See more of the story

Since Franklin Delano Roosevelt's New Deal, Americans have thought that for any new president, the first 100 days are critical, because he has a honeymoon period in which Congress will do what he wants. But in the modern era, the first 50 days are the defining ones. That's when the new executive branch is just taking shape, and the White House has maximal discretion to act entirely on its own — and to turn the government in its preferred directions.

The Trump administration seems primed to exercise that discretion. But in a few months, it is likely to slow down, and for identifiable reasons.

I was privileged to serve in the initial year of the Reagan and Obama administrations — the first as a young lawyer, the second in a more senior position. Nearly 30 years separated the two transitions, but the pattern was identical. In late January, new officials assume their positions with bold plans — new executive orders, new presidential memorandums, new processes, dramatic departures from the paths established by their immediate predecessors.

In the first weeks, the president's top advisers have essentially free rein. Much of the new government isn't even in place. If the White House can control its own Cabinet (a potentially challenging enterprise), it has nearly unlimited room to maneuver.

By the end of February 1981, Reagan issued no fewer than nine executive orders; by the end of February 2009, Obama issued 17. And those numbers give just a glimpse of how rapidly, and how dramatically, both presidents redirected the ship of state in their initial weeks.

But after a few months, two things happen.

First, the number of presidential appointees grows exponentially. The full Cabinet is usually in place, and its members assert themselves; the subcabinet starts to fill in as well. That means that whenever the White House wants to do something, it has to include, and listen to, many more people. It must go through an increasingly formalized clearance process, including people who speak for their "building." Internally, there is a sharp increase in the number of people who can veto or at least constrain what the White House does.

For an executive order, that's an excellent idea, because it reduces the risk of mistake, but it also slows things down. If a Cabinet head dislikes a paragraph or a sentence, or thinks the whole idea is stupid, an executive order might be stopped in its tracks, even if the president himself was initially inclined in its favor. Within the executive branch, an unwelcome word is "nonconcur" — and a nonconcurrence can be a serious obstacle.

True, the president or his chief of staff can insist on a certain direction. But it usually isn't a lot of fun for either of them simply to overrule one of their own people, certainly not in March or April of the first year. Instead of overruling, there is often a discussion, which gums up the works, and which may lead people to give up and to move on to other priorities.

Second, the civil service starts to matter much more. In the first weeks, a new president and his team may well be suspicious of those who recently worked for his predecessor. A new sheriff is in town. The Reagan and Obama teams were courteous and respectful, but in the initial period, the views of civil servants sometimes didn't matter a whit.

As the weeks go by, that changes. The civil service becomes your team, not theirs. A new administration discovers that civil servants know a lot — far more than it does — and surprise: They aren't at all political. Ultimately they will follow your direction. But they often introduce important (and usually helpful) cautionary notes, and they slow you down.

Surprisingly, that can be great. In April 2009, a friend of mine in the Obama administration exclaimed, "I am now thinking that at least half of the things I hate most about the Bush administration are not even true!" He learned that there were good reasons for many of the actions taken by Bush-era officials. A polarized political process had blinded us to those reasons.

Whether or not large-scale change is a good idea, people in the civil service become a stabilizing force. This is not because they are wedded to the status quo (they aren't), but because they offer a relevant perspective — and because it becomes increasingly clear that they need to be consulted.

Just as under Reagan and Obama, Trump's team is moving very quickly to reorient the policy direction of the executive branch. And after the unprecedented nature of Trump's campaign, no one can rule out the possibility that his White House will turn out to be something that we haven't ever seen before. But by April, it is likely to become far more deliberative, and while it will occasionally get frustrated, it will probably embrace what it will come to see as regular order.

Which means that Trump's first 50 days will be defining — the best opportunity for rapid, large-scale change, and also the most likely period for recklessness.

Cass R. Sunstein is a Bloomberg View columnist. He is the author of "The World According to Star Wars" and a co-author of "Nudge: Improving Decisions About Health, Wealth and Happiness."