Because most philosophies that frown on reproduction don't survive.

Sunday, November 22, 2020

What To Make of the Dominion Hacking Claims

 I said I'd write one more post on the various claims about how the election was "stolen" from Donald Trump and address the various claims that center around Dominion Voting Systems, a manufacturer of voting hardware and software which the Trump legal team, particularly lawyer Sidney Powell, have singled out as a suspicious actor in the election.

In the end, it's a bit hard to address, because the claims are many and sometimes vague and thus it's hard to refute them.  In their press conference, Giuliani pitched Powell's claims as showing that many of our ballots are counted outside the country, indeed in Venezuela.  They never provide any evidence to back this up, and indeed I think maybe Giuliani mis-spoke, since Powell never claimed that ballots were transported outside the country for counting.  (In point of fact, ballots are almost invariably counted at the election board of each county.  With just over three thousand counties in the US, this means that the counting operation is incredibly distributed, though obviously a handful of really large counties have a particularly big effect on the outcome of each state.

What Powell did spend a lot of time in her part of the press conference is trying to tie Dominion Voting Systems to another company called Smartmatic.  Smartmatic was founded in Florida, but some of its founders are from Venezuala, and it provided voting systems to Venezuala in 2004 for the referendum on whether or not to recall Hugo Chavez.  How fair that election (which kept Chavez in power) was remains in question.  

However, Smartmatic's only US customer at this point is Los Angeles, which is not one of the ones which Donald Trump is alleged to have come close to winning. The connection to Dominion is that back in the early 2000s Smartmatic briefly owned another voting technology company called Sequoia.  Sequoia was sold to a group of investors in 2007.  In 2010 Dominion (which was founded in Canada back in 2002) bought some of Sequoia's assets before Sequoia finally went bankrupt in 2014.  (You can read this piece from The Dispatch for more detail.)  So the claim is that somehow Venezuelan hacking technology designed for Hugo Chavez was introduced into Sequoia products by Smartmatics and that Dominion then moved that hacked technology into their products, and that now Venezuela and the Democrats are in cahoots to use Dominion to hack the election.

How all this is happening is not very clear from Powell's explanation.  She says in her press conference that it's so easy anyone can hack one of these systems and set it to count the votes according to some ratio. She suggests that they may have been programmed to count every Biden vote as 1.25 while counting every Trump vote as only .75.  However, she says that we might never have known about all this except that the support for Trump was so overwhelming that the vote counting had to be stopped in key states in the middle of the night and the cheating adjusted.  From her comments, it sounds like she believes this both involved introducing fake ballots and also tweaking these ratios by which the vote counters were allegedly mis-counting votes.

Part of the problem is that there are a lot of allegations, but few of them are very specific, and so it's hard to disprove them.  Yes, some minor details are clearly wrong.  For instance, Powell seems to claim that Smartmatic, Sequoia, and Dominion are in fact all one company but they keep changing their name.  This is factually untrue.  But the history of these companies isn't something that people are interested in.  What people are interested in is the vote.

In regards to how the vote was allegedly hacked, Powell provides multiple possible ways and doesn't commit herself to how many votes in which specific jurisdictions she thinks are fake, so without hard claims it's hard to try to refute anything.

One element that did occur to me, however, is that it might be interested to look at examples of counties that were counted by Dominion machines versus other machines.  It so happens that here in Ohio we have just such a situation.  Dominion is one of the approved providers of voting machines, and some counties use them, but other counties don't.  

Powell alleges that Dominion system were probably set on a national basis to overcount Biden votes and undercount Trump votes.  So it stand to reason that if we compared a county using Dominion systems with similar nearby counties that didn't, we'd see a "Biden shift" in the Dominion counties.

First off, in the northwest quarter of the state, let's compare Hancock and Hardin counties which used Dominion technology with Henry, Putnam, and Allen, which used two different types of systems from Election Systems & Software of Omaha, Nebraska.  (If Nebraska isn't enough of a red state bona fide, note that ES&S's address is on John Galt Blvd.)  All five counties are solidly GOP counties.  Hardin voted 76% for Trump in 2020 and 75% for Trump in 2016.  Hancock voted 69% for Trump in 2020 and 72% for Trump in 2016.  Both of those used Dominion machines.  Now to look at neighboring counties using ES&S machines, Henry went 72% for Trump in 2020 and 71% for Trump in 2016.  Putnam went 84% for Trump in both years.  Allen was 70% for Trump both times.

Now let's look in the southeast part of the state where a different set of Dominion voting hardware is used.  We'll look at Harrison and Guemsey which use Dominion and compare them with Tuscarawas which uses ES&S and Carroll which uses technology from Clear Ballot Group, Inc of Boston, MA.  Results?

Harrison (Dominion): 77% Trump 2020, 75% Trump 2016
Guernsey (Dominion): 74% Trump 2020, 72% Trump 2016
Tuscarawas (ES&S): 70% Trump 2020, 69% Trump 2016
Carroll (Clear Ballot): Carroll 77% Trump 2020, 74% Trump 2016

It does not look to me like there is any discernable difference in how Trump did in the counties using Dominion vote counting technology versus those using other companies.  This would seem to be at least some evidence that Dominion machines are not nationally pre-programmed to give Biden an extra advantage, nor hacked to deliver an advantage partway through.  Of course, this doesn't show that such a thing wasn't done elsewhere.  But at least it would seem like some kind of evidence against the claim that this was done nationally.


Paul Zummo said...

I would think the clearest refutation of this conspiracy theory is that Georgia did a manual recount, and not very many votes shifted.

Darwin said...

Good point.

Oddly, most of the people who believe that the voting machines were hacked think the recount also means nothing, though I'm not sure I'm completely clear why. One explanation I heard was because the recount wouldn't look at signatures, though that doesn't seem like it has anything to do with machine counting.

The Dark Canuck said...

I haven't seen Powell's claims directly, but I think either you or she is misrepresenting something. If it's counting Biden votes as being worth 1.25 and Trump votes as being worth 0.75, there would be significant issues with the vote total as well, unless exactly as many people voted for Biden as voted for Trump. Assume 4 Biden voters and 40 Trump voters, for 44 voters total. After counting the votes this way, Biden has 5 votes and Trump has 30 for a total of 35 votes, with 9 votes (~20%) having disappeared. Assume 40 Biden voters and 4 Trump voters. This gives 50 votes for Biden and 3 for Trump, for 53 votes total and 9 more votes than voting population being recorded. Sure, there will be some number of invalid votes that can be used to hide a bit of this discrepancy, but then there should be many more or less invalid votes than normal (depending on how the "real" vote skewed).

Presumably she meant every 4th Trump vote was switched to a Biden vote, which would keep the total population accurate.

The Dark Canuck said...

(To be clear, I think your analysis shows that whatever claim she's making is false. If she's claiming a fixed ratio of vote weighing, it's also falsifiable by looking at the voting population vs total number of votes, and a crazy way to try and fix a vote. If she's claiming every fourth vote, it's still crazy but harder to conclusively disprove)