., 2017; Adler et al., 2019). Though, one needs to be wary of the fact that there is a nuance in clustering, since it can happen on either a macro-level scale (city-wide scale) or the micro-level scale (neighbourhood scale). The role of the entrepreneurial or startup ecosystem in the geography of entrepreneurship is elaborated upon in further sections of this conceptual framework.

2.4. Startups and the locational behaviour of startups

The startup is a form of entrepreneurship, but what constitutes a startup, or often referred to as new venture in the field of economics and business management, is not generally defined in academics. Yet, it is often considered to be a new firm (Kane, 2010), a recently formed firm (Davila et al., 2003) or a firm that is in its early stages of development and growth (Klotz et al, 2013). In their research, others consider a new venture a start-up if it is a firm that is 6 years old or younger, since the first 6 years have been deemed crucial to success or failure (Amason et al, 2006). Within these first years, failure rates of startups is very high (Gorman & Sahlman, 1989, Amason et al., 2006). For instance, Gadet et al. (2016) consider in their research on Amsterdam a startup a firm with a maximum age of 5 years.

Age especially is an important element that characterises what constitutes a start-up. This has everything to do with the fact that age can provide us with an estimate of the startup’s position within their lifecycle. More specifically, a connection with locational behaviour can be made by looking at a startup’s lifecycle. There are different perspectives when it comes to understanding locational behaviour of startups. Using the first perspective, it is important to acknowledge the relationship between location and the development stage in which the startup finds itself. According to Stam (2007), conceptualizing the geography of entrepreneurial firms is not something entirely new. Three dominant theories have influenced past theoretical studies on the location of firms: the neoclassical economic theory, the transaction cost theory and the behavioural theory of the firm. However, these have shortcomings as they have failed to take into account the changing nature of the entrepreneurial firm (ibid.). Successful startups are ever-evolving, not only in size but also in structure. The changing nature of startups can be illustrated by means of the startup life cycle. Salamzadeh & Kesim (2015) state that three stages can be identified that all startups go through: the bootstrapping stage, seed stage and creation stage. Paternoster et al. (2014) also elaborates on the startup development using different, yet similar, stages. The startup stage is where the idea is created and refined until the first sale. The stabilization stage is the second phase that lasts until the new product is stable enough to be sold to new customers, essentially getting the product or service

(12)

11 going. The third stage is the growth phase that begins with a stable product development and ends when market size, share and growth rate have been established. The final stage is the scale-up phase, where product development becomes robust. Stam (2007) produces a comparable view on the lifecycle of startups using slightly different developmental phases: the startup phase, initial survival phase, early growth phase, growth syndrome phase and the accumulation phase. According to Stam (2007, p.30) “…the start-up phase is the period in which an entrepreneur recognizes a

business opportunity and starts to mobilize the resources that are needed to take advantage of that opportunity.” Which can be compared with the seed stage introduced by Salamzadeh and Kesim

(2015). The initial survival phase entails the period in which the firm has sufficient productivity and commercial activity to be able to generate resources. The growth phases are often characterized by profitable exploitation of new market opportunities and the delivery of products to a growing product-market. This growth phase can also show signs of the “artificial” process of acquiring resources, whereby external investors supply financial resources, expecting superior returns in the future (Stam, 2007). Venture capital is part of this external investor’s supply of financial resources. The early growth and growth syndrome phases, as well as the accumulation phase overlap both the seed stage and creation stage defined in Salamzadeh and Kesim (2015). After this a firm enters the growth syndrome phase and accumulation phase. In these stages the firm starts to stand on its own legs and begins accumulating resources.

What these papers show is that there is a whole process going on within a startup before it matures into a “real” firm and exits the startup stage. These processes can influence the organizational structure of the firm, such as the size and the number of employees, which in turn can influence locational behaviour. Yet there is no such thing as an all-encompassing timeline that a startup makes use of, since they differ from each other and the sequence of activities and stages might therefore vary from each other. Consequently this also makes it hard to generate one accepted definition of a startup. Figure 1 is created to visualize the different (overlapping) startup lifecycle stages presented by the different authors as discussed above.

Figure 1: Startup lifecycle stages as presented by Salamzadeh & Kesim (2015) in green, Stam (2007) in grey and Paternoster et al. (2014) in orange.

(13)

12 Moving on, the second perspective that is useful when looking at the locational behaviour of startups is underlined by Stam (2007). He claims that it is important to take into account the role of the entrepreneur in regards to organizational change. For instance, looking at the entrepreneur at a personal level requires an understanding of his relationship with the (physical) environment, but also personal and inter-organizational network relationships. Through this, a connection could be made with locational behaviour.

Thirdly, the economic and social characteristics of a specific location can also influence the locational behaviour of startups. For instance, as research has shown, having a good entrepreneurial ecosystem or startup ecosystem is influential on attracting startups (Suresh & Ramraj, 2012; Stam & Spigel, 2016). Then what exactly constitutes an entrepreneurial or startup ecosystem? The entrepreneurial ecosystem can be categorized into the following six general domains: 1) quality of human capital, 2) the availability of appropriate finance, 3) a conducive culture, 4) venture-friendly markets for products, 5) a range of institutional and infrastructural support and 6) enabling policies and leadership (Isenberg, 2011). Stam and Spigel (2016) also cover some of these attributes. For instance, they underline human capital as a relevant attribute of an entrepreneurial ecosystem. There should be a deep talent pool of employees available at a location. The presence of universities can be helpful as a resource for startup talent. Moreover, they also acknowledge the importance of financial capital. An entrepreneurial ecosystem has different forms of financing such as, but not limited to, a supportive community of venture capitalists and (seed) investors (ibid.). Furthermore, supportive policies that cover economic development, tax and subsidies have to be readily available. The entrepreneurial ecosystem also needs leadership, consisting of other entrepreneurs who are tied to the location, to let others know that this specific location is a good place to start and develop a startup. Additionally, the presence of incubator and accelerator spaces is a prerequisite. Regulation, institutions and norms, infrastructure, city amenities and access to finance vary between regions and cities where new ideas and knowledge reside (Audretsch & Belitski, 2017), showing the importance the entrepreneurial ecosystem. The entrepreneurial ecosystem can thus be defined as:

“…combinations of social, political, economic, and cultural elements within a region that support the development and growth of innovative startups and encourage nascent entrepreneurs and other actors to take the risks of starting, funding, and otherwise assisting high-risk ventures…” (Stam & Spigel, 2014, p.7).

(14)

13 Continuing from the previous, Spigel (2017) has come up with the following categories of entrepreneurial ecosystem attributes: 1) material attributes, 2) social attributes and 3) cultural attributes. They encompass and

summarize the previously

mentioned attributes and their

relationship with each other see figure 2). This visualization is concise and useful to understand what can potentially influence the locational behaviour of startups.

2.5. Agglomeration effects

As aforementioned, economic activity tends to be concentrated within specific geographic locations. Especially urban areas house clusters of economic activity (Bosma, van Stel & Suddle, 2008; Florida & Mellander, 2016). According to Pe`er & Keil (2013), it is exactly these agglomerations which startups are attracted and locate to. Dense (urban) areas have a higher population diversity, which can lead to a higher variety of demand for products and services. In turn, this can promote the emergence of niche markets and attract startups (Bosma, van Stel & Suddle, 2008). Additionally, the clustering of industries can result in positive externalities for firms. Notably, their proximity to each other can lead to information and knowledge spill over, as well as the availability of a specialized labour market (Huisman & van Wissen, 2004). Pe`er and Keil (2013) also discuss labour market pooling, spill overs of technology and knowledge as benefits of agglomeration.

When discussing the concept of agglomeration in regards to the spatial concentration of economic activity or people, two perspectives can be used. The first entails the spatial clustering of different types of industries or local diversification. The phenomenon that people and economic activity concentrate within cities. The advantages gained by this process is referred to as urbanization economies (Huisman & van Wissen, 2004; Malmberg & Maskell, 2002). The second perspective is when firms within the same or closely related industry cluster together. The advantages as a result of this process is referred to as localization economies (ibid.).

However, to what extent startups are affected by agglomeration effects is up for debate. Pe`er and Keil (2013) claim that the benefits as well as drawbacks of agglomeration effects do not affect all startups equally. However, they identify three distinct mechanisms through how clustering can affect a startup’s survival. First, the extensive supply of high skilled labour in dense areas could

Figure 2: Attributes of an entrepreneurial ecosystem and their relationship with each other (Spigel, 2017).

(15)

14 make recruiting potential employees easier. Second, startups can find suppliers and partners to work with much easier in clusters, allowing them to focus more on their own activities. Finally, the market for testing their product and connecting with customers may be easier in clusters.

3. The research aim and questions

A focus point in the field of economic geography has been the clustering of various economic activity in urban areas (Kloosterman & Lambregts, 2001; see Florida & Mellander, 2016; Bosma et al, 2008; Malmberg & Maskell, 2002). Particularly, more attention is now given to the geography of start-ups. A recent study conducted by Adler et al. (2019) shows that the clustering of start-ups can be witnessed on two geographical scales, the macro (city level) and micro (neighbourhood level). They state:

“…tech-startup entrepreneurship (measured here as venture capital-financed startups) is heavily concentrated in a relatively small group of metros that provide assets and capacity in the form of diverse pools of talent, diverse groups of firms, leading-edge research universities and knowledge institutions and other factors. At the micro-geographic scale, within these leading city-regions or metro areas, tech-startup entrepreneurship is also clustered and concentrated in considerably denser and more tightly-woven micro-clusters at the district or neighborhood scale.” (Adler et al., 2019, p. 128).

Economic geographers and urban economists have recognized that the metropolitan region is not the only scale at which entrepreneurial activity occur. After all, Alfred Marshall and Jane Jacobs have stated that entrepreneurship and innovation happens at a much smaller level, such as the district or neighbourhood scale (Adler et al., 2019). Nonetheless, there is less empirical research conducted on this micro-level geography of entrepreneurship, since obtaining data on entrepreneurial activity at this level is more difficult (ibid.). However, Audretsch and Belitski (2017) state that studying entrepreneurial activity at a local context is necessary since that is where decisions take place and research is scarce. Hence, focusing on the social, material and cultural attributes of entrepreneurial ecosystems can aid in filling that gap.

This research aims to add to existing literature on the geography of start-ups, utilizing a multi-scalar approach focusing on the city (macro) level and neighbourhood (micro) level. Using Amsterdam as a case study, the following main question has been formulated:

- To what extent do perceived local socio-economic and physical environmental factors influence the geography of start-ups on the city and neighbourhood scale in Amsterdam?

(16)

15 The following three sub-questions will be used in order to answer the aforementioned question:

- How does the municipality of Amsterdam aim to attract start-ups and the creation of entrepreneurship in the city?

- To what extent are spatial patterns visible in terms of start-ups location within Amsterdam?

- What perceived (neighbourhood) characteristics are influential on the locational behaviour of start-ups?

4. Methodology

4.1. Research design

Amsterdam is used as the case study in this research. Case studies are helpful in understanding the complexity of a particular case and to conduct in depth research (Bryman, 2012). It furthermore has some advantages over other types of research designs. Data and evidence can be collected from an abundance of sources. A researcher can make use of a variety of primary and secondary data such as, but not limited to, literature, observations and interviews (Yin, 2003). Likewise, a case study design provides the opportunity to combine data collections methods. This triangulation of methods can in turn strengthen ones research (Noor, 2008). Hence, triangulation of methods is also made use of in this research. Desk-based secondary data analysis, geographic information system (GIS) analysis and primary data collection through semi-structured interviews are utilized. Thus, the research can be dissected into mainly three parts.

The first part consists of a visualization of startup locations in Amsterdam by making use of the GIS programme ArcMap. For the spatial analysis, start-ups with a maximum age of 5 years are taken into account. A map of start-ups provided by StartupAmsterdam2 helped sorting out the large number of start-ups in the Amsterdam region. The maps generated in ArcGIS are primarily made by focusing on the size of the startup. This visualization lets us know where start-ups are present and whether there is a geographical trend visible in Amsterdam on a neighbourhood level. Start-ups incorporated in the research are venture capital (VC) backed start-ups. The sampling process for the GIS part is elaborated upon in section 4.2.

The second part focuses mainly on the policy, ambition and aims of the municipality of Amsterdam concerning start-ups and the startup environment. This will predominantly be desk-based research consisting of analysing secondary data. Secondary data can be data collected by other researchers and data collected by institutions (Bryman, 2012). Data collected by institutions can be municipal documents such as policy documents, ambition documents and more. These documents

2

StartupMap Amsterdam

(17)

16 produced by the state can be useful for a social researcher (ibid.). However, as an addition, interviews with municipal workers active in the Amsterdam startup environment are also conducted. Notably, a programme manager working for Startup in Residence, a public-private programme hosted by the municipality of Amsterdam and a startup liaison working for StartupAmsterdam, the main organization of the municipality which concerns itself with the creation of a good startup ecosystem in Amsterdam through a public-private cooperation. These interviews are useful since they provide a different perspective on startups in Amsterdam and can aid in acquiring practical insights on and for clarification of abstract policies.

The third part of the research relies predominantly on primary data. Two groups of actors were interviewed. First a number of semi-structured interviews were conducted with high ranking start-up employees (e.g. CEOs and CFOs). High ranking startup employees were chosen because they are often the ones who are directly connected to the locational behaviour of the startup. Furthermore, these high ranking employees are often founders or co-founders who are working in the startup from the very beginning. Hence, the knowledge they have in terms of locational behaviour is more elaborate than the average employee. The second group of interviews were conducted with persons active in the startup ecosystem of Amsterdam such as a startup coach working at an incubator and municipal employees who are in direct contact with startups. The choice for semi-structured interviews is simple. When experts are given the opportunity to talk continuously, provide examples or other forms of exploration, they tend to reveal a lot more interesting information regarding a topic (Meuser & Nagel, 2009). Consequently, the interviews were conducted using open questions, though with a slightly different question for the two groups to accommodate for the difference in knowledge and experiences (see table 1). The startups representatives were asked about their personal experience concerning the location of their startup, while the questions for the other actors were more about their own view and past experiences with startups, in particular on what attracts a startup to Amsterdam. The following table shows the topic guides used for the collection of primary data.

STARTUP RESPRESENTATIVES

ACTORS IN STARTUP ECOSYSTEM (PRIVATE & PUBLIC ORGANIZATIONS

o What type of business model does the startup have?

o What is the role of the municipality in the startup world? (if applicable)

o What stage of the startup lifecycle is the startup currently in?

o What attracts startups to Amsterdam?

o Why choose the Netherlands as location? (only if foreign)

o What are main characteristics of the city/neighbourhood influencing locational behaviour of startups? (local level)

(18)

17 o Why choose amsterdam as location? o To what extent is there a trend visible if we

look at startup locations? o Why choose the specific neighbourhood?

o What role has the government (municipality) played in your locational behaviour?

o How and why have agglomeration effects and clustering influenced decision-making?

Table 1: Topic guide for the interviews with the two interview groups.

In a way, the interviewees can be seen as experts for this research. From a researcher’s point of view, a person is an expert due to his role as an informant (Walter, 1994, p.271, as cited in Meuser and Nagel, 2009). However, Meuser and Nagel (2009, p.18) argue that this definition can be elaborated. They claim that in scientific research someone can be called an expert when a researcher assumes that one can provide knowledge, which is not accessible to anybody in the field of action under study. As such, the strength of expert interviews lies in its ability to provide a knowledge advantage through exclusive information on the research subject (ibid.). It is also important to mention that experts can provide two types of information, namely extrinsic and intrinsic information (Savini, 2018). Extrinsic information is data not produced by the expert, but otherwise difficult to find (e.g. unpublished documents). Intrinsic information is more personal and subjective data. For example, these include opinions on strength and weaknesses of a project, understandings of past events, reasons for particular actions and general feelings about a certain situation (ibid.). The aim of the interviews with startup representative in this research is to find out whether and how

perceived city and neighbourhood characteristics have influenced locational behaviour of the startup.

In addition, these interviews can aid in getting to know the entrepreneur on a personal level. As Stam (2007) mentioned, one way of explaining locational behaviour of startups is to look at the role of the entrepreneur, what motivated them to do certain things. Thus, using the expert interview method is particularly suitable for the research.

4.2. Research sampling

Specifically for the GIS part, a total of 145 startups have been used as a sample for this research. The 5 interviews conducted are from startups from this list as well. The sampling process consisted of two stages, one for the GIS part and the other for the semi-structured interviews.

In the first stage, startups were identified using data and a map provided by StartupAmsterdam, which is an action programme of the municipality of Amsterdam aimed at strengthening Amsterdam’s startup ecosystem. To maintain focus and limit the number of startups, three base criteria were then applied. The first criterion is that a startup has to be venture capital (VC) backed. The relevance of venture capital (VC) backed startups is underlined by Florida and King

(19)

18 (2016), they state that VC can be seen as the fuel that can potentially power high-tech innovative and entrepreneurial start-up companies, that create and define whole industries. However, a direct causal link between VC and innovation has yet to be found (ibid.). While VC is not causally related to creating innovation in the start-up milieu, it has been acknowledged that it can help with the commercialization of innovation according to several studies (Florida & Kenney, 1988; Faria & Barbosa, 2014; Florida & King, 2016). Therefore, VC backed start-ups can in a way be seen as having more potential and thus are taken into account for the research. The second criterion is that the startups had to be no more than 5 years of age. Although, lifecycles of firms differ from each other, 5 years is still relatively young. This diminishes the possibility of taking into account fully matured firms as well. However, the first 4 years of existence of a new firm is generally characterized by a high failure rate (Stam, 2007; Picken, 2017). This means that some startups in this sample do not exist anymore. However, since the GIS part of the research solely visualizes the locational trend of start-ups in the past 5 years, it does not matter whether the startup still exists or not. The last criterion is that startups should have at least 2 employees. This is used to differentiate an actual new firm from a self-employed person labelling him-or herself as a firm. Based on these criteria, a spreadsheet of 145 startups is made using Microsoft Excel. In order for ArcMap to successfully convert and show the Excel spreadsheet on a map, coordinates had to be manually inserted in the Excel file. These coordinates were found using the addresses provided by StartupMap Amsterdam. The Excel file was then imported into GIS, which then produced the maps that visualize locational trends within Amsterdam.

The second stage of sampling consisted of choosing which startups to interview. Interviewees were found using the purposive sampling strategy. Purposive sampling is the selection of units including people, organizations and documents, with direct reference to the research question being asked (Bryman, 2012). Looking back at the research question, the aspect of “perceived characteristics” needs to be taken into consideration and subjective opinions need to be acquired. As a result, interviewees were obviously found by actively seeking persons working for a startup. Yet, it was more important to ensure that they are connected to the decision-making of office locations and have been working at the startup since or close to the beginning. Therefore, I aimed at contacting high ranking employees and founders or co-founders of startups. Potential interviewees were mostly contacted through the business-oriented social networking site LinkedIn. The first reason for this is that most (existing) startups in the sample had LinkedIn pages which helped sort employees based on their position within the startups. Secondly, contacting through LinkedIn seemed more personal than by mail. In the end, roughly 38 people were contacted directly through LinkedIn, though only 4 people replied positively for an interview, with others not replying rejecting or connecting me with others who might have more time or knowledge on the matter.

(20)

19 However, e-mail was also used for contacting people. The main reasons are the fact that some of the potential interviewees had private accounts, which meant they could not be contacted through LinkedIn, and the fact that municipal employees’ e-mail addresses were easier to find online. The final list of interviewees can be found in appendix A.

Lastly, since the dataset of startups takes into account all VC-backed startups from the past 5 years, comparatively new startups with few employees are thrown in the same dataset as older startups, with lots of employees. Wary of this and to ensure that the interviews are representative for all startups in the dataset, a division of three subgroups was made based on the number of employees (size). Startups were considered small-sized if they had 10 or less than 10 employees, medium-sized if they had between 11 and 25 employees and large sized if they had more than 25 employees. Additionally, this division is also used in the GIS maps. Through the use of this method, startups in different stages of the startup lifecycle can be represented. The developmental stage in which the startup finds itself in is important for locational behaviour, as mentioned in the conceptual framework, and the number of employees seems to be a good reference point for development.

4.3. Operationalization

In order to conduct the research, key concepts needed to be operationalised first. The main concept used in this research is entrepreneurship in the form of a startup. In this research a startup is a firm with at least 2 employees, which omits the possibility that a self-employed is also incorporated in the data. In addition, only startups from the period 2014 to 2018 are taken into account. Moreover, it should have venture capital investment, for VC-backed startups can be seen as having more potential in successfully scaling up and innovating.

Secondly, measuring the relevance and the presence of the knowledge economy and urban competitiveness in the city of Amsterdam is done by looking at municipal policy. In particular, whether the municipality take the importance of knowledge, innovation and urban competitiveness into consideration. As will become apparent in the result section, the municipality of Amsterdam indeed discuss the aforementioned.

As previously mentioned in the conceptual framework, when discussing the concept of agglomeration in regards to the spatial concentration of economic activity or people, two perspectives can be used. Operationalising agglomeration effect and clustering in this research is done using the first perspective, which entails the spatial clustering of different types of industries or local diversification, and the concentration of economic activity in particular. The maps in result section 5.2 are used to illustrate this phenomenon, with startups being the representation of the spatial clustering of economic activities in Amsterdam.

(21)

20

4.4. Epistemological considerations

The epistemological consideration for this research is an interpretative one, since it aims to acquire an interpretative understanding of social action in order to generate a causal explanation (Bryman, 2012). The fact that the research is based around policy documents and semi-structured interviews, and one of the aims is to understand the locational behaviour of startups, shows the interpretative nature of the data. Reasons for locational behaviour is subjective and every interviewee has without doubts different personal opinions. In addition, Richardson (1996) underlines that arguments can also be found in politics and policy, based on Foucauldian principles. The shaping of policy is vulnerable to the workings of power, which provides opportunity for manipulation and control, confusion and other distortions. So personal experiences the interviewees have are all influenced to a certain extent, especially the interviewees working with the municipality of Amsterdam. The following quote underlines the interpretative nature of policy and the importance for social researchers to be aware of this: “As politicians know only too well but social scientists too often

forget, public policy is made of language. Whether in written or oral form, argument is central in all stages of the policy process” (Majone, 1989, as cited in Fischer and Forester, 1993, p. 1.). Therefore,

as a researcher a neutral and objective stance is pursued in the research, particularly during the interviews.

4.5. Social research criteria

A crucial part of social research is taking into consideration validity, reliability and replicability (Bryman, 2012), as these three are the most prominent evaluation criteria for social research (Bryman, 2012; Yin, 2003; Yin, 2013). The latter criterion of replicability is very close to reliability and will be explained as part of it. Robert K. Yin is a main contributor when it comes to how validity and reliability should be acquired in social empirical and case study research. Both validity and reliability are challenging aspects when conducting case study research, particularly when there is a limited number of case studies (Yin, 2013). In order to strengthen a case study research, construct validity, internal validity, external validity and reliability need to be tested (Yin, 2003, p. 34). The following few paragraphs are dedicated to explaining each of these concept and how they are integrated in this research.

 Construct validity entails establishing correct operational measures for the concepts being studied (Yin, 2003, p. 34). This means that it is useful to know whether the measure devised of a concept really measures the data that it is supposed measure (Bryman, 2012; Yin, 2003). By having the concepts based on multiple sources of evidence as well as operationalizing the concepts in section 4.3, I aimed to address this issue. Furthermore, the triangulation of methods contributes to the

(22)

21 quality of research and may result in improved confidence in the research findings and validity (Yin, 2003).

 Internal validity is concerned with establishing “…a causal relationship, whereby certain conditions are shown to lead other conditions, as distinguished from spurious relationships” (Yin, 2003, p. 34). In other words, internal validity looks at whether a (proposed) causal relationship between two or more variables holds water (Bryman, 2012). Often in the discussion of causality, independent and dependent variables are used to describe the factor that has a causal impact and the effect respectively. Internal validity however, does not apply for exploratory or descriptive research (ibid.), which is what this research can be categorized as predominantly.

 External validity: this criteria focuses on the question whether the results can be generalized beyond the specific research context (Bryman, 2012; Yin, 2003). Since the research revolves around one case study, this research tries to put a (global) phenomena (i.e. the geography of startups) within a specific context (i.e. Amsterdam). This makes it difficult to generalize all the results since every city is different in terms of characteristics.

 Reliability of a research entails whether the procedures of a study can be repeated by another researcher with the same findings and conclusions (Yin, 2003). In terms of reliability of the data collection, the procedures of sampling have been described as elaborately as possible in order to strengthen reliability. Another researcher could follow the same criteria used for the GIS part and interview part by following the sampling procedure, while also keeping the topic guide close to the original.

5. Results

5.1. The municipality of Amsterdam and startups

The city of Amsterdam has been the staging ground for several successful startups in the past, mentioned in the introduction are examples such as Booking.com, Adyen, WeTransfer and Travelbird. More recently, the e-scooter sharing startup Felyx has become very successful in the city and their scooters are visible almost anywhere in the streets of Amsterdam. These examples also show the diversity of the type of startups in Amsterdam, the examples range from transportation, travel to financial tech (fintech) startups. Why and how then, are startups attracted to Amsterdam and what is the role of the municipality? In order to answer this question, we have to dive deeper into the stance of the municipality regarding innovation and entrepreneurship. Therefore, this first section of the results discusses the transition to the knowledge economy in city policy and the current view of entrepreneurship and its role in Amsterdam by the municipality.

(23)

22

5.1.1. StartupAmsterdam

In the strategic vision document “Structuurvisie Amsterdam 2040”, the municipality underpins the importance of creating and maintaining a leading position within the world economy. The main reason for this aim is to guarantee the well-being and economic prosperity of the citizens of Amsterdam (Gemeente Amsterdam, 2011). In order to do so, the municipality talks about four “grote

bewegingen”, which translate as big movements. In general these are strategic urban development

trends aimed at the creation of a metropolitan and mixed-urban landscape in Amsterdam, through combining residential areas, recreational areas and areas for working and economic activity (ibid.). The first of the four big movements is the expansion of the city centre, in particular the mixed land-use characteristic of the centre. It contains diverse type of land-land-use, from residential areas to areas with economic activities and amenities. The municipality aims to expand this land-use mix towards the periphery of Amsterdam. The second movement entails the incorporation of “green” areas around Amsterdam to the metropolitan region of Amsterdam, for recreational use and to maintain liveability in Amsterdam. The third movement consists of a so called rediscovery of the city’s waterfront areas, the IJ-banks. These waterfronts act as potential economic and housing development areas as well as recreational areas. The final movement acknowledges the internationalisation of the Dutch and Amsterdam’s economy. It aims to attract and maintain a belt of economic activity in the southern flank. Zuidas is already the international business and financial centre of Amsterdam and the municipality aims to make it the location of international business in the Netherlands.

These goals and aims indeed highlight a presence of internationalisation in the Netherlands and Amsterdam, and increased urban competitiveness between Amsterdam and other cities. Though the municipality primarily wants to maintain economic and social wealth for its citizens, it considers increasing urban competitiveness and internationalisation one of the main pillars to do so. Connecting back to the conceptual framework just based on this section, there is thus an increased importance of acquiring urban competitiveness. The municipality of Amsterdam is actually trying to enhance the economic sector it already has, namely the financial and business sector in Zuidas. Continuation and looking at local strengths is crucial. As Musterd & Deurloo (2006) stated, one cannot create a one size fits all policy to increase economic competitiveness. Policies that enhance (urban) competitiveness should not disregard the historic pathway of an area (see Martin & Simmie, 2008).

Another reason why it is relevant to understand what these big movements constitute of, is the fact that a lack of focus on startup and entrepreneurial policy and goals is then notable in this policy document from 2011. Only a small paragraph discusses importance of entrepreneurship for the economy of Amsterdam in a 300 page document. In fact, the word “ondernemerschap” (Dutch

(24)

23 for entrepreneurship) is only used 9 times in this document. However, the paragraph does talk about the need to make Amsterdam an attractive city for young firms. But until recently, the municipality of Amsterdam did not have a wide and suitable arrangement of policies aimed at fast growing small businesses and tech startups (StartupAmsterdam, 2015). With the shift towards a knowledge economy and increased urban competitiveness in mind, these policies are necessary (ibid.). Hence, a point of attention for the municipality was acknowledging that the Amsterdam startup scene was fragmented and without focus. As a reaction to the aforementioned, a public-private cooperation programme StartupAmsterdam was created. StartupAmsterdam is one of the organizations that the municipality is involved in, but it is not an isolated governmental programme. Rather it involves the whole startup ecosystem of Amsterdam with more than 75 partners including accelerators, incubators (e.g. Rockstart, ACE venture Lab StartupBootcamp), knowledge institutions (e.g. Vrije Universiteit Amsterdam, University of Amsterdam), large firms (e.g. Booking.com, TomTom), the Amsterdam Economic board and startup events and initiatives (StartupAmsterdam, 2015).

In the following paragraph, the ambition, goals and the means to reach these goals, of this action program are discussed. Information is predominantly acquired from the StartupAmsterdam action programme (2015) and reflected upon using the concepts introduced in the conceptual framework.

5.1.2. Ambition, goals and means

The ambition of StartupAmsterdam in the short and medium term is to create a good basis in order to accelerate the formation of a strong startup ecosystem in Amsterdam. This development should go to the extent as making the startup ecosystem reach the top three best startup ecosystems in Europe. Moreover, Amsterdam should be compared to top European tech cities such as London and Berlin. In short, Amsterdam should be the place for startups in Europe. The action programme also provided several concrete goals, the most relevant ones are as follows:

1. That more foreign (early stage) startups choose Amsterdam accelerators and incubators 2. That more foreign mid and later stage choose Amsterdam as a hub to scale up their business

practises as well as settling their HQ in Amsterdam.

3. That native Amsterdam startups can more easily recruit employees and talent with outside knowledge, in turn accelerating the development of these local startups.

4. That the position and reputation of Amsterdam tech startups are improved in other (foreign) tech clusters.

5. That more foreign investments are made in the Amsterdam startup scene and to mobilise Dutch financial actors (e.g. banks) to invest in startups and scale ups.

6. To promote strong, coherent and transparent cooperation between all actors in the Amsterdam ecosystem the cooperation within the startup ecosystem of Amsterdam.

(25)

24 Consequently, three general strategies are used to help reach the aforementioned goals. First of all, StartupAmsterdam wants to take advantage of the diverse type startup industries in the city. Instead of focusing on one industry, the municipality aims to maintain its position as a diverse hub and having cross-overs between industries, rationalising this strategy because other successful startup hubs do not have a focus on one industry either. Secondly, StartupAmsterdam discusses the position of Amsterdam as a testing ground for startups. For instance, another startup programme witihin StartupAmsterdam is “Startup in Residence”. This is a six-month programme that provides (early stage) startups the opportunity and means to test out their product in Amsterdam through a number of innovative challenges (Stevens, 2019, personal communication). Their place of origin does not matter. These challenges are often real-life issues in Amsterdam, for example making energy-use more sustainable in (government) buildings. Using this strategy, the municipality of Amsterdam wants to increase the liveability of its citizens, create an attractive startup city and help startups develop their product to scale up and continue its growth (ibid.). This can be partly connected to the third and last strategy, which is to help all stakeholders in the startup ecosystem excel, for if stakeholders grow and are successful, the whole startup ecosystem will be successful.

The course of action for reaching the previously mentioned goals is based on five fundamental foundations a startup requires, as identified by StartupAmsterdam (2015). These include 1) talent 2) customer base 3) content 4) the availability of investment capital and 5) (startup) environment. The following table discusses the five foundations and some exemplary interventions that StartupAmsterdam identified.

Foundations Measures Talent

The goal is to increase and attract startup and tech talent.

1) Online startup jobs advertising

2) Promoting startup & coding academies, traineeships (IT schooling)

3) Increase education on coding in all levels of educational (primary & high schools, universities)

Customers

The goal is to improve startup’s accessibility to potential clients.

1) Invest in corporate partner programs

2) Making the municipality Amsterdam a launching customer and the city of Amsterdam a test bed 3) Promotion of startup (networking) events

Content

The goal is to have one of the most elaborate startup events calendar in Europe. Amsterdam should provide startups new vision, new networks, investors and mentors through activities and events, at the local and

1) Talks at StartupAmsterdam

(26)

25

international level. This helps with marketing Amsterdam as startup hub.

Capital

The goal is to enhance availability of (venture) capital for the development of startups with high potential.

1) Hosting the event “capital on Stage”, which is where venture capitalists can meet founders of startups

2) Increase investments by “traditional” capital (e.g. pension funds) in startups

Environment

The goal is to create a successful startup environment.

1) City branding: advertise Amsterdam as startup city

2) Start relationships with other startup hubs in the world through cooperation

3) Concentrate startup activities within

Amsterdam, through co-working locations, incubators and accelerators and more.

As with the aforementioned policy document, the StartupAmsterdam programme, its ambitions and goals underline the (increased) significance of the knowledge economy and the concept of urban competitiveness on urban policy making in Amsterdam. This action programme for a large part can be seen as a manifestation of what Begg (1999) considers an increase in competition between cities. Furthermore, the global/local duality concept as addressed by Sáez and Periáñez (2015) can be connected to the policy and ambition of the municipality. The city, in this case Amsterdam, can be seen as an incubator for social economic growth. Enhancing Amsterdam’s startup ecosystem can aid in making the city an attractive place for startups. In turn, this can enhance Amsterdam’s position in Europe in terms of being a startup hub. Hence, local initiatives in Amsterdam can create global opportunities for Amsterdam. For instance, it can become a frontrunner on innovation through attracting startups by creating a successful local startup ecosystem. This can be related back to the conceptual framework, where Suresh and Ramraj (2012) and Stam and Spigel (2016) claim that the presence of a well-developed entrepreneurial ecosystem influences the locational behaviour of startup activity. In addition, Begg (1999) already underlined the role the local “environment” well before this action programme was created. In his paper, the local environment is viewed as a factor influencing the willingness of workers as well as businesses to locate in a city. Some examples of this local environment are the existence or absence of clusters of economic activity and even the history and mix of a city’s industries, which might influence the competitiveness of a particular city. With the action programme also discussing the potential clustering of startup activity in Amsterdam, having an overview of the geography startups is useful.

The geography of startups in Amsterdam : insights on the locational behaviour of venture capital backed startups