I am by no means any sort of expert on the financial aspects of electoral politics. On the other hand, I do have a relatively straightforward understanding of basic arithmetic, so I have been doing some sums, based on some relatively well-established facts.

During the 2012 Presidential election, about 129 million Americans cast a vote for President, and, of those 129 million votes,

• About 65.9 million (51.06 percent) were cast for Barack Obama
• About 60.9 million (47.20 percent) were cast for Mitt Romney
• About 2.2 million (1.74 percent) were cast for a wide variety of others

This means that the votes of about 7.2 million Americans actually made the difference in who got elected as president.

You can do the same sums for the previous two elections (in 2008 and 2004) and come up with very similar results.

• In 2008 the votes of about 10 million Americans gave Barack Obama the presidency over John McCain (when 131 million Americans cast votes)
• In 2004 the votes of about 4 million Americans gave George W. Bush the presidency over John Kerry (when about 122 million Americans cast votes)

So, let’s be generous and suggest that it is really (yes, really), at most, only the votes of about 15 million Americans that will make any difference at all to who gets elected President in 2016.

Now, we also know that

• In 2004 the cost of the President election was about \$1.9 billion. That’s \$14.73 per vote cast.
• In 2008 the cost of the Presidential election was about \$2.8 billion. That’s \$21.27 per vote cast.
• In 2012 the cost of the Presidential election was about \$2.6 billion. That’s \$21.31 per vote cast. (Yes, amazingly, the total actually went down a little.)

Current estimates of the cost for the 2016 Presidential election are running somewhere close to \$4.0 billion. The two Koch brothers alone are said to have budgeted just under \$0.9 billion!

So, again, if we were to be conservative, and suggest (a) that the actual cost of the Presidential election in 2016 will be a mere \$3 billion, and (b) that about 130 million Americans will turn out to vote, then in 2016 the cost of the election will be \$23.26 per vote cast.

But hang on a moment … at most only the votes of 15 million Americans are likely to make any difference, so the total cost per vote that will make any difference will actually be only \$23.26 × 15 million = \$348.9 million. That’s a saving of \$2.651 billion, and what is more, most of us whose minds are already made up about who we are going to vote for wouldn’t have to get a single piece of junk mail (let alone junk e-mail).

All we have to do is tell people who aren’t sure who they want to vote for where to register so that they can be badgered by the candidates!

Now … what could we do with \$2.561 billion? Well, clearly I need a bonus for coming up with this simply brilliant idea … and something like \$5 million doesn’t seem overly greedy. (It’s less than 0.2%.) Then we can give the other \$2.546 billion to the National Organization for Rare Disorders to fund research into the 6,000+ rare diseases affecting some 25 to 30 million Americans that have no form of approved treatment whatsoever!

If the real cost of the 2016 Presidential election was to turn out to be nearer to \$4.0 billion, there’d be more like \$3.5 billion to invest in rare disease research — and I’d still settle for a mere \$5 million bonus!

(Yes, I do have a minor conflict of interest. I am a member of the Board of Directors of the National Organization for Rare Disorders!)