Planning errors: the planning fallacy

Planning activities is a very delicate subject in companies. The time planned for a piece of work often has to be extended, for all sorts of reasons. One wonders whether the causes of these delays can always be put down to external circumstances alone. The answer, obviously, is no: numerous experiments have shown how common this type of error is.

In 1994 Roger Buehler, Dale Griffin, and Michael Ross asked 37 psychology students to estimate how long it would take them to complete their theses. The average answer was 33.9 days. They were then asked to estimate the best case and the worst case, thereby creating a range within which they would surely manage to complete the work. This range turned out to run from 27.4 days to 48.6 days (on average). The actual average completion time then turned out to be 55.5 days. Only 30% of the students managed to complete the thesis in the time they had predicted. The study is Buehler, Griffin, and Ross, “Exploring the ‘planning fallacy’: Why people underestimate their task completion times”, Journal of Personality and Social Psychology, 1994.

Other studies, also conducted on students, led to similar conclusions (Buehler, Griffin, and Ross, 1995).

When a group of students was asked to estimate when they would complete their personal academic projects, and in particular the dates by which they believed they would finish with a probability of 50%, 75%, and 99% respectively, the results were:

• 13% of the subjects finished the project by the date to which they had assigned a 50% probability;
• 19% finished by the date to which they had assigned a 75% probability;
• 45% managed to finish by the date defined as the “worst case” (99%).

These figures give an idea of how great the impact of planning errors is, and of how hard it is for companies to allocate people to projects.

These experiments, later confirmed by many others in settings of every kind, show that without a structured planning activity the errors are considerable.

In the study cited, in more than half of the cases (55%) the time defined as the “worst case” (99% certainty) was exceeded. This shows us how uncertain our “certainty” is.

Underestimating the time or resources to devote to a given task can lead to extensions and budget increases, but also to rushed and incomplete (or completely wrong) results.

People’s tendency to underestimate the time needed to carry out an activity is called the “planning fallacy” (Kahneman and Tversky, 1979).

Why is this kind of error so widespread?

In some cases the underestimate is deliberate, a way of presenting a project with modest time and cost, knowing that they will have to be increased along the way. Almost all of us have been on the receiving end of this: in the implementation of an information system, in the building of a house, in a car repair, and so on.

In most cases these planning errors are not intentional (and are therefore harder to control). In particular:

• We fail to consider the time taken in the past to carry out similar activities, because we cannot find points of contact with other projects already completed. Focusing on the planning itself without an overall view leads us to see the differences from other projects rather than the similarities. Valuable information about the past is thus left out.
• “Optimism bias”: a distortion that leads us to believe we are less at risk of suffering a negative event than other people are. This mechanism leads us to underestimate the risk of events that could call our planning into question. According to Daniel Kahneman: “Most of us view the world as more benign than it really is, our own attributes as more favorable than they truly are, and the goals we adopt as more achievable than they are likely to be. We also tend to exaggerate our ability to forecast the future, which fosters overconfidence. In terms of its consequences for decisions, the optimistic bias may well be the most significant cognitive bias.” By not considering possible alternatives to the best case, we underestimate the likelihood of unforeseen (but foreseeable) complications and obstacles. We also do what we can to remember the best cases and forget past difficulties (as already seen in the post “Do we learn from our mistakes?”). Cases of polarization, such as the optimism bias, become more evident when the activity is carried out in a group. Many distorting dynamics come into play in group decisions. One of the most widely observed is the shift of positions toward the extreme, which has a clear negative effect on planning.
• LeBoeuf and Shafir (2006, 2009) studied the mechanism of anchoring: when we take the analysis further, we do so with the (unconscious) aim of staying as close as possible to our initial estimate. We find this mechanism in many other areas as well; once a first impression or assessment has been formed, our effort goes into confirming it rather than evaluating it rationally.
• Once we have an overall view of the project, we need to concentrate on the individual activities. It is easy to overlook some of the key actions that have to be taken and that can significantly affect the planning (and the feasibility). This happens more readily when the person doing the planning experiences a sense of power connected with the task. In four studies, Weick and Guinote (2010) found that people with power make more optimistic and less accurate time predictions, an effect the authors link to the way power narrows attention onto the goal (Journal of Experimental Social Psychology).
• The more desirable we consider the benefits we stand to gain, the less attention we pay to possible problems, and the more errors we make. Obviously the opposite is also true: if we are against a project, its possible downsides will leap out at us immediately. This process is amplified in particular by the incentives (e.g., bonuses) that the company gives to those responsible for delivering the project (incentives that therefore produce the opposite of the intended result).

To avoid planning errors, we have to know that it exists and find adequate countermeasures. The project should therefore be analyzed in all its phases, and it is very important to concentrate on looking for the points that could cause problems. This last step must never be skipped. Looking for similarities with activities already carried out is another element that helps accuracy.

Kahneman proposes three steps for reducing planning errors:

1. Identify an appropriate reference class (the type of project). Treating the project as a member of a class and not as a new activity forces us to measure ourselves against the past.
2. Obtain the statistics of the reference class (time, cost, and errors). Companies often hold information that could be useful to the analysis if only someone consulted it. Use this objective research to generate a baseline prediction.
3. If, despite this process, the planning fallacy is still present, use the historical record to calculate a percentage of error to apply to the baseline prediction. The method, known as “reference class forecasting,” bases predictions on the actual outcomes of a class of comparable projects; its first practical application concerned the costs of large transportation infrastructure projects (Flyvbjerg, 2006).

Awareness of the mechanisms behind the planning fallacy makes it possible to improve accuracy considerably and, as a result, to offer one’s customers more accurate costs and work schedules.

(Alberto Viotto)