Jay Manifold, who blogs a purely technical critique of the Californian photovoltaic regulation," emails some thoughts on productivity:
Re: "The Big (Econonmic) Story" -- kudos for pointing this out, especially Brad DeLong's "becoming smarter about organizing production processes." Sprint's in-house process-improvement training (which we got from somebody else -- we didn't make this up) claims that any process that's been around for a while is susceptible to having as many as three-quarters of its steps removed while remaining capable of achieving the needed result. The potential for increased productivity at any given moment is therefore not so much a percentage as a multiple.
I don't think it's a coincidence that the last four years, with their high-profile difficulties in IT, a recession, and ongoing struggles in (for example) the airline and hospitality industries, have seen striking productivity increases. Harold Kerzner spoke at a PMI chapter meeting here in KC back in '01 (and will be here again this fall; see http://www.kcpmichapter.org/Symposium2004/Symp04home.html, which describes an event I will certainly attend). One of his more memorable remarks was that the recessions of the early '80s and '90s saw the greatest improvements in project management methodologies, and that as for the then-current recession, "I hope it lasts 25 years!"
This must be carefully distinguished, of course, from the notorious "broken window" fallacy, since economic downturns aren't, by themselves, good for the economy. I think that keeping the "triple constraint" of project management in mind (colloquial version: cheap, fast, good -- pick two) resolves this. Assuming the effect of a recession is to force operations to be cheaper, maintaining scope/quality requires that they be redesigned to be somehow faster and better.
Somewhere in the above I segued from talking about processes to talking about projects, and then back again. But the development (or redesign) of a process is itself a project, and the PMBOK devotes an entire chapter to quality management, with special attention to the tools and techniques of quality control. The assumption is that project managers must become intimately familiar with repeatable, high-volume processes.
I am proud of my employer's openness in this area, which distinguishes it from at least some of the "companies that don't want to talk about their operational secrets"; see here.
In the real world, it takes a finite amount of time to redesign processes -- or design them in the first place -- and gain their acceptance and use by those who will be carrying them out. It is nonetheless reasonable to suppose that a sufficiently committed organization could double its productivity every three years. The quality teams and process-improvement initiatives I've participated in have typically paid for themselves in less than six months and eventually resulted in financial gains, or prevented financial losses, equal to several times the annual salaries of all the team members combined. Indeed, a target of 10x one's salary in cost savings is not unheard of.
On the journalistic side of this story, Neiman Watchdog (the site of Brad DeLong's original post), posted an item about my earlier discussion