Both political opponents and reporters who insist Hillary Clinton may have destroyed work-related emails should have to prove their case. (More)

Squirrels aren’t renowned as mathematicians. That’s partly because we have eight fingers and ten toes. Humans have ten of each, so you pretty much agreed to use base-10 numbers, although several cultures used base-20 numbers. But we couldn’t decide whether to count on our fingers and use base-8, or count on our toes and use base-10, or count on both and use base-18, or add our tails and use base-19. On the other hand, or paw, squirrels are very good at hiding things and finding hidden things, so that qualifies me to comment on whether Hillary Clinton has missing emails …

… which is more than I can say for most of the humans who are commenting on that topic. Indeed most of the commentary fits the common conspiracy meme of The Real Evidence Is Hidden: they simply allege that she could (or must) be hiding something, and demand that she prove otherwise. You don’t have to be a lawyer to recognize that Clinton’s critics – and the media – have flipped the burden of proof upside down:

So let’s start with some facts. In her 9-page statement yesterday, Hillary Clinton says she released printed copies of all 30,490 emails that she sent or received as Secretary of State. Or, specifically, from March 18, 2009 – when she switched from the email account she used as a senator to her account – until her last day in office, February 1, 2013.

I’ll save you the math. That’s 30,490 emails divided by 1417 days, for an average of 21.5 work-related emails per day. But is that all of her work-related emails?

There are several evidence-based ways you could explore that question:

  • Find references to ‘missing’ emails – The most direct evidence would be to a reply or other direct reference to an email that isn’t the 30,490 she produced. If someone writes “Per Secretary Clinton’s email to me today…”, and if and her 30,490-email trove includes no email to that person that day, then you have a proven gap to explore. Of course, you still might find she sent an email to that person a day or two earlier, and the person simply didn’t read it until that day.
  • Compare her traffic to other department secretaries – The current presidential cabinet includes 15 department secretaries. You could compare each secretary’s daily email traffic to the traffic of (say) the top 10 officials in that department. For example, if the top 10 officials in the Department of Justice each averaged 50 emails sent or received per day, and Attorney General Eric Holder averaged 25 emails sent or received per day, his daily email traffic would be half the average for top officials in his department. You could do the same calculation for the other 15 department secretaries, then compare their ratios to Clinton’s. You would expect some variance, as some people use email more than others. But if her ratio is more than two standard deviations below the mean for cabinet secretaries, you have at least some evidence that there may be emails missing.
  • Show gaps in her traffic – On Sunday, Rep. Trey Gowdy (R-SC) said “There are gaps of months, and months and months” in then-Sec. Clinton’s emails. That is a testable claim and the media should demand he specify exactly which “months, and months and months” he means. If he won’t, then his claim is no different from Sen. Joe McCarthy’s “I have here in my hand a list of 205, a list of names” bluster. I’ll go on record right now as predicting that Rep. Gowdy will not be able to prove even a single empty month in the Clinton email record. In fact, I doubt he’ll be able to prove a single empty week.
  • Show statistically significant anomalies in her traffic – There will, however, be ups and downs. She’ll have sent and received more emails on some days, and fewer on other days. That’s expected, but statistical analysis might yield some significant anomalies. Statisticians have written extensively about missing data and the Myth of Emptiness. The math is complex, but the gist is that if her email traffic usually correlates closely with some other measurable variable – such as the traffic of the other top 10 State Department officials – statistical models can estimate the likelihood that a large deviation on a given day reflects data that are Missing At Random (MAR) or Missing Not At Random (MNAR). That is, on a given day, does that day’s traffic have fewer-than-expected emails due to some unrelated reason (MAR) – illness, travel, or mere random statistical noise – or because some emails were deleted (MNAR)? But it’s important to note that a high MNAR probability does not prove emails are missing. It may be that more of that day’s emails were classified and not in the public record. And even a 95% MNAR probability means there is a 1-in-20 chance the ‘gap’ is merely statistical noise.

My point is, none of the people commenting on Hillary’s Missing Emails have produced any evidence that there are, in fact, missing emails. No one has said “Here’s a reply to an email that isn’t in her list.” No one has said “Here’s the daily email traffic for other cabinet secretaries, and Hillary Clinton’s is way too low.” No one has said “Her list of 30,490 emails doesn’t include any from this month (or week).” No one has said “Here are five anomalies in her daily email traffic, as compared to the traffic of other senior State Department officials.”

So far, they’re all just insisting there might (or must) be emails missing, and demanding that she prove otherwise. And some are already morphing the missing email charge into equally unproven rumors about the Clinton Foundation, much in the way that Whitewater morphed into Vince Foster and on and on, slinging one charge after another to maintain the perception of ‘corruption’ … even as each charge turned out to be empty.

For those in the media who are “exhausted” by the “Clinton dramas,” here’s a thought: don’t report the charges unless the critics provide solid evidence. And triple-check the evidence they claim to provide. Like, say, a reporter … instead of a gossip columnist.


Good day and good nuts