At the London School of Economics' (LSE) recent Student Strategy & Management Consulting Conference in Westminster, Professor John Kay, columnist for the Financial Times and a former Director of Oxford Saa Business School, warned that "bogus models" are causing huge problems in large-scale project and corporate strategy management. Giving a speech entitled "Knowing what we don't know", Professor Kay cited the much maligned Edinburgh city centre tram project, the risk management systems used by banks at the time of the financial crisis, and a host of private finance initiatives (PFIs) led by the UK government as examples of where such models have led to project implosions on a large scale.
Kay's theory, expanded upon in his 2010 book Obliquity, is that many decisions made in the planning and implementation stages of projects in modern politics, business and finance have been justified by erroneous notions of rationality. He argues that many of the decisions involved in bringing these projects to fruition are made using data with very little empirical evidence to support it. Another common issue is that, because these decisions are often made on the basis of mathematical models, anything that cannot be quantified (like many socioeconomic factors) is deemed to be irrelevant and therefore not accounted for. In addition, he argues that many of these bogus models don't make allowances for future shifts in society, which often have a huge impact on the development of a project.
Off the rails
The construction of the Edinburgh tram system began in 2008, with a budget of £375 million. The most recent costing of the project topped £1 billion, and the original route has been greatly truncated. It's been plagued by budgetary problems, periods of construction downtime and legal wrangling, leaving many Edinburgh residents questioning the initiative's viability and necessity.
Kay claims bogus modelling is at the root of the problem. In order to get the go-ahead for the tram system, unrealistic figures were used in project forecasting. He claims that the planners knew what result they wanted from the feasibility tests, and entered the data accordingly. In addition, the project team failed to take social developments into account in their predictions. Finally, they had few or no contingency plans in place for possible future setbacks. As a result, he argues, the project misled the public, and wreaked havoc on the transport system of central Edinburgh.
Such situations are typical, to a certain extent, of the whole of British industry, says Kay. The solution he prescribes is a reevaluation of how planning is undertaken. Models need to start out simple (the state in which many bogus models remain), but be gradually fleshed out with any details that may impact on the project's lifespan. Wider research needs to be conducted into anything that may possibly affect the project's development.
In his address, Kay used the London Underground map as an example of a model of limited use, and which needs to be augmented with further information. For someone visiting London for the first time, it provides excellent elementary guidance. But for a long-term resident who knows the actual distances involved, it's inadequate. To take the tube from Paddington to Lancaster Gate, for instance, means going back on yourself by almost a mile, when the stations are actually 500 metres apart. A truly representative map would show the distance between them on foot, as well as by tube, bus and train, so that the user could build a full picture of their journey. So for a project model to be successful, it must grow as evidence is collected and the plan develops. Until doing so is the norm, argues Kay, projects up and down the country will continue to haemorrhage money.