Over the last three elections, 95 members of the U.S. House lost general election races. In the three elections before that a total of just 21 incumbents lost.

Those numbers suggests that the period of relative stability that governed House races from 1998 until 2006 — when the incumbent re-election rate never dipped below 96 percent — is over and that we may be entering a time of extended volatility in the lower chamber.

New data from Gallup affirms the sense that voters are restless and that incumbents need to watch their collective backs.

Just 28 percent of registered voters said they thought most members of Congress deserved re-election while a whopping 63 percent said they didn’t in a May Gallup poll.

Those numbers are unchanged from Gallup data in advance of the 2010 election where an equal 28 percent said that most incumbents deserved re-election.

What that consistency seems to suggest is that which side controls the House makes almost no difference to voters when it comes to how they view the institution.

Other data points back up that sentiment.

In the run-up to the 2010 election, 21 percent in Gallup data said they approved of the way Congress was handling its job.

Analysts — present company included — predicted such dissatisfaction would lead to wholesale change in the body. It did, as 63 seats moved from Democrats to Republicans.

And yet, this month, Gallup showed Congressional job approval at 24 percent.

Much of the unpredictability can be traced to independent voters who seem to now almost immediately sour on whichever party happens to be in the House majority.

In the latest Gallup polling, just 23 percent said most members of Congress deserved re-election — numbers below the 26 percent of self-identified Republicans and 36 percent of Democrats who said the same.

That’s in keeping with the behavior of independents in the last two midterm elections. In 2006, independents favored Democratic candidates by 18 points and Democrats picked up 30 seats and the majority. Four years later independents went for Republicans by 19 points and the GOP re-took the House.

Context does matter.

Incumbency rates have typically remained at or above 90 percent even in the last three elections as the 2001 redistricting process largely served to create congressional districts that made it next-to-impossible for members of Congress to lose — barring a major scandal or colossal slipup.

And, Congress has never been all that popular; Gallup’s historical average job approval for Congress is only 34 percent.

Still, it’s important to remember that only 25 seats separate Democrats from the House majority in 2012. That means that turning over only 5.7 percent of the 435 House seats would flip House control for the third time in six years. (That’s particularly remarkable given that Democrats held the House majority for all but four years from 1930 through 1994.)

The slippage in the power of incumbency then needs not to be wholesale for it to make a meaningful difference in who controls the chamber.

Early signs — and it’s important to remember that we are still 537 days away from the 2012 election — suggest that we could be headed for another election in which simply having “Rep.” in front of your name is enough for voters to throw you out of office.

The broader question? If 2012 is another election of wholesale seat change, is it even possible in this political world for either party to build a sustainable House majority?