Description
In this brief file about perceived risks and choices in entrepreneurs new venture decisions.
PERCEIVED RISKS AND
CHOICES IN
ENTREPRENEURS’ NEW
VENTURE DECISIONS
DAVID FORLANI
University of Colorado at Denver, Denver, Colorado
JOHN W. MULLINS
University of Denver, Denver, Colorado
EXECUTIVE
SUMMARY
Though risk plays a central role in most entrepreneurial decision making,
little empirical researchhas explicitly examined howthe elements of risk, risk
perceptions, and entrepreneurs’ propensities to take risks in?uence choices
among potentially risky entrepreneurial ventures. This experimental study
asked a sample of entrepreneurs leading America’s fastest growing ?rms to
make choices among a series of hypothetical new ventures. The results indi-
cate that such choices are in?uenced by the risks inherent in the newventures, as evidenced by the pattern
of outcomes anticipatedin eachventure, the entrepreneurs’ differing perceptions of those risks, anddiffer-
ences in their personal propensities to take risks.
The subjects in our sample of entrepreneurs tended not to choose ventures having a high degree
of variability in their pattern of anticipated outcomes. This avoidance of outcome variability suggests
that the sensitivity analyses commonly prescribed for examining new venture attractiveness may inhibit
risk taking, and may deter potential investors from investing in their ?rms. New approaches to assessing
and presenting new venture risk, other than the traditional best case/expected case/worst case approach,
may be advisable, as well as suf?ciently through market research to provide evidence of the degree to
which market acceptance is likely for the venture’s products or services.
We also found an effect of differences inrisk propensities among entrepreneurs on their newventure
choices. This effect suggests not only that entrepreneurs should be wary of any biases they bring to their
new venture decisions, but that prospective investors should consider the degree to which entrepreneurs
in whom they choose to invest are well-matched to the investors’ own risk-taking propensities.
Finally, while our sample of entrepreneurs tended to shun high levels of variability in their new
venture choices, they appeared willing to accept a considerable degree of hazard, or possible downside,
in their new venture choices, presumably in pursuit of potentially signi?cant gains. Entrepreneurs are
advised to seek a clear understanding of the downside entailed in their proposed ventures, and develop
strategies to mitigate the likelihood of adverse outcomes. Thus they will not jeopardize chances for near
term success and attracting support of investors and others in later stages of the venture or in subse-
quent ventures.
Address correspondence to David Forlani, Assistant Professor, College of Business Administration,
University of Colorado at Denver, Box 165, Campus Box 173364, Denver, CO 80217; (303)556-6616.
Journal of Business Venturing 15, 305–322
© 2000 Elsevier Science Inc. All rights reserved. 0883-9026/00/$–see front matter
655 Avenue of the Americas, New York, NY 10010 PII S0883-9026(98)00017-2
306 D. FORLANI AND J.W. MULLINS
Our research did not attempt to examine how our subjects’ choices would have played out in terms
of performance, but the apparent biases which entrepreneurs’ risk propensities bring to their assessment
of proposed new ventures is a potentially important issue that merits further scrutiny. On one hand, such
biases may lead to patterns of suboptimal decisions. On the other hand, our results suggest that investors
should entrust their newventure investments to entrepreneurs whose risk propensities (and perhaps other
personal characteristics) best matchthe needs of boththe opportunity at handandthe investor’s objectives.
As many venture capitalists attest, the management of a proposed new venture should lie at the heart
of their investment decision. © 2000 Elsevier Science Inc.
INTRODUCTION
The image of entrepreneurs as bold, forward-thinking risk takers is a part of American
business folklore. For entrepreneurs, risk is a central element in a variety of decision
contexts, including those dealing with entry into newventures or newmarkets (Dickson
1992; Timmons 1994), and new product introductions (Devinney 1992). Indeed, the un-
certainnature of consumer andcompetitive responses tomost entrepreneurial decisions
makes consideration of risk an everyday task for most entrepreneurs, as well as for in-
vestors whose funds make possible entrepreneurs’ pursuit of their dreams (Hall and
Hofer 1993; Riquelme and Rickards 1992).
Given the signi?cant failure rate among new ventures (Phillips and Kirchoff 1998;
Reynolds 1986), and the rapidly changing markets in which today’s new ventures are
founded (Dickson 1992), a better understanding of risk and its role in newventure deci-
sion making has the potential to improve the quality of decision making in the risk-
charged environments which most prospective founders of new?rms face. Surprisingly,
however, little empirical research has explicitly explored the role of risk in entrepre-
neurial decision making.
Prior research dealing with risk in entrepreneurial settings has focused largely on
investors’ decisions, and criteria and procedures used to manage investment risk (Sykes
and Dunham 1995) and improve the performance of investment portfolios (e.g., Hall
and Hofer 1993; Riquelme and Rickards 1992). Only recently have researchers begun
to examine entrepreneurial risk-taking from the entrepreneur’s point of view. Palich
and Bagby (1995) found that entrepreneurs tend to view some business situations more
positively than do nonentrepreneurs, perceiving strengths and opportunities where oth-
ers seek weaknesses and threats. Their work suggests that entrepreneurs do not see
themselves as risk takers, but that they pursue opportunities that others do not because
they simply view such opportunities differently. Busenitz and Barney (1997) found that
entrepreneurs tended to employ heuristics and biases to simplify and speed their deci-
sion making in the complex and risky decision environments which typify start-up situa-
tions. We extend this line of researchintothe perceptions anddecision-making behavior
of entrepreneurs to examine entrepreneurs perceptions of risk in new venture settings,
as well as the choices entrepreneurs make among potentially risky new ventures.
PURPOSE
This paper’s purpose is to improve our understanding of both the role and the anteced-
ents of entrepreneurs’ risk perceptions in their new venture decisions. Drawing on es-
tablished traditions in both economics and behavioral decision theory, we ?rst set forth
a framework consisting of several factors likely to in?uence entrepreneurs’ risk percep-
RISKS AND CHOICES IN NEW VENTURES 307
tions. This framework shows how these perceptions and other factors are linked to new
venture decisions. We then experimentally test portions of the framework.
Our study attempts to shed some light on two speci?c research questions: 1) What
factors lead entrepreneurs to perceive new ventures as risky? 2) Why do entrepreneurs
sometimes pursue riskier ventures, and sometimes less risky ones? More speci?cally,
we ask if variability in entrepreneurial risk-taking is due to differences in risk among
different ventures, entrepreneurs’ differing perceptions of such risks, or differences in
the propensity of different entrepreneurs to take risks? Thus, we are concerned in this
paper with both risk perceptions and risk propensities as they in?uence the risky choice
decisions of entrepreneurs. In addition to addressing these researchquestions, our study
seeks to build an interval scale of risky venture choices for use in this and subsequent
empirical research.
A BRIEF OVERVIEW OF THE ROLE OF RISK PERCEPTIONS IN
NEW VENTURE DECISIONS
Previous research involving the notion of risk in business decision-making contexts ar-
ticulates two distinct views of risk (Fisher and Hall 1969; March and Shapira 1987), each
of which holds implications for entrepreneurs’ risk perceptions, and by implication, for
risky decision making. Additionally, recent research suggests that factors other than
the perceived risks associated with the decision alternatives may signi?cantly in?uence
risky choices (Krueger and Dickson 1994; Manimala 1992; Mullins 1996; Sitkin and
Pablo 1992).
Drawing on the work of Sitkin and Pablo (1992), we argue that risk perceptions
of entrepreneurs, venture characteristics, contextual effects, and traits of individual en-
trepreneurs play key roles in entrepreneurs’ decisions to enter newventures (see Figure
1). Stage 1 of our conceptual framework argues that managers’ perceptions of new ven-
ture risks are drivenprincipally bythree sets of factors: a) therelative level of investment
needed to fund the venture; b) variability in the anticipated outcomes of the venture;
and c) any potential losses which may ensue. Perceived new venture risk is expected
to be higher for ventures which entail greater investment (investing available capital
in fewer larger ventures limits opportunities for diversi?cation, and there is more to
lose on a given venture) and for ventures whose anticipated outcomes are either more
uncertain (greater variability in anticipated returns) or entail the possibility of greater
operating losses (March and Shapira 1987).
Stage 2 of our framework indicates that new venture choices are driven by the risk
perceptions and anticipated returns of various alternative ventures, along with differ-
ences in personality traits of entrepreneurs considering such decisions, and a variety
of other contextual factors. Such contextual factors include resource constraints, the
?t of the venture with the entrepreneur’s competencies and interest, and various organi-
zational and environmental conditions operative at the time of the newventure decision
(Baird and Thomas 1985). Space limitations require that we only develop and empiri-
cally test portions of our framework. Thus, in this paper, we focus on an examination of
the effects of anticipated newventure outcomes as well as differences among individual
entrepreneurs on risk perceptions and risk taking behavior. We choose to direct our
attention here because anticipated outcomes lie at the heart of the analysis which entre-
preneurs are encouraged to undertake before launching new ventures (Sahlman 1997;
Timmons 1994).
308 D. FORLANI AND J.W. MULLINS
FIGURE 1 A Framework of the Role of Risk Perceptions in New Venture Selection.
Implicit in our model is the notion that entrepreneurs’ perceptions of risk and deci-
sions involving risk are distinct and separate cognitive processes. This viewis consistent
with an abundant body of research into consumer decision-making that judgments
about products and services and choices among them involve distinct cognitive opera-
tions that do not always work in parallel (cf. Bettman and Park 1980; Johnson and Russo
1984). Examining judgments and choices, or in our case, entrepreneurs’ perceptions of
risk and their decisions involving risk, as fundamentally different notions may offer new
insights into how entrepreneurs view and respond to risk in new venture decision-
making.
THEORETICAL DEVELOPMENT
What Constitutes Risk?
Although considerable attention has been devoted to studying the conceptual and oper-
ational dimensions of risk (MacCrimmon and Wehrung 1986, 1990; Schneider and
RISKS AND CHOICES IN NEW VENTURES 309
Lopes 1986; Shoemaker 1990), much less effort has been devoted to investigating the
dimensions of the risk construct and their in?uences on risky choice behavior. Risk re-
?ects the degree of uncertainty and potential loss associated with the outcomes which
may follow from a given behavior or set of behaviors. Yates and Stone (1992) identify
three elements of the risk construct: potential losses, the signi?cance of those losses,
and the uncertainty of those losses. In risky entrepreneurial contexts, where losses are
almost always possible, it is the signi?cance of any possible losses—or hazard, as we
shall refer to it—and the uncertainty or variability of those losses that are likely to be
most salient in driving risk perceptions and risky decision-making behavior. The hazard
and variability dimensions of risk argue, respectively, that greater potential hazard for
a proposed new venture and greater variability in anticipated returns for a proposed
venture should lead entrepreneurs to view the venture as riskier than one having less
hazard and less variability, all other factors equal. We employ the variability and hazard
dimensions to develop our ?rst two hypotheses.
The Variability Perspective and Entrepreneurs’ Risk Perceptions
The economics literature typically de?nes risk as variability (Armour and Teece 1978;
Fisher and Hall 1969), arguing that greater variability in economic returns constitutes
greater risk. Variability, in turn, is de?ned as the probability of actual returns or out-
comes deviating from the expected return or outcome. For example, investors might
assess the risk of investing in a ?rm in terms of its chances for providing a given return
based on the deviations in prior returns (Armour and Teece 1978; Bowman 1980; Fisher
and Hall 1969). From this perspective, risk is seen as the possibility that an anticipated
level of return will not be realized, and is typically operationalizedas the standard devia-
tion of an investment’s historical returns.
1
New ventures, by de?nition, typically do not have a ?ow of historical returns for
entrepreneurs to examine, however, the risk-as-variability perspective may still be ger-
mane. Entrepreneurs use a variety of procedures for estimating the likely variability
in future returns for proposed new ventures, including critical assumption planning
(Sykes and Dunham 1995), risk analysis (Hertz 1964) and sensitivity analysis (Tim-
mons 1994).
Sometimes such procedures are likely to produce rather subjective estimates of
both the variability and magnitude of future returns for ventures involving new-to-the-
world innovations. However, for less novel ventures, entrepreneurs can often draw on
relevant experience gained fromsimilar ventures in the past and produce risk and sensi-
tivity estimates with greater con?dence. Thus, the variability perspective suggests the
following hypothesis:
H1: The greater the variability in predicted outcomes of a proposed new venture,
the greater will be its perceived risk.
The Hazard Perspective and Entrepreneurs’ Risk Perceptions
March and Shapira (1987) studied executives’ views of risk. They concluded that what
theexecutive perceives as riskis not outcome variability, but hazard—if things gowrong,
1
Although the standard deviation, as an operational measure of risk, encompasses both the magnitude
andthe probability distributionof historical returns, the emphasis inthis theoretical streamis onthelikelihood
of achieving outcomes different from the historical mean.
310 D. FORLANI AND J.W. MULLINS
how much can we lose? “A risky choice is one that contains a threat of a very poor
outcome” (March and Shapira 1987: 1407).
2
March and Shapira found that executives,
in assessing risk, pay little attention to the probabilities associated with alternative out-
comes, for several reasons. First, most managers do not treat uncertainty about positive
outcomes as animportant aspect of risk. Second, for most managers (and most entrepre-
neurs), risk is a concept having to do primarily with loss, not with probabilities. A study
by Shapira (1995: 45) found that 95% of managers surveyed “described risk in terms
of the magnitude of ?nancial loss.” Third, most managers show little desire to reduce
risk to a single quanti?able construct.
Although this perspective incorporates the idea of variability in a range of possible
outcomes, its emphasis is on the magnitude of potential losses. In a newventure context,
March and Shapira’s view leads to the following hypothesis:
H2: The greater the magnitude of a proposed new venture’s largest potential loss,
the greater will be its perceived risk.
The In?uence of Variability and Hazard in Risky Choice Decisions
Expectedutility arguments indicate that decisionalternatives having lower levels of per-
ceived risk, whether due to high levels of variability or hazard, should be preferred to
alternatives having higher levels of risk, other factors being equal (Yates 1990). Consis-
tent with these arguments, March and Shapira’s (1987) interviews found that managers
are likely to avoid choosing decision alternatives for which the chances are high that
the expected outcome will not occur. Indeed, Shapira (1995: 57) summarizes the section
of his work on organizational risk-taking with the advice of an executive he interviewed:
“Avoid risk taking would be my credo.” Thus:
H3A: The greater the variability in predicted outcomes of a proposed new venture,
the less likely it will be selected for funding.
H3B: The greater the magnitude of a new venture’s largest potential loss, the less
likely it will be selected for funding.
Individual Differences, Risk Perceptions, and Risky Choice Decisions
In a recent review of factors which in?uence risky business decisions, Sitkin and Pablo
(1992: 9) posit that the “risk propensity (of the decision maker) dominates the actual
and perceived characteristics of the situation as a determinant of risk behavior.” In the
newventure context, this statement suggests that the riskpropensity of theentrepreneur
making the decision as to which of several proposed ventures to enter is more important
than the returns the alternative new ventures are expected to generate, their risks, and
most other factors commonly considered in the analysis of such ventures. Sitkin and
Pablo’s assertion stands in sharp contrast to the expected utility perspective that domi-
nates research and practice in the new venture decisions area.
Sitkin and Pablo de?ne risk propensity as “the tendency of a decision maker either
to take or to avoid risks” (p. 12). Sitkin and Pablo do not argue that all other factors,
2
March and Shapira (1992: 172), in a different paper, also acknowledge the variability perspective:
“Riskiness is associated with lack of certainty about the precise outcome of a choice and thus with variation
in the probability distribution.”
RISKS AND CHOICES IN NEW VENTURES 311
such as those whose effects we hypothesize above, are inoperative. They simply say
that individual differences among actors in terms of their risk propensities are likely to
explain a greater portion of variance in risky choice behavior. The impact of individual
differences is supported by the work of Lopes (1987), who found that some individuals
tend to base their actions on the upside of a range of possible decision outcomes, while
others tend to act based on the downside. Thus, some individuals are motivated by up-
side potential, while others are motivated by security.
Recent research by Sitkin and Weingart (1995) has found that differences in risk
propensities also in?uence risk perceptions. Individuals of higher risk propensity will
perceive the risks associatedwith a particular decision alternative to be lower than those
having lower risk propensities. Risk perceptions, in turn are expected to in?uence
choices among risky alternatives. Choices among alternatives in a decision set by a deci-
sion maker who perceives the set as less risky are expected to be riskier than for those
who perceive the set as riskier (Yates 1990). The arguments of Sitkin and Pablo (1992);
Sitkin and Weingart (1995); Lopes (1987); and Yates (1990) generate three addi-
tional hypotheses:
H4: The greater the riskpropensity of the entrepreneur, the less will be the perceived
risk associated with a particular new venture.
H5: The greater the risk propensity of the entrepreneur, the more likely he or she
will be to select new ventures having higher levels of risk.
H6: The lower an entrepreneur’s perceived risk across a set of decision alternatives,
the more likely he or she will be to select new ventures having higher levels of risk.
Finally, the argument of Sitkin and Pablo (1992) that risk propensities dominate risky
choice decisions, together with Lopes’ (1987) theory that risk propensity consists largely
of a tendency of individuals to attend to either the upside (i.e., the potential) or the
downside (i.e., the hazard) of a situation, suggests that risk propensity should operate
on March and Shapira’s (1987) hazard conceptualization of risk, rather than on the vari-
ability dimension. If this is so, then for ventures of equal expected value, risk propensity
is expected to in?uence choices among ventures which differ in amount of hazard and
gain, but not necessarily among ventures which differ as to degree of variability. Based
on Sitkin and Pablo’s (1992) argument, this effect should be robust across different lev-
els of variability in anticipated venture outcomes. Thus, for ventures of equal variability
and expected value:
H7: The greater the riskpropensity of the entrepreneur, the greater will be the likeli-
hood of choosing a venture having higher levels of hazard.
METHOD
Design and Procedure
A 2 ?2 full factorial within subjects experimental design was employed to manipulate
the variability and hazard associated with the outcomes of new ventures. We created,
through several pretests, a series of descriptions of predicted outcomes for four new
ventures (see Table 1). All four ventures have equal expected values, described as meet-
ing the entrepreneur’s requirements for return on investment (ROI) for new ventures.
Two of the ventures have higher variability (a 40% chance of meeting target ROI with
312 D. FORLANI AND J.W. MULLINS
a 30% chance of being over target and a 30% chance of being under target versus an
80% chance of meeting target ROI with a 10% chance of being over target and a 10%
chance of being under target), and two have greater hazard (possible outcomes $25 mil-
lion over or under target versus $5 million over or under). In order to maintain equal
expected values for all four ventures, potential for gain above target levels is equal to
the potential for loss below target levels for each venture.
3
Subjects were presented with a scenario which asked them to imagine that they
were about to undertake a new venture. They were presented with the four potential
newventure descriptions, and were told that all four were in the same industry, required
similar and manageable levels of start-up capital, and that all met their target for return
on investment. The ventures were rotated to eliminate possible order effects. The indi-
vidual venture descriptions were repeated, one per page, on the next four pages, along
with instructions for responding to the ?rst dependent measure (see Table 1) provided
on each page. This dependent measure, a three item scale, recorded the amount of risk
the subject perceived in each venture (page ordering was matched to the venture rota-
tion noted above). Next, a second dependent measure asked subjects to indicate which
venture they wouldchoose. Measures of the manipulations’ effectiveness were collected
next, followed by measures of individual differences and demographic items.
Sample
CEOs of the 540 ?rms listed in INC, Fortune, and BusinessWeek magazines on their
combined 1994 and 1996 tabulations of the fastest growing public companies in the
United States were contacted by fax to request their participation in the study. The 210
subjects who agreed to participate in the study (39% of those originally contacted, after
three faxed requests) were then mailed the experimental instruments, and asked to re-
turn them via U.S. mail, to ensure con?dentiality. After ?ve weeks and two faxed re-
minders, 91 instruments (43% of those who had indicated they would participate in the
study) had been returned. After discarding 13 incomplete or unusable instruments, we
were left with a remaining subject pool of 78 entrepreneurs.
The entrepreneurs in the resulting sample had founded from 1 to 12 ?rms (mean
3.0 ?rms), and ranged from 28 to 66 years of age (mean 47.7). All were male. Their
current ?rms ranged in size from 6 to 10,500 employees (mean 1,186), and $0.4 million
to $1.2 billion in sales (mean $147.6 million). Of their current ?rms, 43.6% operate in
TABLE 1 Mean Risk Statistics and Choices Among the Four Ventures
Venture Level of Level of Mean Standard Number of Subjects
Name Variability Hazard Risk Score Deviation Choosing Venture
Green High High 5.57 1.22 10
Yellow High Low 4.53 1.43 3
Purple Low High 3.19 1.34 39
White Low Low 1.97 0.96 26
Total 78
Differences between mean risk scores for all ventures are statistically signi?cant (Tukey HSD test) at the 0.01 level.
3
Given our criterion that expected values of all four ventures be equal, the ventures having greater
hazard also have greater potential for gain. This operational condition permits us to gain insights into Lopes
(1987) contention that differences in risk propensity consist largely of a tendency to pay attention to either
the upside or the downside of an anticipated set of outcomes.
RISKS AND CHOICES IN NEW VENTURES 313
TABLE 2 New Venture Descriptions
Venture Green Venture White
There is a 30% chance of being under target by There is a 10% chance of being under target by
$25 million, a 40% chance of meeting target ROI $5 million, a 80% chance of meeting target ROI
and a 30% chance of going over target by $25 and a 10% chance of going over target by $5
million. Graphically the distribution appears as: million. Graphically the distribution appears as:
Venture Purple Venture Yellow
There is a 10% chance of being under target by There is a 30% chance of being under target by
$25 million, a 80% chance of meeting target ROI $5 million, a 40% chance of meeting target ROI
and a 10% chance of going over target by $25 and a 30% chance of going over target by $5
million. Graphically the distribution appears as: million. Graphically the distribution appears as:
TABLE 3 Scale of Risk Propensity
Please answer the following 5 items by circling the alternative (“a” or “b”) you would feel most comfort-
able with.
1 a) an 80% chance of winning $400, or
b) receiving $320 for sure
2 a) receiving $300 for sure, or
b) a 20% chance of winning $1,500
3 a) a 90% chance of winning $200, or
b) receiving $180 for sure
4 a) receiving $160 for sure, or
b) a 10% chance of winning $1,600
5 a) a 50% chance of winning $500, or
b) receiving $250 for sure
314 D. FORLANI AND J.W. MULLINS
TABLE 4 Scale of Perceived New Venture Risk
For each scale below, kindly circle the number which you feel best assesses the amount of RISK associated
with this scenario:
HIGH 1 2 3 4 5 6 7 LOW
MINIMAL 1 2 3 4 5 6 7 EXTREME
VERY RISKY 1 2 3 4 5 6 7 NOT RISKY
manufacturing and 56.4% in service industries. In order to assess any possible non-
response bias, we compared ?rm demographic data for our sample to mean data from
the published lists from which our sample was drawn. Firms in our sample tended to
be somewhat smaller than those on the published lists, in terms of revenue (mean of
$147.6 million for the sample compared to $217.8 million for the listed ?rms) and num-
ber of employees (mean of 1,186 comparedto 1,753). These data suggest that one should
be cautious in generalizing our results to entrepreneurs who lead very large high-
growth ?rms.
Independent Variables
The two manipulated variables were the variability and the degree of hazard (and gain)
of the ventures’ anticipated outcomes, as described previously (see Table 2). Risk pro-
pensity was operationalized using an adaptation of the established (Schneider and
Lopes 1986) Risk Style Scale, as shown in Table 3. We chose this measure for our study
because it deals with personal propensities toward ?nancial risk taking (as opposed to
other kinds of risks, such as those entailed in sky diving) and because of its ef?cacy in
assessing the construct of interest (Schneider and Lopes 1986).
Dependent Measures
The dependent measure of perceived new venture risk for hypotheses H1, H2, and H4
was a scale of 3, 7-point items (see Table 4). The reliability of this scale is indicated
by a coef?cient alpha for a three item scale of 0.956.
The dependent measure of the riskiness of new venture choice for Hypotheses
H3A, H3B, H5, H6, and H7 was the subject’s selection of one of the four ventures to
fund. The mean perceived risk scores for the four ventures indicate that this scale forms
an approximately interval scale of the riskiness of the four ventures (see Table 5). As
expected, the venture scoring highest inperceived risk(Venture Green: meanperceived
TABLE 5 Manipulation Check Results
Manipulation Check
Independent Variable Mean Values p Value
Variability ? 0.001
High 3.79
Low 2.14
Hazard
1
? 0.001
High 4.32
Low 1.97
Mean values are points on a 5-point scale, with 3 as the midpoint.
1
The manipulation of hazard also manipulated potential gain, in order to maintain equal expected values across ventures
(see Table 2). A manipulation check for gain was also signi?cant (means 4.40 and 1.90, p ? 0.001).
RISKS AND CHOICES IN NEW VENTURES 315
risk score 5.57) is the venture having high levels of both variability and hazard. The
venture perceived as next most risky (Venture Yellow: 4.58) has high variability and
low hazard. The venture perceived as third most risky (Venture Purple: 3.19) has low
variability and high hazard. The venture judged least risky (Venture White: 1.97) is the
venture having lowlevels of both variability and hazard(see Table 5). Pairwise compari-
sons between the perceived risk scores for all of the ventures were signi?cant (Tukey’s
HSD test, p ? 0.01).
Manipulation Checks
The effectiveness of the variability and hazard manipulations was assessed by compar-
ing the mean scores on the manipulation check items (each a 5-point scale) for each
venture across the two levels of each experimental condition. As shown in Table 2, all
of the manipulations were effective.
RESULTS
Antecedents of Risk Perceptions
A repeated measures analysis of variance procedure (Norusis 1990) was run to assess
the effects of variability and hazard on perceptions of new venture risk. Main effects
of both variability (F(1,74) ?307.7, p ?0.001) and hazard (F(1,74) ?94.38, p ?0.001)
were found, thereby providing evidence in support of the impact of both the variability
(H1) and hazard (H2) perspectives on risk perceptions. No signi?cant interaction was
found. To assess the effects of the subjects’ risk propensities on the subjects’ risk percep-
tions of eachventure (H4), we rana regressionanalysis of riskpropensity onriskpercep-
tion. The result was not signi?cant (? ? ?0.116, p ? 0.13).
Variability, Hazard, and New Venture Choices
Evidence to test H3A, which predicts greater likelihood of choice of ventures having
less variability in anticipated returns, is foundby comparing the number of higher versus
low variability ventures chosen for funding (see Table 1). Thirteen subjects chose the
high variability ventures (Ventures Green and Yellow), while 65 subjects chose the low
variability ventures (Purple and White). A chi squared test rejects the null hypothesis
that these differences are due to chance (?
2
(1 df) ?34.67, p ?0.01), thereby supporting
H3A (see Table 1).
Evidence to test H3B, which predicts greater chance of choosing ventures having
less magnitude of anticipated losses, is found by comparing the number of high vs. low
hazard ventures chosen for funding (see Table 1). Forty-nine subjects chose the high
hazard ventures (Green and Purple), while 29 subjects chose the low hazard ventures
(Yellowand White). Thus, contrary to the prediction of H3B, entrepreneurs were more
likely to choose high hazard than low hazard ventures, presumably to obtain the poten-
tial for the greater gains which went along with the high hazard condition. Achi squared
test rejects the null hypothesis that these differences are due to chance (?
2
(1 df) ?5.13,
p ? 0.05).
316 D. FORLANI AND J.W. MULLINS
Risk Perceptions, Propensity, and New Venture Choices
To examine H5 and H6, we ran a regression model in which the dependent variable
was new venture choice, and the independent variables were subjects’ risk propensity
scores and their mean total perceptions of risk across the ventures.
4
H5 predicts a positive relationship between risk propensity and new venture
choices. This prediction was supported (? ? 0.26, p ? 0.05), thereby supporting the
prediction of Sitkin and Pablo (1992). The effect of risk perceptions on new venture
choice (H6) was alsosigni?cant (???0.19, p ?0.05). Thus, we ?ndboth riskpropensity
and risk perceptions in?uencing entrepreneurs’ new venture choices. Given the inabil-
ity, in this experimental research, to ascertain the relative “doses” of differences in risk
perception brought about by our manipulations and differences in risk propensities of
the entrepreneurs in our sample, we can draw no conclusive evidence of the relative
importance of one factor compared to the other.
Finally, H7 predicts that, for ventures of equal expected value and equal variability,
more risk prone entrepreneurs will choose more risky ventures (i.e., those having higher
levels of hazard and gain). For the two ventures having low variability, Ventures White
(chosen by 26 subjects) and Purple (chosen by 39 subjects), the results of a logistic re-
gression do not support this prediction (? ? 0.067, p ? 0.74). Similarly, comparing
choices across the two high variability ventures, Yellow (chosen by three subjects) and
Green (chosen by 10 subjects), the results of a second logistic regression also fail to
support this prediction (? ?2.14, p ?0.22). Though the theories we relied on in devel-
oping our study did not predict that risk perceptions would drive these effects, we did
include risk perception in the regression models to test H7. The effect of risk perception
was signi?cant for the Venture Purple vs. Venture White Comparison (? ? ?0.648,
p ? 0.04), but not for the Venture Green vs. Venture Yellow comparison (? ? 4.75,
p ? 0.18).
DISCUSSION
The results provide evidence which extends the work of March and Shapira (1987). Dif-
ferences in entrepreneurs’ new venture choices were in?uenced not only by differences
in the risks inherent in the patterns of anticipated outcomes for different ventures, but
by differences in our entrepreneurs’ perceptions of those risks, as well as their propensi-
ties totake risk. Further, as MarchandShapira predict, the degree of hazardina decision
alternative does in?uence the degree to which entrepreneurs perceive that alternative
as risky, though greater perceived risk did not deter our subjects from making risky
choices: the entrepreneurs in our study were more likely to choose high hazard than
low hazard ventures (63% chose high hazard ventures, which also entailed high gain),
contrary to our prediction. Apparently, high hazard ventures are acceptable as long as
commensurate gains are suf?ciently likely.
Ontheother hand, entrepreneurs appear tobe more apprehensive about variability
in possible outcomes, having overwhelmingly chosen ventures low in variability (83%
chose either Venture White or Purple). Perhaps they are con?dent that they can inter-
vene to increase their chances of achieving the desired outcome when probabilities of
4
Tests of normality, homogeneity of variance, and multicolinearity were run to ensure that the regres-
sion model adequately ?t the data (Neter, Wassermann, and Kutner 1990). No heteroskedasticity, lack of
normality of multicolinearity problems were found.
RISKS AND CHOICES IN NEW VENTURES 317
unexpected negative outcomes are relatively small. Indeed, a recent study suggests that
the overcon?dence of entrepreneurs enables them to start new ventures before many
of the inherent uncertainties therein have been resolved (Busenitz and Barney 1997).
When probabilities of unexpected negative outcomes are large, however, they may feel
that they lack the control to bring in the venture on target.
As for risk propensities, as Sitkin and Pablo (1992) predicted, entrepreneurs who
have greater risk propensities tend to choose riskier ventures. Interestingly, however,
the subjects’ risk propensities did not signi?cantly in?uence their perceptions of venture
risks, contrary to the prediction of Sitkin and Pablo. Thus, risk propensity appears to
directly impact venture choice behavior, rather than indirectly affecting behavior
through the perceptual process. The absence of an effect of risk propensity on risk per-
ceptions is consistent with the ?ndings of a recent study by Palich and Bagby (1995) that
found a consistently optimistic pattern of categorization of business situations among
entrepreneurs compared to nonentrepreneurs, in spite of no difference in risk propen-
sity among the two groups. It may be that various cognitive patterns and processes of
entrepreneurs are more important determinants of entrepreneurial behavior than their
risk propensity. Indeed, such processes may constitute unobserved variables in our re-
search that may be responsible for the effects we found. Our ?nding that risk propensity
in?uences newventure choice behavior, but not risk perception, is parallel to researchin
consumer choice that judgment and choice tasks involve different cognitive operations
(Bettman and Park 1980; Johnson and Russo 1984).
Our results, taken together with previous research ?ndings by Palich and Bagby
(1995) and Busenitz and Barney (1997), suggest that better understanding is needed
about how entrepreneurs search for and process information about business situations
(Cooper, Folta, and Woo 1995; Manimala 1992), and how such information processing
in?uences entrepreneurial behavior. Kahneman and Lovallo (1993) argue that the rela-
tive balance between what they call isolation errors (over-optimistic forecasting that
ignores the statistics of the past on one hand, and overly timid evaluations of single
risky opportunities that neglect possibilities to pool risks on the other) affect the risk-
taking propensities of individuals. These and other errors or biases, including the well-
known discrepancy (Kahneman and Tversky 1979) between the weights attached to
losses and gains in evaluating risky decisions such as new venture opportunities, may
explain why we found that hazard in anticipated new venture outcomes in?uenced per-
ceptions of risk in our study, but did not deter the entrepreneurs in our sample from
choosing riskier ventures. Additional research into the cognitions of entrepreneurs may
offer additional insights to better explain entrepreneurial behavior.
Finally, our results indicate that the series of four hypothetical new ventures that
we used as our dependent measure of new venture risk (See Exhibit 2) constitutes a
scale having approximately interval properties (see Table 1). The availability of a valid
scale of new venture risk should facilitate future experimental research into the role
of risk in new venture decision making.
LIMITATIONS
Our study, like most experimental studies in business settings, suffers fromseveral limi-
tations. First, entrepreneurs in real situations may not behave as did our subjects in the
hypothetical situations inwhichtheywereplacedinour study. Giventhelikely dif?culty,
ina ?eldstudy, of controlling for the broadarray of factors whichare positedtoin?uence
318 D. FORLANI AND J.W. MULLINS
risky newventure decisions in natural settings, and to separate the effects of risk percep-
tions from“objective” risk, we elected to conduct an experimental study. This approach
has advantages in internal validity for theory testing purposes, but may be criticized
on grounds that the experimental task is not a real one with real payoffs. Given the
early stage of research in this arena, and given our interest in testing theories, some
of which have undergone little empirical scrutiny, we deemed the tradeoff acceptable.
Second, we chose to study decisions among a set of four newventure decision alter-
natives whose investments and expected values were all equal—such precise equality
of investments and expected values across proposed ventures is unlikely in real situa-
tions. An additional limitation is the use of a measure of risk propensity borrowed from
another literature, a measure whichmay not adequately capture the propensity of entre-
preneurs to take risks in new venture situations. Bromiley and Curley (1992) argue that
risk propensities are, to some degree, situation speci?c, and that measures from one
situation may not work well in another situations.
Third, in manipulating hazard we also manipulated gain, in order to preserve the
equality of the expected values in the anticipated outcomes of the four ventures from
which our subjects were asked to choose. Had we not done so, there would have little
motivationfor our subjects tochoose highhazardventures (withlittleprospect for gain),
and differences in choices would have been attributable to differences in expected value
among the ventures, rather than to differences in risk propensity and risk perceptions,
the variables of interest in our study.
Finally, we restricted our examination to only the effects of anticipated outcomes
and differences in risk propensity on the decisions that were made. Potentially impor-
tant contextual factors such as competencies and previous experience of the entrepre-
neur, incentives, and team decision process issues were not examined in our study. Fu-
ture research is needed to explore these and other likely in?uences on new venture
decisions.
CONCLUSIONS AND IMPLICATIONS
Our study holds implications for investors who fund entrepreneurial ventures, as well
as for entrepreneurs themselves, and for educators who train students hoping to join
tomorrow’s cadre of entrepreneurs.
To the extent that potential investors in newventures behave similarly to the entre-
preneurs in our study, the impact of variability in anticipated new venture outcomes
on perceptions of risk entailed in, and choices among, new venture alternatives has im-
plications for how entrepreneurs should employ sensitivity analyses and seek to reduce
perceived outcome variability as they seek investment capital. High levels of hazard
did not deter entrepreneurs from choosing ventures with potentially high levels of gain,
therefore, ventures having high levels of variability, such as that often explicitly detailed
in sensitivity analyses, were less likely to be chosen. Entrepreneurs may ?nd it useful
to?ndnewways, other thanthe traditional best case/expectedcase/worst case approach,
to present to prospective investors the likelihood of deviation in future performance
from desired outcomes. Our results may offer an explanation of why entrepreneurs are
sometimes reluctant to engage in explicit risk and sensitivity analyses for proposed ven-
tures. As Ulrich and Epplinger (1995: 256) suggest, “Often [decision makers] do not
want to confront the true probabilities of bad outcomes.” Rather than attempting to
mask the outcome variability of presenting new ventures, entrepreneurs may be better
RISKS AND CHOICES IN NEW VENTURES 319
advised to invest in suf?cient market research to ascertain the level of market accep-
tance which the products or services of a proposed new venture are likely to enjoy, or
to test the criticality of key assumptions underlying their plans (Sykes and Dunham
1995). Several new techniques of qualitative market research are able to provide tangi-
ble evidence of likely market acceptance, even for new-to-the-world products and ser-
vices, thereby reducing perceptions of outcome variability and helping to build a solid
foundation of customer needs on which to erect the new venture (Grif?n and Hauser
1993; Grif?n 1996; Zaltman 1997).
For entrepreneurs choosing among a set of proposed newventure alternatives, our
results indicate that, where levels of investment and the expected values of returns are
similar, ventures tend to be chosen based on differences in risk propensities among en-
trepreneurs, in addition to the risk entailed in the ventures’ patterns of anticipated re-
turns. This ?nding attests to the importance that venture capitalists and other investors
place in the people who lead the ventures in which they invest (Heilemann 1997; Sahl-
man 1997). Entrepreneurs are advised to explicitly ask themselves whether their assess-
ment of proposed ventures they consider are biased in any way by their propensities
to take risks. Such biases could result in decisions which lengthen the already daunting
odds for new venture success. Various approaches have been identi?ed for overcoming
decision biases, including the use of structured decision aids (Ghosh and Ray 1997) and
treating a particular decision as an instance of a broader class of similar previous deci-
sions about which outcome information is available (Kahneman and Lovallo 1993).
Additionally, entrepreneurs in search of new venture opportunities should think
carefully about strategies for mitigating the hazard they appear willing to accept in
search of potentially attractive gains. Failure to adequately understand and plan for
adverse outcomes not only may jeopardize chances for near term success, but they may
make it more dif?cult to attract support from investors and others for subsequent ?-
nancing (Sahlman 1997).
For teachers whose work it is toprepare a newgeneration of entrepreneurs tomake
wise choices among newventure opportunities and develop successful strategies to pur-
sue them, our study suggests that providing would-be entrepreneurs with tools, tech-
niques, and analytical frameworks for reducing the variability in their forecasts of new
venture outcomes can play an important role in facilitating their pursuit of potentially
attractive, but risky, opportunities. Potentially useful tools and techniques include vari-
ous qualitative and quantitative market research approaches (cf., Grif?n and Hauser
1993; Grif?n 1996; Mahajan, Muller, and Bass 1990; Zaltman 1997); relevant analytical
frameworks include critical assumption planning (Sykes and Dunham 1995) and risk
analysis (Hertz 1964). Use of such tools and frameworks to provide a strong foundation
of evidence on which to build a business plan provides two bene?ts to students as would-
be entrepreneurs: ?rst, stronger evidence supporting the attractiveness of a proposed
venture is likely to enhance the likelihood of obtaining funding for the venture; second,
as the results of our study suggest (recall that our subjects were more likely to choose
ventures having less variability in anticipated outcomes), by eliminating uncertainty and
reducing variability in anticipated outcomes, such evidence will make it more likely that
the entrepreneur will, given favorable evidence, decide to pursue the venture. This psy-
chological role of such evidence should not be underestimated.
Our results raise questions which call for additional research into the risky choice
decisions of entrepreneurs. First, the person versus situation debate in organizational
psychology (Mischel 1977, O’Reilly 1991) has wrestled in recent years with trying to
320 D. FORLANI AND J.W. MULLINS
better identify the kinds of situations inwhichindividual differences tendtobe relatively
more or less important, compared to situational factors, in determining behavior. For
risky new venture decisions such as those we have examined here, what situational or
contextual factors are likely to moderate the importance of risk propensity and other
individual differences in risky choice decisions? March and Shapira (1992) found, in a
simulation study, that recent performance outcomes, accumulated resources, and the
goals or reference points that decision makers attend to in?uence risky choice behavior.
BairdandThomas (1985) articulateda model of strategic risktaking whichincorporated
a broad array of environmental, organizational, industrial, decision maker, and problem
variables. Few of these variables have been empirically studied as to their effects on
risky choice decision making. These works suggest numerous directions for future em-
pirical research.
How our subjects’ venture choices would have played out in terms of new venture
performance was not addressed in our study, but the apparent biases which individual
differences in risk propensity generate in new venture decision making is a potentially
important issue for future research. Viewed from one perspective, our ?ndings suggest
that patterns of suboptimal decisions are likely to be common, as managers’ inertia (Sit-
kin and Pablo 1992) and proneness or aversity toward risk (Lopes 1987; Schneider and
Lopes 1986) lead to consistently risk-prone or risk-avoiding decisions. Viewed from a
more optimistic perspective, however, our results suggest that people matter in such
decisions (Hitt and Tyler 1991). Entrusting new venture investments to individuals
whose risk propensities and other individual characteristics best match the needs of the
market opportunity and a prospective investor’s objectives may help improve portfolio
performance. The considerable effort which venture capitalists and other investors put
forth in evaluating the individuals whose ideas and companies they fund attests to the
importance of individual differences, including differences in risk taking propensity, for
new ?rm success.
REFERENCES
Armour, H.O. and Teece, D.J. 1978. Organization structure and economic performance: a test
of multidivisional hypothesis. Bell Journal of Economics 9:106–122.
Baird, I.S. and Thomas, H. 1985. Toward a contingency model of strategic risk taking. Academy
of Management Review 10(2):230–243.
Bettman, J.R. and Park, C.W. 1980. Effects of prior knowledge and experience and phase of the
choice process on consumer decision processes: a protocol analysis. Journal of Consumer
Research 7:234–248.
Bird, B.J. 1989. Entrepreneurial Behavior. Glenview, IL: Scott, Foresman and Company.
Bowman, E.H. 1980. Arisk/return paradox for strategic management. Sloan Management Review
21 (Spring):17–31.
Bromiley, P. and Curley, S.P. 1992. Individual differences in risk taking. IN J.F. Yates, ed., Risk
Taking Behavior. West Sussex, England: John Wiley and Sons.
Busenitz, L.W. and Barney, J.B. 1997. Differences between entrepreneurs and managers in large
organizations: biases and heuristics in strategic decision-making. Journal of Business Ven-
turing 12:9–30.
Cooper, A.C., Folta, T.B. andWoo, C. 1995. Entrepreneurial information search. Journal of Busi-
ness Venturing 10:107–120.
Devinney, T.M. 1992. New products and ?nancial risk changes. Journal of Product Innovation
Management 3:222–231.
RISKS AND CHOICES IN NEW VENTURES 321
Dickson, P.R. 1992. Toward a general theory of competitive rationality. Journal of Marketing
56:69–83.
Fisher, N. and Hall, G. R. 1969. Risk and corporate rates of return. Quarterly Journal of Econom-
ics 83:79–92.
Ghosh, D. and Ray, M.R. 1997. Risk, ambiguity, and decision choice: some additional evidence.
Decision Sciences 28(1):81–104.
Grif?n, A. 1996. Obtaining customer needs for product development. In M.D. Rosenau, ed., The
PDMA Handbook of New Product Development, New York: John Wiley and Sons, pp.
153–166.
Grif?n, A. and Hauser, J.R. 1993. The voice of the customer. Marketing Science 12(1):1–27.
Hall, J. and Hofer, C.W. 1993. Venture capitalists’ decision criteria in new venture evaluation.
Journal of Business Venturing 8:25–42.
Heilemann, J. 1997. The networker. The New Yorker August 11:28–36.
Hertz, D.B. 1964. Risk analysis in capital investment. Harvard Business Review (January-Febru-
ary): 96–106.
Hitt, M.A. and Tyler, B.B. 1991. Strategic decision models: integrating different perspectives.
Strategic Management Journal 12:327–351.
Johnson, E. and Russo, J.E. 1984. Product familiarity and learning new information. Journal of
Consumer Research 11:542–550.
Kahneman, D. and Lovallo, D. 1993. Timid choices and bold forecasts: a cognitive perspective
on risk taking. Management Science 39(1):17–31.
Kahneman, D. and Tversky, A. 1979. Prospect theory: an analysis of decision under risk. Econo-
metrica 47:263–290.
Krueger, N., Jr. and Dickson, P.R. 1994. How believing in ourselves increases risk taking: per-
ceived self-ef?cacy and opportunity recognition. Decision Sciences 25(3):385–400.
Lopes, L.L. 1987. Betweenhope and fear: the psychology of risk. Advances inExperimental Social
Psychology 20:255–295.
MacCrimmon, K.R. and Wehrung, D.A. 1986. Taking Risks: The Management of Uncertainty.
New York: Free Press.
MacCrimmon, K.R. and Wehrung, D.A. 1990. Characteristics of risk taking executives. Manage-
ment Science 36(4):422–435.
Mahajan, V., Muller, E. andBass, F.M. 1990. Newproduct diffusionmodels inmarketing: a review
and directions for research. Journal of Marketing 54 (January):1–26.
Manimala, M.J. 1992. Entrepreneurial heuristics: a comparison between high PI (pioneering-in-
novative) and low PI ventures. Journal of Business Venturing 7:477–504.
March, J.G. and Shapira, Z. 1987. Managerial perspectives on risk and risk taking. Management
Science 33(11):1404–1418.
March, J.G. andShapira, Z. 1992. Variable riskpreferences andthe focus of attention. Psychologi-
cal Review 99(1):172–183.
Mischel, W. 1977. The interaction of person and situation. In D. Magnusson and N.S. Endler,
eds., Personality at the Crossroads: Current Issues in Interactional Psychology, Hillsdale,
NJ: Erlbaum.
Mullins, J.W. 1996. Early growthdecisionof entrepreneurs: the in?uence of competency andprior
performance under changing market conditions. Journal of Business Venturing 11:89–105.
O’Reilly, C.A. III. 1991. Organizational behavior: where we’ve been, where we’re going. Annual
Review of Psychology 42:427–485.
Palich, L.E. and Bagby, D.R. 1995. Using cognitive theory to explain entrepreneurial risk-taking:
Challenging conventional wisdom. Journal of Business Venturing 10:425–438.
Phillips, B.D. andKirchoff, B.A. 1988. Ananalysis of new?rmsurvival andgrowth. BabsonEntre-
preneurship Research Conference, Calgary.
Reynolds, P. 1986. Predicting contributions and survival. In Ronstadt, et al. eds., Frontiers of
Entrepreneurship Research. Wellesley, MA: Babson College.
322 D. FORLANI AND J.W. MULLINS
Riquelme, H. andRickards, T. 1992. Hybridconjoint analysis: anestimationprobe innewventure
decisions. Journal of Business Venturing 7:505–518.
Sahlman, W.A. 1997. How to write a great business plan. Harvard Business Review (July-
August):98–108.
Schneider, S.L. and Lopes, L.L. 1986. Re?ection in preferences under risk: who and when may
suggest why. Journal of Experimental Psychology: Human Perception and Performance
12(4):535–548.
Shapira, Z. 1995. Risk Taking: A Managerial Perspective. New York: Russell Sage.
Shoemaker, P.J. 1990. Are risk-attitudes related across domains and response modes? Manage-
ment Science 36(12):1451–1463.
Sitkin, S.B. and Pablo, A.L. 1992. Reconceptualizing the determinants of risk behavior. Academy
of Management Review 17(1):9–38.
Sitkin, S.B. and Weingart, L.R. 1995. Determinants of risky decision making behavior: a test of
the mediating role of risk perceptions and risk propensity. Academy of Management Jour-
nal 38(6):1573–1592.
Sykes, H.B. and Dunham, D. 1995. Critical assumption planning: a practical tool for managing
business development risk. Journal of Business Venturing 10:413–424.
Timmons, J.A. 1994. New Venture Creation. Homewood, IL: Irwin.
Ulrich, K.T. and Epplinger, S.D. 1995. Product Design and Development. New York:
Mcgraw-Hill.
Yates, J.F. 1990. Judgment and Decision Making. Englewood Cliffs, NJ: Prentice-Hall.
Yates, J.F. and Stone, E.R. 1992. The risk construct. IN J.F. Yates, ed., Risk Taking Behavior,
West Sussex, England: John Wiley and Son.
Zaltman, G. 1997. Rethinking market research: Putting people back in. Journal of Marketing
34:424–437.
doc_367008731.pdf
In this brief file about perceived risks and choices in entrepreneurs new venture decisions.
PERCEIVED RISKS AND
CHOICES IN
ENTREPRENEURS’ NEW
VENTURE DECISIONS
DAVID FORLANI
University of Colorado at Denver, Denver, Colorado
JOHN W. MULLINS
University of Denver, Denver, Colorado
EXECUTIVE
SUMMARY
Though risk plays a central role in most entrepreneurial decision making,
little empirical researchhas explicitly examined howthe elements of risk, risk
perceptions, and entrepreneurs’ propensities to take risks in?uence choices
among potentially risky entrepreneurial ventures. This experimental study
asked a sample of entrepreneurs leading America’s fastest growing ?rms to
make choices among a series of hypothetical new ventures. The results indi-
cate that such choices are in?uenced by the risks inherent in the newventures, as evidenced by the pattern
of outcomes anticipatedin eachventure, the entrepreneurs’ differing perceptions of those risks, anddiffer-
ences in their personal propensities to take risks.
The subjects in our sample of entrepreneurs tended not to choose ventures having a high degree
of variability in their pattern of anticipated outcomes. This avoidance of outcome variability suggests
that the sensitivity analyses commonly prescribed for examining new venture attractiveness may inhibit
risk taking, and may deter potential investors from investing in their ?rms. New approaches to assessing
and presenting new venture risk, other than the traditional best case/expected case/worst case approach,
may be advisable, as well as suf?ciently through market research to provide evidence of the degree to
which market acceptance is likely for the venture’s products or services.
We also found an effect of differences inrisk propensities among entrepreneurs on their newventure
choices. This effect suggests not only that entrepreneurs should be wary of any biases they bring to their
new venture decisions, but that prospective investors should consider the degree to which entrepreneurs
in whom they choose to invest are well-matched to the investors’ own risk-taking propensities.
Finally, while our sample of entrepreneurs tended to shun high levels of variability in their new
venture choices, they appeared willing to accept a considerable degree of hazard, or possible downside,
in their new venture choices, presumably in pursuit of potentially signi?cant gains. Entrepreneurs are
advised to seek a clear understanding of the downside entailed in their proposed ventures, and develop
strategies to mitigate the likelihood of adverse outcomes. Thus they will not jeopardize chances for near
term success and attracting support of investors and others in later stages of the venture or in subse-
quent ventures.
Address correspondence to David Forlani, Assistant Professor, College of Business Administration,
University of Colorado at Denver, Box 165, Campus Box 173364, Denver, CO 80217; (303)556-6616.
Journal of Business Venturing 15, 305–322
© 2000 Elsevier Science Inc. All rights reserved. 0883-9026/00/$–see front matter
655 Avenue of the Americas, New York, NY 10010 PII S0883-9026(98)00017-2
306 D. FORLANI AND J.W. MULLINS
Our research did not attempt to examine how our subjects’ choices would have played out in terms
of performance, but the apparent biases which entrepreneurs’ risk propensities bring to their assessment
of proposed new ventures is a potentially important issue that merits further scrutiny. On one hand, such
biases may lead to patterns of suboptimal decisions. On the other hand, our results suggest that investors
should entrust their newventure investments to entrepreneurs whose risk propensities (and perhaps other
personal characteristics) best matchthe needs of boththe opportunity at handandthe investor’s objectives.
As many venture capitalists attest, the management of a proposed new venture should lie at the heart
of their investment decision. © 2000 Elsevier Science Inc.
INTRODUCTION
The image of entrepreneurs as bold, forward-thinking risk takers is a part of American
business folklore. For entrepreneurs, risk is a central element in a variety of decision
contexts, including those dealing with entry into newventures or newmarkets (Dickson
1992; Timmons 1994), and new product introductions (Devinney 1992). Indeed, the un-
certainnature of consumer andcompetitive responses tomost entrepreneurial decisions
makes consideration of risk an everyday task for most entrepreneurs, as well as for in-
vestors whose funds make possible entrepreneurs’ pursuit of their dreams (Hall and
Hofer 1993; Riquelme and Rickards 1992).
Given the signi?cant failure rate among new ventures (Phillips and Kirchoff 1998;
Reynolds 1986), and the rapidly changing markets in which today’s new ventures are
founded (Dickson 1992), a better understanding of risk and its role in newventure deci-
sion making has the potential to improve the quality of decision making in the risk-
charged environments which most prospective founders of new?rms face. Surprisingly,
however, little empirical research has explicitly explored the role of risk in entrepre-
neurial decision making.
Prior research dealing with risk in entrepreneurial settings has focused largely on
investors’ decisions, and criteria and procedures used to manage investment risk (Sykes
and Dunham 1995) and improve the performance of investment portfolios (e.g., Hall
and Hofer 1993; Riquelme and Rickards 1992). Only recently have researchers begun
to examine entrepreneurial risk-taking from the entrepreneur’s point of view. Palich
and Bagby (1995) found that entrepreneurs tend to view some business situations more
positively than do nonentrepreneurs, perceiving strengths and opportunities where oth-
ers seek weaknesses and threats. Their work suggests that entrepreneurs do not see
themselves as risk takers, but that they pursue opportunities that others do not because
they simply view such opportunities differently. Busenitz and Barney (1997) found that
entrepreneurs tended to employ heuristics and biases to simplify and speed their deci-
sion making in the complex and risky decision environments which typify start-up situa-
tions. We extend this line of researchintothe perceptions anddecision-making behavior
of entrepreneurs to examine entrepreneurs perceptions of risk in new venture settings,
as well as the choices entrepreneurs make among potentially risky new ventures.
PURPOSE
This paper’s purpose is to improve our understanding of both the role and the anteced-
ents of entrepreneurs’ risk perceptions in their new venture decisions. Drawing on es-
tablished traditions in both economics and behavioral decision theory, we ?rst set forth
a framework consisting of several factors likely to in?uence entrepreneurs’ risk percep-
RISKS AND CHOICES IN NEW VENTURES 307
tions. This framework shows how these perceptions and other factors are linked to new
venture decisions. We then experimentally test portions of the framework.
Our study attempts to shed some light on two speci?c research questions: 1) What
factors lead entrepreneurs to perceive new ventures as risky? 2) Why do entrepreneurs
sometimes pursue riskier ventures, and sometimes less risky ones? More speci?cally,
we ask if variability in entrepreneurial risk-taking is due to differences in risk among
different ventures, entrepreneurs’ differing perceptions of such risks, or differences in
the propensity of different entrepreneurs to take risks? Thus, we are concerned in this
paper with both risk perceptions and risk propensities as they in?uence the risky choice
decisions of entrepreneurs. In addition to addressing these researchquestions, our study
seeks to build an interval scale of risky venture choices for use in this and subsequent
empirical research.
A BRIEF OVERVIEW OF THE ROLE OF RISK PERCEPTIONS IN
NEW VENTURE DECISIONS
Previous research involving the notion of risk in business decision-making contexts ar-
ticulates two distinct views of risk (Fisher and Hall 1969; March and Shapira 1987), each
of which holds implications for entrepreneurs’ risk perceptions, and by implication, for
risky decision making. Additionally, recent research suggests that factors other than
the perceived risks associated with the decision alternatives may signi?cantly in?uence
risky choices (Krueger and Dickson 1994; Manimala 1992; Mullins 1996; Sitkin and
Pablo 1992).
Drawing on the work of Sitkin and Pablo (1992), we argue that risk perceptions
of entrepreneurs, venture characteristics, contextual effects, and traits of individual en-
trepreneurs play key roles in entrepreneurs’ decisions to enter newventures (see Figure
1). Stage 1 of our conceptual framework argues that managers’ perceptions of new ven-
ture risks are drivenprincipally bythree sets of factors: a) therelative level of investment
needed to fund the venture; b) variability in the anticipated outcomes of the venture;
and c) any potential losses which may ensue. Perceived new venture risk is expected
to be higher for ventures which entail greater investment (investing available capital
in fewer larger ventures limits opportunities for diversi?cation, and there is more to
lose on a given venture) and for ventures whose anticipated outcomes are either more
uncertain (greater variability in anticipated returns) or entail the possibility of greater
operating losses (March and Shapira 1987).
Stage 2 of our framework indicates that new venture choices are driven by the risk
perceptions and anticipated returns of various alternative ventures, along with differ-
ences in personality traits of entrepreneurs considering such decisions, and a variety
of other contextual factors. Such contextual factors include resource constraints, the
?t of the venture with the entrepreneur’s competencies and interest, and various organi-
zational and environmental conditions operative at the time of the newventure decision
(Baird and Thomas 1985). Space limitations require that we only develop and empiri-
cally test portions of our framework. Thus, in this paper, we focus on an examination of
the effects of anticipated newventure outcomes as well as differences among individual
entrepreneurs on risk perceptions and risk taking behavior. We choose to direct our
attention here because anticipated outcomes lie at the heart of the analysis which entre-
preneurs are encouraged to undertake before launching new ventures (Sahlman 1997;
Timmons 1994).
308 D. FORLANI AND J.W. MULLINS
FIGURE 1 A Framework of the Role of Risk Perceptions in New Venture Selection.
Implicit in our model is the notion that entrepreneurs’ perceptions of risk and deci-
sions involving risk are distinct and separate cognitive processes. This viewis consistent
with an abundant body of research into consumer decision-making that judgments
about products and services and choices among them involve distinct cognitive opera-
tions that do not always work in parallel (cf. Bettman and Park 1980; Johnson and Russo
1984). Examining judgments and choices, or in our case, entrepreneurs’ perceptions of
risk and their decisions involving risk, as fundamentally different notions may offer new
insights into how entrepreneurs view and respond to risk in new venture decision-
making.
THEORETICAL DEVELOPMENT
What Constitutes Risk?
Although considerable attention has been devoted to studying the conceptual and oper-
ational dimensions of risk (MacCrimmon and Wehrung 1986, 1990; Schneider and
RISKS AND CHOICES IN NEW VENTURES 309
Lopes 1986; Shoemaker 1990), much less effort has been devoted to investigating the
dimensions of the risk construct and their in?uences on risky choice behavior. Risk re-
?ects the degree of uncertainty and potential loss associated with the outcomes which
may follow from a given behavior or set of behaviors. Yates and Stone (1992) identify
three elements of the risk construct: potential losses, the signi?cance of those losses,
and the uncertainty of those losses. In risky entrepreneurial contexts, where losses are
almost always possible, it is the signi?cance of any possible losses—or hazard, as we
shall refer to it—and the uncertainty or variability of those losses that are likely to be
most salient in driving risk perceptions and risky decision-making behavior. The hazard
and variability dimensions of risk argue, respectively, that greater potential hazard for
a proposed new venture and greater variability in anticipated returns for a proposed
venture should lead entrepreneurs to view the venture as riskier than one having less
hazard and less variability, all other factors equal. We employ the variability and hazard
dimensions to develop our ?rst two hypotheses.
The Variability Perspective and Entrepreneurs’ Risk Perceptions
The economics literature typically de?nes risk as variability (Armour and Teece 1978;
Fisher and Hall 1969), arguing that greater variability in economic returns constitutes
greater risk. Variability, in turn, is de?ned as the probability of actual returns or out-
comes deviating from the expected return or outcome. For example, investors might
assess the risk of investing in a ?rm in terms of its chances for providing a given return
based on the deviations in prior returns (Armour and Teece 1978; Bowman 1980; Fisher
and Hall 1969). From this perspective, risk is seen as the possibility that an anticipated
level of return will not be realized, and is typically operationalizedas the standard devia-
tion of an investment’s historical returns.
1
New ventures, by de?nition, typically do not have a ?ow of historical returns for
entrepreneurs to examine, however, the risk-as-variability perspective may still be ger-
mane. Entrepreneurs use a variety of procedures for estimating the likely variability
in future returns for proposed new ventures, including critical assumption planning
(Sykes and Dunham 1995), risk analysis (Hertz 1964) and sensitivity analysis (Tim-
mons 1994).
Sometimes such procedures are likely to produce rather subjective estimates of
both the variability and magnitude of future returns for ventures involving new-to-the-
world innovations. However, for less novel ventures, entrepreneurs can often draw on
relevant experience gained fromsimilar ventures in the past and produce risk and sensi-
tivity estimates with greater con?dence. Thus, the variability perspective suggests the
following hypothesis:
H1: The greater the variability in predicted outcomes of a proposed new venture,
the greater will be its perceived risk.
The Hazard Perspective and Entrepreneurs’ Risk Perceptions
March and Shapira (1987) studied executives’ views of risk. They concluded that what
theexecutive perceives as riskis not outcome variability, but hazard—if things gowrong,
1
Although the standard deviation, as an operational measure of risk, encompasses both the magnitude
andthe probability distributionof historical returns, the emphasis inthis theoretical streamis onthelikelihood
of achieving outcomes different from the historical mean.
310 D. FORLANI AND J.W. MULLINS
how much can we lose? “A risky choice is one that contains a threat of a very poor
outcome” (March and Shapira 1987: 1407).
2
March and Shapira found that executives,
in assessing risk, pay little attention to the probabilities associated with alternative out-
comes, for several reasons. First, most managers do not treat uncertainty about positive
outcomes as animportant aspect of risk. Second, for most managers (and most entrepre-
neurs), risk is a concept having to do primarily with loss, not with probabilities. A study
by Shapira (1995: 45) found that 95% of managers surveyed “described risk in terms
of the magnitude of ?nancial loss.” Third, most managers show little desire to reduce
risk to a single quanti?able construct.
Although this perspective incorporates the idea of variability in a range of possible
outcomes, its emphasis is on the magnitude of potential losses. In a newventure context,
March and Shapira’s view leads to the following hypothesis:
H2: The greater the magnitude of a proposed new venture’s largest potential loss,
the greater will be its perceived risk.
The In?uence of Variability and Hazard in Risky Choice Decisions
Expectedutility arguments indicate that decisionalternatives having lower levels of per-
ceived risk, whether due to high levels of variability or hazard, should be preferred to
alternatives having higher levels of risk, other factors being equal (Yates 1990). Consis-
tent with these arguments, March and Shapira’s (1987) interviews found that managers
are likely to avoid choosing decision alternatives for which the chances are high that
the expected outcome will not occur. Indeed, Shapira (1995: 57) summarizes the section
of his work on organizational risk-taking with the advice of an executive he interviewed:
“Avoid risk taking would be my credo.” Thus:
H3A: The greater the variability in predicted outcomes of a proposed new venture,
the less likely it will be selected for funding.
H3B: The greater the magnitude of a new venture’s largest potential loss, the less
likely it will be selected for funding.
Individual Differences, Risk Perceptions, and Risky Choice Decisions
In a recent review of factors which in?uence risky business decisions, Sitkin and Pablo
(1992: 9) posit that the “risk propensity (of the decision maker) dominates the actual
and perceived characteristics of the situation as a determinant of risk behavior.” In the
newventure context, this statement suggests that the riskpropensity of theentrepreneur
making the decision as to which of several proposed ventures to enter is more important
than the returns the alternative new ventures are expected to generate, their risks, and
most other factors commonly considered in the analysis of such ventures. Sitkin and
Pablo’s assertion stands in sharp contrast to the expected utility perspective that domi-
nates research and practice in the new venture decisions area.
Sitkin and Pablo de?ne risk propensity as “the tendency of a decision maker either
to take or to avoid risks” (p. 12). Sitkin and Pablo do not argue that all other factors,
2
March and Shapira (1992: 172), in a different paper, also acknowledge the variability perspective:
“Riskiness is associated with lack of certainty about the precise outcome of a choice and thus with variation
in the probability distribution.”
RISKS AND CHOICES IN NEW VENTURES 311
such as those whose effects we hypothesize above, are inoperative. They simply say
that individual differences among actors in terms of their risk propensities are likely to
explain a greater portion of variance in risky choice behavior. The impact of individual
differences is supported by the work of Lopes (1987), who found that some individuals
tend to base their actions on the upside of a range of possible decision outcomes, while
others tend to act based on the downside. Thus, some individuals are motivated by up-
side potential, while others are motivated by security.
Recent research by Sitkin and Weingart (1995) has found that differences in risk
propensities also in?uence risk perceptions. Individuals of higher risk propensity will
perceive the risks associatedwith a particular decision alternative to be lower than those
having lower risk propensities. Risk perceptions, in turn are expected to in?uence
choices among risky alternatives. Choices among alternatives in a decision set by a deci-
sion maker who perceives the set as less risky are expected to be riskier than for those
who perceive the set as riskier (Yates 1990). The arguments of Sitkin and Pablo (1992);
Sitkin and Weingart (1995); Lopes (1987); and Yates (1990) generate three addi-
tional hypotheses:
H4: The greater the riskpropensity of the entrepreneur, the less will be the perceived
risk associated with a particular new venture.
H5: The greater the risk propensity of the entrepreneur, the more likely he or she
will be to select new ventures having higher levels of risk.
H6: The lower an entrepreneur’s perceived risk across a set of decision alternatives,
the more likely he or she will be to select new ventures having higher levels of risk.
Finally, the argument of Sitkin and Pablo (1992) that risk propensities dominate risky
choice decisions, together with Lopes’ (1987) theory that risk propensity consists largely
of a tendency of individuals to attend to either the upside (i.e., the potential) or the
downside (i.e., the hazard) of a situation, suggests that risk propensity should operate
on March and Shapira’s (1987) hazard conceptualization of risk, rather than on the vari-
ability dimension. If this is so, then for ventures of equal expected value, risk propensity
is expected to in?uence choices among ventures which differ in amount of hazard and
gain, but not necessarily among ventures which differ as to degree of variability. Based
on Sitkin and Pablo’s (1992) argument, this effect should be robust across different lev-
els of variability in anticipated venture outcomes. Thus, for ventures of equal variability
and expected value:
H7: The greater the riskpropensity of the entrepreneur, the greater will be the likeli-
hood of choosing a venture having higher levels of hazard.
METHOD
Design and Procedure
A 2 ?2 full factorial within subjects experimental design was employed to manipulate
the variability and hazard associated with the outcomes of new ventures. We created,
through several pretests, a series of descriptions of predicted outcomes for four new
ventures (see Table 1). All four ventures have equal expected values, described as meet-
ing the entrepreneur’s requirements for return on investment (ROI) for new ventures.
Two of the ventures have higher variability (a 40% chance of meeting target ROI with
312 D. FORLANI AND J.W. MULLINS
a 30% chance of being over target and a 30% chance of being under target versus an
80% chance of meeting target ROI with a 10% chance of being over target and a 10%
chance of being under target), and two have greater hazard (possible outcomes $25 mil-
lion over or under target versus $5 million over or under). In order to maintain equal
expected values for all four ventures, potential for gain above target levels is equal to
the potential for loss below target levels for each venture.
3
Subjects were presented with a scenario which asked them to imagine that they
were about to undertake a new venture. They were presented with the four potential
newventure descriptions, and were told that all four were in the same industry, required
similar and manageable levels of start-up capital, and that all met their target for return
on investment. The ventures were rotated to eliminate possible order effects. The indi-
vidual venture descriptions were repeated, one per page, on the next four pages, along
with instructions for responding to the ?rst dependent measure (see Table 1) provided
on each page. This dependent measure, a three item scale, recorded the amount of risk
the subject perceived in each venture (page ordering was matched to the venture rota-
tion noted above). Next, a second dependent measure asked subjects to indicate which
venture they wouldchoose. Measures of the manipulations’ effectiveness were collected
next, followed by measures of individual differences and demographic items.
Sample
CEOs of the 540 ?rms listed in INC, Fortune, and BusinessWeek magazines on their
combined 1994 and 1996 tabulations of the fastest growing public companies in the
United States were contacted by fax to request their participation in the study. The 210
subjects who agreed to participate in the study (39% of those originally contacted, after
three faxed requests) were then mailed the experimental instruments, and asked to re-
turn them via U.S. mail, to ensure con?dentiality. After ?ve weeks and two faxed re-
minders, 91 instruments (43% of those who had indicated they would participate in the
study) had been returned. After discarding 13 incomplete or unusable instruments, we
were left with a remaining subject pool of 78 entrepreneurs.
The entrepreneurs in the resulting sample had founded from 1 to 12 ?rms (mean
3.0 ?rms), and ranged from 28 to 66 years of age (mean 47.7). All were male. Their
current ?rms ranged in size from 6 to 10,500 employees (mean 1,186), and $0.4 million
to $1.2 billion in sales (mean $147.6 million). Of their current ?rms, 43.6% operate in
TABLE 1 Mean Risk Statistics and Choices Among the Four Ventures
Venture Level of Level of Mean Standard Number of Subjects
Name Variability Hazard Risk Score Deviation Choosing Venture
Green High High 5.57 1.22 10
Yellow High Low 4.53 1.43 3
Purple Low High 3.19 1.34 39
White Low Low 1.97 0.96 26
Total 78
Differences between mean risk scores for all ventures are statistically signi?cant (Tukey HSD test) at the 0.01 level.
3
Given our criterion that expected values of all four ventures be equal, the ventures having greater
hazard also have greater potential for gain. This operational condition permits us to gain insights into Lopes
(1987) contention that differences in risk propensity consist largely of a tendency to pay attention to either
the upside or the downside of an anticipated set of outcomes.
RISKS AND CHOICES IN NEW VENTURES 313
TABLE 2 New Venture Descriptions
Venture Green Venture White
There is a 30% chance of being under target by There is a 10% chance of being under target by
$25 million, a 40% chance of meeting target ROI $5 million, a 80% chance of meeting target ROI
and a 30% chance of going over target by $25 and a 10% chance of going over target by $5
million. Graphically the distribution appears as: million. Graphically the distribution appears as:
Venture Purple Venture Yellow
There is a 10% chance of being under target by There is a 30% chance of being under target by
$25 million, a 80% chance of meeting target ROI $5 million, a 40% chance of meeting target ROI
and a 10% chance of going over target by $25 and a 30% chance of going over target by $5
million. Graphically the distribution appears as: million. Graphically the distribution appears as:
TABLE 3 Scale of Risk Propensity
Please answer the following 5 items by circling the alternative (“a” or “b”) you would feel most comfort-
able with.
1 a) an 80% chance of winning $400, or
b) receiving $320 for sure
2 a) receiving $300 for sure, or
b) a 20% chance of winning $1,500
3 a) a 90% chance of winning $200, or
b) receiving $180 for sure
4 a) receiving $160 for sure, or
b) a 10% chance of winning $1,600
5 a) a 50% chance of winning $500, or
b) receiving $250 for sure
314 D. FORLANI AND J.W. MULLINS
TABLE 4 Scale of Perceived New Venture Risk
For each scale below, kindly circle the number which you feel best assesses the amount of RISK associated
with this scenario:
HIGH 1 2 3 4 5 6 7 LOW
MINIMAL 1 2 3 4 5 6 7 EXTREME
VERY RISKY 1 2 3 4 5 6 7 NOT RISKY
manufacturing and 56.4% in service industries. In order to assess any possible non-
response bias, we compared ?rm demographic data for our sample to mean data from
the published lists from which our sample was drawn. Firms in our sample tended to
be somewhat smaller than those on the published lists, in terms of revenue (mean of
$147.6 million for the sample compared to $217.8 million for the listed ?rms) and num-
ber of employees (mean of 1,186 comparedto 1,753). These data suggest that one should
be cautious in generalizing our results to entrepreneurs who lead very large high-
growth ?rms.
Independent Variables
The two manipulated variables were the variability and the degree of hazard (and gain)
of the ventures’ anticipated outcomes, as described previously (see Table 2). Risk pro-
pensity was operationalized using an adaptation of the established (Schneider and
Lopes 1986) Risk Style Scale, as shown in Table 3. We chose this measure for our study
because it deals with personal propensities toward ?nancial risk taking (as opposed to
other kinds of risks, such as those entailed in sky diving) and because of its ef?cacy in
assessing the construct of interest (Schneider and Lopes 1986).
Dependent Measures
The dependent measure of perceived new venture risk for hypotheses H1, H2, and H4
was a scale of 3, 7-point items (see Table 4). The reliability of this scale is indicated
by a coef?cient alpha for a three item scale of 0.956.
The dependent measure of the riskiness of new venture choice for Hypotheses
H3A, H3B, H5, H6, and H7 was the subject’s selection of one of the four ventures to
fund. The mean perceived risk scores for the four ventures indicate that this scale forms
an approximately interval scale of the riskiness of the four ventures (see Table 5). As
expected, the venture scoring highest inperceived risk(Venture Green: meanperceived
TABLE 5 Manipulation Check Results
Manipulation Check
Independent Variable Mean Values p Value
Variability ? 0.001
High 3.79
Low 2.14
Hazard
1
? 0.001
High 4.32
Low 1.97
Mean values are points on a 5-point scale, with 3 as the midpoint.
1
The manipulation of hazard also manipulated potential gain, in order to maintain equal expected values across ventures
(see Table 2). A manipulation check for gain was also signi?cant (means 4.40 and 1.90, p ? 0.001).
RISKS AND CHOICES IN NEW VENTURES 315
risk score 5.57) is the venture having high levels of both variability and hazard. The
venture perceived as next most risky (Venture Yellow: 4.58) has high variability and
low hazard. The venture perceived as third most risky (Venture Purple: 3.19) has low
variability and high hazard. The venture judged least risky (Venture White: 1.97) is the
venture having lowlevels of both variability and hazard(see Table 5). Pairwise compari-
sons between the perceived risk scores for all of the ventures were signi?cant (Tukey’s
HSD test, p ? 0.01).
Manipulation Checks
The effectiveness of the variability and hazard manipulations was assessed by compar-
ing the mean scores on the manipulation check items (each a 5-point scale) for each
venture across the two levels of each experimental condition. As shown in Table 2, all
of the manipulations were effective.
RESULTS
Antecedents of Risk Perceptions
A repeated measures analysis of variance procedure (Norusis 1990) was run to assess
the effects of variability and hazard on perceptions of new venture risk. Main effects
of both variability (F(1,74) ?307.7, p ?0.001) and hazard (F(1,74) ?94.38, p ?0.001)
were found, thereby providing evidence in support of the impact of both the variability
(H1) and hazard (H2) perspectives on risk perceptions. No signi?cant interaction was
found. To assess the effects of the subjects’ risk propensities on the subjects’ risk percep-
tions of eachventure (H4), we rana regressionanalysis of riskpropensity onriskpercep-
tion. The result was not signi?cant (? ? ?0.116, p ? 0.13).
Variability, Hazard, and New Venture Choices
Evidence to test H3A, which predicts greater likelihood of choice of ventures having
less variability in anticipated returns, is foundby comparing the number of higher versus
low variability ventures chosen for funding (see Table 1). Thirteen subjects chose the
high variability ventures (Ventures Green and Yellow), while 65 subjects chose the low
variability ventures (Purple and White). A chi squared test rejects the null hypothesis
that these differences are due to chance (?
2
(1 df) ?34.67, p ?0.01), thereby supporting
H3A (see Table 1).
Evidence to test H3B, which predicts greater chance of choosing ventures having
less magnitude of anticipated losses, is found by comparing the number of high vs. low
hazard ventures chosen for funding (see Table 1). Forty-nine subjects chose the high
hazard ventures (Green and Purple), while 29 subjects chose the low hazard ventures
(Yellowand White). Thus, contrary to the prediction of H3B, entrepreneurs were more
likely to choose high hazard than low hazard ventures, presumably to obtain the poten-
tial for the greater gains which went along with the high hazard condition. Achi squared
test rejects the null hypothesis that these differences are due to chance (?
2
(1 df) ?5.13,
p ? 0.05).
316 D. FORLANI AND J.W. MULLINS
Risk Perceptions, Propensity, and New Venture Choices
To examine H5 and H6, we ran a regression model in which the dependent variable
was new venture choice, and the independent variables were subjects’ risk propensity
scores and their mean total perceptions of risk across the ventures.
4
H5 predicts a positive relationship between risk propensity and new venture
choices. This prediction was supported (? ? 0.26, p ? 0.05), thereby supporting the
prediction of Sitkin and Pablo (1992). The effect of risk perceptions on new venture
choice (H6) was alsosigni?cant (???0.19, p ?0.05). Thus, we ?ndboth riskpropensity
and risk perceptions in?uencing entrepreneurs’ new venture choices. Given the inabil-
ity, in this experimental research, to ascertain the relative “doses” of differences in risk
perception brought about by our manipulations and differences in risk propensities of
the entrepreneurs in our sample, we can draw no conclusive evidence of the relative
importance of one factor compared to the other.
Finally, H7 predicts that, for ventures of equal expected value and equal variability,
more risk prone entrepreneurs will choose more risky ventures (i.e., those having higher
levels of hazard and gain). For the two ventures having low variability, Ventures White
(chosen by 26 subjects) and Purple (chosen by 39 subjects), the results of a logistic re-
gression do not support this prediction (? ? 0.067, p ? 0.74). Similarly, comparing
choices across the two high variability ventures, Yellow (chosen by three subjects) and
Green (chosen by 10 subjects), the results of a second logistic regression also fail to
support this prediction (? ?2.14, p ?0.22). Though the theories we relied on in devel-
oping our study did not predict that risk perceptions would drive these effects, we did
include risk perception in the regression models to test H7. The effect of risk perception
was signi?cant for the Venture Purple vs. Venture White Comparison (? ? ?0.648,
p ? 0.04), but not for the Venture Green vs. Venture Yellow comparison (? ? 4.75,
p ? 0.18).
DISCUSSION
The results provide evidence which extends the work of March and Shapira (1987). Dif-
ferences in entrepreneurs’ new venture choices were in?uenced not only by differences
in the risks inherent in the patterns of anticipated outcomes for different ventures, but
by differences in our entrepreneurs’ perceptions of those risks, as well as their propensi-
ties totake risk. Further, as MarchandShapira predict, the degree of hazardina decision
alternative does in?uence the degree to which entrepreneurs perceive that alternative
as risky, though greater perceived risk did not deter our subjects from making risky
choices: the entrepreneurs in our study were more likely to choose high hazard than
low hazard ventures (63% chose high hazard ventures, which also entailed high gain),
contrary to our prediction. Apparently, high hazard ventures are acceptable as long as
commensurate gains are suf?ciently likely.
Ontheother hand, entrepreneurs appear tobe more apprehensive about variability
in possible outcomes, having overwhelmingly chosen ventures low in variability (83%
chose either Venture White or Purple). Perhaps they are con?dent that they can inter-
vene to increase their chances of achieving the desired outcome when probabilities of
4
Tests of normality, homogeneity of variance, and multicolinearity were run to ensure that the regres-
sion model adequately ?t the data (Neter, Wassermann, and Kutner 1990). No heteroskedasticity, lack of
normality of multicolinearity problems were found.
RISKS AND CHOICES IN NEW VENTURES 317
unexpected negative outcomes are relatively small. Indeed, a recent study suggests that
the overcon?dence of entrepreneurs enables them to start new ventures before many
of the inherent uncertainties therein have been resolved (Busenitz and Barney 1997).
When probabilities of unexpected negative outcomes are large, however, they may feel
that they lack the control to bring in the venture on target.
As for risk propensities, as Sitkin and Pablo (1992) predicted, entrepreneurs who
have greater risk propensities tend to choose riskier ventures. Interestingly, however,
the subjects’ risk propensities did not signi?cantly in?uence their perceptions of venture
risks, contrary to the prediction of Sitkin and Pablo. Thus, risk propensity appears to
directly impact venture choice behavior, rather than indirectly affecting behavior
through the perceptual process. The absence of an effect of risk propensity on risk per-
ceptions is consistent with the ?ndings of a recent study by Palich and Bagby (1995) that
found a consistently optimistic pattern of categorization of business situations among
entrepreneurs compared to nonentrepreneurs, in spite of no difference in risk propen-
sity among the two groups. It may be that various cognitive patterns and processes of
entrepreneurs are more important determinants of entrepreneurial behavior than their
risk propensity. Indeed, such processes may constitute unobserved variables in our re-
search that may be responsible for the effects we found. Our ?nding that risk propensity
in?uences newventure choice behavior, but not risk perception, is parallel to researchin
consumer choice that judgment and choice tasks involve different cognitive operations
(Bettman and Park 1980; Johnson and Russo 1984).
Our results, taken together with previous research ?ndings by Palich and Bagby
(1995) and Busenitz and Barney (1997), suggest that better understanding is needed
about how entrepreneurs search for and process information about business situations
(Cooper, Folta, and Woo 1995; Manimala 1992), and how such information processing
in?uences entrepreneurial behavior. Kahneman and Lovallo (1993) argue that the rela-
tive balance between what they call isolation errors (over-optimistic forecasting that
ignores the statistics of the past on one hand, and overly timid evaluations of single
risky opportunities that neglect possibilities to pool risks on the other) affect the risk-
taking propensities of individuals. These and other errors or biases, including the well-
known discrepancy (Kahneman and Tversky 1979) between the weights attached to
losses and gains in evaluating risky decisions such as new venture opportunities, may
explain why we found that hazard in anticipated new venture outcomes in?uenced per-
ceptions of risk in our study, but did not deter the entrepreneurs in our sample from
choosing riskier ventures. Additional research into the cognitions of entrepreneurs may
offer additional insights to better explain entrepreneurial behavior.
Finally, our results indicate that the series of four hypothetical new ventures that
we used as our dependent measure of new venture risk (See Exhibit 2) constitutes a
scale having approximately interval properties (see Table 1). The availability of a valid
scale of new venture risk should facilitate future experimental research into the role
of risk in new venture decision making.
LIMITATIONS
Our study, like most experimental studies in business settings, suffers fromseveral limi-
tations. First, entrepreneurs in real situations may not behave as did our subjects in the
hypothetical situations inwhichtheywereplacedinour study. Giventhelikely dif?culty,
ina ?eldstudy, of controlling for the broadarray of factors whichare positedtoin?uence
318 D. FORLANI AND J.W. MULLINS
risky newventure decisions in natural settings, and to separate the effects of risk percep-
tions from“objective” risk, we elected to conduct an experimental study. This approach
has advantages in internal validity for theory testing purposes, but may be criticized
on grounds that the experimental task is not a real one with real payoffs. Given the
early stage of research in this arena, and given our interest in testing theories, some
of which have undergone little empirical scrutiny, we deemed the tradeoff acceptable.
Second, we chose to study decisions among a set of four newventure decision alter-
natives whose investments and expected values were all equal—such precise equality
of investments and expected values across proposed ventures is unlikely in real situa-
tions. An additional limitation is the use of a measure of risk propensity borrowed from
another literature, a measure whichmay not adequately capture the propensity of entre-
preneurs to take risks in new venture situations. Bromiley and Curley (1992) argue that
risk propensities are, to some degree, situation speci?c, and that measures from one
situation may not work well in another situations.
Third, in manipulating hazard we also manipulated gain, in order to preserve the
equality of the expected values in the anticipated outcomes of the four ventures from
which our subjects were asked to choose. Had we not done so, there would have little
motivationfor our subjects tochoose highhazardventures (withlittleprospect for gain),
and differences in choices would have been attributable to differences in expected value
among the ventures, rather than to differences in risk propensity and risk perceptions,
the variables of interest in our study.
Finally, we restricted our examination to only the effects of anticipated outcomes
and differences in risk propensity on the decisions that were made. Potentially impor-
tant contextual factors such as competencies and previous experience of the entrepre-
neur, incentives, and team decision process issues were not examined in our study. Fu-
ture research is needed to explore these and other likely in?uences on new venture
decisions.
CONCLUSIONS AND IMPLICATIONS
Our study holds implications for investors who fund entrepreneurial ventures, as well
as for entrepreneurs themselves, and for educators who train students hoping to join
tomorrow’s cadre of entrepreneurs.
To the extent that potential investors in newventures behave similarly to the entre-
preneurs in our study, the impact of variability in anticipated new venture outcomes
on perceptions of risk entailed in, and choices among, new venture alternatives has im-
plications for how entrepreneurs should employ sensitivity analyses and seek to reduce
perceived outcome variability as they seek investment capital. High levels of hazard
did not deter entrepreneurs from choosing ventures with potentially high levels of gain,
therefore, ventures having high levels of variability, such as that often explicitly detailed
in sensitivity analyses, were less likely to be chosen. Entrepreneurs may ?nd it useful
to?ndnewways, other thanthe traditional best case/expectedcase/worst case approach,
to present to prospective investors the likelihood of deviation in future performance
from desired outcomes. Our results may offer an explanation of why entrepreneurs are
sometimes reluctant to engage in explicit risk and sensitivity analyses for proposed ven-
tures. As Ulrich and Epplinger (1995: 256) suggest, “Often [decision makers] do not
want to confront the true probabilities of bad outcomes.” Rather than attempting to
mask the outcome variability of presenting new ventures, entrepreneurs may be better
RISKS AND CHOICES IN NEW VENTURES 319
advised to invest in suf?cient market research to ascertain the level of market accep-
tance which the products or services of a proposed new venture are likely to enjoy, or
to test the criticality of key assumptions underlying their plans (Sykes and Dunham
1995). Several new techniques of qualitative market research are able to provide tangi-
ble evidence of likely market acceptance, even for new-to-the-world products and ser-
vices, thereby reducing perceptions of outcome variability and helping to build a solid
foundation of customer needs on which to erect the new venture (Grif?n and Hauser
1993; Grif?n 1996; Zaltman 1997).
For entrepreneurs choosing among a set of proposed newventure alternatives, our
results indicate that, where levels of investment and the expected values of returns are
similar, ventures tend to be chosen based on differences in risk propensities among en-
trepreneurs, in addition to the risk entailed in the ventures’ patterns of anticipated re-
turns. This ?nding attests to the importance that venture capitalists and other investors
place in the people who lead the ventures in which they invest (Heilemann 1997; Sahl-
man 1997). Entrepreneurs are advised to explicitly ask themselves whether their assess-
ment of proposed ventures they consider are biased in any way by their propensities
to take risks. Such biases could result in decisions which lengthen the already daunting
odds for new venture success. Various approaches have been identi?ed for overcoming
decision biases, including the use of structured decision aids (Ghosh and Ray 1997) and
treating a particular decision as an instance of a broader class of similar previous deci-
sions about which outcome information is available (Kahneman and Lovallo 1993).
Additionally, entrepreneurs in search of new venture opportunities should think
carefully about strategies for mitigating the hazard they appear willing to accept in
search of potentially attractive gains. Failure to adequately understand and plan for
adverse outcomes not only may jeopardize chances for near term success, but they may
make it more dif?cult to attract support from investors and others for subsequent ?-
nancing (Sahlman 1997).
For teachers whose work it is toprepare a newgeneration of entrepreneurs tomake
wise choices among newventure opportunities and develop successful strategies to pur-
sue them, our study suggests that providing would-be entrepreneurs with tools, tech-
niques, and analytical frameworks for reducing the variability in their forecasts of new
venture outcomes can play an important role in facilitating their pursuit of potentially
attractive, but risky, opportunities. Potentially useful tools and techniques include vari-
ous qualitative and quantitative market research approaches (cf., Grif?n and Hauser
1993; Grif?n 1996; Mahajan, Muller, and Bass 1990; Zaltman 1997); relevant analytical
frameworks include critical assumption planning (Sykes and Dunham 1995) and risk
analysis (Hertz 1964). Use of such tools and frameworks to provide a strong foundation
of evidence on which to build a business plan provides two bene?ts to students as would-
be entrepreneurs: ?rst, stronger evidence supporting the attractiveness of a proposed
venture is likely to enhance the likelihood of obtaining funding for the venture; second,
as the results of our study suggest (recall that our subjects were more likely to choose
ventures having less variability in anticipated outcomes), by eliminating uncertainty and
reducing variability in anticipated outcomes, such evidence will make it more likely that
the entrepreneur will, given favorable evidence, decide to pursue the venture. This psy-
chological role of such evidence should not be underestimated.
Our results raise questions which call for additional research into the risky choice
decisions of entrepreneurs. First, the person versus situation debate in organizational
psychology (Mischel 1977, O’Reilly 1991) has wrestled in recent years with trying to
320 D. FORLANI AND J.W. MULLINS
better identify the kinds of situations inwhichindividual differences tendtobe relatively
more or less important, compared to situational factors, in determining behavior. For
risky new venture decisions such as those we have examined here, what situational or
contextual factors are likely to moderate the importance of risk propensity and other
individual differences in risky choice decisions? March and Shapira (1992) found, in a
simulation study, that recent performance outcomes, accumulated resources, and the
goals or reference points that decision makers attend to in?uence risky choice behavior.
BairdandThomas (1985) articulateda model of strategic risktaking whichincorporated
a broad array of environmental, organizational, industrial, decision maker, and problem
variables. Few of these variables have been empirically studied as to their effects on
risky choice decision making. These works suggest numerous directions for future em-
pirical research.
How our subjects’ venture choices would have played out in terms of new venture
performance was not addressed in our study, but the apparent biases which individual
differences in risk propensity generate in new venture decision making is a potentially
important issue for future research. Viewed from one perspective, our ?ndings suggest
that patterns of suboptimal decisions are likely to be common, as managers’ inertia (Sit-
kin and Pablo 1992) and proneness or aversity toward risk (Lopes 1987; Schneider and
Lopes 1986) lead to consistently risk-prone or risk-avoiding decisions. Viewed from a
more optimistic perspective, however, our results suggest that people matter in such
decisions (Hitt and Tyler 1991). Entrusting new venture investments to individuals
whose risk propensities and other individual characteristics best match the needs of the
market opportunity and a prospective investor’s objectives may help improve portfolio
performance. The considerable effort which venture capitalists and other investors put
forth in evaluating the individuals whose ideas and companies they fund attests to the
importance of individual differences, including differences in risk taking propensity, for
new ?rm success.
REFERENCES
Armour, H.O. and Teece, D.J. 1978. Organization structure and economic performance: a test
of multidivisional hypothesis. Bell Journal of Economics 9:106–122.
Baird, I.S. and Thomas, H. 1985. Toward a contingency model of strategic risk taking. Academy
of Management Review 10(2):230–243.
Bettman, J.R. and Park, C.W. 1980. Effects of prior knowledge and experience and phase of the
choice process on consumer decision processes: a protocol analysis. Journal of Consumer
Research 7:234–248.
Bird, B.J. 1989. Entrepreneurial Behavior. Glenview, IL: Scott, Foresman and Company.
Bowman, E.H. 1980. Arisk/return paradox for strategic management. Sloan Management Review
21 (Spring):17–31.
Bromiley, P. and Curley, S.P. 1992. Individual differences in risk taking. IN J.F. Yates, ed., Risk
Taking Behavior. West Sussex, England: John Wiley and Sons.
Busenitz, L.W. and Barney, J.B. 1997. Differences between entrepreneurs and managers in large
organizations: biases and heuristics in strategic decision-making. Journal of Business Ven-
turing 12:9–30.
Cooper, A.C., Folta, T.B. andWoo, C. 1995. Entrepreneurial information search. Journal of Busi-
ness Venturing 10:107–120.
Devinney, T.M. 1992. New products and ?nancial risk changes. Journal of Product Innovation
Management 3:222–231.
RISKS AND CHOICES IN NEW VENTURES 321
Dickson, P.R. 1992. Toward a general theory of competitive rationality. Journal of Marketing
56:69–83.
Fisher, N. and Hall, G. R. 1969. Risk and corporate rates of return. Quarterly Journal of Econom-
ics 83:79–92.
Ghosh, D. and Ray, M.R. 1997. Risk, ambiguity, and decision choice: some additional evidence.
Decision Sciences 28(1):81–104.
Grif?n, A. 1996. Obtaining customer needs for product development. In M.D. Rosenau, ed., The
PDMA Handbook of New Product Development, New York: John Wiley and Sons, pp.
153–166.
Grif?n, A. and Hauser, J.R. 1993. The voice of the customer. Marketing Science 12(1):1–27.
Hall, J. and Hofer, C.W. 1993. Venture capitalists’ decision criteria in new venture evaluation.
Journal of Business Venturing 8:25–42.
Heilemann, J. 1997. The networker. The New Yorker August 11:28–36.
Hertz, D.B. 1964. Risk analysis in capital investment. Harvard Business Review (January-Febru-
ary): 96–106.
Hitt, M.A. and Tyler, B.B. 1991. Strategic decision models: integrating different perspectives.
Strategic Management Journal 12:327–351.
Johnson, E. and Russo, J.E. 1984. Product familiarity and learning new information. Journal of
Consumer Research 11:542–550.
Kahneman, D. and Lovallo, D. 1993. Timid choices and bold forecasts: a cognitive perspective
on risk taking. Management Science 39(1):17–31.
Kahneman, D. and Tversky, A. 1979. Prospect theory: an analysis of decision under risk. Econo-
metrica 47:263–290.
Krueger, N., Jr. and Dickson, P.R. 1994. How believing in ourselves increases risk taking: per-
ceived self-ef?cacy and opportunity recognition. Decision Sciences 25(3):385–400.
Lopes, L.L. 1987. Betweenhope and fear: the psychology of risk. Advances inExperimental Social
Psychology 20:255–295.
MacCrimmon, K.R. and Wehrung, D.A. 1986. Taking Risks: The Management of Uncertainty.
New York: Free Press.
MacCrimmon, K.R. and Wehrung, D.A. 1990. Characteristics of risk taking executives. Manage-
ment Science 36(4):422–435.
Mahajan, V., Muller, E. andBass, F.M. 1990. Newproduct diffusionmodels inmarketing: a review
and directions for research. Journal of Marketing 54 (January):1–26.
Manimala, M.J. 1992. Entrepreneurial heuristics: a comparison between high PI (pioneering-in-
novative) and low PI ventures. Journal of Business Venturing 7:477–504.
March, J.G. and Shapira, Z. 1987. Managerial perspectives on risk and risk taking. Management
Science 33(11):1404–1418.
March, J.G. andShapira, Z. 1992. Variable riskpreferences andthe focus of attention. Psychologi-
cal Review 99(1):172–183.
Mischel, W. 1977. The interaction of person and situation. In D. Magnusson and N.S. Endler,
eds., Personality at the Crossroads: Current Issues in Interactional Psychology, Hillsdale,
NJ: Erlbaum.
Mullins, J.W. 1996. Early growthdecisionof entrepreneurs: the in?uence of competency andprior
performance under changing market conditions. Journal of Business Venturing 11:89–105.
O’Reilly, C.A. III. 1991. Organizational behavior: where we’ve been, where we’re going. Annual
Review of Psychology 42:427–485.
Palich, L.E. and Bagby, D.R. 1995. Using cognitive theory to explain entrepreneurial risk-taking:
Challenging conventional wisdom. Journal of Business Venturing 10:425–438.
Phillips, B.D. andKirchoff, B.A. 1988. Ananalysis of new?rmsurvival andgrowth. BabsonEntre-
preneurship Research Conference, Calgary.
Reynolds, P. 1986. Predicting contributions and survival. In Ronstadt, et al. eds., Frontiers of
Entrepreneurship Research. Wellesley, MA: Babson College.
322 D. FORLANI AND J.W. MULLINS
Riquelme, H. andRickards, T. 1992. Hybridconjoint analysis: anestimationprobe innewventure
decisions. Journal of Business Venturing 7:505–518.
Sahlman, W.A. 1997. How to write a great business plan. Harvard Business Review (July-
August):98–108.
Schneider, S.L. and Lopes, L.L. 1986. Re?ection in preferences under risk: who and when may
suggest why. Journal of Experimental Psychology: Human Perception and Performance
12(4):535–548.
Shapira, Z. 1995. Risk Taking: A Managerial Perspective. New York: Russell Sage.
Shoemaker, P.J. 1990. Are risk-attitudes related across domains and response modes? Manage-
ment Science 36(12):1451–1463.
Sitkin, S.B. and Pablo, A.L. 1992. Reconceptualizing the determinants of risk behavior. Academy
of Management Review 17(1):9–38.
Sitkin, S.B. and Weingart, L.R. 1995. Determinants of risky decision making behavior: a test of
the mediating role of risk perceptions and risk propensity. Academy of Management Jour-
nal 38(6):1573–1592.
Sykes, H.B. and Dunham, D. 1995. Critical assumption planning: a practical tool for managing
business development risk. Journal of Business Venturing 10:413–424.
Timmons, J.A. 1994. New Venture Creation. Homewood, IL: Irwin.
Ulrich, K.T. and Epplinger, S.D. 1995. Product Design and Development. New York:
Mcgraw-Hill.
Yates, J.F. 1990. Judgment and Decision Making. Englewood Cliffs, NJ: Prentice-Hall.
Yates, J.F. and Stone, E.R. 1992. The risk construct. IN J.F. Yates, ed., Risk Taking Behavior,
West Sussex, England: John Wiley and Son.
Zaltman, G. 1997. Rethinking market research: Putting people back in. Journal of Marketing
34:424–437.
doc_367008731.pdf