So, how bad is 2018 going to be for Republicans?

The indicators are in line for a wipeout. President Trump’s job approval hovers in the 30s, despite a strong economy and a soaring stock market, and despite several weeks of solid policy and the strong probability of a tax cut. Democratic turnout in post-2016 election has been extraordinarily high:

This doesn’t count the massive Democratic wave in Virginia that handed the governor’s mansion to a Democrat and moved 15 seats from Republicans to Democrats in the House of Delegates.

It’s an off-year election; in off-year elections since World War II, the party in power loses 25 seats on average. And in 2006, 2010, and 2014, the swings have been massive. At this moment, the generic ballot shows Democrats with a 10.3% lead in the RealClearPolitics average; at this time in 2009, Republicans led the generic ballot by 3-4%; they ended up blowing that out to a 6.8% advantage in the 2010 elections, where Republicans flipped the House. In the 2006 cycle, Democrats led by 9-10 points with a year to go; they ended up flipping the House. None of this bodes well.

But two factors militate against Democratic victory here. The first is that the Senate and House map is terrible for Democrats this cycle. Ten Democratic Senate seats are held in states where Trump won by huge numbers. They’d have to hold all of their seats, plus pick up Dean Heller’s seat in Nevada, plus beat Kelli Ward in Arizona, or knock off Ted Cruz in Texas. In the House, the story is similar: the districts have polarized, moving either red or blue. 23 House Republicans are in Clinton districts, but Democrats must defend 12 seats in Republican districts. Heavy Democratic turnout would be the only way to pick up the House.

All of this strongly suggests that Republicans must work to get Trump’s approval rating up — and that they must also create safety for Republican candidates in purple states if Trump’s toxicity drives down turnout. Republicans cannot fall into the trap of thinking that Trump’s defeat of Hillary Clinton means that political gravity no longer applies, or that Trump’s patented strategies are transferrable to other Republicans (Roy Moore learned that the hard way). Republicans might get luck, and Democrats might be stupid enough to run hard-left candidates in purple states — Democrats somehow almost lost to a credibly accused child molester in Alabama by running a pro-abortion extremist — but that’s not a strategy. Republicans had better put some points on the board, and President Trump had better do some hard work to win over more voters, or 2018 will be pretty damn ugly for Republicans.