Lade Inhalt...

A Critical Analysis of Overconfidence as an Explanation for the High Rate of Business Entry Failure

Hausarbeit 2013 24 Seiten

BWL - Unternehmensführung, Management, Organisation

Leseprobe

Contents

List of Figures

1 Introduction

2 Background
2.1 Business Entry Failure
2.2 What is Overconfidence?
2.3 How Overconfidence Arises
2.4 Overconfidence and Excess Entry

3 Analysis
3.1 Critic of Camerer, Lovallo (1999)
3.1.1 The Impact of Skill to Success
3.1.2 The Method/ Contradictory Statements
3.1.3 Beneficial Effects of Overconfidence
3.1.4 How to overcome overconfidence
3.2 To go on

4 Conclusion

References

Appendix

A Additional Aspects
A.1 How to Measure Overconfidence

List of Figures

1 Business Failure (Scott A. Shane, 2008)

2 Firm Survival (Mata and Portugal, 1995)

3 Payoffs (Camerer, Lovallo, 1999)

1 Introduction

Friday, the 19th of October, 2012, the German newspaper ”Handelsblatt” (Stor- beck, 2012) titled: ’’Rethinking Economy”. 15 young economists were intro­duced to be the new hope for economic science. One of them was Ulrike Mal­mendier whose first economic researches deal with managerial hubris. Intro­ducing her is proofing the importance of a new economic domain called ’be­havioural economics”. As a behavioural economist Malmendier criticises the ”homo-economicus”-model and the idea of people acting entirely rational. In fact, people constantly display irrational behaviour that results out of cognitive biases (that distort their perception) and also results out of simple, but biased decision rules (heuristics) that the human mind implies. One of those cogni­tive biases is that people seem to be unreasonable overconfident. The study of overconfidence is originated in the psychological literature. Because of it’s great impact on the behaviour of economic actors, it recently influences economic and finance literature (Benoît, Dubra, 2011) and significantly shapes the research activity in behavioural economics. The task of this essay is to critically analyse the role of overconfidence in a special economic domain: business entry deci­sions. The text is structured as follows: The first section shortly deals with the domain of business entry failures. In the subsequent chapters, overconfidence is introduced as a possible explanation. It is to explain what overconfidence is and how it arises. The text then reviews the paper of Camerer and Lovallo (1999) and critically analyses overconfidence as an explanation for excess entry and high rates of business failure. At last, it will take a further look at the future research tasks.

2 Background

2.1 Business Entry Failure

The survival rates of start-up businesses are usually not very high. A nascent entrepreneur who starts a new firm is making a decision under uncertainty, with a high probability to fail. Furthermore, the costs of business failure are very high, especially for the responsible entrepreneur. For this reason, there has evolved a great amount of literature containing the topic of entrant failure and success that aims for knowledge about the factors that decrease new firm’s hazard rate (the rate of business failure). Factors that seem to have an influence, are among others: owner-manager’s knowledge of the industry, industry size, the growth of industry and the total number of new firms, entering the industry (Mata and Portugal, 1995; Cooper et al., 1994).

illustration not visible in this excerpt

Figure 1: Business Failure (Scott A. Shane, 2008)

A few statistical models are able to describe firm survival and entrant failure. The most commonly used is the Proportional Hazard Model of Sir David Cox. It is a regression function that measures the influence of several variables on the survival rate. The Hazard Function then calculates the probability for the firm to die for each period of time. If overconfidence can be shown in research to highly influence the number of business entries (and thus also to influence the hazard rate), it could be considered in the statistical models and thus inserted as a variable in the Proportional Hazard Function.

The empirical data on how many new businesses fail within a few years highly vary. In Figure 1 you can see an investigation of the percentage of new busi­nesses that are still alive over a period of ten years (from 1992 to 2002). The results are similar to those in the paper of Mata and Portugal (1995) that are displayed in Figure 2. After 10 years in Figure 1 (and after 7 years in Figure 2) less than 30% of the newly entered businesses survived. The severe data makes it necessary to look at possible reasons for the high rate of business failure within the first years.

illustration not visible in this excerpt

Figure 2: Firm Survival (Mata and Portugal, 1995)

This essay deals with overconfidence as one possible explanation. To start with, it is to explain what overconfidence is.

2.2 What is Overconfidence?

In most cases, it is not very difficult to measure people’s knowledge. In contrast, it’s much more complex to check the validity of people’s level of confidence in their own answer (Fischhoff, Slovic, Lichtenstein, 1977). This issue is called ’’metaknowledge”[1]: An estimation of how much we know (Russo, Schoemaker, 1992). Here we can see that people are not very well ’calibrated”. They too strongly believe in the correctness of their answers and are therefore overcon­fident in estimating the probability of their answers to be right (see Pulford, 1996).

One of the first researchers to discover the ”overconfidence-effect” were Alpert and Raiffa (in an unpublished report of the year 1969). They wrote about indi­viduals that should try to rightly assess probability distributions of quantities whose values were unknown (Kahneman, Slovic, Tversky, 1982). They showed the phenomenon that ”[...] when people are p percent sure that they have answered a question correctly or predicted correctly, they are in fact right on average less than p percent of the time” (p4, Bar-Hillel, 2002). So when people shall choose whether Islamabad or Hyderabad has more citizens and have to give a probability to show their certainty in their own answer, it is likely that people overrate the chance that they answered accurate.[2] Speaking generally, people overestimate how much they actually know.

Another form of overconfidence is that people tend to overvalue their ability, knowledge and also exogenous events[3] like fortune, especially in comparison to a group of other people. When Adam Smith in his book ’’Health of Nation” wonders about young people giving little weight to downside risk in their career choices (Smith, 1776), he probably was the first witness of the fact that people can also be overoptimistic about their future (which they can only control in part).

2.3 How Overconfidence Arises

Where does the irrational overestimation of own skill and knowledge and the belief in above-average own fortune come from? May (1987) distinguishes three stages of developing subjective probabilities.

a) process of solving the problem
b) generating the subjective feeling of confidence
c) quantifying the feeling and transfer it into a probability

Three possible explanations for overconfidence are given.

1. People distort the scale of probabilities and always give a higher number when quantifying their feeling in stage 3.[4]
2. Overconfidence is productive, because it motivates people (evolving in stage 2).[5]
3. Overconfidence arises, because people fail to assess the evidence of their answer. They overestimate the value of their background knowledge and their hasty conclusions. The bias lies in the interdependency of stage 2 and 3.

May (1987) refutes explanation 2 and 3 but also found no proof for explanation 1[6]. What seems to have an effect is the type of the question. If the question is misleading[7], the subjects are much more overconfident than if the question leads to the right answer[8] or if the question is neutral[9]. It seems as if over­confidence is higher if the questions are pointing in another direction as the general education background does. These findings raise the question, whether people are keen to believe in the truth of an answer, just because it matches their intuition, based on education and experience?[10]

After Kahneman (2011), ’’subjective confidence in a judgement is not a rea­soned evaluation of the probability that this judgement is correct. Confidence is a feeling [...]” (p212, Kahneman, 2011). This feeling is especially strong if the information is cognitively easy to process. The more coherent a story seems to be, the more confident people are. If it seems plausible, people are willing to believe it, although there might exist empirical evidence that their story is not correct. This is what Kahneman calls the ’Illusion of Validity”: ”We are confident when the story we tell ourselves comes easily to mind, with no contra­diction and no competing scenario. But ease and coherence do not guarantee that belief hold with confidence is true. The associative machine [- the brain -] is set to suppress doubt and to evoke ideas and information that are compatible with the currently dominant story.” (p239, Kahneman, 2011).

Overconfidence also arises, because people are convinced to understand the past and therefore can predict the future (Kahneman, 2011).

It results, because they believe in the data that is apparently available (avail­ability bias)(Bar-Hillel, 2002).

In the example with Islamabad and Hyderabad, people are choosing their an­swer first, before they evaluate the amount of subjective certainty. They may have the tendency to search more for information that confirms their choice than for counter-arguments to their given answer. This is called ’’confirmation bias” in cognitive science and can also lead to overconfidence (Jungermann et. al, 2010). In the end, decisions are based on bounded rationality.

[...]


[1] To confidence as a ’’metakognition” also see Terrace, Son (2009).

[2] (So in fact Hyderabad is an indian city with over 6 million citizens and Islamabad has ’’only” about 690 000, although you might have never heard of the first.)

[3] Malmendier and Tate (2005b) suggest to use the ”overconfidence”-term to underline the exces­sive confidence in personal abilities and to use the ”overoptimism”-term to describe general optimism about exogenous events. This essay follows their notation to distinguish the two sorts of overconfidence.

[4] This could explain overconfidence, but it can’t explain underconfidence that also arises some­times.

[5] This explanation can’t explain underconfidence either.

[6] Because it can’t explain, why the subjects were overconfident in each single question, but didn’t overestimate the overall rate of right own answers in the test.

[7] Like: Which US-streetname is used more? Right: Chestnut, wrong: Main

[8] Like: Which US-Streetname is used more? Right: Washington, wrong: Wilson.

[9] Like: In Freiburg, there is a street called... Right: Annengassle, wrong: Probstgassle.

[10] To the importance of intuition and how it is beaten by statistical algorithms also read: Paul Meehl, Clinical vs. Statistical Prediction: A Theoretical Analysis and a Review of the Evi­dence, 1978.
”[...] you will [only] label that unease an intuition if it is followed by a bad experience” (Kahneman, 2011)
1 Compare to the explanation of Camerer and Lovallo (1999) in chapter 2.4.

Details

Seiten
24
Jahr
2013
ISBN (eBook)
9783656423065
ISBN (Buch)
9783656423423
Dateigröße
562 KB
Sprache
Deutsch
Katalognummer
v213717
Institution / Hochschule
Christian-Albrechts-Universität Kiel – Applied economics of the firm
Note
1,0
Schlagworte
Camerer Lovallo 1999 Overconfidence Excess Entry Entrepreneurship entrepreneur entry decision behavioral economics experimental economics neuroeconomics

Autor

Zurück

Titel: A Critical Analysis of Overconfidence as an Explanation for the High Rate of Business Entry Failure