ule No. 1: Always have a cover story. The ostensible purpose of the Bush administration's plan to open up 850,000 federal jobs to private competition is to promote efficiency. Competitive vigor, we're told, will end bureaucratic sloth; costs will go down, and everyone — except for a handful of overpaid union members — will be better off.

And who knows? Here and there the reform may actually save a few dollars. But I doubt that there's a single politician or journalist in Washington who believes that privatizing much of the federal government — a step that the administration says it can take without any new legislation — is really motivated by a desire to reduce costs.

After all, there's a lot of experience with privatization by governments at all levels — state, federal, and local; that record doesn't support extravagant claims about improved efficiency. Sometimes there are significant cost reductions, but all too often the promised savings turn out to be a mirage. In particular, it's common for private contractors to bid low to get the business, then push their prices up once the government work force has been disbanded. Projections of a 20 or 30 percent cost saving across the board are silly — and one suspects that the officials making those projections know that.

So what's this about?

First, it's about providing political cover. In the face of budget deficits as far as the eye can see, the administration — determined to expand, not reconsider the program of tax cuts it initially justified with projections of huge surpluses — must make a show of cutting spending. Yet what can it cut? The great bulk of public spending is either for essential services like defense and the justice system, or for middle-class entitlements like Social Security and Medicare that the administration doesn't dare attack openly.

Privatizing federal jobs is a perfect answer to this dilemma. It's not a real answer — the pay of those threatened employees is only about 2 percent of the federal budget, so efficiency gains from privatization, even if they happen, will make almost no dent in overall spending. For a few years, however, talk of privatization will give the impression that the administration is doing something about the deficit.

But distracting the public from the reality of deficits is, we can be sure, just an incidental payoff. So, too, is the fact that privatization is a way to break one of the last remaining strongholds of union power. Karl Rove is after much bigger game.

A few months ago Mr. Rove compared his boss to Andrew Jackson. As some of us noted at the time, one of Jackson's key legacies was the "spoils system," under which federal jobs were reserved for political supporters. The federal civil service, with its careful protection of workers from political pressure, was created specifically to bring the spoils system to an end; but now the administration has found a way around those constraints.

We don't have to speculate about what will follow, because Jeb Bush has already blazed the trail. Florida's governor has been an aggressive privatizer, and as The Miami Herald put it after a careful study of state records, "his bold experiment has been a success — at least for him and the Republican Party, records show. The policy has spawned a network of contractors who have given him, other Republican politicians and the Florida G.O.P. millions of dollars in campaign donations."

What's interesting about this network of contractors isn't just the way that big contributions are linked to big contracts; it's the end of the traditional practice in which businesses hedge their bets by giving to both parties. The big winners in Mr. Bush's Florida are companies that give little or nothing to Democrats. Strange, isn't it? It's as if firms seeking business with the state of Florida are subject to a loyalty test.

So am I saying that we are going back to the days of Boss Tweed and Mark Hanna? Gosh, no — those guys were pikers. One-party control of today's government offers opportunities to reward friends and punish enemies that the old machine politicians never dreamed of.

How far can the new spoils system be pushed? To what extent will it be used to lock in a permanent political advantage for the ruling party? Stay tuned; I'm sure we'll soon find out.