Almost every president since Herbert Hoover has tried to reorganize the executive branch of the federal government; none has thought he accomplished much. President Reagan brings a new feature to this quadrennial enterprise: he has proposed to "reorganize" by abolishing various agencies. The Department of Education and the Department of Energy have been the most conspicuous targets of his reorganizational enthusiasm. I think he should forget it.

Creating a new Cabinet agency or reorganizing an existing one is very costly in scarce political capital. Joseph Califano, a key aide to President Lyndon Johnson, later reported that the effort to create the Department of Transportation was one of the most difficult, time-consuming tasks undertaken by that administration. Creating the Department of Housing and Urban Development was no easier. President Carter managed to create the departments of energy and education, but not without a serious fight.

When I was chairman of the National Council for Drug Abuse Prevention, an agency created by Congress in the early 1970s to help "coordinate" the campaign against heroin addiction, I was involved in the effort to merge the old Bureau of Narcotics and Dangerous Drugs with part of the Customs Service, to create the "new" Drug Enforcement Administration. It was enormously difficult -- Congress was reluctant to approve the change, many parts of the bureaucracy fought it tooth and nail, various labor unions representing affected employees were upset by the proposed personnel transfers, and the administrative confusion accompanying the merger sapped the energy of the DEA for years while it tried to get its bearings. And at the street level -- down where the narcotics agents and customs inspectors work in the gritty struggle against drug traffickers -- nothing really changed. The job remained pretty much the same; only the obstacles to doing the job changed. They got worse -- at least in the short run.

Reagan may suppose that the costs and frustrations associated with creating or merging an agency will not accompany the abolition of one. But breaking up an agency is not like wrecking a building -- it can't be done with dynamite and sledgehammers. Most of the constituent parts of any Cabinet department will -- must -- survive, and therfore they will have to be moved somewhere.

Relocating the bits and pieces of a dismantled agency is akin, as someome once said, to moving a cemetery -- it must be done cautiously, in the dead of night, and with due regard for the feelings of the bereaved.

In fact, most of the reorganizing that goes on is not necessary, and much of it is not even useful. What counts is not organization but policy -- what the government does, at the street level, on a daily basis. Now, political scientists, such as myself, like to think that policy is powerfully shaped by the organization in which it is embedded. I think we exaggerate a good bit: after all, we are paid to study organizations, and so not surprisingly we tend to think them important.

But even if organization does matter -- and in some sense, of course, it does -- the problem facing a president is not to discover ways of molding all governmental policies so that they correspond to his preferences. Some presidents pretend that is their goal, but the facts of life are against them. The government is too big, too complex to be managed from the center or even from a strategic outpost. Organizationally, the best a president can do is find able subordinates and motivate them to accept responsibility for their own actions. The presidency is not a manageral job. It is an executive job -- that is, it entails policy direction. Besides, Congress and the courts have something to say about the daily work of the government, and no president can possibly win all of the battles he may have with these institutions.

There is one institution the president must manage, and that is his own White House. And that is hard enough, without trying to manage other people's houses. If there are two books I would recommend that any president -- or at least any president's chief advisor -- should read, it would be these:

First, "Presidential Agency" by Herman Somers, an account, published many years ago, of the various unworkable devices tried by Franklin D. Roosevelt in his efforts to mobilize the domestic economy to face the challenges of World War II. Roosevelt, surely one of the worst administrators (and most brilliant leaders) ever to occupy the White House, did almost everything wrong before turning the job over to Jimmy Byrnes, whose Officer of War Mobilization was a great success. (It employed 10 -- that's right, 10 -- persons.)

The second is Roger Porter's "Presidential Decision-Making," recently published, that describes the successful effort to get advice on ecomomic policy to President Ford in ways that brought the key Cabinet leaders together in a useful setting.

How the president gets advice on policies he must make is of crucial importance, as Somers and Porter make clear. But how he sets about organizing the executive branch is much less important, as the experience of most presidents will confirm.

The president's attitude toward the Department of Energy or the Department of Education or whatever should be selective and policy-related. He should ask himself these questions: What do these departments now do that I wish them to stop doing, or do differently? And - just as important -- which of these changes that I would like to make can I afford, politically, to make? And finally, given what I want to do and what am I willing to try to do, what blunt instrument, what sharp tool, can I find that has the great chance of forcing the change? It may be a budget cut or ceiling. It may be a new administrator or an executive order. It may require new legislation altering policy. Rarely will it require -- or even be helped by -- a Cabinet reorganization.