This sounds like an interesting study. I found a couple hits on the google, can you link which one you're referring to?
McCarty, Poole, and Rosenthal, 2009, Does Gerrymandering Cause Polarization? The money shot... "But in our search to uncover the smoking gun, the case has crumbled. True, the sorting effect has increased over time, as shown in Table 2. But the secular increase in sorting does not appear to be linked to census-triggered redistricting that would reflect gerrymanders."
Or you could say the current GOP is the most dysfunctional party in history, which would be saying a lot considering the Democratic Party in about 1966. And the reason is they shot themselves in the foot with their tea party/FAUX PR campaign. It's kinda like Saudi Wahabis regretting funding the start of ISIS, throw gas on a fire any you might get burned yourself.
I would be curious to see this study done again after 2010 - that was the year the GOP won a lot of state legislatures and a lot of the craziest recent structural changes were made. We saw the effects in 2012 when Dems actually won many more total votes for their candidates but didn't come anywhere close to taking back the House. It is now structurally virtually impossible for Dems to win the House outside of a giant wave election, whereas that was not the case before 2008. A lot of the crazy tea partiers were elected in swing districts (removing moderate Dems in 2010) and are now in much safer districts, from my understanding.