How Business Opportunities Constrain Young Technology-Based Firms from Growing into Medium-Sized Firms

How Business Opportunities Constrain Young Technology-Based Firms from

Rögnvaldur SaemundssonGrowing into Medium-Sized Firms Åsa Lindholm Dahlstrand

ABSTRACT. This paper analyses how the novelty of businessopportunities at start-up constrains young technology-basedfirms from attaining substantial growth and becomingmedium-sized. Data from 262 young Swedish technology-based firms are used to estimate a logit regression modelrelating different types of opportunities to the probability ofbecoming medium-sized. The results show that firms whichseek to exploit opportunities based on new market knowledgeare less likely to attain substantial growth than firms that seekto exploit opportunities based on existing market knowledge.The former class of firms can nevertheless increase theprobability of such growth by actively seeking externalfinancing.

Introduction

Many studies have examined barriers to growth inyoung and small firms and, generally, foundgrowth constraints to be of three types: lack ofmanagement competence or motivation, lack ofresources, and lack of business opportunities (e.g.Barber et al., 1989).

The purpose of this paper is to analyse how thenovelty of business opportunities at start-up con-strains young technology-based firms fromattaining substantial growth. Through deductiveanalysis of the effects of differences in novelty ofbusiness opportunities on the build-up of man-agement competence and on the need for externalfinancing, two sets of hypotheses are proposed.These hypotheses are empirically tested by usinga logit regression model.

Young technology-based firms (YTBFs) arerecently established firms where the novelty ofstart-up opportunities is likely to affect whetherthey attain substantial growth or not. YTBFs arelikely to be involved in activities of very diversenovelty. They take part in the creation and diffu-sion of new technology, the creation and earlydevelopment of new markets and industries as wellas the renewal of old ones (Rickne, 2000). Thedegree of novelty of the business opportunity atstart-up is likely to affect the obstacles that thefounders face when pursuing its profitableexploitation, and hence, to affect their chances ofsurvival and substantial growth.

Young technology-based firms are also likelyto be sensitive to start-up conditions regarding thescope of the business opportunities that theirfounders seek to exploit. This scope is highly con-strained by the founders’ technical expertise. Dueto these constraints, founders are unlikely to makefundamental changes in the scope of the opportu-nities that they seek, even if conditions turn outto be unfavourable (Oakey and Cooper, 1991).

The paper is structured as follows. First, theconstraining influences of differences in thenovelty of opportunities at start-up will be derived,resulting in two sets of hypotheses. Second, thesample used for estimating the logit regressionmodel is described, as are the variables used in themodel. Third, the results are presented, in termsof both the estimation of the model and analysisof complementary descriptive data. Finally, theresults are discussed and further research isoutlined.

Final version accepted on March 7, 2003

Industrial DynamicsChalmers University of Technology412 96 GöteborgSwedenE-mail: [email protected]@mot.chalmers.se

.5C© 2005 Springer

0Small Business Economics 24: 113–129, 2 0

Frame of reference and the development ofhypotheses

This section is organized into two parts. First, thenature of business opportunities in young tech-nology-based firms (YTBFs) and how theirnovelty can differ will be discussed. Second, itwill be analysed, using Penrose’s (1959) theory ofgrowth and results from recent research, how thesedifferences in novelty influence the likelihood ofbeing constrained from attaining substantialgrowth.

Business opportunities in young technology-based firms

At start-up the focus of activity in technology-based firms is on the exploitation of a businessopportunity that is based upon the technical exper-tise of the founders (Oakey and Cooper, 1991).This activity can lead to innovation, which insmall firms usually is in the form of an introduc-tion of a new product or service (Kirchhoff, 1994).

A business opportunity consists of productivepossibilities that the firms’ founders can identify,take advantage of, and are willing to act upon(Penrose, 1959). A business opportunity isbelieved to be profitable, but its future profitabilityis impossible to determine beforehand (Shackle,1955, p. 81). This uncertainty is exaggerated intechnology-based firms because of the uncertaintyinherent in any technological development thatmay be needed before entering the market with anew product or service (Garnsey, 1995).

The novelty of the business opportunity, and thesubsequent innovation, may differ greatly amongfirms. Some firms may introduce products andservices that are new to the world, while othersmay introduce refinements of existing ones, orsimilar products at a lower price.

The novelty of the business opportunity isclosely related to the novelty of the founders’knowledge base. As the opportunities thatfounders can identify are based on their priorknowledge (Penrose, 1959; Shane, 2000), novelbusiness opportunities cannot be identified withoutnovel knowledge. Similarly, it is difficult toidentify opportunities for refinement withoutknowledge of what is to be refined.

Two dimensions of the founders’ knowledge

base are important determinants of the novelty ofthe opportunities they seek to exploit: technicalknowledge and market knowledge (Abell, 1980;Autio and Lumme, 1998).

The technical knowledge that the opportunityis based on can be well known and common to alarge population of experts, e.g. a particular fieldof engineering. It can also be less well known, atleast to the population of experts that are servingparticular markets. This may be due to novelty ofthe knowledge as such, or due to geographical orindustrial differences in diffusion. When technicalknowledge is less well known, universities mayhave an important role for both producing anddiffusing new technical knowledge (Gibbons et al.,1994).

In the same way, the market knowledge that theopportunity is based on can be more or less wellknown. The founders, and others, are less likelyto have knowledge of markets and industries thatare emerging. On the other hand, previous workingexperience from within existing firms is an impor-tant source of knowledge of existing markets andindustries (Cooper, 1986).

Based on the above two dimensions of knowl-edge, one can analytically discern four types ofopportunities in young technology-based firms,each representing a different degree and type ofnovelty (Figure 1).

First, opportunities can be based on existingtechnical knowledge and existing market knowl-edge (ET-EM). This type of opportunity is likelyto lead to incremental innovations, if any innova-tions at all. Second, opportunities can be basedon new technical knowledge but existing marketknowledge (NT-EM). This type of opportunityleverages new technology to existing markets,

Rögnvaldur Saemundsson and Åsa Lindholm Dahlstrand

Figure 1. Types of business opportunity in young technology-based firms.

114

possibly adding new features and improving theperformance of previous products. Third, oppor-tunities can be based on existing technicalknowledge but new market knowledge (ET-NM).This type of opportunity uses existing technicalknowledge to address a new set of customerrequirements. It may be related to the emergenceof new industries or changes in existing ones.Fourth, opportunities can be based on new tech-nical knowledge and new market knowledge(NT-NM). This type of opportunity addresses anew set of customer requirements by using newtechnical knowledge, and may lead to radicalinnovation if successful.

So far, four different types of opportunitieshave been identified. The novelty of these typesranges from low, when opportunities are onlybased on existing knowledge, to high, when oppor-tunities are only based on new knowledge. Thenext step is to investigate how these different typesof opportunities at start-up may constrain youngtechnology-based firms from attaining substantialgrowth. A useful point of departure is Penrose’s(1959) theory of the growth of the firm. Herobjective was to analyse the limits to growth bycreating a theory of the process of firm growth.

Business opportunities and substantial growth

Firm growth is a process of development leadingto an increase in size (Penrose, 1959, p. 1). Thepattern of development, as well as the growth rate,can vary greatly across time periods for individualfirms and across firm size, age and industry affil-iation (Delmar et al., 2002; Audretsch, 1995).

Many studies of growth in young and smallfirms seek to explain differences in growth rates(Delmar, 1997). Instead, this study aims atexplaining what constrains new firms fromattaining substantial growth. To attain substantialgrowth is to make the transition from being a smallfounder-managed firm focused on innovation intobeing a professionally managed firm whereinternal pressures are exerted for further growth(Hofer and Charan, 1984; Garnsey, 1998).

Growth in any single time period, according toPenrose (1959), is made possible by opportuni-ties identified by the firms’ entrepreneurs. In orderto accomplish growth, the firms’ managers needto be willing to act upon these opportunities, to

plan for their exploitation, and to obtain thenecessary resources to carry out the plans. Aftera period of expansion, increased resources andknowledge will provide incentives, as well asopportunities, for further expansion. The devel-opment of knowledge is cumulative and newopportunities for growth are therefore likely to berelated to previous opportunities.

Based on assumptions about the general avail-ability of resources and of profitable opportuni-ties, Penrose (1959, pp. 43–49) argued that thelimit to how fast a firm can grow in any timeperiod is set by the capacity of the firm’s man-agerial resources. She argued that new manage-ment capacity could only be developed within thefirm by using current managerial resources. Onlyby being a member of the management team couldnew managers learn what is needed for creatingand executing expansion plans for a particularfirm. Hence, the capacity of the current manage-ment team will constrain both the degree of expan-sion possible during the current period and theincrease in new management capacity during thenext period.

The capacity of management in each firm isdependent on the amount of managerial resourcesand the knowledge that these human resourcespossess. This collection of knowledge is the foun-dation for identifying business opportunities andpursuing them in a profitable manner leading togrowth (Penrose, 1959; Shane, 2000).

At start-up, managerial resources tend to bescarce and limited to the founding team. The man-agement capacity in the new firm is therefore builtupon the knowledge of the founders – the sameknowledge that is the foundation of the opportu-nity they have selected to exploit. As the noveltyof that knowledge is likely to affect the difficultyof finding a way to pursue the opportunity in aprofitable manner, the type of opportunity is likelyto affect the time it takes to build up the manage-ment capacity that is needed for growth. The moretime it takes to build up the management capacity,the more growth-constrained the firm becomes,and the less likely to attain substantial growth.

Various researchers have observed how dif-ferent “lead times” may affect the growth ofYTBFs. Oakey and Cooper (1991) observe thatfirms in different high-tech industries, such asbiotechnology and software, differ greatly in how

How Business Opportunities Constrain Young Technology-Based Firms 115

long it takes them to get a new product to themarket, and these differences affect the firms’ability to attain fast growth. Roberts (1991) alsoobserves that success in MIT spin-offs is relatedto the amount of technology transfer from earlieremployers. He argues that the technology is moremature and will take less time to develop intoproducts when the technology transfer is high.

Other researchers have used business similarityof parent organization (previous employer) as ameasure of lead times. Feeser and Willard (1990)found that new technology-based firms whichobtained high growth were similar to their parentsin both the technology utilized and the marketsserved. Chandler (1996) also found that task envi-ronment similarity (same customers, industry,competitors, and technology) between the parentorganization and the new venture positively mod-erated the relationship between pre-ownershipexperience and sales growth.

Research on small and young technology-basedfirms tends to emphasize the uncertainty relatedto development of new technical knowledge (e.g.Garnsey, 1995; Oakey and Cooper, 1991; Roberts,1991). But this uncertainty is likely to be furtherexaggerated by uncertainty related to the devel-opment of new markets and industries. In Klepperand Graddy’s (1990) study, the average time fromthe creation of a new industry to its stability was29 years. It takes time to spread knowledge abouta new set of customer requirements and to gainlegitimacy for their relevance (Aldrich and Fiol,1994). Additionally, firms may be dependent onthe development of complementary technologiesand institutions, which takes time and involves anumber of external actors of different types(Nelson, 1994; Rickne, 2000; Lindmark, 2002).

There is some evidence that ‘lead times’ arelonger when opportunities are based on newmarket knowledge, as compared to opportunitiesbased on new technical knowledge. Eisenhardt andShoonhoven (1990) found that semiconductorfirms were less likely to grow in emerging marketsas compared to existing markets, but found norelationship between growth and the newness oftechnology. Autio and Lumme (1998) also foundin their sample of Finnish new technology-basedfirms that firms entering new markets withexisting technology had obtained significantly lesssales growth than firms entering existing markets

with new technology. Autio and Lumme’s resultsare not very definite as they found no significantdifferences in employment growth between thesetwo types of firms, and did not control for othervariables that may affect growth, such as age.

The above reasoning leads to the conclusionthat young technology-based firms seeking oppor-tunities based only on existing knowledge are lesslikely to be growth-constrained than firms seekingopportunities based on new knowledge. Addi-tionally, firms seeking to exploit opportunitiesbased on new market knowledge and existingtechnical knowledge are more likely to be growth-constrained than firms seeking to exploit oppor-tunities based on new technical knowledge andexisting market knowledge. Finally, firms seekingto exploit opportunities based on both new marketknowledge and new technical knowledge are themost likely to be growth-constrained. Theseconclusions lead to the first set of hypotheses:

Hypothesis 1: The novelty of business oppor-tunities at start-up will affect the likelihoodthat an YTBF will attain substantial growth.

Hypothesis 1a: YTBFs seeking to exploitopportunities based on new knowledge areless likely to attain substantial growth thanYTBFs seeking to exploit opportunitiesbased on existing knowledge.

Hypothesis 1b: YTBFs seeking to exploitopportunities based on new market knowl-edge and existing technical knowledge areless likely to attain substantial growth thanYTBFs seeking to exploit opportunitiesbased on new technical knowledge andexisting market knowledge.

Hypothesis 1c: YTBFs seeking to exploitopportunities based on new market knowl-edge and new technical knowledge are lesslikely to attain substantial growth than otherYTBFs.

The time it takes to pursue an opportunity in aprofitable manner is highly related to the amountand the quality of available resources. The longertime it takes before firms become profitable, themore firms have to rely on external sources ofresources, especially financial resources. Hence,firms seeking to exploit opportunities that have

Rögnvaldur Saemundsson and Åsa Lindholm Dahlstrand116

long “lead times” may also face financial limita-tions that may further constrain them fromattaining substantial growth. The degree of theselimitations is determined by the difficulty inobtaining external financing and how sensitivefirms are to such difficulties. For example, firmsthat need to finance extensive technologicaldevelopment before generating any revenues aremore growth constrained by difficulties in gettingaccess to financing than are firms that are able toquickly generate revenues.

In young technology-based firms, differencesin opportunities are likely to influence both thedifficulty in obtaining external financing, and howgrowth constraining this difficulty is for the firm.Previous research has found that YTBFs areespecially prone to financial constraints (Garnsey,1995; Murray and Lott, 1995; Lockett et al., 2002)and that the degree of those constraints is relatedto their technical sophistication (Westhead andStorey, 1997). The main argument is that theconstraints experienced by YTBFs are due to thelack of willingness, competence, or both, ofexisting financial markets to provide financialcapital to firms facing technological uncertainty.This results in a failure of borrowers and lendersto reach an agreement on the price and conditionsfor financial support.

If the knowledge on which firms base theiropportunity is widely known, it may be easier toget access to financial resources on favourableterms. Financiers may be more confident in esti-mating risks when opportunities are based onexisting knowledge, and founders may be moreproficient in convincing the financier about theviability of the opportunity. The opposite is truefor opportunities based on new knowledge, espe-cially new market knowledge. When markets areemerging, or do not even exist, it is likely to bemore difficult to convince financiers about theviability of an opportunity, as compared to oppor-tunities were the size of the potential market mightbe known.

The type of opportunity is unlikely to be theonly factor explaining the difficulties facing theyoung technology-based firms when trying toaccess external sources of financing. There maybe individual differences in the ability and will-ingness to seek external financing (Penrose, 1959),and environments differ in their capacity of

providing financial support (Aldrich, 1979). Forwhatever reasons firms have difficulties inaccessing external financing, the type of opportu-nity is likely to influence how growth constrainingthese difficulties are. Firms that seek to exploitopportunities based on new knowledge are likelyto be more dependent on external financing thanfirms seeking to exploit opportunities based onexisting knowledge. It is likely to take more timefor them to generate own revenues and to obtainthe management competence needed to attainsubstantial growth. As above, one can expect firmsthat seek to exploit opportunities based on newmarket knowledge to be more dependent onexternal financing than firms depending on newtechnical knowledge.

Thus, it can be argued that firms seekingopportunities based only on existing knowledgeare less likely to be growth-constrained due todifficulties in getting access to external financingthan firms seeking opportunities based on newknowledge. Additionally, firms seeking to exploitopportunities based on new market knowledge andexisting technical knowledge are more likely to begrowth-constrained due to such difficulties thanfirms seeking to exploit opportunities based onnew technical knowledge and existing marketknowledge. Finally, firms seeking to exploitopportunities based on both new market knowl-edge and new technical knowledge are the mostlikely to be growth-constrained due to such diffi-culties. These conclusions lead to the second setof hypotheses:

Hypothesis 2: The novelty of business oppor-tunities at start-up will moderate how diffi-culties in getting access to external financinginfluence the likelihood that a YTBF willattain substantial growth.

Hypothesis 2a: Difficulties in getting access toexternal financing are more likely to con-strain YTBFs seeking to exploit opportuni-ties based on new knowledge from attainingsubstantial growth than YTBFs seeking toexploit opportunities based on existingknowledge.

Hypothesis 2b: Difficulties in getting access toexternal financing are more likely to con-strain YTBFs seeking to exploit opportuni-


ties based on new market knowledge andexisting technical knowledge than YTBFsseeking to exploit opportunities based onnew technical knowledge and existingmarket knowledge.

Hypothesis 2c: Difficulties in getting access tofinancing are more likely to constrainYTBFs seeking to exploit opportunitiesbased on both new market knowledge andnew technical knowledge from attaining sub-stantial growth than other YTBFs.

Figure 2 summarizes the proposed relationshipsbetween differences in the type of opportunity, dif-ficulty in getting access to external financing, andthe probability of attaining substantial growth.

The type of opportunity is hypothesized toinfluence the probability that firms attain sub-stantial growth in two ways. First, differences inthe type of opportunity is hypothesized to influ-ence this probability directly due to differencesin the time it takes to build up sufficient manage-ment capacity to exploit the opportunity in aprofitable manner (H1). Second, the type of oppor-tunity is hypothesized to influence the probabilityof attaining substantial growth indirectly throughits moderating influence on the effects of diffi-culties in getting access to external financing (H2).

It was also argued that the type of opportuni-ties would influence the difficulty in getting accessto financing. Hence, it could be expected thatdifferences in the type of opportunities could alsoinfluence the probability that firms attain sub-

stantial growth through influences on these diffi-culties. Do reduce the complexity of the logitregression model this relationship was notincluded in the model. Instead, complementarydata on external financing of the firms in sampleis analysed to investigate the assumptions madeabout differences in the need for external financingand difficulties in obtaining such financing.

In the next section the methods and sample usedto test the two sets of hypotheses are described.

Method and sample

To test the hypotheses, a logit regression modelis estimated for a sample of YTBFs1 in Sweden.In this section the identification of the sample isdescribed, as are the variables and measuresincluded in the model.

Sample

Based on information from the Swedish Bureau ofStatistics (SCB) all firms fulfilling the followingcriteria were included in a population of SwedishYTBFs:

• The firms should be founded between 1975 and1993.

• In 1993 the firms should have at least threeemployees, of whom at least one should havea university degree in engineering, naturalscience or medicine.

• The firms should not have been established as


Figure 2. Summary of the proposed relationships between differences in the type of opportunities, difficulties in getting accessto external financing, and the probability of attaining substantial growth.

118

a foreign direct investment or as a diversifica-tion from a larger firm.2

• The firms should belong to industries, bothmanufacturing and industry-related services,that can be termed knowledge-intensive.3

In 1998 a postal questionnaire was sent to allsurviving firms (1,190) belonging to the popula-tion. After a single reminder, 400 firms hadresponded to the survey, of which 344 hadanswered the questionnaire. Additionally 20 firmsanswered it after a telephone reminder. Theseanswers were used to test response bias. Since nosignificant difference was found between respon-dents and non-respondents regarding age and sizedistribution, the 20 firms were added to thesample. Thus, the total sample of firms havinganswered the postal questionnaire consists of 364firms.

Of these 364 firms, only 262 provided completeanswers that could be used for model estimation,while 102 answers had missing values for at leastone of the variables used in the model. Since thesize and age distribution of the 102 firms havingmissing values did not differ significantly from thedistribution of the total sample of 364, it was con-sidered unlikely that removing the answers withthe missing values from the sample would bias theresults. Hence, the final sample of YTBFs used formodel estimation consists of 262 firms, or 22%of the identified population. Of these, 36 weremedium-sized or larger.4

Variables and measures

Substantial growth. The hypotheses proposed inthe previous section are concerned with how dif-ferences in the types of business opportunities atstart-up may constrain young technology-basedfirms from obtaining substantial growth. To attainsubstantial growth is to make the transition frombeing a small founder-managed firm focused oninnovation into being a professionally managedfirm where internal pressures are exerted forfurther growth (Hofer and Charan, 1984; Garnsey,1998).

This transition is related to changes in organi-zational structure, which are related to the numberof employees. The number of employees is there-fore an appropriate measure for estimating when

firms have attained substantial growth. While itis impossible to determine the exact point oftransition that fits all firms, the EU definition ofmedium-sized firms (50–249 employees) was usedin this study. Firms that had become medium-sizedor larger were therefore considered to haveattained substantial growth.

The dependent variable of substantial growth inthe model is therefore a categorical variablehaving the value of 1 for those firms having 50employees or more, and 0 for those having lessthan 50 employees.

Difficulties in getting access to external financing.In the postal questionnaire, the respondents wereasked to report the difficulty they had experiencedin obtaining external financing, both at start-upand later. Difficulty was reported on a scale from1 to 5, ranging from very easy to very difficult,for different financial sources. The differentsources included self-financing, business angels,venture capital, various sorts of loans, governmentsupport, and “other” sources of financing.

The variable used in the model was constructedby calculating a single average from all thereported external sources, both at start-up andlater. For those firms which reported that theywere wholly self-financed, which can be consid-ered as the lowest degree of difficulty in gettingaccess to external financing, the variable was setto zero.

Opportunities. Two measures were used toquantify the newness of market and technicalknowledge.

First, if the new firm had any business relationsat start-up with former employers, the businessopportunity was considered to be based onexisting market knowledge. If there were nobusiness relations at start-up, or if the founders didnot have any previous employment, the businessopportunity was considered to be based on newmarket knowledge. Founders coming directly fromthe university belong to the latter category.

This measure of market knowledge captureswell whether opportunities are based on knowl-edge of existing markets and industries. Themeasure may nevertheless overestimate thenewness of market knowledge in some of the caseswhere founders had no business relationships with


previous employers, or had no previous employer.It is therefore possible that the results are biasedtowards less difference between new and existingknowledge.

Second, if the respondent reported that univer-sities or research institutes had been important atstart-up, the business opportunity was consideredto be based on new technical knowledge. If therespondent did not report this, the businessopportunity was considered to be based onexisting technical knowledge.

The measure of technical knowledge captureswell those cases where the technical knowledgeis new due to strong links to science. The measuremay nevertheless underestimate the newness oftechnical knowledge in some of the cases wherethere is no relationship with universities orresearch institutes at start-up. Firms, especiallylarge firms, may develop new technology that canbe used by independent spin-offs without any linkswith universities or research institutes at start-up.It is therefore possible that the results are biasedtowards less difference between new and existingknowledge.

Based on the above two measures of marketknowledge and technical knowledge at start-up,the four categories of opportunities were created:

1) ET-EM: Existing technical knowledge andexisting market knowledge

2) NT-EM: New technical knowledge andexisting market knowledge

3) ET-NM: Existing technical knowledge and newmarket knowledge

4) NT-NM: New technical knowledge and newmarket knowledge

Control variables. Three variables were includedin the model as control variables: growth inten-tions, age and industry.

The lack of growth intentions has been foundto constrain the growth of many small firms(Davidsson, 1989), including technology-basedfirms (Oakey, 1994). Growth intentions weremeasured as self-reported intentions about futuregrowth. The growth intentions were reported on ascale of 1–4, ranging from reducing the number ofemployees to greatly increasing the number ofemployees.

As the process of growth takes time, older firms

should be more likely to have obtained substantialgrowth, all other things equal. Age in years istherefore included as a control variable.

Previous research has shown that growth ratesof new firms vary between industries. In his studyof manufacturing industries, Audretsch (1995)found relationships between growth rates of firmson one hand, and industry innovation rate, industrygrowth rate, and minimum efficient scale in anindustry, on the other. Differences in innovationrates were not controlled for here, as the wholesample in this study belongs to industries withrelatively high rates of innovation. A controlvariable measuring differences in industry growthin Sweden at the two-digit level was constructedand tested, but was found to have no effect, eitheron the coefficients in the model or on the modelfit. It was therefore not included in the finalmodel. A variable that classifies firms asbelonging to either manufacturing industries orindustry-related services was used to control fordifferences in minimum efficient scale betweenthese two types of activities. This variable has thevalue of one for service firms, and the value ofzero for manufacturing firms.

Descriptive statistics and correlation. The descrip-tive statistics and correlations for the non-cate-gorical independent variables are presented inTable I. There is little correlation between thevariables, reducing the risk of disturbing effectsdue to multi-colinearity.

In Table II the descriptive statistics and fre-quencies for the independent variables in the logitmodel are displayed for YTBFs that have attainedsubstantial growth (growth firms) and YTBFs thathave not (constrained firms).

Analysis and model estimation

The four categories of opportunities were includedin the model by using three dummy variables. Theinteraction effects between types of opportunitiesand difficulty in getting access to externalfinancing (the opportunity and difficulty variables)were modelled by adding three variables, each amultiplication of the difficulty variable and thecorresponding opportunity variable. In effect, alogit model including the difficulty variable andthe control variables was simultaneously estimated


for the four different types of opportunities,allowing for differences between groups in boththe coefficient of difficulty variable and theconstant term. In that way it could be testedwhether the relationship between difficulties inaccessing external financing and attaining sub-stantial growth is different depending on the typeof opportunity, and whether there is a direct rela-tionship between the type of opportunity andattaining substantial growth.

A logit regression model with eleven parame-ters was estimated by using the STATA softwarepackage. First, a baseline model, including onlythe control variables, was estimated for compar-ison with the full model. Second, three configu-rations of the full model were estimated, eachincluding different sets of opportunity variables.The presentation of three different configurationsmakes the comparison between the relative effectsof different opportunities easier. Each configura-

tion represents the same model, but the opportu-nity that is the point of reference when interpretingthe coefficients is different in each configuration.

In addition to estimating the model, comple-mentary data on external financing in the selectedsample were analysed. The purpose of this addi-tional analysis was to increase the robustness ofthe conclusions of the study. Investigating theassumptions made about external financing whenderiving the hypotheses provides a better under-standing of the results from the model.

A limitation of the study that is important tokeep in mind when interpreting the results is thesurvivor bias inherent in the data. Because the dataonly include survivors, they can only answerquestions about the characteristics of the survivingfirms, not questions related to the characteristicsof all firms belonging to the original population.Hence, we can only make inferences from our dataabout what constrains the surviving firms from

How Business Opportunities Constrain Young Technology-Based Firms

TABLE IThe descriptive statistics and correlations for the non-categorical independent variables

Variable Descriptive statistics Correlations

Mean S.D. 1. 2. 3.

1. Difficulties in accessing external financing 02.24 1.44 01.002. Growth intentions 03.12 0.71 00.10 01.003. Age 11.66 4.85 –0.07 –0.12 1.00

TABLE IIDescriptive statistics (means and SDs) and frequencies of the variables in the logit model for YTBFs that have attained sub-

stantial growth (growth firms) and those YTBFs that have not attained substantial growth (constrained firms)

Variables Constrained firms Growth firmsN = 224 N = 36

Difficulties in accessing ext. financing 02.2 (1.4) 02.4 (1.5)Growth intentions 03.1 (0.70) 03.5 (0.69)Age 11.3 (4.0) 13.3 (8.0)

IndustryManufacturing 25% 42%Service 75% 58%

OpportunitiesET-EM 50% 50%NT-EM 17% 21%ET-NM 15% 18%NT-NM 18% 11%

ET = Existing Technical Knowledge, EM = Existing Market Knowledge, NT = New Technical Knowledge, NM = New MarketKnowledge.

121

attaining substantial growth, not what constrainsfirms in general from doing so.

Empirical results

The empirical results will be presented in twoparts. The first part will present the estimatedmodel. The second part will present complemen-tary data on differences in external financing foryoung technology-based firms seeking differenttypes of opportunities.

Estimation of logit regression model

The estimated logit regression model is presentedin Table III, including both the baseline model andthe three configurations of the full model (ModelsA, B, and C).

The baseline model including age, industryaffiliation (manufacturing or service) and growth

intentions is highly significant, and age andgrowth intentions have significant influence on theprobability of becoming medium-sized. This isconsistent with previous research which predictsthat younger firms, and firms with lower growthorientation, are less likely to have becomemedium-sized than older firms and firms withhigher growth orientation.

The addition of the investigated variablesdoubles the explanatory power of the model, sup-porting the idea that the investigated variableshave a significant influence on the likelihood ofbecoming medium-sized. In the full model theindustry control variable is significant, indicatingthat service firms are less likely than manufac-turing firms to grow into medium-sized firms. Thisis consistent with previous research which predictsthat the minimum efficient scale is higher in man-ufacturing industries than in service industries.


TABLE IIIThe estimated logit regression model. The baseline model is a model including only the control variables. Models A, B, and Care the same full model with different dummy variables left out of the model to facilitate comparisons between different types

of opportunities

Variables Baseline model Model A Model B Model C

Intercept 0–6.52*** 0–6.36*** 0–6.36*** 0–6.36***Age (A) 000.103** 000.104*** 000.104*** 000104***Industry (I) 0–0.703 0–1.04** 0–1.04** 0–1.04**Growth intentions (GI) 001.20*** 001.43*** 001.43*** 001.43***Difficulty in access to financing (DAF) 0–0.330 0–0.163 001.08*ET–EM (O1) 00– 0–0.0491 003.90*NT–EM (O2) 000.0491 00– 003.95*ET–NM (O3) 0–3.90** 0–3.95* 00–NT–NM (O4) 0–7.19** 0–7.24* 0–3.29O1*DAF 00– 00–0.167 0–1.42*O2*DAF 000.167 00– 0–1.25*O3*DAF 001.42** 001.24* 00–O4*DAF 002.07** 001.90* 000.655

Log likelihood –95.3 –85.7 –85.7 –85.7Model χ2 026.3*** 045.50*** 045.50*** 045.50***Pseudo R2 000.121 000.210 000.210 000.210

N = 262, * = p < 0.05, ** = p < 0.01, *** = p < 0.001. H1 (using Model A): O2 = O3 = O 4 = 0, p < 0.05.H2 (using Model A): O2*FC = O3*FC = O4*FC = 0, p < 0.05.O1: Opportunities based on existing technical knowledge and existing market knowledge (ET-EM).O2: Opportunities based on new technical knowledge and existing market knowledge (NT-EM).O3: Opportunities based on existing technical knowledge and new market knowledge (ET-NM).O4: Opportunities based on new technical knowledge and new market knowledge (NT-NM).

122

Hypothesis 1: Direct influence of differenttypes of opportunities

Using post-estimation statistical tests, it was foundthat the coefficients of the dummy variablesrepresenting the different types of opportunities(O2-O4 in Model A) were, as a group, signifi-cantly different from zero (p < 0.05). This supportsHypothesis 1, which stated that differences in thetype of opportunities at start-up influence theprobability of attaining substantial growth.

According to Hypothesis 1a, firms seeking toexploit opportunities based on existing technicalknowledge and existing market knowledge (O1)should be less constrained than firms seeking othertypes of opportunities (O2-O4). This is supportedfor O3 and O4, i.e. opportunities that are basedon new market knowledge, but not for O2, whichis based on new technical knowledge and existingmarket knowledge.

According to Hypothesis 1b, firms seeking toexploit opportunities based on new technicalknowledge and existing market knowledge (O2)should be more likely to attain substantial growththan firms seeking to exploit opportunities basedon existing technical knowledge and new marketknowledge (O3). The model supports this hypoth-esis.

According to Hypothesis 1c, firms seeking toexploit opportunities based on new technicalknowledge and new market knowledge (O4)should be less likely to attain substantial growththan firms seeking other opportunities (O1-O3).This is only partially supported by the model.Firms seeking to exploit O4 were found to besignificantly less likely to become medium-sizedthan firms seeking to exploit existing knowledge(O1) or new technical knowledge and existingmarket knowledge (O2) as predicted, but were notfound to be significantly less likely to becomemedium-sized than firms seeking to exploitopportunities based on existing technical knowl-edge and new market knowledge (O3).

Hypothesis 2: Indirect influence of differenttypes of opportunities

Using post-estimation statistical tests, it was foundthat the coefficients of the variables representingthe interaction between types of opportunities and

difficulty in accessing external financing(O2*DAF, O3*DAF, and O3*DAF in Model A)were, as a group, significantly different from zero(p < 0.05). This supports Hypothesis 2, whichstated that differences in the type of opportunitiesat start-up moderate how the difficulty in gettingaccess to external financing influences the proba-bility of attaining substantial growth.

According to Hypothesis 2a, firms having dif-ficulties in getting access to external financeshould be more likely to attain substantial growthif they seek to exploit opportunities based onexisting knowledge (O1) compared to other typesof opportunities (O2-O4). This hypothesis is notsupported. There is no difference found betweenfirms seeking to exploit O1 and firms seeking toexploit opportunities based on new technicalknowledge and existing market knowledge (O2).For both types of opportunities, the degree ofdifficulties in getting access to external financinghas no significant effects on the probability ofbecoming medium-sized, even if it has a negativesign (see the coefficients for DAF in models A andB). There is a significant difference between firmsseeking to exploit O1 and firms seeking to exploitopportunities based on existing technical knowl-edge and new market knowledge (O3) as well asfirms seeking to exploit opportunities based onnew technical knowledge and new market knowl-edge (O4). But instead of the expected negativerelationship between difficulties of getting accessto external financing and the probability ofbecoming medium-sized, the relationship is sig-nificantly positive for both O3 and O4. Hence, forfirms seeking to exploit O3 and O4, those firmsthat report less difficulty in getting access toexternal financing are more likely to stay small.

Because of the positive relationship betweenthe difficulty of getting access to externalfinancing and the probability of attaining sub-stantial growth for firms seeking to exploit O3 andO4, Hypothesis 2b and Hypothesis 2c are not sup-ported. While firms seeking to exploit O3 differsignificantly from firms seeking to exploit O2, therelationship between financial constraints andgrowth is positive instead of negative. As forHypothesis 1, there is no significant differencebetween O3 and O4.


External financing

Investigating differences in external financingacross different types of opportunities may help tounderstand the reasons for the lack of support forHypotheses 2a–2c. In this section complementarydata are provided on the sources of financingobtained by the firms in sample (Table IV), theimportance of each source (Table V), and thedifficulties in accessing them (Table VI).

Table IV shows that the average number ofsources of external financing increases withincreased novelty of the opportunity, with theexception of opportunities based on new technicalknowledge and existing market knowledge. Theseresults are in line with the argument made earlierthat firms seeking to exploit opportunities basedon new knowledge are more dependent on externalfinancing than firms seeking to exploit opportu-nities based on existing knowledge. This is espe-cially true for firms seeking opportunities basedon new market knowledge.

Table IV also shows that the degree of bankfinancing is similar for all types of opportunities,where around 60% of the firms have obtainedbank financing. The degree of prepayment or loansfrom customers is also similar across differenttypes of opportunities, even for the firms that seekto exploit new market knowledge.

Financing from private persons is three timesas common in firms seeking to exploit opportuni-

ties based on both new technical knowledge andnew market knowledge as for firms that seek toexploit opportunities based on existing knowledge.Financing by private persons includes investmentby business angels.

Table IV also shows the importance of the“other” types of financing, especially for firmsseeking to exploit opportunities based on both newtechnical knowledge and new market knowledge.Almost half of the firms have obtained financingfrom other sources than those that are consideredtraditional sources of venture financing. Table Valso shows that this source of financing is of highimportance for these firms.

Table V show the relative importance of govern-ment support and venture capital for firms seekingto exploit opportunities based on new knowledge.Interestingly, government support seems to beespecially important for firms seeking to exploitopportunities based on new technology, whileventure capital firms are especially important forfirms seeking to exploit new market knowledge.

Table V shows that the importance of externalfinancing is higher for firms that seek opportuni-ties based on new knowledge than for firms thatseek opportunities based on existing knowledge.The importance of external financing is alsohighest for firms that seek to exploit opportunitiesbased on both new technical knowledge and newmarket knowledge.

Table V also shows the relative importance of


TABLE IVSources of external financing in young technology-based firms seeking different types of opportunities

External sources of financing Opportunities

ET-EM NT-EM ET-NM NT-NMN = 130 N = 47 N = 40 N = 45

Private persons 06% 11% 08% 20%Prepayments or loans from customers 22% 13% 18% 24%Government grants or loans 27% 34% 28% 42%Bank loans 62% 57% 60% 67%Venture capital 18% 23% 33% 44%Other 32% 26% 33% 47%

Average number of sources per firm 01.66 01.64 01.78 02.44

ET-EM: Opportunities based on existing technical knowledge and existing market knowledge.NT-EM: Opportunities based on new technical knowledge and existing market knowledge.ET-NM: Opportunities based on existing technical knowledge and new market knowledge.NT-NM: Opportunities based on new technical knowledge and new market knowledge.

124

loans and prepayments of customers for firms thatseek to exploit opportunities based on new marketknowledge. This finding indicates not only that thecustomers provide financial assistance, but thatcontacts with potential customers are importantin order to learn about customer needs and require-ments.

Table VI provides data about the difficulties ingetting access to different sources of externalfinancing. It shows that firms which seek to

exploit opportunities based on new knowledgereport more difficulties in getting access toexternal sources of financing than firms seekingto exploit opportunities based on existing knowl-edge. Interestingly, the firms that seek to exploitopportunities based on both new technical knowl-edge and new market knowledge report the leastdifficulty among firms that seek to exploit oppor-tunities based on new knowledge.

Table VI also shows that firms seeking to

How Business Opportunities Constrain Young Technology-Based Firms

TABLE VImportance of different external sources of financing for young technology-based firms seeking different types of opportunities

(average and standard deviation)



Private persons 3.00 (0.93) 2.60 (1.82) 2.83 (1.89) 3.67 (1.41)Prepayments or loans from customers 3.73 (1.16) 3.44 (1.19) 4.19 (0.90) 3.95 (1.42)Government grants or loans 2.78 (1.56) 2.97 (1.19) 2.45 (1.06) 3.60 (1.45)Bank loans 3.63 (1.19) 4.02 (1.01) 4.03 (1.14) 3.76 (1.20)Venture capital 4.02 (1.12) 3.03 (1.57) 4.09 (1.26) 4.22 (1.15)Other 3.68 (1.22) 3.60 (1.32) 3.68 (1.19) 4.19 (1.10)

Total 3.63 (1.14) 3.80 (1.03) 3.74 (1.05) 3.99 (1.08)

ET-EM: Opportunities based on existing technical knowledge and existing market knowledge.NT-EM: Opportunities based on new technical knowledge and existing market knowledge.ET-NM: Opportunities based on existing technical knowledge and new market knowledge.NT-NM: Opportunities based on new technical knowledge and new market knowledge.Importance is measured on a scale 1 to 5, where 1 is very little importance and 5 is very high importance.

TABLE VIDifficulties in getting access to external sources of financing for young technology-based firms seeking different types of

opportunities (average and standard deviation)



Private persons 3.06 (0.42) 4.33 (0.62) 3.00 (2.00) 2.91 (1.26)Prepayments or loans from customers 2.29 (1.08) 2.69 (0.91) 3.14 (1.57) 2.88 (1.20)Government grants or loans 2.85 (1.46) 3.62 (1.05) 3.05 (1.11) 3.30 (1.23)Bank loans 2.43 (1.27) 2.85 (1.26) 2.63 (1.07) 2.68 (1.22)Venture capital 3.26 (1.20) 4.03 (1.03) 3.40 (1.23) 3.13 (0.95)Other 2.12 (1.58) 2.42 (0.96) 3.00 (1.41) 2.63 (1.41)

Total 2.47 (1.17) 3.18 (1.12) 2.87 (0.98) 2.74 (1.04)

ET-EM: Opportunities based on existing technical knowledge and existing market knowledge.NT-EM: Opportunities based on new technical knowledge and existing market knowledge.ET-NM: Opportunities based on existing technical knowledge and new market knowledge.NT-NM: Opportunities based on new technical knowledge and new market knowledge.Difficulty is measured on a scale 1–5, where 1 is very easy and 5 is very difficult.

125

exploit opportunities based on new technicalknowledge and existing market knowledge reportthe highest difficulty in getting access to externalfinancing. This finding is interesting because itcontradicts the relative difference between newmarket knowledge and new technical knowledgefound in the regression model. It is also interestingto note that these firms have great difficulties inobtaining financing from venture capital investors,private investors and government, while at thesame time financing from venture capital investorsand private investors is of relatively low impor-tance for those who get it. It is difficult to inter-pret this finding, but it might indicate the lack ofcommitment of these investors to this particularclass of firms.

The results of the analysis of complementarydata on the degree, importance and difficulty ofexternal financing in young technology-basedfirms seem to support the results from previousresearch about the role of external financing fordifferent types of opportunities. Firms that seek toexploit opportunities based on new knowledgeneed more external financing than firms that seekto exploit opportunities based on existing knowl-edge, and also experience more difficulty ingetting access to external sources of financing.The high number of different sources that thesefirms are able to access indicated that the diffi-culty in accessing financing might be more dueto greater needs than to the impossibility ofaccessing a particular source of financing. Forexample, banks do not seem to be sensitive todifferent types of opportunities. This might thoughnot be true for firms that seek to exploit opportu-nities based on new technical knowledge andexisting market knowledge. These firms seem tohave difficulties getting access to, and takingadvantage of, certain sources of financing, mostnotably private investors and venture capitalists.

It is nevertheless interesting to note the impor-tance of venture capital financing, governmentsupport, support from customers, and otherunspecified sources of financing. These types offinancing seem to have a special role for partic-ular types of opportunities, especially for oppor-tunities based on new knowledge. This role couldbe related to complementary resources that theseactors provide, which could be important forattaining substantial growth.

Discussion and conclusions

The aim of this paper was to investigate howbusiness opportunities at start-up constrain youngtechnology-based firms from attaining substantialgrowth. The results of the empirical analysissupport the basic hypothesis of the paper that thetype of business opportunities at start-up signifi-cantly influences the probability of attaining sub-stantial growth, both directly and by moderatingthe effects of difficulties in getting access toexternal financing. The support for the hypothesisabout the specific direction of this influence wasmore mixed.

The results confirmed that firms seeking toexploit opportunities based on existing knowledgeare least likely to stay small. The results also con-firmed that firms seeking to exploit opportunitiesbased on new market knowledge are less likely toattain substantial growth than firms seeking toexploit opportunities based on new technicalknowledge.

The results did not confirm that firms seekingopportunities based on new technical knowledge,irrespective of the newness of market knowledge,were less likely to attain substantial growth thanfirms seeking to exploit opportunities based onexisting technical knowledge. This finding isunexpected because previous studies of youngtechnology-based firms (e.g. Oakey and Cooper,1991; Roberts, 1991) have emphasized the impor-tance of low technological uncertainty for successand growth.

A possible explanation of this discrepancy isthat studies emphasizing the importance of lowtechnological uncertainty were only studying firmsthat were operating in new markets as well. Thenthe uncertainty about the user problem is a partof the technological uncertainty.

Another possible explanation has to do withmeasurement problems. The measure used in thisstudy might underestimate the difference betweennew and existing knowledge. This study is never-theless not the only study to report a weak, or no,relationship between newness of technology andgrowth. Eisenhardt and Schoonhoven (1990)found no relationship between innovativeness offirms’ technology and growth, and Autio andLumme (1998) found no difference in growthbetween market innovators and paradigm innova-


tors. Hence, it seems reasonable to argue that thedifference between existing and new marketknowledge has much stronger constraining influ-ence on growth than does the difference betweenexisting and new technical knowledge. It is simplymore difficult to learn how to profitably exploitopportunities based on new market knowledge,probably because this will always involve learningabout both markets and technology.

While the results confirmed the basic hypoth-esis that the type of opportunities at start-upmoderates the relationship between difficulties inaccessing external financing and the probability ofattaining substantial growth, the results did notconfirm the expected interaction as stated in thehypotheses. In a similar fashion as for the directeffects of differences in the novelty of opportuni-ties, no effects were found for new technicalknowledge. For opportunities based on existingmarket knowledge, irrespective of the newness oftechnical knowledge, there was no significantrelationship between difficulties in getting accessto external financing and the probability ofbecoming medium-sized. For opportunities basedon new market knowledge, irrespective of thenewness of technical knowledge, there was asignificant relationship between difficulties ingetting access to external financing and the prob-ability of becoming medium-sized. But instead ofthis relationship being negative, i.e. that difficul-ties in getting access to finance would be moreconstraining for firms seeking opportunities basedon new market knowledge, the relationship ispositive. Thus, for firms seeking to exploit oppor-tunities based on new market knowledge, the firmsthat have more difficulty in getting access tofinance are more likely to attain substantialgrowth.

One possible explanation for this positiverelationship may be due to measurement problems.There might be a confounding effect of howactively the owner-managers seek externalfinancing in the variable that measures the degreeof difficulty in getting access to external financing.Hence, growth-constrained firms, i.e. firms that donot attain substantial growth, may not activelyseek external financing even if they need to do soin order to grow. This could be due to the lack ofgrowth intentions, or because owner-managershave a particular reason to avoid seeking external

financing. The latter is more probable, as differ-ences in growth intentions are controlled for in themodel. The reasons to avoid seeking externalfinancing could be reluctance to give up controlor ownership to equity investors, or reluctance touse debt financing.

Another possible explanation for the lack ofsupport for the proposed hypothesis, about themoderating effects of the type of opportunity onthe relationship between the difficulties in gettingaccess to external financing and the probabilityof attaining substantial growth, is the survivor biasinherent in the data. Firms that experience severefinancial limitations due to lack of access toexternal financing are more likely to fail anddisappear from the sample. Hence, firms thatexperience difficulties in accessing externalfinancing may be underrepresented in a sample ofsurvivors. The effects of the resulting survivor biasmight differ for different types of opportunities,being strongest for firms seeking opportunities ofhighest novelty which are, as shown in theprevious section, likely to be most dependent onexternal financing.

Taken together, the results of this study pointto the important distinction between young tech-nology-based firms that seek to exploit opportu-nities based on new market knowledge and firmsthat seek to exploit opportunities based on existingmarket knowledge. Firms that seek to exploitopportunities based on new market knowledge areless likely to attain substantial growth than firmsthat seek to exploit opportunities based on existingmarket knowledge. Furthermore, the results fromthe model, combined with the analysis of com-plementary data, suggest that firms seeking toexploit opportunities based on new market knowl-edge are more likely to attain substantial growthif they actively seek external financing. Seekingexternal financing from multiple sources,including e.g. customers and venture capitalists,not only ensures the financial resources needed forsurvival, but also may help to obtain other com-plementary resources and knowledge needed tobuild up the necessary management competencefor attaining substantial growth.

In this paper it has been argued, based onprevious theory, that different types of opportuni-ties represent different degrees of learning chal-lenges which face the firm. Further research is


needed in order to understand these differentchallenges and how firms respond to them. Dofirms refine their original product idea, concen-trating on building up the complementaryresources needed for successful commercializa-tion? Or do firms “innovate away” from theirproblems, pursuing new product ideas that theyjudge to be more likely to be accepted in themarket? While the acquisition of complementaryresources is important for early growth (Garnsey,1998), the ability to identify new opportunities iscrucial for continued growth (Penrose, 1959).Balancing the two is a critical question for man-agement. Without the former, the firm might neverenter on the path of growth – and without thelatter, the firm is not likely to get very far.

Acknowledgements

First, the authors wish to extend thanks for thefinancial support provided by the L. E. LundbergFoundation. Without this financial assistance itwould not have been possible to conduct thisresearch. The study is also linked to a largerresearch program in the CREATE group, gener-ously financed by the Ruben Rausing Fund, atIMIT and Industrial Dynamics, ChalmersUniversity of Technology. The support from theCREATE group has been very important for us,and is hereby gratefully acknowledged. Finally,the authors want to extend thanks to the editor andtwo anonymous reviews, whose constructivecomments and critique has greatly improved thequality of this paper.

Notes1 The concept of the young technology-based firm refers toa recently established firm whose competitive strength comesfrom the knowledge and skills of the employees within thefields of the natural sciences, engineering and medicine, andthe subsequent transformation of this knowledge into productsand services that can be sold on a market (Klofsten, 1992;Rickne and Jacobsson, 1999). This includes manufacturingfirms as well as firms providing industry-related services.2 The firms were independent at start-up, but could have beenacquired by another firm at the time of data collection.3 These industries were ISIC 341, 35, 37, 38, 6112, 72002,8323, 83249, 83292, 83299 and 932. Special rules wereapplied in ISIC 61120 and in ISIC 83292 and ISIC 83299 forselecting the relevant firms. See Rickne and Jacobsson (1999)for further details.

4 The relatively low number of medium-sized firms raisestwo questions. First, are the medium-sized firms under-represented in the sample? A follow-up study on the sizedistribution of the original population in 1999 found that 14%of the firms that still existed were medium-sized or larger.This is the same percentage as obtained in the final sampleused for estimating the model. Second, how will this relativelylow number of medium-sized firms affect the maximumnumber of parameters in the model? Hosmer and Lemeshow(2000, pp. 346–347) argue that the issue of sample size inlogistic regression models is complex and dependent on thecharacteristics of the data. They suggest as a rule of thumb,which may be too stringent, that 10 events of the least frequentoutcome are needed for each parameter. In our case this wouldmean that the model in this study is constrained to 3–4 para-meters. The full model presented in this paper includes 11parameters, which is 3–4 times larger than recommended bythe rule of thumb. To test the robustness of the model, a modelwith seven parameters (only two categories of opportunities)was tested. The results were stable, suggesting that theproblem of underestimation or overestimation due to largemodel size is not serious.

References

Abell, D. F., 1980, Defining the Business: The Starting Pointof Strategic Planning. Englewood Cliffs, NJ: Prentice-Hall.

Aldrich, H. and C. M. Fiol, 1994, ‘Fools Rush In? TheInstitutional Context of Industry Creation’, Academy ofManagement Review 19(4), 645–670.

Audretsch, D., 1995, ‘Innovation, Growth and Survival’,International Journal of Industrial Organization 13,441–457.

Autio, E. and A. Lumme, 1998, ‘Does the Innovator RoleAffect the Perceived Potential for Growth? Analysis ofFour Types of New, Technology-Based Firms’,Technology Analysis & Strategic Management 10(1),41–54.

Barber, J., S. Metcalfe and M. Porteous, 1989, ‘Barriers toGrowth: The Acard Study’, in J. Barber, S. Metcalfe andM. Porteous (eds.), Barriers to Growth in Small Firms.London: Routledge, pp. 1–19.

Chandler, G. N., 1996, ‘Business Similarity as a Moderatorof the Relationship between Pre-Ownership Experienceand Venture Performance’, Entrepreneurship Theory &Practice 20 (Spring), 51–65.

Cooper, A. C., 1986, ‘Entrepreneurship and HighTechnology’, in D. L. Sexton and R. W. Smilor (eds.), TheArt and Science of Entrepreneurship. Cambridge, MA:Ballinger, pp. 153–168.

Davidsson, P., 1989, ‘Entrepreneurship – and After? A Studyof Growth Willingness in Small Firms’, Journal ofBusiness Venturing 4, 211–226.

Delmar, F., 1997, ‘Measuring Growth: MethodologicalConsiderations and Empirical Results’, in R. Donckels andA. Miettinen (eds.), Entrepreneurship and SME Research:On its Way to the Next Millennium. Aldershot: Ashgate,pp. 199–215.

Delmar, F., P. Davidsson and W. Gartner, W. (forthcoming),


‘Arriving at the High-Growth Firm’, Journal of BusinessVenturing.

Eisenhardt, K. M. and C. B. Schoonhoven, 1990,‘Organizational Growth: Linking Founding Team,Strategy, Environment, and Growth among U.S.Semiconductor Ventures, 1978–1988’, AdministrativeScience Quarterly 35, 504–529.

Feeser, H. R. and G. E. Willard, 1990, ‘Founding Strategy andPerformance: A Comparison of High and Low GrowthHigh Tech Firms’, Strategic Management Journal 11,87–98.

Garnsey, E., 1995, ‘High Technology Renewal and the U.K.Investment Problem’, Journal of General Management20,(4), 1–22.

Garnsey, E., 1998, ‘A Theory of the Early Growth of theFirm’, Industrial and Corporate Change 13(3), 523–556.

Gibbons, M., C. Limoges, H. Nowotny, S. Schwartzman, P.Scott and M. Trow, 1994, The New Production ofKnowledge: The Dynamics of Science and Research inContemporary Societies. London: SAGE Publications.

Hofer, C. W. and R. Charan, 1984, ‘The Transition toProfessional Management: Mission Impossible’, AmericanJournal of Small Business 9(1), 1–11.

Hosmer, D. W. and S. Lemeshow, 2000, Applied LogisticRegression. New York: John Wiley & Sons.

Kirchhoff, B. A., 1994, Entrepreneurship and DynamicCapitalism. Westport: Praeger.

Klepper, S. and E. Graddy, 1990, ‘The Evolution of NewIndustries and the Determinants of Market Structure’,RAND Journal of Economics 21, 27–44.

Klofsten, M., 1992, Tidiga utvecklingsprocesser iteknikbaserade företag. Ph.D. Thesis, Department ofManagement and Economics. Linköping, LinköpingUniversity.

Lindmark, S., 2002, Evolution of Techno-Economic Systems.Ph.D. Thesis, Department of Industrial Management andEconomics. Göteborg, Chalmers University of Technology.

Lockett, A., G. Murray and M. Wright, 2002, ‘Do U.K.Venture Capitalists Still Have Bias against Investment inNew Technology Firms’, Research Policy 31, 1009–1030.

Murray, G. C. and J. Lott, 1995, ‘Have U.K. VentureCapitalists a Bias against Investment in New Technology-Based Firms?’, Research Policy 24, 283–299.

Nelson, R. R., 1994, ‘The Co-Evolution of Technology,Industrial Structure and Supporting Institutions’, Industrialand Corporate Change 3(1), 47–63.

Oakey, R., 1994, New Technology-Based Firms in the 1990s.London: Paul Chapman Publishing Company.

Oakey, R. P. and S. Y. Cooper, 1991, ‘The Relationshipbetween Product Technology and Innovation Performancein High Technology Small Firms’, Technovation 11(2),79–92.

Penrose, E. T., 1959, The Theory of the Growth of the Firm.Oxford: Basil Blackwell.

Rickne, A., 2000, New Technology-Based Firms andIndustrial Dynamics. Ph.D. Thesis, Industrial Dynamics.Göteborg, Chalmers University of Technology.

Rickne, A. and S. Jacobsson, 1999, ‘New Technology-BasedFirms in Sweden: A Study of Their Direct Impact onIndustrial Renewal’, Economics of Innovation and NewTechnology 8, 197–223.

Roberts, E. B., 1991, Entrepreneurs in High Technology:Lessons from MIT and Beyond. New York: OxfordUniversity Press.

Shackle, G. L. S., 1955, Uncertainty in Economics and OtherReflections. Cambridge: Cambridge University Press.

Shane, S., 2000, ‘Prior Knowledge and the Discovery ofEntrepreneurial Opportunities’, Organization Science11(4), 448–469.

Westhead, P. and D. J. Storey, 1997, ‘Financial Constraintson the Growth of High Technology Small Firms in theUnited Kingdom’, Applied Financial Economics 7,197–201.


How Business Opportunities Constrain Young Technology-Based Firms from Growing into Medium-Sized Firms

Documents