My favourite saying about organizational effectiveness is:

"To optimize the whole you have to suboptimize the parts."

I've stolen this from Bob Lewis and although it sounds trite it covers a whole lot of assertions like:

• organisations should capture knowledge locally to facilitate its reuse globally;
• a benchmark is only as useful as what it actually measures (rather than claims to measure); and
• money spent on training is more than repaid by productivity improvements.

This also provides an explanation and legitimacy for "soft factors" like happiness even if we accept that the sole purpose of a business is to make money. Employee happiness becomes a 'suboptimal' decision which nevertheless contributes to long-term whole-of-organization effectiveness (higher morale, lower absenteeism, etc).

And when it comes down to it, what is the ultimate role of a government? To ensure the happiness of its citizens (which to date has largely been equated with peace and prosperity). Any public service department becomes only a small part of the government 'machine'...

While on the surface the Public Service often appears grossly ineffective, I have a slightly crackpot theory that a large part of the Public Service's role may simply be to add inertia. By slowing change, it's easier to avoid catastrophic consequences of poor policy. Again, a local suboptimization may in fact be the best thing for the country in the long run.