Recounts Should Not be the Agenda Alone; Its About the Audit

The goal of a voting system is not to convince the winner and their supporters that they won, but rather to convince the loser and their supporters that they lost, fair and square.  With that in mind, we appreciate the effort to seek a recount, but not because we believe a recount is in of itself, deserved or to reverse results. That's done, and the outcome is likely correct. 

No, we appreciate the effort only because is raises the real issue that needs to be addressed:  audits of elections. And while statewide audits are sometimes conducted anyway, whether it happens and how varies from state to state.  Audits should be thought of like referee reviews of all touchdowns scored in American professional (NFL) football.  Its simply to provide the level of assurance citizens deserve from their electoral process. 

Two of the nation's leading election integrity professionals, Ron Rivest and Phil Stark published an Op-Ed in the USA Today last week providing a simple explanation of election audits -- their purpose and process.  It's worth a read.  Our CTO also lays out a simple explanation of the importance of audits (vs. recounts) in a must-read column at the TrustTheVote Blog.

I'm more interested in a digital forensic audit of voting machinery and a machine used in the back-office running an App called an EMS ("Election Management System"), which I consider below first, before I turn to a concern I have with this particular recount initiative.. 

A Real Machine Concern, Not About Ballot Casting Devices

For me, a digital forensic audit including examination of the machinery and the cast ballots (something not currently performed in the ordinary course of statewide audits) would, in this case, put to rest a concern at least I have about the vulnerable EMS. 

My concern is not that a bunch of ballot casting machines were hacked.  We've repeatedly explained that's very unlikely.  Voting machinery is not connected to the Internet, and physically accessing individual ballot machines is simply not worth the effort or the scale required to affect an election.  However, what is possible is that a vulnerable, unsecured, desktop PC in the back office running a vital App that configures the voting machines and poll books used in the polling places could be compromised without a trace.  That machine, running a particular App called an "EMS" (and unlike voting machines is sometimes connected to the Internet for general office activities) is a weak link in voting systems as currently configured (and managed).  For me, that's an issue that should be examined in an audit process.

I posit that because for this election, there was just too much at stake in this election, not just for domestic agendas, but for foreign state interests.  America experienced plenty of poking, prodding, and penetrating of various systems related to election administration this year.  I argue this was enough "casing of the joint" to raise the question whether the (surprisingly simple) next step could've been taken to compromise that "weak link" (the EMS) in effort to advance foreign state agendas.  We don't know, but its worth examining.

Again, we think the outcome is probably correct and don't expect any recount to change that.  But this does illuminate the importance of audits.  To be sure, WI at least is already undertaking its standard statewide machine audit. And that's good.  But, I'm talking about a more in depth examination of the machinery and the ballots in a manner that may not be what this recount effort allows.  For instance, hand examination of the ballots may require judicial approval and time is short.  However, as the election integrity community has observed, risk limiting audits can also achieve the objective. If implemented in a nationwide uniform manner, these audits would likely obviate the need for close examination of ballots and machinery in all but a very few limited exceptions.

In summary, auditing to confirm outcomes is imperative.  It's a matter of increasing confidence in elections and their outcomes, but not because they could or should change those outcomes.  For the TrustTheVote Project, audits can provide more hard evidence to prove the absolute necessity to replace our failing election infrastructure with a fully evidenced based, verifiable, accurate, secure, and transparent update and upgrade.  It's a matter of preserving our democracy (or "Republic," whatever you want to call our great nation), and frankly at some point it may well become a matter of national security.

The Trouble With This Specific Recount Approach

Back to this specific recount initiative, I'm a bit concerned, notwithstanding my hope that actual machinery and ballots will be thoroughly investigated (the motivation explained in this article).

Here's the the main thing.  If any recounting or auditing is to happen in the name of understanding just how accurate or inaccurate final results are, we strongly suggest adding two to three other States where results were equally close for Sec. Clinton.  New Hampshire, Nevada, and Minnesota come to mind immediately.

The issue is not to try to overturn the election; the issue is to understand the extent to which results in a recount were at variance with results as initially certified and/or (for us more importantly)  whether any machinery malfunctioned through accidental, natural, or intentional acts. 

Second, the issue of when each State does a final certification is playing out here too, in the impatience of some in not seeing more "closure" sooner.  It varies, State to State -- and if certification in some of these key States such as MI, WI, PA, etc. -- happens later in the cycle, it ends up creating the appearance of unjustified changes. The lack of patience on part of lots of folks -- and the need to better educate the press, which should have been noting this from the start -- is to an extent, telling.

From Nov 9th onward, there should have been a clear tally of ballots not yet processed in each and every State.  Even today, the vote totals for Trump vs Hillary are about 123 million but we know that 135 million votes were cast nationally.  Yes, Stein, Johnson, et al earned a few million; but even as of yesterday the 23rd November, might there be 2 or as many as 5 million more votes still to count?  That's hard to know for sure, because reporters haven't paid any attention to this particular metric.

Given that Stein's recount campaign impressively succeeded in raising the $4 million necessary to file for recounts, they should not stop there.  They need to step up with additional money to look at six (6) close States, not just the three swing states that appear partisan. We cannot stress this enough: this is an American issue not a political one.  And in absence of examining all close States, its no wonder the media is characterizing this as a Clinton issue.

Those pressing for the recount need to let the chips fall where they eventually may, because the goal should not be about overturning a result you don't like, but learning from 2016 to improve 2018 and beyond.

And ponder this: consider the exact same result, except flipped: where Trump wins the popular vote by 2 million votes (or even 3 million eventually when all the dust settles) but Clinton wins the Electoral College.  We should not kid ourselves about what an outcry would be happening right now in the Trump camp, matched by the resistance that the Clinton camp would raise alleging the GOP as "sore losers," "conspiracy crazies," etc.

I'll add this (in a shameless self-plug for our work at the TrustTheVote Project): I hope those same individual donors who helped raise that record-breaking $4M to mount the MI, WI, and PA recount initiative, will consider what that same fund raising effort would do to finish the election technology we're building right now that will definitively put an end to the lack of confidence in elections and their outcomes.  375,000 contributors nationwide, with a one time (tax-deductible) donation of $25 and ElectOS would be completed. And that means freely and publicly available election technology to deliver lower cost replacement systems by 2020, with better results and audit capabilities for the midterm of 2018.  And this publicly available technology would also revitalize the necessary commercial industry needed to deliver it.  Think about it.  And let me know if you have ideas.  Let's finish something of, by, and for the people to solve for this mess that will only grow worse next time.