The Republican Party’s ongoing civil war has dominated media coverage this primary season, but there’s plenty of interesting stuff going on in Democratic primaries. One to watch is the showdown in California’s 17th District—Silicon Valley, basically—between seven-term incumbent Democrat Representative Mike Honda and former Obama Commerce Department official Ro Khanna. For my money, it’s the most interesting Democratic race of the day.
The race is notable for many reasons. Khanna is young (37); Honda is old (73 later this month). Khanna has won the support of a swath of Silicon Valley tech titans; Honda is backed by the traditional liberal interest groups. Khanna has raised boatloads of cash (more than Honda), which is unusual for an intra-party challenger taking on a veteran incumbent. And as I detailed in April 2013, Khanna has used that money to hire much of Barack Obama’s tech team in order to deploy at the congressional level the same sophisticated turnout and persuasion techniques used in the 2012 presidential campaign. As I put it last year, “It’s as if Bill Belichick and the staff of the New England Patriots decided to coach a high school football team.” Will it make a difference? We’ll find out. The fact that this is happening in Silicon Valley, the most-wired place in the country, makes it all the more intriguing to watch.
But what’s especially interesting about the race is that it’s being conducted under the auspices of California’s new “jungle primary” system. In 2010, California voters passed Proposition 14, which established an open primary in which voters can pull the lever for any candidate, from any party, and the top two finishers—regardless of party—will then face off in the November general election. Because CA-17 is heavily Democratic (it voted 72 percent for Obama in 2012) the expectation is that Honda and Khanna will be the top two finishers, although that is not guaranteed. As the San Francisco Chronicle revealed last week, Honda’s supporters in the labor movement have also been donating to a Republican candidate, Vanila Singh, in hopes of boosting her into second place and knocking Khanna out of a runoff.
In addition to establishing who will appear on the ballot in November, today’s results should indicate whether or not this will be a close race with real potential for an upset. Honda’s supporters take pains to point out that there’s no obvious reason why a Democratic district would oust a seven-term incumbent who isn’t plagued by scandal. (Honda is not.) Khanna’s supporters claim that, as a tech attorney, he’s a better fit for the district, which houses the headquarters of Apple, Intel, and Cisco. And they hint that Honda’s advanced age and modest record of accomplishments warrant a change.
One argument simmering among political junkies is how close Khanna needs to finish behind Honda to have a shot of knocking him off. As a general rule, an incumbent who doesn’t finish well ahead in a multi-candidate primary faces serious jeopardy in a two-person runoff because the other candidates’ supporters have already declined to give their vote. For an example, look no further than the neighboring 15th District where, in 2012, a young Democratic upstart, Eric Swalwell, finished the primary within six points (42-36) of the 20-term incumbent, Democratic Representative Pete Stark, and then went on to beat Stark 53-47 in November.
So how close does Khanna need to be tonight? It’s a good question with no clear answer. According to Eric McGhee, a research fellow at the Public Policy Institute of California who has written extensively on the new primary system, “It depends on how many partisans you have in the district, how many candidates are running on each side, and any other wildcards that may be present in the race.”
Most journalists and political consultants seem to think that a challenger needs to finish within 10 points or so of the incumbent to have a realistic shot in November, although no one I spoke with could cite any academic studies to back that up. California’s primaries are especially difficult to handicap because so few races have occurred under the new rules. “This has only existed since 2012,” McGhee points out. A May 23 Survey USA poll provided one of few glimpses of public opinion in CA-17. It found Honda at 40 percent, followed by Khanna at 21 percent, Singh at 8 percent, and a second Republican, Joel VanLandingham, at 6 percent. Some 24 percent of voters were still undecided.
Incumbents polling below 50 percent are usually considered to be in hot water. Still, Survey USA found Khanna a healthy 19 points behind—not exactly breathing down Honda’s neck. Perhaps not coincidentally, Khanna’s people seem to think that if he finishes within 20 points, he’ll be in good shape to win in November. Does McGhee think that’s plausible? “For that seat,” he says, “I could see it. There are enough Republicans and enough independents. The complexion of that district is not as rock-solid Democratic as some others that have had same-party runoffs.”
Part of the rationale here is that Khanna, who’s a notch or two to the right of the liberal Honda, will be the natural choice of Republicans and indies lacking other options—provided they bother to show up in November, which is a further question mark. “Statewide, independents are 10 to 15 percent of the electorate in the primary,” says McGhee. “They’re about 20-some percent of all registrants [in the 17th District].”
But Honda probably shouldn’t bank on a surge of Democratic support in November, either. In 2012, the first year of the jungle primary, McGhee studied same-party races and found a big drop-off in participation. “There clearly was less interest: about a 10 percent drop-off for these same-party races, on average,” he says. Many voters who showed up to cast presidential ballots didn’t bother to vote in down-ballot races. Presumably, they didn’t see any difference between two candidates of the same party—at least not enough to bother choosing one over the other. Without the White House at stake this year, it’s not clear that they’ll show up at all. That’s yet another variable that has to be factored in to any projections. As McGhee notes, “They call it ‘jungle primary’ for a reason. It’s very confusing.”