Every few months another major infrastructure project appears in the headlines having exceeded its budget by hundreds of millions, sometimes billions, of dollars. The reaction is usually predictable. Politicians blame inflation, supply chains, labour shortages or unforeseen complexity.

Critics point to incompetence, poor governance and failed planning. The assumption underpinning both arguments is that the original budget was intended to be an accurate estimate of the final cost. But was it?

That question interests me because it shifts the inquiry away from project management and towards incentives. We assume the purpose of a budget is to forecast reality. Yet in large public projects the first challenge is often not delivery. It is approval. Before a tunnel can be built, a railway extended or a hospital expanded, someone must convince government, Treasury and ultimately the public that the project is worthwhile. Which raises a provocative possibility.

What if the budget is not merely a forecast? What if it is also a sales document?

Imagine a project that genuinely costs $20 billion. If that figure is presented upfront, does the project proceed? Perhaps. But perhaps not. There are competing priorities, fiscal constraints and political realities. A lower figure, however, may make the project appear affordable, achievable and worthy of support. Once approved, the project develops a momentum of its own. Contracts are signed. Construction begins. Jobs are created. Communities adjust their expectations. Political capital is invested.

Years later, when the cost inevitably rises, the conversation is no longer whether the project should commence. It is whether it should be abandoned. At that point billions may already have been spent. The sunk cost effect takes hold. Nobody wants to explain why a half-finished tunnel should remain half-finished. Nobody wants to defend a cancelled railway line after years of disruption. Continuing becomes easier than stopping. Which makes me wonder whether some cost overruns are less about forecasting failure and more about the realities of getting major projects approved in the first place.

Of course, that is not to suggest every cost overrun is deliberate or that corruption, industrial relations, governance failures, scope creep, inflation and labour shortages do not exist. They clearly do. In Victoria, for example, recent scrutiny of construction industry practices and allegations surrounding the CFMEU have added another dimension to the discussion. But even if corruption is a factor in some projects, it does not fully answer the broader question.

Why do large projects across different governments, jurisdictions and decades so often exceed their original budgets?

The common thread may not be the individuals involved. It may be the system itself. Optimism is rewarded. Ambition is rewarded. Big visions are rewarded. The person arguing for caution, contingencies and worst-case scenarios is often competing against someone promising transformation, growth and prosperity. Under those conditions, should we be surprised when estimates lean towards optimism?

Perhaps the most interesting question is how we define success.

A tunnel that costs twice its original budget may still transform a city for the next fifty years. A railway line may exceed every forecast and still generate enormous economic and social value. Conversely, a project delivered precisely on budget may fail to produce any meaningful benefit at all. Yet public debate tends to focus overwhelmingly on the variance between forecast and actual cost. We treat the budget as the scoreboard. Maybe it isn’t. Maybe it is merely one measure among many.

Which brings me back to the original question.

When a major project exceeds its budget by billions of dollars, are we witnessing inept budgeting? Or are we witnessing a political process in which projects must first become acceptable before they can become reality? Because if the latter is even partly true, then the cost overrun may not be the story. It may simply be the consequence of a much earlier decision about what was required to get the project underway.