That's not too surprising. It's called incumbency. Most incumbents win easily and a majority of incumbents int he house were republican. People have a tendency to think Congress is doing a terrible job but keep sending the same politicians back because "My congressman seems like a nice guy. He's not the problem." or something like that.
If someone could upload a Clinton vs House democrats map, I think you'll notice that many democrat incumbents outperformed her. I know Matt Cartwright, who represents Scranton-Wilkes Barre, Monroe county, and Schuylkill county PA sure did.
You got a point there. But it does not take away from the fact that Trump under-performed the GOP even in places where the Dem congressional candidates should have the advantage of incumbency. A good example would be MD. That is a state where the GOP and Dem contested all CD which makes a good apples-to-apples comparison. In MD Trump lost 60.33%-33.91% but with 8 out of 9 CDs having Dem incumbents the GOP House candidates lost the PV to Dem only 60.43%-35.53% which meant they over-performed Trump by over 1% despite a clear disadvantage of incumbency. Another one would be NJ where Dems hold 7 out of 12 CD. In NJ Trump lost 54.99%-41.00% but the GOP House candidates managed to hold the Dems to 52.60%-44.51% in the PV despite a slight Dem incumbancy advantage. If you look at a lot of CD in CA you see the same thing, Trump under-performing GOP candidates even with a Dem incumbent. Now CA NJ and MD are poor fits for Trump but it is clear that in metropole suburban areas Clinton is capturing part of the GOP vote. But that is the whole point: Large number of GOP voters in Metropole suburban areas voted for Clinton but mostly voted GOP in House races.
In fact out of the races that Dem House candidates won, I found a total of 156 seats where the GOP ran a candidate AND I have access to Trump and Clinton performance (NC does not have number yet) Out of this 156 such seats Clinton outperformed the Dem candidate in terms of victory margin in 63 out of those seats which is quite impressive. Out of the races that the GOP candidate won, I found a total of 202 seats where the Dem ran a candidate AND I have access to Trump and Clinton performance (again NC does not have data yet) Out of this 202 seats Trump outperformed the GOP candidate in terms of victory margin in only 31 out of these seats. I agree that the missing CD data in NC makes this not airtight but there are plenty of evidence of Clinton over-performance the generic House candidate even after taking incumbency into account.
In addition there were 37 House Seats the GOP failed to nominate a candidate (a lot of them were in CA where the GOP did not make it into the second round) and only 30 seats where the Dems failed to nominate a candidate. Had the GOP and Dem been able to nominates in all these seats the GOP PV lead of 0.9% (I include the DC non-voting congressional race to make it apples-to-apples comparison) would be actually larger. I did a back-of-envelope guess of what that would look like (using the PV share in the Prez election) and concluded that in such a scenario it would have been a GOP House PV victory of around 1.2%
