While he’s writing about Ohio, I think this by John Scalzi sums up the democracy problem in the U.S. (boldface mine):
On the topic of “Liberals/Democrats strategically voting in the Republican primary,” it be at least some of the reason that the Democratic senatorial primary vote was only 48% of the GOP senate primary vote, with a very similar percentage for the gubernatorial primaries. It was also because there was far less drama involved; everyone expected Tim Ryan to win it, which he did, handily, with nearly 70% of the primary vote — Ryan in fact received more primary votes (nearly 356k) than Vance, the Republican primary winner (nearly 341k).
Is there something to be readily gleaned from these numbers, when it comes to the senatorial race in November? Maybe but possibly not. Ohio has more registered Democrats (947k) than Republicans (836k) and both of these numbers are easily swallowed by the number of unaffiliated voters (6.2m), and across the state only 18.8% of eligible voters turned out. Which means that our senatorial candidates were decided by roughly 4.4% of our electorate in both cases. So that’s great, and also leaves lots of room for things to happen in the general, in which possibly 36% of our electorate will vote.
Sad, but very true, though it does mean if Democrats rally the troops–and Republicans don’t–victories, even in reddish-leaning areas, are quite possible.