4 Years Later: President Trump and Global Catastrophic Risk

Four years ago on 18 November 2016 I posted a piece called President Trump as a Global Catastrophic Risk in which I argued:

“Because of his character, his stated policies, and uncertainty about what he will do as President, Trump likely increases the risk of a global catastrophe. He likely increases two general risks, or drivers of risk: increased international tension and a rise of authoritarianism. He also likely increases four specific risks: climate change, nuclear war, pandemics and risks from emerging technologies.”

Four long years later, in this post I review some of the claims in that piece. Unfortunately, I find that many of these concerns have indeed been borne out. A global catastrophe is typically thought of as an event with hundreds of millions of deaths, possibly 10% of the global population. That has not come to pass, but the risk of that has increased. The risk that has been highlighted most visibly is of course a pandemic, but I am as concerned about the backwards progress on nuclear weapons and climate change.

Trump has increased global catastrophic risk over the last four years. But he could increase it even further in a second term. He would continue to increase international tensions, encourage authoritarians, block action on climate change, allow the last major nuclear agreement with Russia to end, mishandle the risks from emerging technologies, and poorly manage another—possibly even worse—pandemic.


I suggested that the risk increases were driven by Trump’s character, policies, and the uncertainty about his actions.

“Donald Trump seems to have a worrying character. He seems temperamental, volatile and unpredictable. He also seems thin-skinned and to have a big ego. He seems to have a short attention span and little willingness to pay attention. He has clear authoritarian tendencies. He has a past of somewhat shady business dealings and political donations—he is currently on trial for fraud. He has also been accused of sexual assault and of racial discrimination.”

The worries about his character have been proven true. As many have said, the Presidency ‘doesn’t change your character, it reveals it’.

Carlos Lozada’s recently reviewed 150 non-fiction books on the Trump administration, which he calls ‘chaos chronicles’, including: Fire and Fury, Fear, Rage, A Very Stable Genius, Holding the Line, Team of Vipers, Audience of One, Devil’s Bargain, A Warning, A Higher Loyalty: Truth, Lies, and Leadership, Disloyal, American Carnage, The Toddler in Chief, Trumpocalypse, True Crimes and Misdemeanors, The Fifth Risk, and Unmaking the Presidency. His conclusion was that: “It seems clear that many of the president’s senior advisers, especially those in the national security realm, are extremely concerned with his erratic nature, his relative ignorance, his inability to learn, as well as what they consider his dangerous views.”

Trump’s stated policies have made the global catastrophic risk situation worse, as I will discuss below.

We faced some uncertainty in 2016. Almost uniquely in US history, someone with no political or military experience had become President. It was unclear what would happen—I noted that “it could be that the real risk of Trump is low. We will find out more over the course of the next few years.” We have now found out.

We face new uncertainties over the next four years. Trump has been even less clear than in 2016 as to his policy platform—what he would do on domestic and foreign policy. Trump’s character flaws will likely worsen. He is now 74, and there are legitimate concerns about his health (especially following his infection with COVID-19) and general capability. The Presidency is a tiring, exhausting job (even if 60% of Trump’s day is ‘executive time’): years of pressure and deadlines, isolated from the real world. In his second term he would be freed from the incentives to try and be reelected, and vindicated by a 2020 election he was expected to lose, one would expect Trump to lean into his prejudices and intuitions even more.

Overall then, Trump’s character and policies (and uncertainty about those) are the key drivers of increasing global catastrophic risk, and these would likely worsen in a second term.

General risks

Increased instability

Trump’s protectionism, isolationism and indifference to authoritarianism has undermined the three pillars of the postwar liberal global order: trade, security alliances and liberal democratic values. The result has been increased international instability, increased international tensions, and decreased international cooperation.

For example, on trade, Trump pulled out of the Trans-Pacific Partnership, replaced NAFTA with the United States-Mexico-Canada Agreement (USMCA) and raised tensions with the EU. Most importantly, he started a trade war with China, which has had the predicted effects on bilateral tensions.

However, things could get worse over another four years. Trump has not pulled out of NATO or withdrawn from other security alliances, but could carry through on his threats in a second term. As discussed in the nuclear section below, a renewed arms race is possible.

The counterargument I considered in 2016 was that Trump: “is a dealmaker, and when he is President there will be lower international tensions with other countries, especially Russia and China.” However I did and do not find this counter-argument persuasive. International tensions have not decreased, especially with Russia and China. He has broken more deals than he has made.


Authoritarians in democratic countries have been emboldened by Trump.

Over the last four years we have seen the continued rise of the authoritarian far-right across Europe. In the 2017 French Presidential elections, Le Pen received 34% of the vote. In the 2017 German federal election the AfD became the third-largest party. In Spain’s November 2019 election, Vox received 15% of the vote. However it can be seen most clearly in the rise of Salvini in Italy—following the 2018 general election, in which he led the largest party, he spent a year as Deputy Prime Minister.

This pattern is also borne out on the EU’s borders. In Turkey, Erdogan continued oppression of the opposition has worsened. In Israel, Netanyahu clings to power despite three inconclusive elections (2019-2020) and an ongoing corruption case. However, current protests in Belarus are encouraging.

Modi was reelected in the 2019 election in the world’s largest democracy, India. He has proven willing to break down institutional safeguards and disregard the rights of minorities and the opposition.

How much this can be attributed directly to Trump is unclear. The general rise of right-wing populism has many causes from social media to income inequality. However, Trump had a clear ‘demonstration effect’ - that his election was possible. Unlike previous Presidents who might have at least criticised these authoritarians, Trump has been silent or encouraging. Moreover, people associated with the Trump campaign—most notably Steve Bannon—have collaborated closely with European authoritarians.

In 2016, I noted that “authoritarians in authoritarian countries – especially China and Russia – are reportedly celebrating his win.” Over the last four years authoritarians have become more repressive, most notably in Hong-Kong and Xinjiang. Overall, Freedom House found that 2019 was the 14th consecutive year of decline in global freedom. In both democratic and authoritarian countries, things could get worse over another four years.

On the situation in the US itself, I said:

There is some small risk that Trump takes the US down an authoritarian, anti-democratic route. He seems to have little regard for constitutional restraints. In office, signs of him actually trying to subvert democracy could be attempting to extend his term limits, or undermine free and fair elections. However I doubt whether Trump would even want to set some sort of dictatorship. If he were to try, he would likely face strong opposition from Republicans, the states and the courts.

I continue to believe that a military dictatorship, coup or Trump outlasting his term limits is unlikely. There have been a number of books published on this subject over recent years: How Democracy Ends, How Democracies Die, The People vs. Democracy and On Tyranny to name but a few.

However, I understated how antidemocratic the system could be without it becoming a dictatorship. In particular, I underestimated how committed to minority rule and voter suppression many Republican politicians at all levels are.

This has become increasingly clear following Justice Ginsburg’s death. Trump recently stated that “I think it’s very important that we have nine justices because I think [the election] will end up in the Supreme Court”. This raises the real possibility of a minority-rule Senate appointing a minority-rule Supreme Court that will hand the election to someone who again lost the popular vote—and perhaps the electoral college.

Specific risks

Climate Change

Trump has increased global catastrophic risk from climate change. He followed through on his campaign promises on climate—withdrawing from the Paris Agreement and replacing the Clean Power Plan with the much weaker “Affordable Clean Energy (ACE) rule”.

The US’ withdrawal has slowed momentum, but not collapsed the Agreement. Indeed we may be able to make up for this temporary break. Optimism on this subject can be drawn from continued technological improvements such as the drastic reductions in solar and battery prices, the vocal support for climate action across the West, and political commitments, especially by the EU. Nevertheless, momentum has dramatically slowed at a crucial time.

However, things could get worse over another four years. The Agreement is likely to continue, with the first ‘global stocktake’ of collective progress on emissions due in 2023. But sustaining progress without the world’s leading power will be hard. Moreover, four more years of domestic inaction on climate from the biggest per-capita emitter would limit global efforts.

Nuclear war

Trump has increased global catastrophic risk from nuclear weapons. This can be split into the risk of use, failure to reduce stockpiles, and proliferation.

The sole good news is that there has not been any use of nuclear weapons. We have not faced a “Cuban Missile-style crisis with Russia”, and there has not been any “limited” (or tactical/​battlefield) use of nuclear weapons. We should not be complacent as the history of nuclear weapons is littered with close calls and crises.

However, I would argue we are in the worst nuclear situation since at least the end of the Cold War. Trump has withdrawn from almost all the nuclear treaties: the Iran Deal, the INT (Intermediate-Range Nuclear Forces Treaty), and the Open Skies Treaty. Trump officials have discussed resuming nuclear testing, which would break the Comprehensive Test Ban Treaty (signed but not ratified by the US). The New Start Treaty is the only significant arms control treaty left, and that expires automatically in 2021 if it is not extended. Trump has suggested he will not extend it.

The second risk is that “Trump could fail to continue the path of previous Presidents in reducing stockpiles”. Trump has become the first President since Nixon not to cut the number of nuclear warheads the US possesses. This is itself a catastrophic failure. But it could get worse; there is a real possibility of a new arms race with Russia and an increase in the size of the arsenal. Functionally the ‘modernisation’ of the nuclear arsenal has made those weapons more destructive. Furthermore, new technology such as the use of machine learning in early warning systems is potentially destabilising nuclear deterrence.

The third risk is nuclear proliferation. Despite very visible talks, North Korea has not slowed its nuclear program at all. Trump’s withdrawal from the Iran Deal has led to the resumption of uranium enrichment. Given a further four years, both countries could progress much further—which could spark proliferation across their respective regions.

We should not expect this to go ‘back to normal’ if Trump leaves. In a number of conversations I have had with nuclear arms control experts, they have said there is no going back to normal. Trump has shown that progress can be undone and that the US can break its commitments. In the view of some I spoke to, trust is now permanently broken, and countries will adopt a policy of ‘hedging’ from now on.


Trump has increased global catastrophic risk from pandemics.

The update on this particular risk will be unfortunately familiar for many. I will quote the 2016 post in detail:

“First there is likely to be less international cooperation. Monitoring and preparing for pandemics relies on extensive international coordination and trust. Trump seems less willing than other Presidents to participate meaningfully in cooperative systems like the World Health Organisation. Additionally, the US is likely to give less international aid. This would mean less help to build up developing world health systems and developing world disease monitoring systems.

Second Trump is seemingly inconsistent, volatile, and does not respect scientific conclusions. He reacted poorly to the Ebola outbreak – exaggerating fears and proposing populist solutions. He seems to not respect science, as is also shown by his climate change position. He might, for example, react to reports of an emerging disease in ways that raised the risk of global catastrophe, for example by mandating that most vaccines produced for the disease be kept in the US, rather than used to prevent the early spread.”

Much of this came to pass: Trump has contributed to less international cooperation on pandemics, and he has reacted poorly to the COVID-19 pandemic.

As part of a general withdrawal from developing world health systems and developing world disease monitoring systems, Trump withdrew CDC officials from a Wuhan lab, which could have provided a crucial early warning. He ignored the pandemic plan prepared by the Obama Administration, and did not write one of his own. Trump did not prevent the early spread of COVID-19 within the US, holding off on lockdown thus allowing it to spread unchecked for those crucial weeks and months. Trump withdrew from the World Health Organisation, and has refused to join the international effort to develop and distribute vaccines.

The risk of man-made pandemics has not been made significantly worse by the Trump administration. While I was worried that Trump was “less hesitant to use or develop biological weapons”, the law and norms around the use or development of biological weapons in the US seem to remain strong. However, the situation remains completely inadequate: the Biological Weapons Convention has a staff of four and a smaller budget than an average McDonalds restaurant. The norms against chemical weapons have been weakened, with the Russian Novichok attacks in the UK in 2018 and on Russia’s leading opposition politician in 2020.

However, things could get worse over another four years. Though it might seem provocative, and I do not want to minimise the real hurt and loss COVID-19 has caused, as a disease it is far from the worst pandemic we could face. Future natural pandemics could be worse on several dimensions: even more transmissible, even deadlier, even less amenable to medical countermeasures and vaccine development, and with an even longer incubation period. And that is just natural pandemics—deliberately engineered pandemics could be even worse. This will not be the last pandemic we face.

Emerging technologies

Trump has increased global catastrophic risk from emerging technology.

I argued that:

“responsible innovation is likely to require extensive cooperation and goodwill between researchers, companies and countries. If cooperation decreases and tension (and the risk of arms races) increases then we will manage the emergence of these technologies worse”

I would suggest Trump has particularly failed to manage the increasing capabilities of artificial intelligence (AI) properly. There are no significant international arms control or proliferation agreements for AI in defence contexts, either offensive cyber operations or lethal autonomous weapons (LAWS). The US has nothing like the EU’s proposal for mandatory safety checks by independent testing centres for AI systems used in high-risk application areas like transport and health. Alongside the general increase in tensions with China, there has been specific focus on technology. This poorly named ‘tech cold war’ has made cooperation with Chinese researchers and companies harder.

The proliferation and democratisation of new biotechnologies has continued towards a situation where small groups are able to create engineered pandemics. The Obama-era moratorium on ‘gain-of-function’ research was lifted under Trump.

However, things could get worse over another four years. As capabilities increase in both AI and biotechnology, we need cooperation between researchers, companies and governments—both in the US (which leads on R&D) and internationally. Trump is likely to continue not to provide that leadership.


In 2016, I wrote:

“I think it is more likely than not that he will last four years.”

“I think it is more likely than not that he will be the Republican nominee in 2020.”

There were four ways out before the 2020 election: death, impeachment, his not standing again, or the nomination of another candidate. Trump has not died, he stood again, and is the Republican nominee. He was impeached, but Republican Senators acquitted him. Those ways out did not happen.

Trump has increased global catastrophic risk over the last four years, and could increase it even further if he was allowed a second term.

Haydn Belfield is a Research Associate at the University of Cambridge’s Centre for the Study of Existential Risk. He is writing in a personal capacity, and his views do not necessarily reflect his employer’s.