As the 2020 presidential elections get underway, a new risk is emerging, of “overlearning” the lesson of three years ago and being too generous regarding Trump’s chances of reelection.

The shockwave brought on by Donald Trump’s bewildering victory in 2016 is still being felt today. Analysts, commentators and pollsters, both in Canada and in the U.S., were so dramatically mistaken that it is easy to understand their fear of repeating the same mistake as the 2020 election kicks off.

Yet, we run the opposite risk too: of “overlearning” our lesson from three years ago and therefore overestimating Trump’s chances of being reelected.

The reality is, with a little over a year to go before voting begins, the incumbent president’s position is far from strong.

First and foremost, the majority of Americans are dissatisfied with his record. As we have talked about previously, presidential campaigns involving an incumbent president generally become an appraisal of their term. In this light, Trump is a poor figure. If it’s true that his approval rating has remained at a remarkably stable level of 40-45% since the beginning of his term, his disapproval rating, on the other hand, has changed. It’s gone up. At the beginning of his term, it was around 45%, already higher than normal for the “honeymoon” period. Now, it is nudging toward 55%.

This is important: a portion of voters seemed, at least initially, prepared to give him a chance. Almost three years later, the verdict is in, and it isn’t particularly positive. In the latest poll from NBC and the Wall Street Journal, the percentage of voters who strongly disapprove of the president’s record is higher than the number of voters who strongly or somewhat approve of his record.

In other words, not only is the opinion of many Americans negative, but it also seems to be more and more embedded. In this context, the best weapon in Trump’s armory would be for the Democrats to choose a candidate who voters find totally unacceptable. Of course, this is perfectly possible for a party that was happy to roll out the red carpet for a woman being investigated by the FBI.

At this point, Trump finds himself in the same position as many of his predecessors at this stage of their presidencies — candidates who were ultimately reelected. Ronald Reagan and Barack Obama actually had similar approval ratings to Trump’s at the end of the summer before their reelection year.

Yet, here’s the thing: these rates were not the same once voting day arrived a year later — they had increased. In fact, it’s a fairly good bet that, particularly in Obama’s case, he would have been defeated if voting had taken place a year earlier. Polls carried out at the time supported this theory. In the summer of 2011, they gave a “generic” Republican candidate an eight-point lead on Obama. In the wake of the crisis concerning raising the debt ceiling, several sample polls showed that Obama was clearly vulnerable in the face of hypothetical rivals, even the most marginal ones.

By 2012, the political situation had altered enough for Obama to be back on track, just like Reagan in 1984. Does the same fate await Trump in 2020? Maybe. That said, public opinion towards him seems more fixed than it was towards his predecessors.

The shock of 2016 taught us to exercise a higher level of caution. That makes sense. Trump, like any candidate, will not be reelected or defeated until the votes have been counted.

That should not stop us from stating that although he won a first term against all expectations, the president should currently be considered as (marginally) underestimated to win a second.