DANGEROUS CURES, THE ARITHMETIC

Politics, Voting Politics

Dangerous Cures: A Final Comment 

One final thought on this issue:  The arithmetic.

To recall:

  • In 42 years, there were only 28 convictions for “Impersonation Fraud” at the polls
    (less than 1 per year).
  • In 42 years, there were only 289 convictions for “Fraudulent use of Absentee Ballots”
    (less than 7 per year).
  • In 42 years, there were only 71 cases of “Alien Ineligible Voting”
    (less than 2 per year).

That comes to a total of 10 invalid/illegal votes per year.

We vote in Federal elections every second year. 

So, the determined rate of error (invalid voting) is around 20 per federal election, across the entire country. 

One argument that has been made is that we are missing a lot of bad actors and, therefore, underestimating the actual number of invalid votes. 

How does the arithmetic work?

If the convictions represent only a fraction of the total number of invalid/illegal votes, what is the magnitude of that error? 

If we are off by an order of magnitude (a power of 10), the error rate would be 200 votes per federal election, per year, across the entire country.

If we are off by TWO orders of magnitude, the error rate would 2,000 votes per year.

If we are off by THREE orders of magnitude, the error rate would be 20,000 votes per year.

If we are off by FOUR orders of magnitude (we undercount by 10,000 times), the error rate would be 200,000 votes per federal election, per year.

 

Being off in any measurement by 4 orders of magnitude is hard to even contemplate; but, if we are off by 4 orders of magnitude, what does that mean?

 

  • We know that in every federal election there are, on average, more than one vote submitted.
    At least one vote is cast each election for a member of the house,
    In every other election a vote is cast for the President,
    and once every three elections, a vote is cast for a senator. 
    So, for every 12 election (24 years) we cast 12 votes for a member of the House, 3 votes for President and 2 votes for a member of the Senate.  

    Every 12 elections we cast 22 Federal votes, which comes out to 1.83 votes cast per federal election.

  • Since each individual casts 1.83 votes per federal election, the 200,000 invalid votes would represent only about 110,000 individuals.
  • We often make the mistake of thinking that ALL the invalid votes must have been made for the candidate that we oppose. But this is clearly not the case.  Invalid voters are always split between the two candidates.  And even if we assume that split to be 60/40 in favor of the candidate that we oppose (a far greater split than is ever seen in Federal elections), this would represent a total of 66,000 invalid individuals casting votes for the candidate you oppose, each election, versus 44,000 voting for your chosen candidate. 
  • This results in a total of around 22,000 additional votes being cast for the candidate you oppose. Just 22,000 per election, across the entire country.
  • AND this assumes that the number of invalid votes uncovered is off by 4 ORDERS OF MAGNITUDE.
    That means that for every uncovered incidence of invalid voting, there are 10,000 incidents that are undiscovered.
    This would mean that we are only finding 0.01% of all fraudulent votes.
  • Finally, let us assume the above is actually true – that we are only finding 1 out of every 10,000 fraudulently cast votes – and that if we had found all of the guilty partes, that it would have resulted in an additional 22,000 votes for the candidate that is “rigging” the election.
    We also know that this has been a topic of wide interest over the past decade.  And we also know that there are avid researchers on both sides of the political spectrum that have invested time, money, and personnel in attempts to demonstrate how many invalid votes are cast. 
    There have been economic studies, vote recounts, hand counts, ballot investigations, etc.  And yet, after all of this work, there have been NO studies that show that the detection rates were even 1 or 2 orders of magnitude greater than what has been found, prosecuted, and convicted.  

    And if voting fraud was actually sufficient to affect any outcomes, it would have to be at least greater than 4 orders of magnitude than what we have seen. 
    GREATER THAN 10,000 INVALID VOTES FOR EACH AND EVERY ONE DETECTED.        

 The bottom line here is that the arithmetic does not support any conclusion that there is a problem with invalid or illegal voting.