In 1960, Richard Neustadt published "Presidential Power." He argued that the office of the presidency was inherently weak, and that presidents should attempt to accumulate as much power as possible, which would help them and result in good public policy.

Fifteen years later, one president who seemingly did everything possible to accumulate power had fled from office amid policy disaster in Vietnam. His successor, who also apparently had single-mindedly sought to aggrandize the presidency, resigned ahead of certain impeachment and conviction.

Yet what Lyndon Johnson and Richard Nixon did wrong wasn't their quest for power. It was how they went about it.

Power, for Neustadt, was always transactional. He called what he wanted presidents to do "persuasion," but that's misleading, as are many of his terms. It's not that he wanted presidents to use strong logical or emotional arguments to persuade people to do things. He wanted them to use the large number of bargaining chips inherent in the job to cut deals. By doing so, and by impressing people of his skill in finding plus-sum bargains, he would make each subsequent round of bargaining easier.

This is imposed by a system of government (of "separated institutions sharing power") that gives presidents almost no authority that isn't ultimately shared: with Congress, the courts, the bureaucracy, the states and local governments, and even those outside the government, such as political parties, interest groups and the press.

In such a system, influence is only possible through skilled bargaining, which requires all sides to trust one another, and for presidents to see their goal as getting things they want, but not by defeating other legitimate players, and certainly not by crushing them. To get things done, presidents need others to bargain with.

advertisement | advertise on newsday

Therefore the first substantive chapter of "Presidential Power" is about how giving orders (not just to Congress - to the bureaucracy above all, but also to all those who share authority with the president) doesn't work very well. First of all, unless the circumstances are right, presidential orders are usually ignored. But even when they are followed, the price tends to be high in terms of future bargaining abilities.

Which brings us back to Nixon and Watergate. Seen one way, Watergate, and the Nixon presidency in general, were experiments in attempting to govern by command, rather than by bargaining. Take, for example, the establishment of the "Plumbers," the unit set up in the White House to gather intelligence and disrupt Nixon enemies. The Plumbers only existed because Nixon failed to persuade the Federal Bureau of Investigation to do the job, and then, frustrated, ordered something to be done about it.

This kind of thing wasn't limited to the illegal, unethical, or just plain sordid activities we know as Watergate. It was Nixon's standard operating procedure. Think, for example, of the opening to China - kept secret from the State Department, including the secretary of state, until the last minute. It's so much easier, it seemed to Nixon, to just cut out those who would have to be swayed than it would be to win their support.

Because Nixon acted, some of what he wanted happened. Command, it seemed, worked. And yet ultimately, as Neustadt predicted, command had enormous costs. Agencies that Nixon worked around instead of working with saw no reason to support the president when he needed it. Nor did, eventually, anyone else in the system. Skilled bargaining presidents make it in the interests of as many people as possible to have the presidency succeed. For Nixon, almost no one (at least outside the White House) believed Nixon's success was critical.

The Nixon presidency, then, only confirmed what Neustadt said. Moreover, it confirmed that when it comes to presidential power, there are no shortcuts - or at least, none except ones that potentially carry tremendous costs.

A quick note on President Barack Obama and executive action, since it's been in the news all week. Executive action doesn't necessarily mean acting alone, or cutting out peoples who are normally part of bargaining. It all depends on how presidents go about those actions. Do they, for example, work with a wide range of important players in the relevant policy areas, both inside and outside the government?

If so, working around Congress may be a perfectly acceptable form of bargaining (and threatening to take executive action can almost always be an acceptable way to use presidential bargaining chips). If not, executive action can start to pile up the costs Neustadt talks about.

Jonathan Bernstein is a Bloomberg View columnist covering U.S. politics.