Another special election, another near upset for Democrats. Running against a slew of Republican competitors to replace Health and Human Services Secretary Tom Price in Georgia's 6th congressional district, Democrat Jon Ossoff fell just short of the majority needed to avoid a runoff. He will now face the second-place finisher, the infamous Karen Handel, on June 20.

It's a good sign for Democrats — and a good test case for the areas of the country where Hillary Clinton did unusually well. But they should be wary before using Ossoff's campaign strategy as a template for the rest of the country. To win in districts President Trump dominated in 2016, Democrats would be well advised to adopt better class politics.

The 6th district is a wealthy suburb of Atlanta — Newt Gingrich's old seat, though the district boundaries have been adjusted since then — and has been reliably electing Republicans for decades. Tom Price won the 2016 election by 23 points (though his opponent did not campaign and maybe did not exist at all).

The swing from the previous election is similar in magnitude to that of Kansas' 4th district, which Democrat James Thompson narrowly lost earlier this month. However, there is one key difference: Where Ossoff only did slightly better than Hillary Clinton did, Thompson bested her margin by 20 points, despite running on a pro-choice platform in about the most fervently pro-life city in the entire country. You can chalk that up to Thompson's class politics, which were miles better than Clinton's and which allowed him to credibly criticize the wretched Republican economic performance in the state.

Clinton's campaign was by and for upper-class white-collar professionals, and Ossoff (perhaps wisely) followed her lead. His extremely bland campaign was about responsible governance, an end to corruption, and bipartisanship. In a wealthy, largely white Georgia suburb, that plays pretty well — as it probably will in similar places, like Texas' 32nd and 7th districts, which Clinton actually won. Run a more competent and energetic version of her basic campaign strategy, and Democrats ought to be able to scoop up those seats without breaking a sweat. (Colin Allred, a former NFL player, has already declared for the 32nd, which Democrats did not even contest in 2016.)

However, as the election of President Trump shows, catering mainly to upper-class professionals does not play well in lower-class communities — especially those that have been harmed by neoliberal deregulation and trade deals supported by Clintonite Democrats. Personally taking gobs of cash from big banks and spending half the campaign fundraising from the ultra-wealthy makes working-class people suspect (probably correctly) that you are not particularly interested in meaningfully improving their lot in life.

For Trump-friendly states, Sanders-style class politics is a far better approach, as his primary performance (where he generally rolled up huge margins in Trump country) and Thompson's campaign shows.

Unfortunately for Clinton, one campaign cannot cater to both the working class and the upper class. Clinton chose the latter and lost. But individual candidates can do that. A national party should contest every single seat, and that means tailoring candidates to fit each district. Running Clinton-style candidates in places like Orange County makes decent sense — but so does running fervent populists in Trump country (incidentally, this is where most of the Senate Democrats who are up for re-election in 2018 are).

I would bet a substantial sum that extant Democratic Party elites are going to focus overwhelmingly on the rich districts, where people like themselves live. Luckily, there is large and growing alternative machinery to help Sanders Democrats run. Conditions look good for a very strong showing for Democrats in 2018, even if outsiders have to drag the incompetent party leadership kicking and screaming into it.