Human beings are bad at scale. Our civilization progressed from small groups that hunted and gathered together into towns and then cities and then, suddenly, into an interconnected global behemoth in which random people can send you a tweet complaining about something you wrote. It's magical.
But this is probably why it can be so hard for us to differentiate between "what a few guys I know think" and "what everyone thinks." It's why we're overly skeptical of polls (not that we always shouldn't be) and quick to dismiss things that contradict the evidence that we ourselves have gathered. Anecdotes are powerful and easily transmitted. A lot of supporters of basically every campaign find it hard to believe that supporters of opposing campaigns actually exist without being part of some inscrutable and devious plot. I can assert this because they send me tweets complaining about things I write.
As the primary process has gotten more complex and more heated, there's been a resurgence of the idea that partisans on one side or the other are putting their fingers on the scales to throw the election one way or the other. The term "strategic voting" is thrown around, almost as much as the term "push polling," intimating the existence of clever plots that -- in the most common iteration -- are handing Donald Trump victory after victory so that Hillary Clinton has a clear path against him in November.
We can start by revisiting the statement with which we started: Human beings are bad at scale. Anyone who has ever spent 30 minutes with a campaign that's trying to make sure its own voters go to the polls will know that it's goofy to think that a secret grass roots movement could exist to derail a candidate by voting a particular way. Turnout is hard. Campaigns spend gobs of money and hire tons of staff to try and make sure that people who 1) they know support them and 2) want to vote actually go out and do it. It takes more than emailing someone and saying, "Hey, you should go vote for Donald Trump."
You can see that in practice this week. The "Stop Trump" effort is an explicit attempt to mess up Trump, and Trump won three of four states on Tuesday. Marco Rubio is begging Trump opponents to strategically support his candidacy in Florida -- and he's still trailing badly. Getting 10 Democrats to go vote for Trump is simple. Getting so many to do so that he is actually guaranteed to win? That's a whole different order of magnitude. Scale.
But we can do better than simply raising an eyebrow. We can look at numbers.
In states for which we have Republican exit poll data this year (largely via CNN) and exit poll data from 2012, we can compare the percentage of the electorate in those primaries and caucuses which was Democratic. As you might expect if you weren't expecting a conspiracy, some states have seen more Democrats turning out to vote and some have seen fewer. More have seen fewer, in fact. In Michigan, one of the states where "strategic voting" has been alleged, nine percent of the 2012 electorate was Democratic. This year, seven percent was.
If you look at both Democrats and independents, same deal. In some states, more independents are turning out than in 2012. In some states, they aren't.
(There's an ancillary theory that Trump is only winning because of independents. He isn't; in the 15 contests where we have data, Trump does as well or better with Republicans in 12.)
Even if there were a plot to turn out Democrats to hand victories to Trump, the percentage of the vote that's Democratic is almost always smaller than the margin of victory in each state. The exceptions are marked with stars.
And that's assuming that literally every single Democrat who turns out to vote is turning out to strategically vote for Trump.
Put another way: Trump's average margin of victory is 12 points. The average density of the Democratic vote is 4.1 percent. Even if he were getting every Democratic vote, he usually wouldn't need them.
The most obvious reason to think that Trump's not winning because of a Democratic strategy would be that the signs of it would be pretty obvious. It would be obvious in poll data, it would be obvious in the results. It isn't.
But, yeah, I know. You know a guy who knows a guy. Maybe you are that guy. Well, good for you. But in 2016, it takes far, far more than a village to win an election.