OK, so I don't actually love the Electoral College -- I think nobody does -- but at least this time we were spared the experience of a Strangelove-esque near-meltdown with a one-state margin in the Electoral College, and severe electoral dysfunction in key states. But as usual this time every four years, I'm seeing a lot of stuff about how bad the electoral college is, and why it should be changed. I'm personally sympathetic to some -- I wish more than 54% of my California county turned out to vote on important ballot measures on taxes even if the presidential election were not in doubt -- but disagree with others. Perhaps my biggest disagreement is the position of folks of the "I don't trust election officials" stripe.
Here's my summary of that position: State and local election officials want to use their powers to swing an election to their own political party. In swing states where a few thousand votes represent that margin of victory for the state's electoral college votes, it is quite feasible to "steal" the election by a combination of administrative actions like voter-roll purges, precinct-pollbook errors, uncounted provisional ballots, and selective rejection of absentee ballots -- to say nothing of voter ID laws and intimidation by poll workers.
As a result -- this theory goes -- the actions of a handuful of election officials can deliver a key state and decide the presidential election. And hence we should ditch the electoral college in favor of National Popular Vote. I disagree on two grounds. First of all, I don't actually believe that election officials often consciously abuse their powers for direct political gain, despite the fact that legislators certainly do, and even (at least this year in PA) avowedly so during voter ID laws passed with the belief of a selective dis-enfranchising effect.
But leaving aside beliefs about what goes on in the minds of election officials, I also don't believe that National Popular Vote (NPV) would be an improvement over the EC, in terms of foiling hypothetically nefarious election officials (HNEOs). Here's why. With EC, an HNEO can only "deliver" one state; that state's EC votes are the limit of their scope for how much of the election they can effect. Yes, HNEOs in swing states have an advantage as compared to non-swing states. But the EC acts a firewall that limits the effects of election dysfunction (accidental or intentional) within one state.
By contrast with NPV, HNEOs can adopt a different strategy. Instead of suppressing voters (reducing the state's effect on a national election), they could inflate votes. If you really believe in HNEOs, then their powers also include mis-reporting vote totals, for example, with tabulation errors in double-counting lopsided precincts (which we seen happen accidentally in real life). Just as an HNEO in the EC system can suppress thousands of votes to deliver a single state in close state election, HNEOs in the NPV system can inflate tens of thousands of votes to deliver a close national election. HNEOs in large states have the advantage there.
My point is that every system can be gamed, and switching from one system to another doesn't change the fundamental game-ability. I'm not in charge of decisions about EC Vs. NPV, thankfully, but I and my colleagues in both election technology and election integrity do have the ability to do something that's helpful in either system, or any -- promote election operation transparency at a detailed level so that all can see what EOs are doing, and decide on the basis of facts-and-figures whether there is any nefariousness in the management of voter rolls or the disposition of provisional or absentee ballots. That's where the technology we're developing for detailed logging and data mining, can act as that proverbial dis-infectant sunshine to counter the suspicions of the HNEO-worriers.