Top Ten Behavioral Biases in Project Management

Abstract

This wiki article describes key notions on the research on bias’ and fallacies’ influence on decision making in project management conducted by Bent Flyvbjerg. Bent Flyvbjerg is an esteemed professor within the field of project management, especially specialized in the management of megaprojects. Bent Flyvbjerg highlighted in 2021 what he considers to be the ten most important biases and fallacies influencing project management. This selection build on his extensive research within this area. In the discussion of the ten most important biases, he highlights the distinction between political biases and cognitive biases, and how the approach to mitigate the two types should differ. in This selection of biases build on his extensive research on the subject of behavioral sciences influence on project management. Flyvbjerg has developed a method that project managers can apply to mitigate political- and cognitive biases called reference class forecasting, which suggest that project managers apply an outside view on their project forecasting. This is done by identifying similar, completed projects and analyze the probability of a forecast based on the performance of these past projects. Flyvbjerg’s research can be seen as complimentary to project management guides such as the PMBOK guide, but seems to some degree to be based on a different ontological interpretation of the landscape of project management than the PMBOK guide. Application of reference class forecasting entails some limitations, for example the availability of valid data, as well as the sometimes limited autonomy of the individual project manager to change procedures.

Introduction

According to the PMBOK guide, project management can be defined as the application of knowledge, skills, tools and techniques to project activities to meet project requirements (Project Management Institute, 2021, p. 4). This definition places a large responsibility on the cognition of a project manager to ensure success. Bent Flyvbjerg’s article Top Ten Behavioral Biases in Project Management: An overview (2021) challenges the rationality of decision making, as he highlights the ten cognitive and political biases he believes influence project management the most. This wiki article presents the 10 behavioral biases identified by Flyvbjerg, the method Flyvbjerg has developed to reduce the impact of biases in project management, with a focus on the biases involved in forecasting. Limitations to the view on planning and uncertainty is also proposed in the article.

Bent Flyvbjerg is an esteemed professor, currently positioned at the Danish institute for IT Program management at ITU. He is conducting research within the realms of social sciences, including research on megaproject and their planning, as well as decision making. For more information about Bent Flyvbjerg’s collected contributions, see the wiki page for Bent Flyvbjerg. Flyvbjerg’s research is especially concerned with what is referred to as the iron law of projects: “over time, over budget, under benefits, over and over again” (Flyvbjerg, 2017). The ten biases highlighted in Flyvbjerg’s article (2021) Ten Behavioral Biases in Project Management: An Overview is an exploration of some of the underlying causes of the Iron Law.

A quick definition of bias and fallacy

Biases are according to the APA defined as “…any tendency or preference”, whereas a fallacy is defined as; “an error in reasoning or argument that leads to a conclusion that may appear valid but is actually invalid”(APA). The two psychologists Daniel Kahneman and Amos Tversky played a significant role in pushing forward research that investigates the effect of biases and fallacies in judgment and decision making (Flyvbjerg, 2021). Their research has had a particular significant role in the field of behavioral economics, and Kahneman won in 2002 the nobel prize for their research contributions to the field of economy. For more information regarding the history of biases in behavioral economics, please see the Wikipedia articles on Daniel Kahneman, Amos Tversky and Behavioral Economics.

Bent Flyvbjerg build on top of Kahnemann and Tversky’s research, but emphasizes a distinction between a political- and a cognitive bias. The former is a tendency which root cause is political, and where the subject is aware of the deception. The latter is a tendency which root cause is psychological, and where the bias leads to self-deception (Flyvbjerg, 2021, p. 532).

The ten most important biases in project management according to Flyvbjerg (2021)

The following table is a direct excerpt from Flyvbjerg (2021, p. 531)

Bias	Description
Strategic misrepresentation	The tendency to deliberately and systematically distort or misstate information or strategic purposes. Aka political bias.
Optimism bias	The tendency to be overly optimistic about the outcome of planned actions, including overestimation of the frequency and size of positive events and underestimation of the frequency and size of negative ones.
Uniqueness bias	The tendency to see ones project as more singular than it actually is.
Planning fallacy	The tendency to underestimate cost, schedule and risk and overestimate benefits and opportunities.
Overconfidence bias	The tendency to have excessive confidence in one’s own answers to questions.
Hindsight bias	The tendency to see past events as being predictable at the time these events happened. Also known as the knew-it-all-along effect.
Availability bias	The tendency to overestimate the likelihood of events with greater ease of retrieval (availability) in memory.
Base rate fallacy	The tendency to ignore generic base rate information and focus on specific information pertaining to a certain case or small sample.
Anchoring	The tendency to rely too heavily, or “anchor” one trait or piece of information when making decisions, typically the first piece of information acquired on the relevant subject.
Escalation of commitment	The tendency to justify increased investment in a decision, based on the cumulative prior investment, despite new evidence suggesting the decision may be wrong. Also known as the sunk cost fallacy.

With what purpose is it relevant to highlight the ten most important biases in project management?

The interpretation of the root cause for phenomena such as cost overruns, delays, and benefit short-runs in investment decisions greatly affects the proper response to these phenomena, and is therefore a central question for project managers.

Flyvbjerg (2021) claims the role of cognitive biases is too emphasized in project management, in relation to political biases. He highlights strategic misrepresentation as the most important bias in forecasting in megaprojects, where the political-organizational pressures are high, whereas cognitive biases, such as optimism bias, are mainly explanatory when outside pressures remain low.

Application: Mitigating biases in management forecasting

One of the management tasks vulnerable to biases and fallacies is forecasting of project prize, demand and other impacts, as this is a task that typically entails a high level of uncertainty. Flyvbjerg has developed a method to mitigate both cognitive and political biases called reference class forecasting, which is described in his article From Nobel Prize to Project Management: Getting Risks Right (2006).

Flyvbjerg (2006) argues that biases are mainly to blame for inaccurate forecasting in megaprojects, as collected data shows that inaccuracies are not normally distributed around zero, and professions in which megaprojects are conducted does not seem to show increased accuracy over time.

The reference class forecasting method consists of three steps (Flyvbjerg, 2006, p. 7):

• 1. Identify a relevant reference class of past, similar projects.
• 2. Establish a probability distribution for the selected reference class.
• 3. Compare the specific project with the reference class distribution, in order to establish the most likely outcome for the specific project.

The purpose of this method is to incorporate an outside view on the project at hand. Flyvbjerg explains why this method works to mitigate biases: “In the outside view project managers and forecasters are not required to make scenarios, imagine events, or gauge their own and others' levels of ability and control, so they cannot get all these things wrong. Human bias is bypassed” (Flyvbjerg, 2006, p. 9).

The method requires that there is a sufficient amount of empirical data to make sufficiently robust conclusions regarding the probability distribution for the reference class. If political bias is the most dominating course of inaccurate forecasting, a more fundamental change in the forecasting climate across organizations will likely be needed, to a system where accuracy is rewarded (Flyvbjerg, 2006).

A comparison of Flyvbjerg (2021, 2006) and the PMBOK Guide

The PMBOK standard and guide (2021) lists forecasting as an activity under the planning performance domain, one of eight project performance domains. Here, several types of quantitative forecasting are listed, including estimate to complete, variance at completion and throughput analysis. These forecasting methods relies as standard on data collected within the project in question, to facilitate learning from own experiences. In this regard, reference class forecasting is a radically different approach to what is suggested in the PMBOK guide.

The PMBOK guide (2021) also lists uncertainty as a project performance domain, critical for the delivery of desired project outcomes. Bias and fallacy mitigation is relevant for performance under uncertainty, as biases and fallacies in its nature is a (faulty) response to uncertainty.

The uncertainty performance domain (PMBOK, p. 117) includes uncertainty (a lack of understanding and awareness of issues), ambiguity (a state of being unclear), complexity (a difficulty to manage due to human behavior, system behavior and ambiguity), volatility (rapid and unpredictable change) and risk (an uncertain event that will effect a project objective).

The ten biases listed by Flyvbjerg (2021) relates to different aspects of uncertainty, with most of them relating to uncertainty and ambiguity. Examples of biases/fallacies with a clear relation to uncertainty includes optimism bias and planning fallacy; These responses result in the actor not being aware that there could be potential errors in their judgement. Ambiguity is related to phenomena such as anchoring and base rate fallacy, as these describe responses that are decreasing the experienced unclearness of a situation.

The political bias political misrepresentation does not have as clear a link to the uncertainty domain as the other biases highlighted by Flyvbjerg. As uncertainty in its broadest sense is a state of not knowing or unpredictability (PMBOK, p. 117), a bias where the actor is aware of the deception might not fit the category. This bias can be interpreted as a stakeholder management response, as the bias is a deliberate strategic distortion and is prevalent in big organizational contexts (Flyvbjerg, 2021, p. 532). It can in this regard be linked to one of the listed activities under the stakeholder performance domain (PMBOK), namely understand and analyze; the bias is activated if the analysis of the political-organizational pressures are interpreted as to require a distortion of information.

The PMBOK guide recommends to respond to uncertainty by:

• Gather information
• Prepare for multiple outcomes
• Develop a set based design
• Build in resilience

The desired content or criteria for the gathered information is not explicitly specified in the guide. The guide suggest to prepare for multiple outcomes, when only a few possible outcomes are possible. The biases listed by Flyvbjerg (2021) might question a project managers ability to interpret how many different outcomes could be possible; phenomena such as overconfidence bias and availability bias could distort a project managers overview of possible options.

To develop a set based design, the PMBOK guide suggest that multiple designs are explored in the early phases of a project, and analyze trade-off’s of different solutions, in terms of key performance parameters such as quality vs. cost or risk vs. schedule. If the different scenarios are created with data from other, similar projects, then this step resembles Flyvbjerg’s reference class forecasting (2006). It is however not specified in the guide how the scenarios are created, or what they contain. Assessing the impact of different scenarios could be vulnerable to significant uncertainty, and the analysis could therefore be vulnerable to biases such as base rate fallacy and availability bias; overemphasizing the unique and prevalent information and downplaying the most likely scenarios.

To build in resilience in the project, the PMBOK guide suggests to design processes that makes the project team learn, adapt and respond quickly to unexpected changes. Learning from past experiences is in line with Flyvbjerg’s notion that the intuition of project managers might not be reliable - but the applicability of past experiences on a current project may differ.

It should be mentioned that a key difference between Flyvbjerg’s description of biases in project management and the PMBOK guide is the purpose of the two texts. Flyvbjerg’s aim is to describe which underlying factors influence decision making, whereas the PMBOK guide describes ideal processes and priorities for a project manager. The article and guide therefore also have the potential to supplement the other, for example, Flyvbjerg’s notion of reference class forecast (2006) can be seen as a description of which kind of data is relevant in the pursuit to mitigate uncertainty. In some aspects, Flyvbjerg’s research (2021, 2006) questions fundamental aspects of the PMBOK guide, such as to which degree a manager should trust his/hers ability to make choices based on logic, in loosely structured decision processes.

Limitations to Flyvbjerg (2021, 2006)

Flyvbjerg (2006) points to the differences of the potential of reference class forecasting when distinguishing between forecasting errors caused by respectively optimism bias and strategic misrepresentation. Whereas the former is unintentional and has a high potential to be mitigated by reference class forecasting, the latter is more consciously engaged by the actor, and the demand for accuracy tends to be low, combined with high barriers to reduce this bias (Flyvbjerg, 2006). Another risk that Flyvbjerg (2006) addresses is that a more realistic forecast of expenses might result in overspending, if the funds set aside for the project turns out to more than compensate the actual cost.

Another limitation in reference class forecasting is the availability of valid data sets, especially outside of a research setting (Flyvbjerg, 2008). One could imagine that companies might be hesitant to provide their competitors with data on their past projects, especially if this information is sensitive – for example if the project in question faced serious budget overspending in comparison with the original forecasting. Another aspect of the reference class forecasting is to analyze which projects could be similar enough to be effectively integrated into a reference class. Flyvbjerg & Cowi (2004) describes the first project where reference class forecasting where for the first time applied in practice (Flyvbjerg, 2006) for the treasure and Department of Transport. In this instance, the forecasting was focused on the parameter cost overrun vs. budget. One could imagine that adding complexity to the forecast, for example by adding other dimensions of realized key performance parameters to the forecast, it would also increase the difficulty involved in attaining the appropriate data input.

One thing is to be aware of one’s own biases as a project manager, and to be interested in making more accurate forecasts. But what about the situation where there is no incentive to improve accuracy? This kind of behaviour can be motivated by the desire to get a project approved (Flyvbjerg, 2009). As such, there can be a lack of incitement within an organizations, and collegaues might question project managers that are enforcing more realistic forecasting.

The biases might be imbedded in corporate culture. Flyvbjerg does acknowledge that the treatment of especially political biases will likely go through changing the entire system, to reward better forecasting and punish bad ones, as Flyvbjerg (2006) highlights in this paragraph: “Forecasters and promoters should be made to carry the full risks of their forecasts. Their work should be reviewed by independent bodies such as national auditors or independent analysts, and such bodies would need reference class forecasting to do their work. Projects with inflated benefit-cost ratios should be stopped or placed on hold. Professional and even criminal penalties should be considered for people who consistently produce misleading forecasts.” - Flyvbjerg, 2006, p. 17

To realize the scenario described above will likely entail fundamental changes to the political-organizational climate in most sectors. It would also entail changes in legislation – at least if criminal penalties were used as a tool to punish (too) inaccurate forecasts. In order to be able to conduct a reference class forecasting, it might be necessary to establish a data platform where organizations can retrieve data on projects that are suitable to include in their reference class, which as well would be a systemic change that must be realized in cross organizational settings. System change might also be needed in relation to cognitive biases, however perhaps with a focus on organizational culture and climate. In 2017, the Danish book Jytte fra marketing er desværre gået for i dag (Translated: Jytte from marketing has unfortunately left the building) gained immense popularity in a Danish context, as it provide tips to tackles biases related to organizational work culture. The popularity of the book might suggest a mature climate to work on biases within an organizational context, and point to a need for these changes.

Annotated bibliography

• Flyvbjerg, B. (2021). Top Ten Behavioral Biases in Project Management: An Overview. Project Management Journal, Volume 52, Issue 6, December 2021, Pages 531-546. The main article described in this wiki article, identifying the ten most important biases for project management, their impact on project management. Includes a reflection on what can be done to mitigate the effects, thus heighten the chances of successful management.
• Project Management Institute (2021). The standard for project management and a guide to the project management body of knowledge (PMBOK guide). Seventh edition. Newtown Square, Pennsylvania: Project Management Institute, Inc A collection of fundamental concepts and constructs of the project management profession, divided into 12 principles of project management and eight project performance domains.
• Tversky, A. & Kahneman, D. (1974). Judgment Under Uncertainty: Heuristics and Biases. Science, New Series, Vol. 185, No. 4157. (Sep. 27, 1974), pp. 1124-1131. A walkthrough of three types of heuristics employed to assess probabilities and the biases that occur in decision making as a result.
• Flyvbjerg, B. (2006) From Nobel Prize to Project Management: Getting Risk Right. Project Management Journal, vol. 37, no. 3, August 2006, pp. 5-15. An approach to mitigating risk in forecasting projects, focusing on analyzing a group of similar projects as preparation.
• Flyvbjerg, B. (2008) Curbing optimism bias and strategic misrepresentation in planning: Reference class forecasting in practicein European Planning Studies — 2008, Volume 16, Issue 1, pp. 3-21. DOI: 10.1080/09654310701747936. Provides two examples of how reference class forecasting has been conducted in practice, as well as a discussion of the limitations and positbilities of the method.

Other sources

• Bent Flyvbjerg ITU: https://pure.itu.dk/da/persons/bent-flyvbjerg
• APA Dictionary of Psychology: https://dictionary.apa.org/bias
• Flyvbjerg, Bent and Glenting, Carsten and Rønnest, Arne (2004) Procedures for Dealing with Optimism Bias in Transport Planning. London: The British Department for Transport, Guidance Document, June 2004, Available at SSRN: https://ssrn.com/abstract=2278346
• Flyvbjerg, Bent(2017), Introduction: The Iron Law of Megaproject Management, in Bent Flyvbjerg, ed., The Oxford Handbook of Megaproject Management(Oxford: Oxford University Press), pp. 1-18.
• Bent Flyvbjerg (2009): Survival of the unfittest: why the worst infrastructure gets built—and what we can do about it, Oxford Review of Economic Policy, Volume 25, Issue 3, Autumn 2009, Pages 344–367, https://doi.org/10.1093/oxrep/grp024