Let me background that generalization a bit.  Back before 1913, when the state legislatures chose U.S. senators, the House did shift party control at least eight times without the Senate doing so.  During the Reagan era the Senate shifted control twice without the House doing so:  an opposite pattern.  Also, you need to look carefully to see the pattern I am referring to.  It is true, and well-known, that the Democrats took over the House but not the Senate in 1931 after a bad midterm for the Republicans under Hoover as the Depression economy unraveled.  But the Grim Reaper, not the midterm, engineered that House takeover.  The direct result of the 1930 election was a Senate of 48 Republicans, 47 Democrats and one Farmer-Laborite, and a House of 218 Republicans, 216 Democrats and one Farmer-Laborite.  It was deaths of House victors followed by special elections that cemented control of that chamber for the Democrats once the new Congress met for the first time in December 1931.     

But there it is.  Since 1913, the Senate has been at least as change-prone as the House in the sense specified here, its elections rocked just as much by new impulses of voter opinion.  In a constitutional sense, that is a surprise.  Staggered terms were supposed to foster a certain laggardness in the Senate membership. 

The key to this sameness pattern seems to be the precariousness of individual Senate seats.  When an opinion gust comes along, the 33 or so Senate seats up for renewal need to supply as much party turnover, relative to the full membership of the Senate, as do the always up 435 House seats supply for the full membership of that chamber.  Generally speaking, that seems to happen.  In 2008, for example, the Democrats netted 8 new Senate seats among the 33 seats up.  This was a gain of 24% of the seats up and 8% of the full Senate.  In the 2008 House elections, the Democrats netted 21 seats:  a gain of 5% of the full House.  Various familiar reasons underpin this special precariousness of the individual Senate seats:  better challengers, better funding of challengers, incumbents out of practice after six years, constituencies that aren’t as naturally safe, and so on.  

But this year looks like an exception to the sameness rule.  The House may change party control but the Senate may stay put.  If this happens, there is a good reason.  In play is an extreme dissonance, which will get more extreme, in the political complexion of the group of Senate seats up this year compared to those not up.  Of the 37 seats up, 19 are now held by Democrats and 18 by Republicans.  Most of these senators sailed in with Bush in 2004.  If the Republicans gain something like 8 new seats they will convert this 18-19, for them, balance into a 26-11 edge, an extraordinary disparity.  It will be pushing the envelope.  Of the 63 seats not up, 40 of them are held by Democrats (including Bernard Sanders and Joseph Lieberman).  The Iraq war and the Wall Street crash of 2008, which are vanishing issue assets for the Democrats these days, boosted many of these members of the classes of 2006 and 2008 to the Senate.  Even with a good deal of seat turnover, these background numbers have made it hard for the Republicans to take the Senate this year.   

So what will the new Senate be like?  For one thing, taking into account within-party replacements well as party turnovers, an usually large Republican freshman class of probably 12 to 15 members will enter the Senate and supply some notable zest.  For another, roughly half of the 40 holdover Democrats will be taking a look at new home political contexts that would have given them real headaches if they had run again themselves this year.  Democratic senators like Max BaucusMax Sieben BaucusOvernight Defense: McCain honored in Capitol ceremony | Mattis extends border deployment | Trump to embark on four-country trip after midterms Congress gives McCain the highest honor Judge boots Green Party from Montana ballot in boost to Tester MORE and Ben Nelson have some serious home rewiring to do.  This will not be an environment in which big White House policy initiatives like those pressed in 2009-10 are likely to find much favor.     

David R. Mayhew is a Sterling Professor of Political Science at Yale University.