Back to
The 13 Keys to the White House | American University, Washington, D.C. - Allan Lichtman's theory that Presidential elections are referenda on the incumbent's party. Let's look at this election so far.
The problem with these analyses is that their sample size is inevitably small. There have only been 46 different presidents, and about 60 presidential elections, going back to the 18th century. Any election before maybe 1948 or so (and probably even later) has a country that is way too different for that data to be useful in drawing any conclusions about 2024.
It's 60 since the first one in 1788, and 42 since 1860, both counting this year's election.
A good halfway point in AL's sample is the New Deal era. One could split at 1936, the first Presidential election that was a referendum on FDR's New Deal, giving 1860 - 1932 and 1936 - present. So one could fit to the first half and check it on the second half, and vice versa.
This current kerfluffle in the Democratic Party threatens to falsify key 2, and if President Biden steps down, that will falsify key 3. Thus giving him a strong risk of defeat. So the Dems ought to settle their differences and accept the necessity of supporting him a few months more.
These keys do not consider the risk of incumbency, because we have never had an 81 year old incumbent. The fact that Biden is so old and so clearly shows his age diminishes the value of that key almost entirely I think.
List of presidents of the United States by age - Joe Biden is the oldest, but not by much. He seems old because he sometimes seems very feeble. Donald Trump doesn't seem very feeble, or else he is better at hiding his feebleness. Ronald Reagan didn't seem very feeble either. As to hiding disabilities, FDR was very big on that, hiding his wheelchair and leg braces.
I concede that the keys model could use some improvement, like using continuous instead of binary scoring where feasible, like how well the President's party did in the previous midterm as compared to the midterm before that one.