Cogito Ergo Non Serviam
GOP Electoral College Advantage Shrinking?
In the last eight presidential elections, the Democrats have won the popular vote seven times. However, only five times was the next president a Democrat. The flaw is in the winner-take-all approach used in 48 states in filling the seats in the electoral college. Because power is skewed in the US federal system toward rural states, it is relatively easy for a Republican to take the White House while finishing second in the hearts of voters. Mathematically, the Democrats have to win the popular vote by about 3-6% (depending on the model used) to carry the electoral college. Evidence suggests that number is falling and the GOP advantage is fading.
At the heart of the problem is the electoral college. It was a folly by the founders to think that men appointed to vote for the chief executive would do so without a thought given to politics. They wanted the electors to act as philosopher-kings choosing the wisest and most able leader. The fact that the leader would want something contrary to the interests of some electors never seems to have occurred to them.
Indirect elections are problematic, but they are not inherently bad. The issue of agency is always a problem in human activities, but in politics in this century, there can be sufficient transparency that the matter is a small one when it does arise.
What is at issue here is the winner-take-all nature of elector selection in all the states but Nebraska and Maine. A candidate can win a state by a single vote or by 3 million, but in either instance, he or she gets all the electors. This can result in a ridiculous situation where the popular vote is close but the electoral college result is a landslide.
The polling question here is how much does the Democrat have to make up in popular votes in order to secure the 270 electoral votes needed to be elected? This is due to the fact that there is not much state-level polling outside the battleground states. Statisticians have had to resort of the national polls and then engage in some mathematical manipulations to get the "right" answer. In doing so, they must make some assumptions that are, at best, suspect.
For instance, Mr. Trump is destined to lose California. No one disputes that. The question is whether he loses by 100,000 votes or 1,000,000. In the electoral college, it will not matter, but the popular vote result is significantly affected. Mr. Trump could suddenly rally and gain 2,000,000 votes in California, or rather polls at a level suggesting the election day result will be 2,000,000 higher for Mr. Trump than was previously predicted. That will alter the popular vote overall for Mr. Trump while having zero impact on the electoral college.
If he were to pick up just 50,000 in Wisconsin, that would be a blip on the popular vote count compared to the much higher California count. But it would swing the state into the Republican column. So, it is easy to see how relying on national polls and statistical maneuvers can mislead.
This all presumes that the polls are drawing appropriate samples, which they probably are not doing. The Tilt newsletter from the Washington Post reported:
By a two-to-one margin, the polls that you see nowadays are weighted by "recall vote." This is a little wonky, but it means that the number of respondents who say they voted for Mr. Biden and Mr. Trump is adjusted to match the actual result of the last election. Whatever the merits of this approach, it has the consequence of forcing many state polls into very close alignment with the 2020 result. The polls that do not weight by past vote, however, show results that correlate as much with the midterm vote as the last presidential election.
Both cannot be right. It is best to ignore the polls from here. Some people have already voted, and the Get Out the Vote operations are what matter. Ms. Harris is winning based on field offices in operation and paid staff working with volunteers.
© Copyright 2024 by The Kensington Review, Jeff Myhre, PhD, Editor. No part of this publication may be reproduced without written consent. Produced using Ubuntu Linux.
|
|
|
