Taylor & Francis Publishers Award for Excellence in Research on the Topic of Venture Capital
THE STAGING OF VENTURE CAPITAL INVESTMENTS AND REINVESTMENT DECISIONS
Carolyn
Birmingham, University of Oklahoma
Lowell Busenitz, University of Oklahoma
Jonathan Arthurs, University of Oklahoma
While venture capitalists are often considered to be expert decision makers, errors in their judgment are widely evident. To be successful and profitable on a continuing basis, VCs need to be able to cut their losses in a timely manner and get out of an investment that is not materializing. This study examines VC investment patterns among 330 early stage ventures. Patterns of escalation, that is reinvestment in ventures headed towards failure, were found among 66 funded ventures. The number of VCs participating in the first round of funding, and any increase or decrease in the number of VCs investing in a venture over time, as well as the timing of first round funding all had a significant impact on the tendency to engage in escalation of commitment. Theoretical implications are discussed.
Venture capitalists (VCs) are investors who specialize in financing and assisting entrepreneurs with their startup ventures (Bygrave & Timmons, 1992). Without access to institutional lenders and public financial markets, entrepreneurs of high growth potential ventures often look to VCs for capital and other forms of assistance (Sapienza, 1992). To deal with the risk and uncertainty inherent in these ventures, VCs usually stage their financing and form syndicates by including other VCs in their investment. Staged financing involves periodic investments in a venture that allow it to grow while limiting access to excess cash. The staging of capital infusions allows VCs to monitor a firm’s progress and gather additional information as it becomes available.
Given their focus and experience with entrepreneurial ventures, VCs are often considered to be expert decision-makers. However, it is clearly evident that VCs frequently make errors in judgment about a venture’s potential (Zacharakis & Meyer, 1998). Approximately one-third of all the ventures in which they invest make it to an initial public offering (IPO) with the other VC-backed ventures being acquired, going out of business, or eking out a meager existence known as “living dead.”
With ventures that do not materialize as expected, VCs should ideally cut their losses in a timely manner and exit the investment. This study examines the escalation of commitment among venture capitalist syndicates investing in the earlier stages of entrepreneurial ventures. The contributions of this study are at least twofold. First, although support for escalation of commitment in decision-making has been found in numerous laboratory studies, this study confirms this phenomenon in a large-sample study over time involving real-world decisions. Second, this study indicates that VCs sometimes escalate in their venture investments and identifies the conditions under which escalation is more likely to occur.
Most VC decision-making studies have focused on the initial decision to invest. Several studies have also looked at the exit decisions such as the timing of an IPO offering or acquisition by another company (Bygrave & Timmons, 1992; Lam, 1991; Lerner, 1994; Norton, 1993). Very little is known about VC decision-making from the time of initial investment to the exit decision. This study contributes to this important gap by examining the escalation of commitment phenomenon and the staging of VC investments in entrepreneurial ventures.
Escalation of commitment to a failing course of action occurs when individuals or groups decide to commit additional resources to a project that is on a failing course (Staw, 1976). Once decision-makers commit to an initial decision, the tendency is for them to become overly committed and ignore subsequent information, particularly negative information (Brockner, 1992). Four stages in the escalation of commitment decision process have been identified. First, the initial decision to invest is made when there is promise of future favorable outcomes and so the investment is made. Second, feedback is received that is ambiguous but that brings into question whether or not the original goals will be able to be met. Investors may consider this a temporary setback rather than an indicator of an error in judgement (e.g. investing in a ‘dog’). Third, they continue to invest, however, progressively more negative feedback is received. Fourth, the investor continues to invest despite this negative feedback, in what becomes a failing course of action (Brockner, 1992; Staw, 1981, 1976).
Most of the research on escalation of commitment has taken place in the laboratory or by analyzing case studies. These studies document that escalation is more likely to occur when the investor responsible for making the original decision to invest is the one making the decision to reinvest. The investor is loss framed (Rutledge, 1995) when there would be a negative outcome to the original investor for pulling out of the project. Both Prospect theory and Self-Justification theory have been employed to provide explanations for this escalation phenomenon (Brockner, 1992; Ross & Staw, 1993; Staw & Ross, 1987; Rutledge, 1995; Staw, Koput, & Barsafe, 1997: Staw, 1981, 1976).
In an attempt to validate findings of escalation outside the laboratory, Staw, et al. (1997) conducted a field study to examine bad loan write-offs in the banking industry. They found that the lack of personal responsibility for granting a bad loan resulted in decreased escalation (e.g. more willingness to write off the bad loan). In another large-scale study, entrepreneurs were found to escalate when they started their own firms and were over-confident concerning the likelihood for success (McCarthy, Schoorman, & Cooper, 1993).
We argue that the VC investment process represents an important setting in which to further examine whether escalation of commitment does exist. If this decision-making phenomenon were present among VCs, we would expect to see them make repeated investments in failing or low-growth potential ventures (commonly referred to as “living dead” among VCs). This study allows us to clarify whether or not escalation of commitment occurs among VCs and provides theoretical direction to VC decision-making after the initial decision to invest. Furthermore, it allows us to probe why VCs sometimes make bad decisions.
The staging of rounds of funding and the reinvestment decisions that VCs make provide an ideal setting to examine the potential for escalation of commitment. Instead of providing all the capital initially, VCs practice staged financing which provides the opportunity to make repeated funding decisions based on the progress of the venture over time. While there are many positive aspects to the staged financing process, it may also give rise to the possibility of escalation of commitment. VCs often get personally involved in the venture by giving advice and input. Furthermore, the progress and performance of a startup firm is rarely linear. Instead, venture performance is commonly up and down with numerous difficulties and hurdles to overcome. Consequently, the staged funding process will often be surrounded by ambiguous and negative feedback. Since the new-venture investment setting is ideal for the potential for escalation of commitment, it may be quite commonplace among VCs.
Hypothesis 1: Escalation of commitment by VCs will be evident by a greater number of rounds of financing in ventures that eventually go out of business or remain “living dead” compared to those that are acquired or go to IPO.
Building from the assumption that escalation of commitment does occur among VCs, we now develop arguments and hypotheses for why specific factors are likely to impact VC decision-making. We suspect that there are patterns to the timing and investment behavior of those who escalate versus those who do not. We then develop a series of hypotheses regarding VC syndication patterns and how they may impact the tendency for VCs to escalate.
Investors can either base their initial investment decision on the reputation of the entrepreneur or the perceived worthiness of the entrepreneurial idea/product. VCs sometimes talk about investing in an A-level management team even though they may only have a B-level idea (Bygrave & Timmons, 1992). In a series of case studies, Ryan (1994; 1995) observed that VCs who invested in a person rather than an idea were more likely to escalate. When investors invest in the entrepreneur, a company must first be created to pursue the entrepreneur’s idea. The felt responsibility of the investors when they not only invest in the person but also create a new company is more likely to be higher than when they invest in a pre-existing entrepreneurial company. Consistent with past research (Staw, et al. 1997), felt responsibility increases commitment to the project and therefore increases the chances for escalation. In addition, McCarthy et al. (1994) found that independent opinions from advisors who felt personally responsible for the venture were less likely to be objective in their assessment. We suspect that when VCs choose to invest in a firm within the first year of the venture’s existence, they are in all likelihood, investing in the founding team.
Furthermore, when VCs invest in a venture within its first year of existence, exuberance about the venture will be less checked by negative feedback and problems. Once a decision to invest in a venture has been made public, there is a tendency to more strongly support the initial decision (Staw, et al. 1997). VCs will believe that it is a good investment and that the venture will one day attain success. Although these ventures often suffer setbacks, the VC may view this feedback as equivocal. These arguments lead to the following hypothesis:
Hypothesis 2: VCs are more likely to escalate in their staging of investment decisions if they invest in a venture that is in its first year of existence.
VC Syndication and Escalation of Commitment
Venture capitalists usually invest in new ventures as groups or syndicates to improve their decision-making and to diversify their risk (Bygrave, 1987). Once a VC has identified an entrepreneurial venture that they think has high potential, they typically send the proposal to other VCs for their consideration and review. When additional VCs are willing to invest in a new venture, this sends an important signal and confirmation to the lead VC that it is a legitimate investment (Lerner, 1994). The syndication process provides an opportunity for VCs to check their own evaluation of a venture against the insights and judgments of other VCs. These arguments suggest that when a group of VCs invest together, they are likely to make better decisions and not escalate in later rounds of funding.
More specifically, Lerner (1994) argued that VC syndication was particularly important for first round funding and the decision of whether to invest in risky firms. His results were consistent with the view that established VC firms typically syndicate together for first round funding. We suspect that the greater the number of VCs in the syndicate for first round funding, the better will be the information regarding the investment decision. Furthermore, we believe that the greater the number of VCs investing in the first round of funding, the less likely that escalation will occur in subsequent rounds.
Hypothesis 3: The greater the number of VCs involved in first round funding of a venture, the less likely it is that VCs will engage in escalating decision-making in subsequent rounds.
The size of a VCs syndicate may be good only to a point. When the number of VCs in a syndicate gets too large, we suspect that the probability of escalating will increase. Earlier research suggests that larger groups tend to have a more difficult time making good decisions (McGrath, 1983). Individual VCs may be more likely to raise probing questions about the viability of a venture among a smaller group. However, with a larger group, it may be that VCs will succumb to groupthink or the opinions of the lead VC. Thus, we hypothesize that there will be curvilinear relationship between the number of VCs in the syndicate and escalation of commitment.
Hypothesis 4: The number of VCs per round will decrease escalation of commitment to a point followed by an increase in probability of escalation of commitment.
The composition of VC syndicates changes from round to round. Some VC firms tend to specialize and only invest in later rounds or vice versa. The needs of entrepreneurial ventures change over time creating a need to bring new VCs into the syndicate with different skill sets. Other VCs that invested in the early rounds may choose not to reinvest in subsequent rounds because they think that the venture does not hold the promise that it once did. Either way, when new VCs are added to or subtracted from a syndicate, new input and evaluation regarding the viability of the venture will be communicated. This new input is valuable because once the decision is made to invest, the investors have come to a common understanding of the situation and as a result there is less diversity of opinion. This could create problems after the initial investment decision because diversity of opinion leads to better decisions due to the additional information and viewpoints available (Jehn, 1997; Arrow & McGrath, 1994; McGrath, 1983; Birmingham & Michaelson, 1999). Stated differently, when the number of VCs in the syndicate remains constant, re-investment decisions are less likely to be challenged. Information that formed the foundation for earlier decisions is likely to remain in place and new negative information will be ignored.
In addition, when the number of VCs in the syndicate does not change, the VC group should become more cohesive. Cohesiveness increases the danger of homogeneity of opinion about the interpretation of a series of events potentially leading to inferior decisions (Arrow & McGrath, 1994; McGrath, 1983). As a result, a lack of fresh insight via VCs leaving or joining the syndicate will increase a group’s tendency to escalate (Bazerman, Giuliano & Appelman, 1984; Brockner, 1992; Rutledge, 1995). These arguments lead to the following hypothesis:
Hypothesis 5: Increases and decreases in the number of VCs involved in later rounds of funding will decrease the probability of escalating in VC re-investment decisions.
The role of original members of a group or syndicate may also impact decision-making. For example, when the original decision-maker is removed from the decision process, a new decision-maker is less likely to escalate (Staw, et al. 1997). Lead and first round VC’s have more public commitment to an investment decision. They bring in other VC investors (Bygrave & Timmons, 1992, Lerner, 1994) and so they may have greater felt responsibility for the outcome. In addition the lead investor is often more intimately associated with the management of the venture (Sapienza, 1992). First round VCs usually are involved in the management of the new venture and so have additional public commitment and responsibility to the venture. If VC-backed ventures do not make the projected targets, the VC syndicate can become loss framed and increase the risks taken to meet the targets (Rutledge, 1995). These factors would increase the likelihood of escalation leading to the following hypothesis.
Hypothesis 6: When at least one VC from the first round of funding continues to invest in all subsequent rounds of funding, escalation is more likely.
Data
To examine the possibility of escalation of commitment among VCs in their decisions to fund new ventures, we needed investment information regarding the rounds of funding. The VentureXpert Web database offered by Thompson Financial/Venture Economics provided us this information. From this database, we identified a total of 485 companies in the high velocity environments of computer hardware, software, telecommunications and semiconductor industries that received VC funding during the 1990-2000 timeframe. Of these, 330 received first round funding during the 1990-1995 timeframe. The exit status of these companies was observed in 2000 (IPO, acquired, alive but not public or successful, bankrupt, out of business, and presumed dead).
Dependent Variable
Based on previous research (Ruhnka, Feldman, and Dean 1992; Gladstone 1989), we categorized the final disposition of VC-funded ventures in the following manner: (1) out-of-business, (2) still-private, (3) merged or acquired, and (4) IPOs. This categorical approach was chosen because continuous measures of performance are readily available only for IPO firms, thus eliminating the inclusion of out-of-business, still-private, as well as ventures that have been acquired by another firm. This categorization of venture outcomes is consistent with the VC literature (Ruhnka, Feldman, and Dean 1992; Gladstone 1989) and it may be the best way of capturing the outcomes of all ventures that have received VC funding.
To determine a venture’s exit status, we searched the Lexis-Nexis business database, specifically the business news and company financial sections. Lexis-Nexis catalogs stories about both publicly and privately held ventures from such sources as news articles, wire and transcript articles, magazines and trade magazines, newsletters, journals, disclosure reports, and bankruptcy reports. We also recorded the year that a change in a venture exit status occurred. We classified a firm as out-of-business if a report/article indexed by Lexis-Nexis explicitly stated this to be the case or if no trace of information was found on the specified firm for at least two consecutive years. As a final check, we examined the online business phone directory at “www.switchboard.com” for a possible listing before a firm was categorized as out-of-business.
In studying escalation of commitment, it is generally impossible to determine a priori which series of decisions will result in escalation of commitment (e.g. will be bad decisions) and which ones will be good decisions (Staw, et al., 1997). With regard to VC investments in high-growth, high-risk ventures, we argue that there are two fundamentally different categories of good investment decision-making. The first is knowing when to continue investing in a venture that is going to be successful (e.g. the venture makes it to IPO or acquisition). The second type of good decision is knowing when to stop investing in a venture that is not going to be successful, thereby minimizing investment losses. We considered it good decision-making or non-escalating when funding for a living dead or dead firm was limited to either one or two rounds of funding. Apparently the VCs recognized early on that they were backing a company in serious trouble and cut their losses by bailing from further investments.
We identified escalation of commitment among VCs as a series of decisions involving three or more rounds of funding with the venture ultimately going out of business or attaining a “living dead” status. In these cases, the VCs involved apparently failed to recognize that they were backing a loser and failed to cut their losses early. When the escalation of decision making processes is at work, VCs fail to recognize it as such and are more likely to interpret any bad news as only a bump in the road to success (Brockner, 1992; Staw, 1981, 1976).
Independent and Control Variables
In order to understand the dynamics of VC funding, we included the following seven independent variables in our analysis: 1) total number of VCs investing in the venture, 2) the presence of a first-round VC investing in every subsequent round (dummy coded 1 or 0), 3) the number of VCs investing in the first round of funding, 4) the number of VCs investing in the first round of funding squared (in order to identify any curvilinear relationship), 5) any increase in the number of VCs investing from round to round (dummy coded 1 or 0), 6) any decrease in the number of VCs investing from round to round (dummy coded 1 or 0), and 7) ventures that were first funded by VC’s within the first 12 months after they were founded (dummy coded 1 or 0).
We expect that in the escalating ventures, the number of VCs would remain constant. Those syndicates that had new VCs join or leave the investment, and therefore additional information or signals about the direction the venture is headed, should have less potential for groupthink (Janis, 1982) that would lead to escalation. For those ventures that were funded the year they were founded, we would expect a greater potential for escalation of commitment from the VCs. Those ventures that were founded earlier (more than 12 months before the year they were funded) will have had some operating history that would assist the VCs in making a prudent decision concerning the likelihood of success or failure. Without knowledge of any historical information concerning the operations of a venture, the VCs are investing in the entrepreneur and his or her idea. These first positive impressions concerning the entrepreneur and his or her venture idea may subsequently cloud the VC’s future judgment.
We included three control variables: 1) industry code, 2) the year of founding of the venture, and 3) the total dollars invested by VCs. The first two variables will control for any industry or general economy effects that would bias the results. The total dollars invested variable will assist in controlling for the level of funding that would differ based on the strategy of the venture. For example, those that were seeking a highly profitable niche would require less funding than those seeking to dominate a category.
Data Analysis
The analysis of data over time requires special care. For example, some ventures that were still private could change to out of business, merged, or an IPO, which would create a right censoring problem. In addition, a change in the timing of the status changes of only a few ventures could have an impact on the interpretation of results. Finally, a firm that undergoes an IPO after only two years is more likely to be a profitable than one that takes nine years. We used an event history analysis, more specifically Cox regression, to accommodate the time sensitive nature of the longitudinal data (Allison, 1984; Yamaguchi, 1999).
In event history analysis, the dependent variable is the hazard rate, which is a function of the probability that a firm in the risk set will have a particular outcome during a particular time interval. For this study, we calculated the hazard rate by multiplying a venture’s status by the number of years from the beginning of the study to the year of change.
Our findings suggest that VCs do indeed sometimes escalate in their decision making, that is, they continue to make investments in firms that are on a failing course. The cross tabulation results shown in Table 1 indicate a total of 66 ventures to be in the escalating dead and living dead categories. With this class of firms, it is interesting to note that not only did the VCs continue to invest in these firms even though they were on a failing course of action, but the drop-off in investments in later rounds of funding was particularly slow. For example, of the 100 firms that eventually went public, only 36 needed fourth-round funding. Out of the 66 escalating firms, 57 received fourth-round funding. In order to test Hypothesis 1, we examined the sample of firms that received more than two rounds of funding. We split this sample into two groups; group 1 (N = 102) included the ventures that eventually went to IPO or were acquired and group 2 (N = 66) included those ventures that went to living dead or dead status. We wanted to check whether there was a significant difference between the mean number of rounds of financing between the two groups. Group 1 (non-escalating) had a mean of 4.39 rounds of funding (sd = 1.18) whereas group 2 (escalating) had a mean of 4.82 rounds of funding (sd = 1.20). We conducted a Welch-Satterthwaite two independent sample t-test to compare the means. The results indicate that there is a significant difference between the two groups in the average number of rounds funding; t = 2.26; p = .025. Thus, it appears that VCs do indeed have the capacity to escalate and persist in funding ventures on a failing course.
Hypothesis 2 proposed that when VCs made first round investments within twelve months of founding, the probability of escalation would significantly increase. Weak support was found for this hypothesis in the escalating versus IPO/acquired class. The hazard ratio suggests that the probabilities go up about 40% in this model. No support was found in the escalation versus dog analysis. The conjecture that VCs committing venture funding within the first year of the life of a venture encourages escalation received modest support.
Hypotheses 3 and 4 examined the effects of the number of VCs getting involved in first-round funding. The greater the number of VCs investing in the first round, the greater the likelihood that escalation will occur. Hypothesis 3 was supported in both the IPO/acquired model (.05 level) and the “dog” model (.10 level). Hypothesis 4 proposed a possible curvilinear relationship between the number of VCs involved and the likelihood for escalation. That is, when the number of investing VCs is too high, it was posited that escalation will be more likely. The IPO/acquired model shows support at the .10 level with no support in the “dog” model. Only weak support was found for hypothesis 4.
Hypothesis 5 examined the effects of any increases or decreases in the number of VCs in the investment syndicate. As shown in Table 2, both increasing and decreasing the number of VCs appears to have significant effects as reflected in both models. This suggests that the addition or subtraction of VCs from the syndicate team stimulates further inquiry or perhaps more careful scrutiny of the reinvestment decision. The hazard ratio indicates that for each increase or decrease in the number of VCs to the syndicate, the probability of escalating decreases from 40 to 80 percent.
We also hypothesized that those ventures in which VCs involved in the first round funding invested in all subsequent rounds would have a greater likelihood of experiencing escalating decision making. However, as reflected in Table 2, hypothesis 6 was not supported in either model.
In this research we explored escalation of commitment of VCs when they invest in early stage ventures using staged funding. This study provides further evidence beyond the laboratory that escalation of commitment exists even with expert decision-makers such as VCs.
People are more likely to search for and interpret information to support their existing preferences and beliefs (Nisbett & Ross, 1980). VCs believe that the venture they have backed is a potential winner as reflected in their financial and public commitment to the firm. VCs who don’t take steps to counter their biases and information processing errors are more likely to miss problems and negative signals leading to escalating decisions. In order to avoid escalation of commitment to a failing course of action, the VC group needs to be aware of it (Montealegre & Keil, 2000). The findings of this study suggest when VCs move in and out of a syndicate, escalation of commitment is less likely to occur.
When groups such as VC syndicates make decisions, a common problem is focusing attention on new or unique information. Failure to do so often results in inferior decisions (Arrow & McGrath, 1994; McGrath, 1983). Conflict can help get new and unique information and views out into the open and fully discussed and considered (Jehn, 1997). When there is less cognitive conflict, other ways are needed to make it more likely that VCs will consider alternative opinions. When a VC leaves a venture, this is likely to send a signal that there are alternative assessments of the venture (e.g., that the venture is failing).
Another way to draw attention to new information that does not fit the existing VC consensus is to bring in a new contributor to the group. In other contexts an independent evaluation of a project decreases the likelihood of escalation of commitment (Keil & Robey, 1999; Ghosh, 1997). After the first round of funding when a new VC is invited to join the group, presumably this new VC will conduct an independent evaluation of the venture and discuss his or her assessment with at least the lead VC. As a result, alternative interpretations of the situation and new information are more likely to be brought to the forefront. If any of the information is potentially negative, action can be taken to help make better investment decisions. While it is unlikely that a VC would join a venture that he/she assessed as likely to fail, the insights that the new VC has may well improve the potential of the venture. This needs further study.
Adding too many VCs may create additional problems. As groups get larger their dynamics change. There can be diffusion of responsibility (Janis, 1982) and increasingly more unresolved conflict (Jehn, 1997, Birmingham & Michaelsen, 1999). As a result the quality of the oversight of the venture may decline leading to escalation of commitment. Another possibility is that when groups get too large, they spread their risk in a venture without giving the venture careful and repeated evaluations.
Further research is needed on whether there are differences between outcomes when VCs back primarily an entrepreneurial team versus a business concept. When VCs funded a venture within the first 12 months of founding, we took this to indicate that they were first and foremost funding the entrepreneurial team rather than an idea. However, we suggest that future research should examine this more explicitly. Furthermore, the experience of the VCs investing in a venture and how VC experience affects the assembling of a syndicate also needs further investigation.
Because this study used secondary data, we were unable to examine some of the actual processes that took place during the VC funding decision process. Certainly it would be worth examining more directly how VCs get into an escalating situation as well as how they prevent it. While theory can be used to offer reasonable explanations for the results, process data needs to be collected in order to see if these explanations are accurate.
CONTACT: Lowell Busenitz, University of Oklahoma, Management, Norman, OK 73019-4006; (T) 405-325-2653; (F) 405-325-1957; Busenitz@ou.edu
Allison, P.D. Event history analysis: Regression for longitudinal event data. Newbury Park, CA: Sage. 1984.
Arrow, H. & McGrath, J.E. (1993). Membership matters: How member change and continuity affect group structure, process, and performance. Small Group Research, 24, 334–361
Bazerman, M.X., Giuliano, T., Appelman, A. (1984). Escalation of commitment in individual and group decision making. Organizational Behavior and Human Performance, 33(2), 141–152.
Birmingham, C., & Michaelson, L.M. (1999). The use of cooperation as a conflict resolution strategy in longitudinal groups. Proceedings—Midwest Academy of Management, Chicago.
Brockner, J. (1992). The escalation of commitment to a failing course of action: Toward theoretical progress. Academy of Management Review, 17(1), 39–61.
Bygrave, W.D. (1987). Syndicated investments by venture capital firms: A networking perspective. Journal of Business Venturing, 2, 139–154.
Bygrave,W.D., & Timmons, J.A. (1992). Venture Capital at the Crossroads. Boston: Harvard Business School Press.
Gladstone, D. (1989). Venture capital investing, Englewood Cliffs, NJ: Prentice Hall.
Ghosh, D. (1997). De-escalation strategies: Some experimental evidence Behavioral. Research in Accounting, 9, 88–112.
Janis, E. (1982). Victims of groupthink. Boston:Houghton Mifflin.
Jenh, K.A. (1997). A qualitative analysis of conflict types and dimensions in organizational groups. Administrative Science Quarterly, 42(3), 530–58.
Keil, M. & Robey, D. (1999). Turning around troubled software projects: An exploratory study of the deescalation of commitment to failing courses of action. Journal of Management Information Systems, 15(4), 63–87.
Lam, S. (1991). Venture capital financing: A conceptual framework. Journal of Business Finance & Accounting, 18(2), 137–149.
Lerner, J. (1994). The syndication of venture capital investments. Financial Management, 23(3), 16–27.
McCarthy, Anne M; Schoorman, F. David; Cooper, Arnold C. (1993). Reinvestment decisions by entrepreneurs: Rational decision-making or escalation of commitment? Journal of Business Venturing, 8(1), 9–25.
McGrath, J.E. (1983). Groups: Interaction and Performance. Englewood Cliffs, N.J.: Prentice Hall.
Montealegre, R. & Keil, M. (2000). De-escalating information technology projects: Lessons from the Denver International Airport. MIS Quarterly, 24(3), 417–447.
Nisbett, R.E. & Ross, L. (1980) Human inference: Strategies and shortcomings of social judgement. Englewood Cliffs, N.J.:Prentice Hall.
Norton, E. (1993). Venture capital finance: review and synthesis. Advances in Quantitative Analysis of Finance and Accounting, vol. 2, part B, 141–165.
Ross, J. & Staw, B.M. (1993). Organizational escalation and exit: Lessons from the Shoreham nuclear power plant. Academy of Management Journal, 36(4), 701–732.
Ruhnka, J. C., Feldman, H. D., & Dean, T. J. (1992). The ‘living dead’ phenomenon in venture capital investments. Journal of Business Venturing, 7, 137–155.
Rutledge, R.W. (1995). Escalation of commitment in groups and the effects of information framing. Journal of Applied Business Research, 11(2), 17–24.
Ryan, P. M. (1995). Why some venture capitalists escalate and others don’t in the same investment situation. Best papers Proceedings—Academy of Management, 55th Annual meeting Vancouver, Canada.
Ryan, P. M. (1994). Escalating commitment: A study of contrasts in the venture capital industry. Unpublished dissertation, University of Texas, Austin.
Sapienza, H.J. 1992. When do venture capitalists add value? Journal of Business Venturing, 7, 9–27.
Staw, B.M. (1976). Knee-deep in the Big Muddy: A study of escalating commitment to a chosen course of action. Organizational Behavior and Human Performance, 16(1), 27–44.
Staw, B.M. (1981). The escalation of commitment to a course of action, Academy of Management Review, 6(4), 577–587.
Staw, B.M., Koput, K.W., & Barsafe, S.G. (1997). Escalation at the credit window: A longitudinal study of bank executives’ recognition and write-off of problem loans. Journal of Applied Psychology, 82(1), 130–142.
Staw, B.M. & Ross, J. (1987). Understanding escalation situations: Antecedents, prototypes, and solutions. In B.M. Staw & L.L. Cummings (Eds.)., Research in organizational behavior, vol. 7g: 39–78, Greenwich, CT. JAI Press.
Yamaguchi, K. (1991) Event history analysis. Newbury Park, CA: Sage.
Zacharakis, A.L., and Meyer, G.D. (1998) A lack of insight: Do venture capitalists really understand their own decision process? Journal of Business Venturing, 13, 57–76.
2002 Babson College. All Rights Reserved. Last Updated March 2003.