Mankiw's post on how Research Assistants estimate their time is eerily similar to how state agencies estimate time. The way professors estimate time is also similar to someones way of thinking, that of a private project manager.
State agencies greatly underestimate the necessary time for so many tasks that in the end they have to defend by their projects are not on time. ACA is only the latest big example, although a very obvious one and on a very grand scale. In Germany, I could name at least 3 projects that are very similar: BER airport, Stuttgart 21 and the Karlsruhe Subway. All three projects run late and in the case of the Berlin airport, it actually displays all the signs of dysfunctional government planning. Commitées that fought each other, quotations that were selected on the wrong basis and timings that did not include a lot of reserve times in case of iterational development.
When it comes to private project, the project is often not quicker, but the initial sample timing already includes a lot of back-up time so that the executive comitee can already decide on a more conservative plan. Sometimes this means the death of a project, but in that case probably it is justified. Perhaps the same would happen with government projects, the question is would that really be a bad thing???