Bloody Shovel writes some complementary things about my Employment Policy post.
The comments run off into the direction of the pathologies of organisations, Nydwracu citing Robert Anton Wilson.
The SNAFU principle is real enough, but the central issue Shovel is emphasizing is not that; organisations like the education system fail at their nominal purpose not primarily because they are uninformed or incompetent (though they generally are), but because their real purpose is not their nominal purpose. Their real purpose is to survive and grow. C. Northcote Parkinson is more to the point than RAW.
(That's not a complete opposition; the reason that RAW says that communication is only possible between equals is because otherwise the interests of the subordinate are not the interests of the superior. Whichever way you put it, the root problem is the agency problem.)
In the modern world, we have two treatments for Parkinson's Law: one which sometimes works and one which never does.
The one that sometimes works is market competition. An organisation which has to fund itself in the market must succeed in satisfying its market or go out of business and be replaced. Big, powerful organisations have crumbled in the face of competition, and many others have reformed themselves effectively to avoid that fate.
But not every organisation has to compete. Government power exists and will always exist, and is a far more reliable nourishment. Not only do the inevitable organisations of power —- the parties, the security forces, the tax-collectors —- feed on it, but organisations which would otherwise have to compete in a market seek to secure government lifelines. Bankruptcy, Carlyle wrote, will bring down all falsehoods, but in the case of governments, rather too late for most of us.
To restrict government-fed organisations to their proper purpose, therefore, another treatment must be used. The one we rely on is to create a second organisation to regulate the first. This never works, for reasons too obvious to labour. First you create (or take over) a school. Then you create a board to control the school. Then an education authority to control the board. Then an inspectorate to control the education authority. The end result is you have four organisations pursuing their own agendas instead of one.
I can suggest possible better solutions to the problem, but not without a context. The normal context of any policy suggestion is the framework of organisations that make up modern government. The policy suggestion is then aimed at some organisation in that framework to carry out. In this case, that is obviously nonsensical. Appointing a single absolute ruler is not in itself a solution to the problem of organisations, but it is at least a context in which solutions can be meaningfully suggested. That, then, is the context I assume in making suggestions. As far as what "we" do now, in the current context, my answer is passivism: we merely remark that the problems are not treatable in this context.
The alternative I suggest, in the context of absolute rule, is to do, as far as possible, without organisations.
Tim Worstall made a revealing little post about the GSM cellular communications standard. An Observer article remarked that the group of European telecommunication companies which established the standard was "the kind of intergovernmental initiative that drives Ukip nuts". Worstall, a (former, I think) UKIP press officer, is baffled by the suggestion that opponents of the EU, a permanent transnational government, would be opposed to an ad-hoc agreement (including non-EU members such as Norway and 1993 Finland) to carry out a single task. It should not be baffling; the institution-centric worldview of the political mainstream simply does not allow it to see the vital difference. That is a sufficient explanation for the state we are now in.
A wise ruler would not delegate permanent power or independence to any organisation. The principle should be that any organisation can be abolished by one person, and most should have a defined life. Obviously, an organisation can be abolished in theory and in reality just reconstituted under another name: the important principle is not to be dependent on an organisation so that it is able to do that. It is not likely that a ruler can avoid depending on anything, but it is better to depend on a person than on an organisation. Let the depended-on person build and destroy ad-hoc organisations the way the ruler does; the responsibility stays with him. It is better to depend on a lieutenant you know (and ideally have chosen) rather than on bureaucrats you cannot even name.
That sounds almost impossible, but we are conditioned by a world of large organisations dedicated to surviving and growing. It obviously entails a sacrifice to not have large permanent organisations, but the benefits could be still larger. After all, in commerce, there is an enormous sacrifice of efficiency involved in the duplication of functions by competing firms, and more sacrifice of efficiency in the destruction caused by bankruptcy of failed competitors. But in commercial fields the benefits of limiting the growth of cancerous dysfunctional organisations seem to consistently outweigh the very significant costs. The same may be true of government by temporary ad-hoc organisation.