President Joe Biden this month blocked a Japanese company from buying the struggling US Steel – a decision to which, reporting from the Wall Street Journal and Washington Post indicates, most of the president’s senior foreign policy advisers objected. On top of several other suboptimal choices, among them the chaotic withdrawal from Afghanistan and the slow-rolled support and weapons-use authorisations for Ukraine, this episode justifies the judgement that, by its own criteria, the Biden administration’s foreign policies have failed.

Observers for years have been analysing why a well-intentioned man with almost half a century’s experience of international affairs has chosen so poorly, so consistently. Bad presidential judgement surely plays a role here – former Defense Secretary and CIA Director Robert Gates in 2014 claimed that Biden had been ‘wrong on nearly every major foreign policy and national security issue over the past four decades’ – but this can be only an incomplete explanation. No presidents, after all, have perfect instincts all the time.

That’s why they assemble teams of advisers to help them develop their best ideas and to talk them out of their worst. How they do so, it turns out, helps to explain Biden’s deficient foreign policy outcomes – and deepens our understanding of why some other presidents have found success despite their own deficiencies in experience or judgment.

Chief executives can structure their closest circle of advisers in many ways, but the options at each end of a spectrum stand out. Each pole has strengths and weaknesses.

The first approach is a ‘team of rivals’ model, which Abraham Lincoln employed to great effect in the 1860s and which many others have used in various ways, if not to the same extreme, before and since. This features a collection of advisers chosen more for their independent standing or the constituencies they can deliver than for their strict alignment with the president’s views.

As Doris Kearns Goodwin described, Lincoln assembled a cabinet of men who did not care much for the president or for each other. William Seward at State, Salmon Chase at Treasury, and Edward Bates at Justice brought instead independent judgement that they would express energetically to the president, even when he did not want to hear it. And their political standing – not tied primarily or even largely to Lincoln – gave him good reason to listen.

The second approach is a ‘staffer’ model, which is more common in the modern presidency. This features a team around the chief executive dominated by loyal advisers, men and women selected less for their separate political contacts and bases of support than for their dedicated service to the president and, generally, alignment with his views.

Biden went in this direction in building a body of foreign policy advisers, including Secretary of State Antony Blinken, National Security Advisor Jake Sullivan, Secretary of Defense Lloyd Austin, and Deputy National Security Advisor Jon Finer, who had all worked closely with Biden, at least back to his time as vice president, and none of whom had an independent political base.

Most teams of advisers land somewhere on the spectrum between strict versions of each approach. Such a balance – some loyal, committed staffers and some independent, political rivals – tends to work well with most presidential personalities. Take George H.W. Bush, who as president assembled a cadre that combined people with long experience working with him (exemplified by Secretary of State Jim Baker and Secretary of Commerce Robert Mosbacher), but also took care to include strong personalities with their own political base (such as Secretary of Defense Dick Cheney and Chief of Staff John Sununu).

And that makes sense. Going all in on one or the other archetype only works well under rather uncommon circumstances; more often, the polar models’ pathologies will outweigh their benefits.

The team of rivals worked fairly well for Lincoln. Of course, he was a highly unusual president, bringing a combination of minimal military and government experience and an almost superhuman willingness to listen to others, whom he knew to be of different persuasions, and thereby consider policies not of his choosing. Others who have attempted it since then, such as George W. Bush in his first term or Barack Obama soon after, didn’t fare so well. Bush had power players Secretary of Defense Donald Rumsfeld and Vice President Dick Cheney teaming up against Secretary of State Colin Powell. Obama’s first-term foreign policy team included such experienced political actors as Vice President Joe Biden, Secretary of State Hillary Clinton, and Secretary of Defense Leon Panetta, who all found themselves and their aides battling in the press as well as with the commander in chief.

Even such a public fan of the Lincoln model as Obama moved away from it as he relied more on the staffer type in his second term. For presidents who are neither preternaturally persuasive themselves nor liable to be convinced by aggressive argumentation – which turns out to be most of them – a radical team of rivals falls short.

Most chief executives, however, find dramatic benefit from having at least one heavily political figure in the inner circle of foreign policy advisers. While long-time loyal staffers often find themselves thinking similarly to the president they serve, and struggle to affect his judgement because their reputations depend largely or fully on him, strong voices with independent constituencies represent, and have the backing of, centres of power less reliant on the current president. They bring a political price to the president who does not listen seriously to, and sometimes go along with, their advice. Good policy doesn’t necessarily emerge from this, but bad policies nursed by groupthink are less likely to make it through. And incentives to loyally defend bad policies are few.

That is the core problem with over-reliance on the staffer model. Presidents need to be challenged – yet a top team composed entirely, or almost entirely, of staffers is the least likely to perform this function. A president like Biden pays no price for failing to heed the collective advice of his own secretary of state, deputy secretary of state, national security advisor, deputy national security advisor, treasury secretary, and others – as the recent reporting suggests he did in the Japan/US Steel decision – because he has no compelling reason to listen to those who depend more on him than he does on them. Overruling their counsel is all too easy.

It’s one of the (many) reasons why George Shultz was a more effective secretary of state for President Reagan than his predecessor at Foggy Bottom, Al Haig. Although not a formerly elected politician with a traditional constituency, Shultz was firmly outside the follow-Reagan-from-California-to-Washington clique. He had already served as secretary of labor, director of the office of management and budget, and treasury secretary – all before Reagan even ran for the highest office in the land. His clout was independent of Reagan, so when he argued forcefully on policy or even on process, to the point of threatening to resign multiple times, the chief executive usually listened.

To be sure, the staffer side of the spectrum makes a president’s policy easier to formulate and execute than a typically stressful, potentially chaotic, team of rivals approach. The staffer model minimises friction in the policy process; loyalty to the president is its signature characteristic. And the president pays no political price for overruling their counsel because they have no power bases to call on. But policy, even high-level staffers must sometimes be reminded, is inescapably a political act.

Over-reliance on a staffer model makes it all too clear that bad decisions coming out of the White House are the president’s own. And for a person like Joe Biden, with at least one eye on his legacy, that probably stings a lot more than difficult policy roundtables with independent political actors would have. And, based on his nominations so far, President-elect Trump is about to make the same mistake.