• No results found

THE EFFECTIVENESS OF THE DUTCH SEED CAPITAL PROGRAM ON FIRM GROWTH

N/A
N/A
Protected

Academic year: 2021

Share "THE EFFECTIVENESS OF THE DUTCH SEED CAPITAL PROGRAM ON FIRM GROWTH"

Copied!
51
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

THE EFFECTIVENESS OF THE DUTCH SEED

CAPITAL PROGRAM ON FIRM GROWTH

MASTER THESIS

By: Tim Zandvoort

S3541959

MSc BA Small Business & Entrepreneurship

Faculty of Economics & Business, University of Groningen, Netherlands Word Count: 11,664 (Body text)

Supervisor: S. Murtinu Co-assessor: M. Wyrwich

(2)

“The days in which anyone with a good idea could visit the bank

manager and leave with a loan are no more.”

(3)

ABSTRACT

(4)

Definitions & Abbreviations

• SCP – Seed capital program: This program was founded by the Dutch Ministry of Economic affairs to stimulate VC funds to invest in the early stages of innovative start-ups.

• Valley of death – Defined as the shortage of (venture) capital, firms are able to attract caused by the lack of information/competition in the capital market (Bax et al., 2019) • Individual venture capitalists – Venture capitalists that make no use of governmental

subsidies.

• Techno starters – A person or group that drives an enterprise or prepares the start on the basis of technological innovations or new ways of performing existing technology. This is all about the sale and delivery of products, processes and or services (no advisory).

• High-tech startups – Young and small firms with high-tech products that internationalize and grow faster than average (Preece et al., 1999).

• DVI – Dutch venture initiative: Governmental program in late-stage development of high-tech startups.

• GF – Growth facility: Governmental program to support SMEs with growth ambitions. • Market failure – Occurs when promising technologies fail to emerge due to weak

incentives for investment due to risks, uncertainty and the need for large investments (Nemet et al., 2018)

(5)

Table of Contents

1. INTRODUCTION ... 6

2. THEORETICAL FRAMEWORK ... 9

2.1.FINANCING HIGH-TECH STARTUPS ... 9

2.2.APPROACHES TOWARDS HIGH-TECH START-UP POLICIES ... 10

2.2.1.HOLISTIC APPROACHES ... 12

2.3.THE SEED CAPITAL PROGRAM ... 13

3. METHODOLOGY ... 15

3.1.EMPIRICAL MODEL ... 15

3.2.DEPENDENT VARIABLES ... 15

3.3.INDEPENDENT VARIABLES ... 16

3.4.CONTROL VARIABLES ... 17

3.5.LINEAR REGRESSION MODEL ... 18

4. DATA & DESCRIPTIVE STATISTICS ... 19

4.1.DATA COLLECTION ... 19

4.2.SAMPLE SELECTION PROCEDURE ... 20

4.3.SELECTION BIAS ... 20

4.4.DESCRIPTIVE STATISTICS ... 22

5. ANALYSIS ... 24

5.1.REGRESSION RESULTS ... 27

5.1.ADDITIONAL ANALYSES ... 28

6. DISCUSSION & CONCLUSION ... 35

7. LIMITATIONS & FUTURE RESEARCH ... 38

7. REFERENCES ... 40

8. APPENDIX ... 46

APPENDIX A:SAMPLE STATISTICS ... 46

APPENDIX B:DESCRIPTIVE STATISTICS TREATMENT GROUP ... 48

APPENDIX C:DESCRIPTIVE STATISTICS CONTROL GROUP ... 49

(6)

1. INTRODUCTION

In recent years, many high-tech firms and especially small firms are likely to face financing constraints. This is in most cases due to lack of collateral and asymmetric information which causes adverse selection and moral hazard problems, leading to large funding gaps (Carpenter & Petersen, 2002b). Since 2000, numerous governments worldwide have started programs to stimulate innovative start-ups & small- and medium sized enterprises (SMEs) to overcome this market failure. According to Henk Kamp, former minister of economic affairs, banks have become reluctant in providing finance, especially if it concerns start-ups with new products or concepts (Ministerie van Economische Zaken, 2016). Banks are not able to assess the quality and potential of innovations, which results in barriers for start-ups (Bonilla & Cancino, 2011). Therefore, entrepreneurs will have to rely on other sources of finance, because without finance, these enterprises won’t survive the ‘valley of death’. This term is widely used to express the need for finance after the start-up enters the seed-stage, which is marked by early growth (Nemet et al., 2018). Venture capital is often used by start-ups to achieve this necessary growth. Venture capitalists (VCs) are involved in the development of early stage businesses and funding, preferably in knowledge-based high-tech sectors (Cumming, 2007). It was estimated that 90% of all investments made in Canada and the USA were in high-technology sectors. The importance of these firms has also gained attention of governments, which led to stimulation of these firms with subsidies, direct venture capital funds and tax policies (Poterba, 1989; Lerner 1999; Gompers & Lerner, 2001; Keuschnigg & Nielsen, 2001; Kanniainen & Keuschnigg, 2004; Keuschigg, 2003).

(7)

firms and venture capital. In order to illustrate the scope of this research, the different programs are illustrated in table 1 which also shows the different stages of the lifecycle the programs are aimed towards. In most programs, there is overlap and therefore it is wise to illustrate the goals of the different programs. Although Table 1 shows 8 different programs, only 3 are focused at enlarging the venture capital market. These are the SCP, Dutch Venture initiative (DVI) and Growth facility (GF).

Dutch Venture initiative

The Dutch venture initiative is a Fund of Funds which supports VC funds that invest in fast-growing innovative firms. (Tweede kamer der Staten-Generaal, 2018). The DVI is a collaboration between the ministry of economic affairs and regional development agency Oost NL and the European investment fund (EIF). It differs from the SCP and GF since it did not start at the ministry. It focuses on later stages of the firm lifecycle, with the aim to make it easier for fast-growing SMEs to acquire financing (Bax et al., 2019).

Growth Facility

The growth facility program started in 2006 to help SMEs in obtaining venture capital. The subsidies are granted to venture capitalists or banks to stimulate SMEs’ with growth ambitions. These firms may want to expand internationally, take over other businesses, deal with growing demand, but are having a hard time obtaining financial resources due to their risk-profile. The GF’ subsidies reach up to 5 million euros. These investments are not attractive to traditional investors due to high costs and calculating the risks. The market above the 5 million appears to work properly, no help is needed in this stage currently (Staatscourant, 2018).

(8)

• The growth of employment after the initial investment was made • The growth of total assets after the initial investment was made These growth measures are used to answer the following research question:

What is the effect of the Seed Capital Program on the growth of firms that received funding? The remainder of this thesis is structured as follows. Chapter 2 describes the theoretical framework that helps to understand the underlying phenomena. Chapter 3 describes our methodology. Chapter 4 includes the data collection process & descriptive statistics. In chapter 5, the regression analysis was performed and the results are explained. Also, the descriptive statistics are explained here. Chapter 6 includes a discussion of the results followed by a conclusion. Finally, in chapter 7 the limitations of this thesis are discussed as well as suggestions for future research.

Table 1 – Subsidies of the Ministry of Economic Affairs (Bax et al., 2018)

Preparation/ Start of Start-up Growth of Start-up Consolidation Reorganization/balance

Ideas for

commercialization From idea to financed business plan From business plan to first customer and turnover From turnover to profit From turnover

to fast growth Repositioning and reorganization

Early phase financing (VVF)

Innovation credit

Seed capital program (SCP)

SME Guarantee DTIF/DGGF

(International expansion) Dutch Venture Initiative (DVI)

Guarantee entrepreneurial financing Growth facility (GF)

(9)

2. THEORETICAL FRAMEWORK

This chapter includes a theoretical background on financing high-tech startups, an overview of the approaches towards high-tech startup policies and ends with a description of the seed capital program issued by the Dutch Ministry of Economic affairs.

2.1. Financing high-tech startups

Schumpeter (1911) stated that entrepreneurial activities in sectors classified as highly technological are deemed to be a major source of innovation, qualified employment and economic growth. Nonetheless, financing these activities has been a hard task because of multiple market & governmental failures. Both VCs and banks face the problem of information asymmetry when evaluating investment opportunities and assessing lending applications (Binks & Ennew, 1996; Westhead & Storey, 1994). However, the risks for VCs and banks are not equal. If investments in a high-tech startup fail, banks have sufficient equity to cover this up, however bankers have to deal with two distinct types of risk. Adverse selection (1) and moral hazard (2). The first risk concerns providing finance for failing businesses (error 1) or not providing finance for (a potential) successful business (error 2). The second type of risk is moral hazard, since banks are not able to monitor the activities and expenditures of high-tech startups after the loan has been awarded (Parker, 2002; Binks & Ennew, 1997). Therefore, to mitigate these risks, banks need collateral and prove that the company is able to generate cash flows to pay back the loans and interests (Wynant et al., 1991). Due to the fact that VC’s do not face moral hazard and adverse selection, it is questionable whether banks and VCs target the same type of high-tech startups, especially since VC’s highly value the capability of management team, potential returns, market & product characteristics (Muzyka et al., 1996; Manigart et al, 1997) whereas banks put more value in collateral and cashflows.

Also, high-tech firms receive external financing from VCs (?) in a much earlier stage compared to banks (Gompers & Lerner, 1999). Because of these types of risks that banks face, high-tech startups rely primarily on venture capitalists as the source of financing. Venture capital currently suits the market imperfections best, also since they monitor the firms closely and have the right skills to overcome agency and information programs partially (Carpenter & Petersen, 2002a).

(10)

(Jacobs & Theeuwes, 2004). Young enterprises often do not have a track-record to convince investors of their idea or innovation. This results in a loss of potential good investments because of the estimated risks and costs to get possession of the necessary information. Furthermore, investors are having a hard time assessing the quality of the management team and assessing the risk of the investment. This slows down investments in sectors characterized by innovation (Jacobs & Theeuwes, 2004).

2.2. Approaches towards high-tech start-up policies

(11)

gap. Other issues have to be addressed as well in order to stimulate highly skilled and experienced founders in sectors characterized by innovation, governments should be aware of this fact.

There are many ways in which governments can stimulate venture capital markets. According to Denis (2004), governments stimulate start-ups by tax policy, development of venture financing markets, regulatory restrictions on investment, stock market development, development of angel networks & public financing of entrepreneurial ventures. In general, these methods could be classified into two distinct forms (Poterba, 1989; Gombers & Lerner, 1998). These are (1) law and (2) direct investment schemes by the government. Taxes are an important legal factor in the stimulation of venture capitalists. According to Poterba (1989), US VC fundraising rose from $68.200.000 in 1977 to $2.100.000.000 in 1982 due to the reduction of the corporate tax rate from 35% (1977) to 20% (1982). The main goal of most venture capitalists is a successful exit in terms of an IPO or sale of shares, since most start-ups are not able to pay dividends on equity and therefore an exit is the most profitable option. IPO’s and acquisitions are the most profitable exits for high-tech start-ups (Gompers & Lerner, 1999; Cumming & MacIntosh, 2003; Cochrane, 2005). This indicates that tax policies are an important instrument in the market of venture capitalists.

(12)

is associated with another measure of growth – growth of equity value of startups after multiple investment rounds. Their results show that changes in employment growth are positive and significantly associated with equity growth. In many studies accounting data is used to measure growth, however this type of data is usually not available for start-ups. This illustrates that employment growth, a variable that is often more accessible than equity values, can be used as a good proxy for the growth of startups. Also, Alemany & Martí (2005) found that VC-backed firms outperform Non-VC-backed firms in terms of employment growth. Growth of total assets is a commonly used measure in VC literature, since multiple studies found a positive relationship between VC investment and asset growth (Martí et al., 2013).

2.2.1. Holistic approaches

Governments tend to create their own programs towards financing startups, however there are many different forms in which startups are subsidized. Two holistic approaches in similar markets in Western-Europe are the enterprise capital fund (UK) and the High-Tech Gründerfonds (Germany). Both policies are in comparable markets and have comparable features, however both schemes some distinct features which are highlighted:

The Enterprise capital fund (ECF) represents the UK government’s flagship in addressing

the financing gap for young high-tech businesses (Bertoni et al., 2015). The policy was introduced in 2006 to improve the market for GVC (government-backed venture capital) whereas it previously relied on Regional Venture Capital Funds (Colombo et al., 2014). The ECF has two key objectives: first, provide funding for high-tech startups in the early and seed phase that require £250.000 - £2.000.000. Second, providing a demonstration model in order to get VC’s to invest in startups that are currently in the early/seed stage (Lerner et al., 2005). From the start of 2019, a greater emphasis lies on seed and earlier stage VC. Funds are required to contribute at least one third private investment with a maximum private individual investment of 50%. It is believed the ECF has met the needs of the viable seed and early stage ‘equity gap’. They encouraged domestic and foreign VC’s to invest in seed and early stage startups. Furthermore, the market has increased in size and scale and economic impact has been generated, leading to indirect job creation, sales and innovation (Baldock, 2016).

(13)

government provided 240 million to the fund and all other partners participated for a total amount of 32 million which gives the fund a total of 272 million. The main reason to start with the fund dates back to the Dotcom-bubble which burst in 2000. After this moment, VCs withdrew from seed capital investments which led to a standstill in this segment of the market. The period of the HTGF is twelve years including a five-year investment followed by a seven-year disinvestment phase. It focuses on small R&D-based firms with less than 50 employees (Geyer et al., 2010). Besides, it focusses on spin-offs from universities and large corporations which implement new technologies. According to Geyer et al. (2010) the HTGF has been successful in targeting the high-tech firms in need of finance. Both founders and seed fund managers rated the goals and activities of the HTGF positively. The HTGF is unique since it encourages large corporations to participate in financing in the seed and early stages of the firm life cycle.

2.3. The Seed capital program

The Dutch policy environment stimulates private investments and the market is characterized as small, open and relying on international trade with a high dependency on an advanced knowledge base and a labour force that is highly educated. A large portion of its policies aims to reduce CO2 adhering to the EU targets (Adenfelt et al., 2013). Almost all policies towards entrepreneurship were created by the Ministry of economic affairs. Compared to Anglo-Saxon markets, the Dutch VC market can be described as a ‘thin market’ for entrepreneurial finance (Bertoni et al., 2015; Lerner & Tag, 2013; Li & Zahra, 2012; Nightingale et al, 2009).

(14)

affairs without having to pay any interest. With this approach, it will stimulate the foundation of private seed funds by private parties (VCs, business angels and large corporations) in order to enlarge the market for venture capital in the Netherlands. Resulting in the market mechanism that will stay intact.

The Seed capital program (SCP) is an indirect subsidy for techno starters in innovative industries. The SCP is governed by the RVO (Government service for Entrepreneurs). Loans are given to funds that follow a tender procedure. In 2016 and 2017 there were also sector specific tenders for the agri-horti-food sector and eHealth sector. The application process is divided into three faces: idea, concept and application. Investment funds that want to participate in Seed funds will have to create a fund plan to qualify for the loan. Industry experts and an advisory board will judge the fund plan and potential grounds for rejection. The fund plans are judged on the basis of the following criteria:

- Extent to which the funds can rely on experience and expertise

- Extent to which the funds contribute to successful business via techno starters - Extent to which the funds is build up efficiently

The funds are also known as ‘closed end funds’. This means that the funds only invest in firms by the rules of the SCP. The loans are granted to the funds as a loan with a maximum of 6 million euros (since 2014). The fund has to contribute at least 50% of the investment from private capital. The seed fund invests venture capital in exchange for shares in the techno starter. From the revenues created by the fund, a part has to be refunded to the ministry and the remains to private parties that invested in the fund. The refund scheme depends on the results and exists of 3 separate periods.

- Period A: From the moment the fund creates revenue, 20% is refunded to the ministry and 80% is refunded to the private parties until the investment paid back itself.

- Period B: After the point that investors earned back their investment 50% of the revenues are refunded to the ministry and 50% to investors until the ministry also earned back their initial investment.

- Period C: After this point, 80% of the revenue streams flow into private parties and 20% into the ministry.

(15)

Furthermore, the investment period of the fund is 6 years maximum after which the funds have 6 years to alienate the participations. This means that 12 years after the fund started it is liquidated, in practice this is extended in some cases. The seed funds each have their own investment strategy and focus area. The industries of the funds are ICT, Life sciences & Health, energy, High-tech, logistics, agri-horti-food and/or a combination of these sectors (multi-sector). The funds are free to invest as long as it fits within the criteria of the SCP. Another important aspect of the seed funds besides providing venture capital is guiding the techno starters in which they invest.

3. METHODOLOGY

In the previous section, we explained how subsidies are used by governments in order to stimulate high-tech startups as well as the struggle high-tech startups are dealing with in order to close the financing gap and overcome the ‘valley of death’. This chapter explains the empirical models used to answer the research question as well as an explanation of the variables and operationalizations of the variables. An overview of all variables can be found in Table 3.

3.1. Empirical model

In order to give an answer to the research question, an empirical model is used in order to assess the influence of the Seed-capital program. We measure firm performance by the growth of employees and total assets by making use of the following difference-in-difference models:

"#$%&'($#)*+,'-*. = 01+ 034 + 0567 + 89:*+ ;:*

<='*>&>??#*+,'-*.= 01+ 034 + 0567 + 89:*+ ;:* Where @ = 1,……,n, & t = 1,…..n,

Growth of employees and total assets are widely used in academic research on firm performance and will be elaborated in section 3.2. The variable 034 is a dummy variable that equals 1 if the firm received treatment by receiving funding from the SCP program and 0 otherwise. It shows the additional effect of treatment compared to variable 0567. Variable 0567 is a dummy variable that equals 1 if the firm received VC investment in that specific

year. 8D:* include all control variables used. Finally, ;:* is the standard error term.

3.2. Dependent variables

(16)

Erasmus center for Entrepreneurship and Dialogic by request of the Ministry of Economic affairs. Direct results of the SCP include the access to venture capital for entrepreneurs, making it easier for investment funds to obtain financial resources, improving the market for Venture capital in the Netherlands (Bax et al., 2019). Since we focus on indirect results of the SCP, we’ll look at growth measures of the treated firms in the sample. We do this by looking at the growth of employment and growth of total assets as parameters of growth. These measures are often used in policy evaluation and assessing firm growth (Evans, 1987; Hall, 1987).

For both growth measures we took the logarithms as this allows us to transform the variables into normalized datasets which is a common method to handle skewed variables.1

Robustness analyses:

In performing robustness checks and further analyses, we used 3 additional dependent variables. We used tangible assets, intangible assets and solvency ratio % (Asset Based). These values are also provided by Orbis and in order to standardize these values, we use the log function for tangible & intangible assets. This is not possible for solvency ratio, since it is not possible to take a logarithm from negative values. Tangible assets are an organization’s equipment and infrastructure, both physical & technological. (Durnev et al., 2004; Calabrese et al., 2005). Bottazzi & Da Rin (2002) explained intangible assets as the stock of goodwill, patents, software & advertising. Solvency ratio is the ratio calculated by dividing the shareholders equity by total assets. Furthermore, to test the robustness of our findings we account for fixed time effects. A dummy variable was created for all years in which firms received investment from venture capitalists from 2010 – 2019 (Bell & Jones, 2014). We did not include this variable in the main analysis, since it showed signs of multicollinearity, but since difference-in-difference models require fixed time effects, we included it in the robustness analysis to show the effect on our dependent variables.

3.3. Independent variables

The independent variables used in the model are all dummy variables to measure the effect of treatment & venture capital investment. The first independent variable is 034 and denotes 1 if the firm received treatment in the form of SCP funding and 0 otherwise. The second variable

1 A total of 9 years has been used for the calculation of growth. We calculated the logs from 2011 – 2019. Year

(17)

is 0567 and equals 1 if the firm received VC investment and 0 otherwise. For the treated firms,

this means these values are equal, since treated firms received VC investment in the same year as their treatment.

3.4. Control variables

Additional to the independent variable, we include multiple control variables. We include dummy variables for region and industry and a continuous variable for firm age to control for cross-sectional differences between sectors, age and region. The reason for this is the large number of sectors, regions and age and relatively small sample.

SECTOR – First of all, sector data by BvD sector was obtained after which a dummy variable

was created. It includes 1 for manufacturing sector and 0 for services. The BvD sectors provided by Orbis are explained in table 2.

Table 2 – Industry BvD sectors

1. Agriculture, Horticulture & Food 15. Mining & Extraction 2. Banking, Insurance & Financial

Services

16. Miscellaneous Manufacturing 3. Biotechnology and Life Sciences 17. Printing & Publishing

4. Business Services 18. Property Services

5. Chemicals, Petroleum, Rubber & Plastic 19. Public Administration, Education, Health Social Services

6. Communications 20. Retail

7. Computer Hardware 21. Textiles & Clothing Manufacturing 8. Computer Software 22. Transport Manufacturing

9. Construction 23. Transport, Freight & Storage 10. Food & Tobacco Manufacturing 24. Travel, Personal & Leisure 11 Industrial, Electric & Electronic

Machinery

25. Utilities

12. Leather, Stone, Clay & Glass products 26. Waste Management & Treatment 13. Media & Broadcasting 27. Wholesale

14. Metals & Metal products

The sectors had to be decoded and combined into 1 for manufacturing a 0 for services, we combined sectors 2, 4, 6, 8, 13, 18, 19 & 24 and decoded it in 0 for services and all other sectors were decoded in 1 for manufacturing.

REGION – A region dummy was included with 1 for the Randstad and 0 for other provinces.

(18)

receive 1. Provinces Drenthe, Groningen, Zeeland, Noord-Brabant, Limburg, Zeeland, Gelderland, Friesland & Flevoland are the other regions and received 0. The reason for this is because the Randstad is the region with the highest production per hectare relative to its inhabitants (Oevering, 2019).

FIRM AGE – Age of the company measures in years at the time of foundation until 2019.

This data is obtained from the Orbis database. We used 2019 and not 2020, since the data was available until 2019.

3.5. Linear regression model

(19)

Table 3– Definition and source of variables

Variable Type of

Variable

Description Structure

Growth of employees DV Logarithm of the number of employees Scale Growth of Total

Assets

DV Logarithm of total assets calculated as follows

Scale Dummy SCP IV Equals 1 if firm received SCP funding

and 0 otherwise

Nominal Dummy VC IV Equals 1 if firm received VC investment

and 0 otherwise

Nominal Industry dummy CV Equals 1 for Manufacturing and 0 for

service industry

Nominal Firm Age CV Age of the firm measured by subtracting

year of foundation with the current year 2019.

Scale Region dummy CV Equals 1 for Randstad and 0 otherwise Nominal

Note: DV = dependent variable, IV = Independent variable, CV = Control variable

4. DATA & DESCRIPTIVE STATISTICS

4.1. Data Collection

The data used in this study are obtained from database Orbis by Bureau van Dijk (BvD). Orbis is a database that collects company data from over 360 million firms around the globe. Typically, Eikon is used for research on VC investments, however Eikon has more detailed information on the US-market and does not (fully) cover seed-capital investments in Western-Europa. Eikon collects data on round-by-round investments for both VC’s as well as the firms they invest in. In order to measure firm growth, Orbis was deemed as a better alternative, since it collects data on employment and asset growth over time, which are the dependent variables used to measure firm growth.

(20)

was imported into Orbis, resulting in data on the economic variables used in the analysis. We employ a matching procedure using firm names and chamber of commerce (CoC) numbers to get the dataset out of Orbis resulting in a dataset of 315 firms that received treatment.

4.2. Sample selection procedure

After the first sample was created, a match had to be made with a control group of firms that received VC investment, but not from the SCP scheme and matched the same criteria of the firms in the treatment group. The dataset provided by the NVP was also used to make the control group. First, only firms were selected that received VC investment in their early stage, seed stage or other early stage. After this, all the other criteria from the SCP are met.

These criteria are:

- The firm started less than 7 years ago with the sale of their products/services - The firm meets the definition of an SME as set by the Dutch ministry.

- The firm produces products, services or processes but its key tasks are not advisory related.

- The core business of the firm is about innovations or new ways to do existing business practices.

We only selected those firms that meet these criteria by eliminating firms from the dataset that: - were older than 7 years before the investment was made

- were labelled as advisor by BvD sector

- had more than 250 employees before receiving investment

This resulted in a control group of 522 firms that met the criteria and received investment in the early stage, seed stage or other early stage. After this step, the chamber of commerce numbers were matched with the Orbis database to get the firm specific data out of the program.

4.3. Selection Bias

(21)

tested. The propensity score method (PSM) is chosen to match the treatment and control group. The method was first introduced by Rosenbaum & Rubin (1983). The PSM is a probability score of receiving treatment based on the chosen covariates (Thoemmes, 2012):

I(D) = K(L = 1|X)

In this formula, P is the probability, Z=1 which is the treatment indicator with 0 for control & 1 for treatment. The ′|′ is a conditional sign and X is the set of the covariates used for matching (Thoemmes, 2012). The purpose of this procedure is to create a balanced sample of treated and untreated firms based on the propensity score, which makes sure the treatment effect cannot be biased by pre-test covariates. There is one major limitation of the PSM, because it does not control for the unobservables on the selection of firms that receive SCP support (Chemmanur, 2011). We matched the treatment group and control group based on region, industry & firm age. The fact that we only took these measures limits the effect of the PSM, however limited information was available, since Orbis only provides data from 2010 – 2019 on these variables2. In order to perform this matching procedure, we had to take care of the missing

values in the covariates used. According to Choi et al. (2018), there are multiple methods to handle missing values in the propensity analysis. These methods are a complete case analysis, missing indicator method, multiple imputation and a combination of the missing indicator method and multiple imputation method. The method we used is the complete case analysis, since it is believed to show valid causal treatment effects and even partially correct for unmeasured confounding factors (Choi et al., 2018). Furthermore, these variables were available for >95% of the sample.

After the observations with missing values were excluded from the sample, the PSM was performed. We used a five-step analytical approach towards PSM suggested by Thoemmes (2012). The first step is to determine the set of covariates that are important determinants of performance. In this paper we focus on the growth of number of employees and total assets, however it is important to look at firm characteristics in order to estimate the propensity score of treated and untreated firms. Therefore, we include three firm characteristics: firm age, region

2 Only Dutch investments are included. Also, only Dutch funds are included in the sample. The focus of the

(22)

and industry. The second step is estimating the propensity score by using logistic regression and choosing treatment as the outcome variable and the covariates as the predictors. The third step is the actual matching procedure, this can be done in multiple ways. According to Thoemmes and Kim (2011), nearest neighbor matching is the simplest way. This means that a single untreated firm is matched to a single treated firm. After this step, statistics of the sample are checked to make sure balance is created on the covariates. We checked whether the mean of the covariates was close to 0 and the variance ratio was close to 1. Finally, the treatment effect is estimated by making use of a linear regression model which is elaborated in section 3.5. This resulted in a control group of 315 firms and a treatment group of 315, totaling 630 firms that were analyzed in section 5. Table 4 presents the results of the PSM analysis. The means of both the treated & control group are not statistically different in the years before the treatment for number of employees. It shows a weak evidence for total assets (p = 0.065).

Table 4 – Matched sample comparison of Log of employment growth after treatment

Variable Statistic Treatment

group Control group Difference T-score Average Log growth of Employees Mean 1.00*** 0.89 0.11 4.84*** St. Dev. 0.48 0.44 0.04 - Observations 532 1112 580 - Average Log growth of employees years prior to treatment Mean 0.84 0.79 0.05 1.31 St. Dev. 0.38 0.38 0.00 - Observations 316 274 42 - Average Log growth of Total Assets Mean 5.83*** 5.72 0.11 2.877*** St. Dev. 0.80 0.86 0.06 - Observations 873 1039 166 - Average Log growth of total assets in years prior to treatment Mean 5.79* 5.61 0.18 1.84* St. Dev. 0.70 0.87 0.17 - Observations 501 505 4 - * Significant at .10 level ** Significant at .05 level *** Significant at .001 level 4.4. Descriptive statistics

(23)

assets, & firm age are higher in the treatment group. Furthermore, there is a large difference in the regions. From the treatment group, 61% of the firms are based in the Randstad region (Noord-Holland, Zuid-Holland & Utrecht) & 39% in all other provinces. For the treatment group this is the other way around, only 28% of the firms are based in the Randstad & 72% of the firms are located in other provinces. The firms in the control group seem more spread out over the country where this is not the case for the treated firms. Furthermore, the control group includes more manufacturing firms compared to the treatment groups (42% versus 33%).

Table 5 - Descriptive Statistics treatment group

Variable Mean Stand. Dev. Max. Min. N.

Growth of employees 0.949 0.459 2.41 0.301 852

Growth of Total Assets 5.789 0.771 8.76 1.26 1374

Dummy SCP3 0.640 0.479 1.000 0.000 2835

Dummy VC 0.640 0.479 1.000 0.000 2835

Industry dummy 0.330 0.470 1.000 0.000 2835

Firm Age 7.710 3.921 18.00 0.000 2835

Region dummy 0.610 0.487 1.000 0.000 2835

3 Dummy SCP & VC are equal here since VC investment for the treatment group is equal to SCP

investment by a VC fund.

4 SCP is not included, since the control group did not receive treatment / investment from a VC backed by

the SCP program.

Table 6 - Descriptive Statistics control group

Variable Mean Stand. Dev. Max. Min. N

Growth of employees 0.909 0.461 2.39 0.301 792

Growth of Total Assets 5.680 0.868 8.12 1.30 1544

Dummy SCP4 - - - - -

Dummy VC 0.690 0.464 1.000 0.000 2835

Industry dummy 0.420 0.494 1.000 0.000 2835

Firm Age 8.530 4.111 18.00 1.000 2835

(24)

5. ANALYSIS

In this chapter the results of the regression analysis are explained as well as the additional analysis to provide evidence of the robustness of our analysis. Before we analyze the results, we show the results of the Pearson correlation matrix to address the possible issue of multi-collinearity. As can be seen in Table 7, most coefficients are rather low if we take in mind the book of Hair et al., (2006) who explained that correlations between 0 - .20 are ‘none’; 0.21 – 0.35 are weak; 0.36 – 0.60 are moderate; 0.61 – 0.80 are strong and 0.81 – 1 are very strong. This also holds for negative coefficients. Since there are no issues with multicollinearity, we show the results of the linear regression models in tables 8 & 9 followed by an explanation of the models.

Table 7 – Correlation Matrix (Pearson)

(25)

Table 8: Linear regression of the effect of SCP treatment on growth of employment

Model 1 Model 2 Model 3

Variable B SE t B SE t B SE t Constant .950 0.31 31.107*** .905 .031 29.625*** 0.909 0.31 29.681*** Control variables Firm Age .001 .003 .178 -0.011 0.003 -3.365*** -0.011 0.003 -3.268*** Dummy Region -0.048 .023 -2.059** -0.037 -0.040 -1.644* -0.050 0.024 -2.090** Dummy Industry -0.007 0.024 -0.293 -0.003 0.023 -0.115 -0.002 0.023 -0.073 Independent variables VC Investment - - - 0.207 0.026 7.935*** 0.179 0.031 5.860*** SCP treatment - - - 0.050 0.029 1.698* F 1.440 16.863*** 14.083*** R Square 0.003 0.040 0.041 Adjusted R square 0.001 0.037 0.038

Note: B = Unstandardized Beta coefficient | SE = Standard error | t =T-score

(26)

Table 9: Linear regression of the effect of SCP treatment on growth of Total Assets

Model 1 Model 2 Model 3

Variable B SE t B SE t B SE t Constant 5.827 0.041 142.463*** 5.786 0.042 139.409*** 5.795 0.042 139.517*** Control variables Firm Age -0.008 0.004 -2.073** -0.019 0.004 -4.232*** -0.018 0.004 -4.138*** Dummy Region -0.058 0.031 -1.857* -0.054 0.031 -1.722* -0.081 0.032 -2.525** Dummy Industry 0.003 0.032 0.095 0.008 0.032 -0.253 0.012 0.032 0.389 Independent variables VC Investment - - - 0.188 0.036 5.200*** 0.126 0.041 3.103*** SCP treatment - - - 0.127 0.039 3.242*** F 2.407 8.581 8.990 R Square 0.002 0.012 0.015 Adjusted R square 0.001 0.010 0.014

Note: B = Unstandardized Beta coefficient | SE = Standard error | t=T-score

(27)

5.1. Regression results

Table 8 & 9 show the results of the linear regression model in which we look at the effect of the SCP program. For each regression analysis three separate models are shown to point out the additional value of the SCP program. The first model includes only the control variables (Region, Firm Age & Industry), in the second model we introduce the independent variable ‘VC Investment’. In the third model we include ‘SCP treatment’ as well to show the additional effect of treatment on the growth measures and whether this additional effect is significant or not. Table 8 shows the effect of the program and VC investment on growth of employment whereas table 9 shows the effect on growth of total assets. In order to assess the fit of the model, the F-value, R-square and adjusted R-Square were added to the output values within the tables. The R-Square & Adjusted R-square values in both tables show relatively low values, this indicates that there are (multiple) other variables influencing the effect on our growth measures. The results we find in table 8 concerning the growth of employment are consistent with the findings of Honjo & Harada (2006) who found a that firm age is an important predictor of growth, since young firms are more likely to grow harder than older firms. This is also consistent with the learning model of Jovanovic (1982). Although the additional effect is weak (p = 0.090), this can be partly explained by the relatively high level of missing values and wide variance of the variable.

(28)

the VC, which could lead to an additional effect of the program besides the regular VC’s. Another possible explanation of the additional effect is that we have mitigated the selection bias based by performing propensity score matching on the sample, but there could be a selection bias in the VCs selected by the government to provide management support and experience to the start-ups. One might expect that VCs with more experience and skills could improve the performance of start-ups better and might be able to take more risks once they are aware of the fact that they are outperforming other VCs in terms of performance and exits. The Ministry of Economic affairs selects the funds carefully based on their fund plans and experience, this could be the underlying reason for the additional effect.

5.1. Additional analyses

(29)

Besides looking at growth of employment, total assets, tangible assets & intangible assets, previous research on the effect of venture capital investment has also focused on other performance indicators (Kelly & Hankook, 2013; Paglia & Harjoto, 2014). One of these indicators is solvency ratio and indicates the financial healthiness of companies. The ratio is calculated by dividing the shareholders equity by total assets., if this ratio is low, companies have little equity compared to their liabilities which can cause financial distress and in some cases bankruptcy (Loi et al, 2012). Savaneviciene et al. (2015), found that the solvency ratio significantly increases after receiving venture capital investment, since venture capital affects portfolio companies’ growth by stabilizing its financial capability6. Therefore, we performed

another linear regression using Solvency Ratio % (Asset Based) as dependent variable to test the consistency of our findings. The results of the regression are illustrated in table 12. The results in the table shows there is a significant additional effect of SCP treatment besides the significant influence of venture capital investments on growth of solvency ratio. This result is in line with the findings of Savaneviciene et al. (2015). Although we found multicollinearity when we included fixed time effects, this is not unusual since we’re using dummy variables that highly fluctuate, therefore we also show the output of the regression results when fixed time effects where included in the model since this is a common procedure in performing difference-in-difference models to account for possible discrepancies in the year in which the firms received their investment. The results in table 13 & 14 show similar results regarding the effect of SCP treatment on growth of employment and total assets, providing more evidence that the SCP has an additional effect on firm growth.

6 The paper by Savaneviciene et al. (2015) used as case study with 4 cases in which they found a

(30)

Table 10: Linear regression of the effect of SCP treatment on growth Tangible Assets

Model 1 Model 2 Model 3

Variable B SE t B SE t B SE t

Constant 4.629 0.060 77.019*** 4.566 0.061 74.345*** 4.559 0.061 74.166***

Control variables

Firm Age -1.157E-5 0.006 -0.002 -0.013 0.007 -1.986** -0.014 0.007 -2.129**

Dummy Region -0.066 0.047 -1.396 -0.055 0.047 -1.170 -0.027 0.049 -0.558 Dummy Industry 0.187 0.049 3.848*** 0.192 0.048 3.969*** 0.189 0.048 3.920*** Independent variables VC Investment - - - 0.246 0.054 4.586*** 0.302 0.060 5.011*** SCP treatment - - - 0.122 0.060 2.030** F 6.492*** 10.174*** 8.976*** R Square 0.009 0.019 0.021 Adjusted R square 0.008 0.017 0.019

Note: B = Unstandardized Beta coefficient | SE = Standard error | t=T-score

(31)

Table 11: Linear regression of the effect of SCP treatment on growth of Intangible Assets

Model 1 Model 2 Model 3

Variable B SE t B SE t B SE t Constant 5.332 0.079 67.340*** 5.304 0.080 66.424*** 5.310 0.080 66.407*** Control variables Firm Age -0.010 0.008 -1.282 -0.021 0.009 -2.343** -0.021 0.009 -2.257** Dummy Region -0.150 0.063 -2.397** -0.153 0.063 -2.440** -0.174 0.065 -2.691*** Dummy Industry 0.008 0.063 0.127 0.013 0.063 0.212 0.013 0.063 0.206 Independent variables VC Investment - - - 0.178 0.072 2.474** 0.131 0.081 1.623 SCP treatment - - - 0.100 0.078 1.278 F 2.285** 3.252*** 2.930** R Square 0.006 0.012 0.013 Adjusted R square 0.004 0.008 0.009

Note: B = Unstandardized Beta coefficient | SE = Standard error | t=T-score

* Significant at .10 level ** Significant at .05 level *** Significant at .001 level

(32)

Table 12: Linear regression of the effect of SCP treatment on Solvency Ratio% (Asset Based)

Model 1 Model 2 Model 3

Variable B SE t B SE t B SE t Constant 25.533 2.644 9.658*** 25.731 2.710 9.494*** 26.243 2.710 9.683*** Control variables Firm Age -0.174 0.253 -0.688 -0.133 0.282 -0.473 -0.085 0.282 -0.300 Dummy Region 3.161 2.002 1.579 3.121 2.006 1.556 1.457 2.072 0.703 Dummy Industry -5.580 2.033 -2.744** -5.605 2.035 -2.754*** -5.454 2.032 -2.684*** Independent variables VC Investment - - - -0.768 2.306 -0.333 -4.641 2.616 -1.774* SCP treatment - - - - - - 7.887 2.528 3.120*** F 4.109*** 3.109** 4.443*** R Square 0.005 0.005 0.010 Adjusted R square 0.004 0.004 0.007

Note: B = Unstandardized Beta coefficient | SE = Standard error | t=T-score

(33)

Table 13: Linear regression of the effect of SCP treatment on Total Assets (Including fixed time effects)

Model 1 Model 2 Model 3

Variable B SE t B SE t B SE t Constant 6.087 0.085 71.585*** 5.891 0.096 61.549*** 5.893 0.096 61.647*** Control variables Firm Age -0.024 0.006 -4.045*** -0.024 0.006 -4.000*** -0.023 0.006 -3.879*** Dummy Region -0.059 0.031 -1.870* -0.061 0.031 -1.929* -0.086 0.033 -2.641*** Dummy Industry 0.012 0.032 0.386 0.014 0.032 0.446 0.018 0.032 0.553 YOF 2010 0.174 0.078 2.244** 0.175 0.077 2.265** 0.148 0.078 1.906* YOF 2011 -0.079 0.066 -1.209 -0.078 0.065 -1.195 -0.077 0.065 -1.184 YOF 2012 -0.322 0.075 -4.318*** -0.305 0.074 -4.106*** -0.304 0.074 -4.089*** YOF 2013 -0.135 0.072 -1.876* -0.094 0.073 -1.296 -0.097 0.072 -1.338 YOF 2014 -0.331 0.068 -4.833*** -0.278 0.069 -4.004*** -0.274 0.069 -3.951*** YOF 2015 -0.100 0.069 -1.440 -0.013 0.072 -0.187 -0.012 0.072 -0.163 YOF 2016 -0.140 0.067 -2.075** -0.024 0.072 -0.333 -0.018 0.072 -0.245 YOF 2017 -0.120 0.076 -1.581 0.023 0.083 0.283 0.030 0.082 0.369 YOF 2018 -0.399 0.083 -4.810*** -0.226 0.092 -2.466** -0.223 0.091 -2.442*** YOF 2019 -0.235 0.103 -2.289** -0.040 0.112 -3.56 -0.029 0.111 -0.262 IV VC Investment - - - 0.192 0.043 4.418*** 0.142 0.027 3.032*** SCP treatment - - - - - - 0.113 0.039 2.860*** F 5.913*** 6.919** 7.019*** R Square 0.026 0.032 0.035 Adjusted R square 0.021 0.028 0.030

(34)

Table 14: Linear regression of the effect of SCP treatment on Employment (Including fixed time effects)

Model 1 Model 2 Model 3

Variable B SE t B SE t B SE t Constant 1.232 0.062 20.001*** 1.021 0.069 14.814*** 1.022 0.069 14.838*** Control variables Firm Age -0.016 0.004 -3.773*** -0.016 0.004 -3.744*** -0.015 0.004 -3.624*** Dummy Region -0.038 0.023 -1.615 -0.037 0.023 -1.629 -0.051 0.024 -2.091** Dummy Industry 0.003 0.023 0.143 0.004 0.023 0.161 0.005 0.023 0.197*** YOF 2010 -0.026 0.053 -0.480 -0.025 0.053 -0.469 -0.033 0.053 -0.626 YOF 2011 -0.145 0.048 -3.005*** -0.144 0.048 -3.023 -0.141 0.048 -2.978*** YOF 2012 -0.162 0.054 -3.015*** -0.145 0.053 -2.738*** -0.144 0.053 -2.706*** YOF 2013 -0.274 0.057 -4.813*** -0.221 0.057 -3.884*** -0.221 0.057 -3.882*** YOF 2014 -0.164 0.051 -3.212*** -0.097 0.051 -1.877* -0.094 0.051 -1.832* YOF 2015 -0.147 0.051 -2.874*** -0.047 0.053 -0.880 -0.045 0.053 -0.860 YOF 2016 -0.212 0.049 -4.298*** -0.084 0.052 -1.606 -0.082 0.052 -1.557 YOF 2017 -0.130 0.055 -2.368** 0.027 0.059 0.455 0.031 0.059 -.524 YOF 2018 -0.402 0.059 -6.796*** -0.215 0.065 -3.297*** -0.214 0.065 -3.285*** YOF 2019 -0.330 0.077 -4.280*** -0.121 0.083 -1.461 -0.114 0.083 -1.375 IV VC Investment - - - 0.207 0.032 6.518*** 0.181 0.035 5.152*** SCP treatment - - - - - - 0.050 0.029 1.700* F 5.459*** 8.232*** 7.885*** R Square 0.042 0.066 0.068 Adjusted R square 0.034 0.058 0.059

(35)

6. DISCUSSION & CONCLUSION

The goal of this research is to estimate the effect of the seed capital program on firm growth. With this we aim to answer the research question:

What is the effect of the Seed Capital Program on the growth of firms that received funding? This program was specifically selected, since no earlier attempts were made to analyse the program by looking at economic variables that are deemed to indicate firm growth. In this study, we built on existing literature concerning governmental support schemes and venture These findings are important for policymakers, venture capitalists & start-ups. Previous empirical papers examined the effect of VC’s & governmental support programs on growth, but are still inconclusive.

This paper contributes to the existing literature in multiple ways. First of all, this paper focuses on numerical data gathered from multiple objective sources rather than basing the analysis on other qualitative methods. Second, to address the selection bias, we performed the propensity score method in which we showed that there is no significant difference in means between the treated and the control group before the investment. This showed us that these firms did not significantly performed better prior to the investment as they did after the investment. Furthermore, the impact of the Seed capital program is measured by making use of two variables, number of employees & total assets. We performed additional analysis to check whether the findings are consistent by using other measures of growth. Finally, this paper is a first attempt to evaluate the Seed Capital Program in terms of quantitative measures whereas previous evaluations of the program relied on qualitative measures and did not look at firm growth compared to investments by individual venture capitalists.

(36)
(37)

line with the findings of Savaneviciene et al. (2015) who showed that the solvency ratio increased after the involvement of venture capitalists and in our case even more when SCP treatment was added to the model. A possible explanation of this could be that venture capitalists are willing to take more risk and invest larger amounts than they would without the governmental support program, thus ultimately leading to an increase of the solvency ratio. Furthermore, including the fixed time effects in the regression model led to similar results for the outcome on both growth of employment and total assets.

(38)

7. LIMITATIONS & FUTURE RESEARCH

(39)

outcome to the Netherlands. Other countries and policymakers can use these findings with caution and take the SCP as an example in setting up effective governmental support programs. Lastly, the R-square and Adjusted R-square can be increased to increase the fit of the model by adding other variables and specifying variable such as sector into multiple sectors and region by multiple regions. A suggestion for future research is to look at the composition of the seed capital funds. There is a lot of diversity in terms of funds, some funds are founded by business angels that are somewhat informal investors whereas other funds are founded by professional institutions such as banks and airlines or venture capitalists that also handle large funds and participate in other governmental policies to follow-up on the seed stage. Especially in the early 2000’s, Dutch CVC was a relatively small which appears to have grown, partially because of the seed capital program (Vollaard, 2019). Furthermore, future research should look at the influence of multiple investors & investment rounds on the growth of these firms. Also, practitioners could examine the results based on the VC investors that invested in the specific firm to see whether certain funds outperform others in terms of growth.

(40)

7. REFERENCES

Adenfelt, M., Sanders, M. & Stavlöt, U. (2013). Mobilising Private Funds for the Transition to a Sustainable Economy, Brussels: European Liberal Forum. Available at:

http://fores.se/wp-content/uploads/2014/05/ELF_2013_final.pdf

Alemany, L. & Martí, J. (2005). “Unbiased Estimation of Economic Impact of Venture Capital Backed Firms”, EFA 2005 Moscow Meetings Paper.

Bonilla, A. Cancino, A., (2011). The Impact of the Seed Capital Program of SERCOTEC in Chile.

Baldock, R. (2016). An assessment of the business impacts of the UK’s Enterprise Capital Funds. Environment and Planning C: Government and Policy, 34(8), 1556–1581. Baum, J.A.C. and Silverman B.S. (2004). “Picking Winners or Building Them? Alliance,

Intellectual, and Human Capital As Selection Criteria in Venture Financing and Performance of Biotechnology Start-ups", Journal of Business Venturing, 19 (3), pp. 411-436.

Bax, M., Bongers, F., Groot Beumer, T., Janssen, M., Keizer, C., & Roosenboom, P. (2018). Evaluatie Seed capital regeling, Groeifaciliteit & Dutch Venture Initiative. Retrieved from: https://www.dialogic.nl/projecten/evaluatie-seed-capital-regeling groeifaciliteit-en-dutch-venture-initiative/

Bell, A., & Jones, K. (2015). Explaining Fixed Effects: Random Effects Modeling of Time Series Cross-Sectional and Panel Data. Political Science Research and Methods, 3(1), 133-153. doi:10.1017/psrm.2014.7

Bertoni, F., Colombo, M.G. & Quas, A. (2015). The patterns of venture capital investment in Europe. Small Business Economics 45, 543–560. https://doi:10.1007/s11187-0196620 Bertoni, F., Colombo, M.G., & Grilli, L. (2011). Venture capital financing and the growth of

high-tech start-ups: Disentangling treatment from selection effects. Research Policy, 40(7), 1028–1043. https://doi.org/10.1016/j.respol.2011.03.008 Binks, M., and C. Ennew. (1996). ‘‘Growing Firms and the Credit Constraint.’’ Small

Business Economics 8 (1): 17–25.

Bottazzi, L., & Da Rin, M. (2002). Venture capital in Europe and the financing of innovative companies. Economic policy, 17(34), 229-270.

Calabrese, A., Gastaldi, M., & Levialdi, N. (2005). Real options model to evaluate infrastructure flexibility: An application to photovoltaic technology, International Journal of Technology Management, 29(1-2), 173-191.

(41)

Carpenter, R. E., & Petersen, B. C. (2002b). Is the Growth of Small Firms Constrained by Internal Finance? Review of Economics and Statistics, 84(2), 298–309.

https://doi.org/10.1162/003465302317411541

Chandler, G. N., McKelvie, A., & Davidsson, P. (2009). Asset specificity and behavioral uncertainty as moderators of the sales growth - Employment growth relationship in emerging ventures. Journal of Business Venturing, 24(4), 373–387.

https://doi.org/10.1016/j.jbusvent.2008.04.002

Cosh, A., Fu, X., and Hughes A., (2012). Organization structure and innovation performance in different environments. Small Business Economics, 39(2):301– 317.

Chemmanur, T, Karthik K., and Debarshi N, (2007). “How does Venture Capital Financing Improve Efficiency in Private Firms? A Look beneath the Surface.” Working paper, Boston College and York University.

Choi, J., Dekkers, O.M., le Cessie, S. (2019). A comparison of different methods to handle missing data in the context of propensity score analysis. Eur. J. Epidemiol. 34, 23 36. https://doi.org/10.1007/s10654-018-0447-z

Cochrane, J. (2005). The risk and return of venture capital. Journal of Financial Economics 75, 3–52.

Colombo, M.G., Cumming DJ and Vismara M (2014). Governmental venture capital for innovative young firms. Journal of Technology Transfer. Available online at: http://papers.ssrn.com/sol3/papers.cfm?abstract_id1⁄42471469.

Cruickshank, D. (2000). Competition in UK Banking. London: HMSO.

Cumming, D.J. (2003). The structure, governance and performance of U.K. venture capital trusts. Journal of Corporate Law Studies 3, 401–427.

Cumming, D. (2007). Government policy towards entrepreneurial finance: innovation investment funds. J. Bus. Venturing 22 (2), 193–235.

https://doi.org/10.1016/j.jbusvent.2005.12.002.

Cumming, D.J., MacIntosh, J.G., (2003). A cross-country comparison of full and partial venture capital exits. Journal of Banking and Finance 27, 511–548.

Cumming, D.J., MacIntosh, J.G., in press. Crowding out private equity: Canadian evidence. Journal of Business Venturing.

Davila, A., Foster, G., & Gupta, M. (2003). Venture capital financing and the growth of startup firms. Journal of Business Venturing, 18(6), 689–708.

https://doi.org/10.1016/s0883-9026(02)00127-1

(42)

Durnev, A., Morck, R., & Yeung, B. (2004). Value-enhancing capital budgeting and firm specific stock return variation. The Journal of Finance, 59(1), 65-105.

Estay, C., Durrieu, F., & Akhter, M. (2013). Entrepreneurship: From motivation to start up. Journal of International Entrepreneurship, 11(3), 243–267.

https://doi.org/10.1007/s10843-013-0109-x

Evans, D. S. (1987). The relationship between firm growth, size, and age: Estimates for 100 manufacturing industries. The Journal of Industrial Economics 35: 567-581.

Geyer, A., Heimer, T., Holscher, L., & Schalast, C. (2010). Evaluierung des High-Tech Gründerfonds (1). Retrieved from http://www.technopolis-group.com/wp content/uploads/2014/05/htgf.pdf

Gilson, R.J., (2003). Engineering a venture capital market: lessons from the American experience. Stanford Law Review 55, 1067–1103.

Gompers, P.A., Lerner, J., (1999). The Venture Capital Cycle. MIT Press, Cambridge. Gompers, Paul, & Josh Lerner. (2001). "The Venture Capital Revolution." Journal of

Economic Perspectives, 15 (2): 145-168.

Gompers, P.A., Lerner, J., Blair, M.M., & Hellmann T., (1998). "What Drives Venture Capital Fundraising?" Brookings Papers on Economic Activity. Microeconomics, vol 1998.

Gregory, F., Nemet & Martina Kraus & Vera Zipperer, (2016). "The Valley of Death, the Technology Pork Barrel, and Public Support for Large Demonstration Projects," Discussion Papers of DIW Berlin 1601, DIW Berlin, German Institute for Economic Research.

Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, R. L. (1998). Multivariate data analysis (Vol. 5, No. 3, pp. 207-219). Upper Saddle River, NJ: Prentice hall. Hall, B. H. (1987). The relationship between firm size and firm growth in the US

manufacturing sector. Journal of Industrial Economics 35: 583-606.

Honjo, Y., & Harada, N. (2006). SME Policy, Financial Structure and Firm Growth: Evidence From Japan. Small Business Economics, 27(4–5), 289–300.

https://doi.org/10.1007/s11187-005-6703-0

Heirman, A., & Clarysse, B. (2007). Which Tangible and Intangible Assets Matter for Innovation Speed in Start-Ups? Journal of Product Innovation Management, 24(4), 303–315. https://doi.org/10.1111/j.1540-5885.2007.00253.x

Hirsch-Kreinsen, H., & Schwinge, I. (2014). Knowledge-Intensive Entrepreneurship in Low Tech Industries. Retrieved from

(43)

Jääskeläinen, M., Maula, M., & Murray, G. (2007). Profit distribution and compensation structures in publicly and privately funded hybrid venture capital funds. Research Policy, 36(7), 913–929. https://doi.org/10.1016/j.respol.2007.02.021

Jacobs, B., Theeuwes, J.J.M. (2004). Innovatie in Nederland: de markt draalt en de overheid faalt. In: ESB.

Jovanovic, B. (1982). Selection and the Evolution of Industry. Econometrica, 50(3), 649. https://doi.org/10.2307/1912606

Kang H. (2013). The prevention and handling of the missing data. Korean journal of anesthesiology, 64(5), 402–406. doi:10.4097/kjae.2013.64.5.402

Kalkuhl, M. & Fuss, M., & Nemet G., and Steckel, J, (2016). The role of political uncertainty for climate change mitigation. in preparation.

Kanniainen, V., Keuschnigg, C., (2004). Start-up investment with scarce venture capital support. Journal of Banking and Finance 28, 1935–1959.

Kelly, R., & Hankook, K. (2013). Venture Capital as a Catalyst for High Growth. http://www.ic.gc.ca/eic/site/eas-aes.nsf/eng/h_ra02218.html, Ottawa: Industry Canada.

Keuschnigg, C., (2003). Optimal public policy for venture capital backed innovation. CEPR Working Paper, vol. 3850.

Keuschnigg, C., Nielsen, S.B., (2001). Public policy for venture capital. International Tax and Public Finance 8, 557 – 572.

Kim, J. D. (2018). Is there a startup wage premium? Evidence from MIT graduates. Research Policy, 47(3), 637–649. https://doi.org/10.1016/j.respol.2018.01.010

Koninkrijk der Nederlanden. (2018). Staatscourant (24020). Retrieved from https://zoek.officielebekendmakingen.nl/stcrt-2018-24020.html

Lerner, J. (1999). “The Government as Venture Capitalist: The Long-run Effects of the SBIR Program.” Journal of Business. 72, pp. 285–318.

Lerner, J., (2002). When bureaucrats meet entrepreneurs: the design of effective public venture capital programs. Econ. J. 112 (477), F73–F84. https://doi.org/10.1111/ 1468-0297.00684.

Lerner, J., Moore, D., and Shepherd, S., (2005). A study of New Zealand’s venture capital market and implications for public policy. Report to the Ministry of Research Science and Technology, Auckland, LECG Ltd.

(44)

Li, Y. & Zahra, S.A., (2012). Formal institutions, culture, and venture capital activity: A cross country analysis. Journal of Business Venturing, 27(1), pp.95–111. Loi, T.H., & Khan, A.A. (2012). Determinants of firm growth: evidence from Belgian companies.

Manigart, S., Wright, M., Robbie, K., Desbrieres, P., De Waele, K., (1997). Venture capitalists’ appraisal of investment projects: An empirical European study. Entrepreneurship Theory Practice 21 (4), 29–44.

Martí, J., Menéndez-Requejo, S., & Rottke, O. M. (2013). The impact of venture capital on family businesses: Evidence from Spain. Journal of World Business, 48(3), 420–430. https://doi.org/10.1016/j.jwb.2012.07.025

Milewicz, J., & Herbig, P. (1994). Evaluating the Brand Extension Decision Using a Model of Reputation Building. Journal of Product & Brand Management, 3(1), 39–47. https://doi.org/10.1108/10610429410053077

Ming, K., & Rosenbaum, P., (2000). Substantial gains in bias reduction from matching with a variable number of controls. Biometrics, 56, 118-124.

Ministerie van economische zaken. (2016). De oogst van 10 jaar Seed capital 2005-2015 (1). Destination: https://www.rvo.nl/subsidies-regelingen/seed-capital

Muzyka, D., Birley, S., Leleux, B., (1996). Trade-offs in the investment decisions of European venture capitalists. J. Bus. Venturing 11 (4), 273–287.

https://doi.org/10.1016/0883 9026(95)00126-3.

Myers, S. C. (1977). Determinants of corporate borrowing. Journal of Financial Economics, 5(2), 147–175. https://doi.org/10.1016/0304-405x(77)90015-0 Nemet, G. F, Zipperer, V., & Kraus, M. (2018). The valley of death, the technology pork

barrel, and public support for large demonstration projects. Energy policy, 119, 154 167. doi: 10.1016/j.enpol.2018.04.008

Nightingale, P. (2009). From funding gaps to thin markets: UK Government support for early-stage venture capital, London: Nesta.

OECD (1997). Government Venture Capital for Technology-Based Firms, OCDE/GD (97) 201. Organization for Economic Cooperation and Development, Paris.

Oevering, F. (2019). Economische verschillen tussen Nederlandse provincies (1). Received from: https://economie.rabobank.com/publicaties/2019/mei/economische-verschillen tussen-nederlandse-provincies/#

(45)

Parker, S.C., (2002). Do banks ration credit to new enterprises? And should governments intervene? Scottish J. Polit. Econ. 49 (2), 162–195. https://doi.org/10.1111/1467 9485.00227.

Poterba, J.M., (1989). Capital gains tax policy toward entrepreneurship. National Tax Journal, 42(3), 375–389.

Preece, S.B., Miles, G. and Baetz, M.C. (1999). “Explaining the international intensity and global diversity of early stage technology-based firms”, Journal of Business

Venturing, Vol. 14 No. 3, pp. 259-281.

Rosenbaum, P. R., & Rubin, D. B. (1983). The central role of the propensity score in observational studies for causal effects. Biometrika, 70, 41-55.

Savaneviciene, A., Venckuviene, V., & Girdauskiene, L. (2015). Venture Capital a Catalyst for Start-Ups to Overcome the “Valley of Death”: Lithuanian Case. Procedia

Economics and Finance, 26, 1052–1059. https://doi.org/10.1016/s2212 5671(15)00929-6

Schumpeter, J.A. (1911). The Theory of Economic Development. Harvard University Press, Cambridge.

Technopartners. (2012). Eindevaluatie TechnoPartner (1). Retrieved from:

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=2a Thoemmes, F. (2012). Propensity score matching in SPSS. arXiv. 1201.

Thoemmes, F., & Kim, E. S. (2011). A Systematic Review of Propensity Score Methods in the Social Sciences. Multivariate Behavioral Research, 46, 90-118.

Tucker, J. & Lean J. (2003). Small firm finance and public policy. Small Business and Enterprise Development Journal, 10(1), 50–61.

Tweede Kamer der Staten-Generaal. (2018). Bedrijvenbeleid – Zelfstandig Ondernemerschap (318). Retrieved from

https://www.eerstekamer.nl/commissievergadering/20180710_iwo_en_ezk_lnv Vollaard, H. (2019). Corporate Venture Capital status in The Netherlands. Retrieved on 12th

of december 2019, retrieved from

https://rabofrontierventures.com/corporateventurecapitalstatus-in-the-netherlands/ Westhead, P., and Storey. D.J., (1997). ‘‘Financial Constraints on the Growth of High

Technology Small Firms in the United Kingdom.’’ Applied Financial Economics 7: 197– 201.

Wynant, L., Hatch, J., Shadbolt, J., Thibeault, A., (1991). Banks and Small Business

(46)

8. APPENDIX

Appendix A: Sample statistics

Table 1- Descriptive Statistics total sample

Variable N Minimum Maximum Mean Std. Deviation

(47)
(48)

Appendix B: Descriptive statistics treatment group Table 2: Descriptive statistics treatment group

Variable N Minimum Maximum Mean Std. Deviation

(49)

Total assets th EUR 2019 0 - - - - Total assets th EUR 2018 299 .11300000 0000000 578458.0000 0000000000 0 6056.018336 000000000 35729.5084300 00000000 Total assets th EUR 2017 469 .11300000 0000000 370908.0000 0000000000 3980.439340 000000000 19970.8940600 00002000 Total assets th EUR 2016 470 .02000000 0000000 125915.0000 0000000000 3273.578590 000000000 11873.7888100 00002000 Total assets th EUR 2015 417 .01800000 0000000 464715.0990 0000005000 4181.705575 999999000 25204.0397300 00000000 Total assets th EUR 2014 378 .14900000 0000000 72398.21600 0000000000 2686.497435 999999700 9114.07813299 9999000 Total assets th EUR 2013 339 .00100000 0000000 86321.33700 0000010000 1863.616301 000000000 5883.57285900 0000000 Total assets th EUR 2012 295 2.8500000 00000000 103244.3290 0000000000 1791.116450 999999900 6743.58261600 0000000 Total assets th EUR 2011 252 .69800000 0000000 111505.3630 0000001000 2195.057228 000000000 8159.98584799 9999000 Total assets th EUR 2010 220 .00100000 0000000 130744.4030 0000000000 2427.135755 000000000 9743.73859100 0000000 Predicted probability 630 .08195 .67794 .3830977 .16098615 Valid N (listwise) 0

Appendix C: Descriptive statistics control group

Variable N Minimum Maximum Mean

(50)

Referenties

GERELATEERDE DOCUMENTEN

In the subsample of negative total gains in Panel C, the correlation too is positive and significant, this is an unexpected result as we had assumed agency would be a dominant

The results show that the negative relation between leverage and investment, with and without growth opportunities, stay significantly the same for one measure of leverage

This Act, declares the state-aided school to be a juristic person, and that the governing body shall be constituted to manage and control the state-aided

This may suggest that the results from the previous research on the relationship between board gender diversity and the quality of earnings do not apply to a setting where gender

  Figure 5. Overview results questionnaire 

Eight deductive codes were developed: 'impact of program adaptability ', 'impact of program embeddedness', 'impact of co-workership role of the technostructure',

Do employees communicate more, does involvement influence their attitude concerning the cultural change and does training and the workplace design lead to more

After realizing these exits the whole fund should be liquidated within 12 years after the start of the fund (Bureau Bartels, 2011). A period of 12 years is expected to be