The Imperfect Science of Construction Estimates

Posted on May 27, 2011


This week I heard from a CEO that a medical office building (MOB) his hospital system was planning came in so under-budget compared to their strategic plan, they decided to buy two: one for their main campus and a duplicate for a satellite location.

Good for him:  leverage the power of money to its fullest in an underwhelming economy. In construction parlance this is called a favorable bidding environment.

More interestingly, though, is why the strategic plan was so off (luckily, to his benefit) in its project cost prediction in the first place? The obvious answer: bad information. Next question to answer to assist in future strategic planning exercises: how can a more accurate project estimate be found?

As a very base economic exercise, construction is a short-selling business. A contractor will eventually assemble all the labor, parts and pieces that go into a building—but not when he is asked to price the project. When someone asks for a price, the contractor uses past experience and provides a back-of-the-envelope estimate.

And when it comes time for actually signing on the dotted line, he makes his best guess at what everything will cost—but he still does not buy it. Why? Because no one has $x million to buy all the parts at once. Also, once purchased, where would all of the…material go until it is needed? So even when signing a contract for a guaranteed maximum price (GMP), it is still an estimate as to what things will really cost, albeit a very high-level, well-thought out estimate.

Construction estimates are difficult because they involve many variables that require prediction, and we know anything that requires knowledge of the future has a lot of uncertainty. To complicate matters, purchase of building materials and labor occurs in the future—and over a long stretch of time.

When pricing a job, a design-builder must look at when the actual purchase of the specific item, aka “buy-out”, will happen and account for this time lag. On Wall Street, stock traders can make or lose a lot of money on this time lag, i.e. speculating that something will go up and buying now to sell at a profit in the future; or in some cases, borrowing to sell something now with the requirement to buy it back in the future, ideally at a cheaper price. In some cases, prices will go up; in other cases, prices go down.

A major item of distinction:  in design-build construction, if prices go up, the design-builder is at risk for escalation, that is, the difference between the price in the GMP and what he had to actually pay for the goods. If a hospital is utilizing design-bid-build or CM-at-Risk, the owner is at risk for this escalation.  If prices go down, the design-builder may be able to save money on the buy-out because goods may cost less in the future than what was allocated in the GMP.

Three factors contribute to a construction project cost: labor, materials and overhead. Overhead is easy to predict because it is constant and controlled by the design-builder; think:  a flat line on a graph. Labor is not as easy because it generally goes up over time, though the rate is never known; think:  a constantly rising slope from left to right on a graph. Materials are the most difficult because they are diverse in cost and their prices are all over the place—trending down, spiking up—all based on regional, national and global economies and events; think:  a roller coaster on a graph.

Even with all the real-time knowledge of the 21st century at a design-builder’s fingertips—elaborate software, historical databases, material futures contracts—and 30 or 60-day bid locks from subcontractors, estimates are still a large part guess. Some design-builders are better at it than others, which leads me to a point from a previous postfind a reliable information source.

So, about that strategic plan….