Eli5: why are estimated budgets for big construction projects always wrong and end up spending more than expected? Why can’t we get better at these estimates?
In: Economics
What if there were penalty clauses for getting estimates wrong? £1m over budget, £2m fine.
Every such project is a one-off and in the US, at least, governments are loath to undertake any kind of large public works project. How many times was Boston’s “Big Dig” built? Once, of course. It was a really fantastic idea that, before it was even halfway built, became a particularly egregious example of poor management, lack of long-term thinking, incompetence in both government and the private sector, and corruption. It will be on anyone’s mind for generations whenever a big public project is proposed, and that alone will prevent a lot of projects from getting off the ground. As a result, any lessons that could be learned are likely to be forgotten by the time anyone else decides to bury a freeway under a major US city.
New York’s Second Avenue Subway is another good example in certain different ways. A major driving factor there is that it’s been almost a century since any significant subway construction was undertaken in New York City. As a result, there were no experienced contractors to bid on most major aspects of the project, and so it’s not a big surprise that bids turned out to be nonsense. Even if records from the 1930s were perfect, today’s construction has many more requirements for health and safety; there’s more stuff buried underneath the city than back then; and the customer wants a nicer finished product.
A final major factor is political. Let’s say today’s administration has a Big Idea for a public works project. Step One is to convince the voters that the project is a good idea and get funding. This may happen before today’s administration gets voted out. If not, the project probably doesn’t happen. In the event that the new administration takes up the Big Idea (or in the event that some outside force makes the project a must-have), Step Two is putting together an RFP and getting bids. Again, it could take until everyone involved is out of a job, and then the project probably dies. In the increasingly unlikely event of Step Three, the least realistic bid is accepted, and the project begins. By now, it’s years later and the project requirements are guaranteed to have changed: increased population needs increased capacity; the big airlines merged and now the airport serves 40% fewer passengers and doesn’t need that runway; the FAA changes regulations and now there’s a need for TWO new runways. Could be any number of things that have made the Big Idea almost unrecognizable, and all the bids are now bullshit, but if you try to start over at Step One, then literally nothing will EVER get done. The best option is to soldier on and suck it up on the cost overruns and hope you’re either out of office before word gets out, or hope you come into office at the right time to shit on the last guy and pretend to “right the ship.” Everyone who’s else, who’s NOT the face of the administration, has a really bad taste in their mouth and will work to discourage any future Big Ideas.
Most public projects are low bid – the participating bidders don’t know what everyone else is bidding, lowest bid wins. In that setting the contractor bids can be wildly different from each other, sometimes an order of magnitude.
Estimates try to predict what that low bid will be, but it’s very difficult when even the people trying to win the job are pricing things that different from each other.
Some items – like paving a road- are commodities and are very easy to get close based on the quantity of yards, tons, etc.
Others, like a unique building or some work that could be done using specialty equipment are really hard to guess what the contractor plans to do.
Finally, timing is everything. Sometimes a contractor already doing a job down the street could save money mobilizing. Maybe the contractors in the area that do the type of work are all busy and would only take the job at a premium. And maybe the permits for the project took longer than anticipated and the project costs another year of inflation in prices.
TLDR it’s not easy
Source – engineer that is responsible for estimates and public bids
Gov: Hey can some one build us a bridge?
Builder: Could you tell me more about….
Gov: WE DO NOT ANSWER QUESTIONS!
Because the initial estimate of schedule and cost are always success-oriented in order to get the project green-lit. If you include too much margin for problems, you get told that you’re a pessimist and to cut it down or the project won’t be approved. Then once you’re in execution and unforseen problems arise, you start to slip schedule and have cost overruns. The larger and more complex the project, the number of issues and their impact goes up exponentially.
Because every estimate, whether for cost or schedule, is trying to do something nobody is very good at: predicting the future. Estimating for these projects involves making guesses and assumptions that may not be able to be validated for *years*. Plus there are the completely unpredictable, random events that nobody sees coming. Do you think anyone estimating a 5 year project in 2018 anticipated a 2-year delay due to COVID? Or that a ship would crash into a bridge, shutting down a major port and delaying materials?
Prices are pretty much best case estimates (if not underestimated) because for public bidding the lowest price usually wins.
So there’s no incentive to be accurate if that puts you in the “too expensive” bracket.
Basically how most of these projects work is that there are bids. Companies will give their plan and their budget and then the client chooses which one they want.
The two big issues this leads to is
1. People underestimate the budget so that they are more likely to win the bid. They want the client to choose their cheap option and by the time they have to go over budget, it’s too late to switch. The budgets are usually somewhat accurate but don’t take into account some smaller factors or are built off the idea of no mistakes, which is unrealistic
2. They have to put together these plans quickly and generally most of the people that bid on something don’t get the job, so it’s hours of work for no pay. They often kinda rush through the plan and if they win they come back and refine it. As you take into account more factors, it rarely decreases the projected costs
Big jobs of any sort require a LOT of moving parts coming together. The time between quote and start of work can be months. That’s a lot of time for prices to fluctuate, schedules to change, regulations to change, and so on and so on.
There’s plenty of illegitimate reasons as well. Overscheduling, understaffing, flat out lying to get the job in the first place, etc…
Overall, because there’s so many moving parts, there’s a lot of room for both unexpected, and ill-intentioned, changes to the original quote.
Because a lot of the public in America has this idea that the private construction company that does the project, should be the one that assumes the risk for the project (despite it being a government project that the public or government is in favor of).
The reasonable approach to having to assume risk, is to estimate a budget that includes such risk in it.
Alon Levy has a blog post about why, despite it going against some people’s common sense, it has to be the other way around and the government has to assume the risk:
https://pedestrianobservations.com/2020/11/30/who-should-bear-the-risk-in-infrastructure-projects/
Latest Answers