Sunny With a Chance of Manipulation: Keeping an Eye on the Bias Brew

The hidden dangers of political and cognitive biases in major project management

This week I re-visited the paper “Top Ten Behavioural Biases in Project Management:An Overview” (Flyvberg, 2021). Bent Flyvberg is a professor, a scholar, and a pioneer in the field of major programme management. He has revealed how political and cognitive biases can lead to disastrous outcomes in large-scale projects, and how to avoid them.

In major project management, biases can combine to create a toxic brew resulting in hugely underestimated costs and overestimated benefits. Two types of bias play a role - political bias from power, and cognitive bias from human psychology. Both are underestimated, but the underestimation of political bias is particularly detrimental. As professionals working on big programmes, we must be aware of how power and bias interact at all project scales to understand and advise leaders on reducing bias and improving forecasting.

The Power-Bias Spectrum

Many large projects end up with substantial cost overruns. Megaprojects (over $1 billion) like the Channel Tunnel, Boston’s Big Dig, and Berlin Brandenburg Airport have experienced overruns of 50-100%. But cost overruns plague projects of all sizes, it is what Bent calls “The Iron Law of Programme Management”.

Traditionally, experts attributed overruns to technical factors. However, research over the past 20 years reveals human bias as a major cause. Two key biases are at play:

Political bias: This stems from power, incentives, and interests leading to intentional distortion of information to get projects approved and funded. It is driven by organisational politics and competition for resources.

Cognitive bias: This results from limits in psychology causing unintentional but systematic judgment errors. Over-optimism about planned projects is a universal effect.

Importantly, political and cognitive bias exist on a spectrum. As project scale and impact grows, political pressures and incentives for misrepresentation normally increase. On small projects, cognitive biases like optimism may dominate. But for major projects, political bias often becomes more pivotal in driving misrepresentation and distortion by powerful sponsors to get pet projects funded.

Internal Pressures 

Political bias should not be associated only with government megaprojects. Private organisations have their own versions. Statements like “must win” programme and tying promotions to winning flagship projects create strong incentives for distorting estimates to gain advantage. Power and political bias exist inside firms too.

As projects grow larger in scope and impact, the pressures grow for underestimating costs and risks while overestimating benefits and margins in order to gain approval. However, deliberate strategic misrepresentation driven by political bias at any scale should not be underestimated.

The Toxic Brew

Together, political and cognitive bias form a toxic brew. Power amplifies inherent cognitive biases through a feedback loop:

Power → Strategic Misrepresentation → Distortion

Distortion → Exaggerated Cognitive Biases → More Distortion  

Self-serving strategic misrepresentation creates initial distortion. But this also interacts with and boosts the leader’s innate cognitive biases like optimism and availability bias. The result is an escalating spiral that takes project plans further from reality.

Unless diffused, this mix of emboldened political bias and amplified cognitive bias will severely distort forecasting, leading to unfounded project approvals.

The Ecological Blind Spot

Equally worrying is the blind spot around recognising how power exacerbates bias. Cognitive psychology studies can be devoid of context through isolated lab experiments. But decisions involve social and political environments where power changes cognition. This ecological reality is ignored.

Likewise, in project forecasting, cognitive biases get the focus while political drivers go unmentioned. But political power and incentives often intentionally distort estimates before cognitive biases further skew them. This ecological influence of power is neglected.

Leaders themselves frequently underestimate their own susceptibility, seeing bias as something that only affects others. In truth, power worsens bias. This blindness allows the toxic brew to go unaddressed.

From Risks to Remedies

There are ways leaders and advisors can help neutralise the toxic power-bias blend and improve forecasting:

- Build a learning culture: Make asking questions and challenging assumptions the norm, not a rarity.

- Hire divergent thinkers: Ensure different perspectives to balance optimism and groupthink.

- Demand realistic estimates: Require justifiable forecasts backed by hard data, not intuition.

- Do premortems: Imagine failure and work backward to surface hidden risks.

- Separate sunk costs: Track already-spent funds apart from future budgeting.

- Reward truth-telling: Praise those who deliver bad news early instead of shooting messengers.

- Thank critics: Recognise constructive opposition as strengthening, not weakening.

- Admit uncertainty: Leaders should be open about past forecast errors and limitations.

- Limit power: Diffuse and distribute power to hamper misrepresentation.

These steps aim to reshape incentives, bring in outside views, and counteract blind spots. While challenging, they can neutralise the toxic brew of power and bias.

Owning Our Duty

As leaders and advisors, we have a duty to promote realism over comfort. This requires courage to speak truth to power when necessary.

Sincere forecasting grounded in reality is vital for excellence. If we care about outcomes, we must address biases, leverage expertise, and reshape systems that incentivise distortion. This responsibility starts with awareness.

The power-bias brew will always exist. But its toxicity can be reduced through humble but firm engagement. We must recognise the ecological reality and help leaders name and tame the elephant in the room – the unstable mixture of power and bias.

Flyvbjerg, B. (2021). Top Ten Behavioral Biases in Project Management: An Overview. Project Management Journal, 52(6), 531–546. https://doi.org/10.1177/87569728211049046