The Rob Anders – Donna Kennedy-Glans spectacle in Calgary West is only the most obvious example of the ramifications of perceived incumbency advantage.

For those who haven't been following it, Rob Anders is widely perceived as a poor MP, having been called by Don Martin "Canada's worst MP." He has said and done things that make even the most rock-ribbed Conservative cringe. Ms. Kennedy-Glans wants an opportunity to contest the nomination for the Conservative Party, but the party executive says there will only be a nomination fight in ridings where two-thirds of the riding executive asks for one.

In effect, this means the bar for replacing a sitting Conservative MP as the nominated candidate is not 50 per cent plus one, but in fact whatever level of popular support is required to gain the support of two-thirds of the executive at an annual general meeting AND 50 per cent plus one in a separate nomination fight.

This is actually an improvement over the past. In 2006, the CPC did not allow any contest against sitting MPs. In 2008, the window for contesting nominations was peculiarly short and no nomination vote was actually held in Calgary West.

But the Conservatives are not the only party that goes out of their way to protect incumbents. For instance, the Liberals automatically re-nominated all incumbents in the 2000 election, citing electoral urgency because the Prime Minister wanted to call an election.

Why do parties of all stripes go to incredible lengths to protect incumbents, no matter their merits or damage to the party brand?

Because incumbents win. Period.

The conventional wisdom is that incumbents are inherently better candidates, and decades of academic work supports that finding.

In the United States, the re-election rate for senators is 88 per cent. For the House of Representatives, it is an astonishing 96 per cent.

In Canada, the rate is less dramatic, as changes in government tend to produce wholesale removal of numerous incumbents in a sweep. But there is still a significant advantage for the incumbent.

But effect does not dictate cause and the reasons are less than clear. 

In the United States, the rise of television was long credited for the incumbency advantage. The two phenomena rose in concert, with 1958 the first year the number of televisions equaled the number of American homes, and that same year the power of incumbency took off in the U.S. The power of TV demanded it be the centre of any political campaign, but TV news coverage favoured incumbency and the cost of ads favoured the fundraising advantages incumbency holds,

However, a recent study out of MIT finds that television in fact has little impact on the incumbency advantage. By looking at counties with out-of-state media markets (ie, parts of Massachusetts are not in the Boston media market, but in closer Albany or Providence), they theorized that if television was the cause, the incumbent should have lost some of their advantage in those counties. But that theory did not hold up in testing.

An earlier paper from the same department at MIT tested the theory that the advantage is unique to legislators and must therefore be related to things like redistricting and gerrymandering, legislative irresponsibility or pork barrel politics. In fact, they found that state executives (governors, attorneys general and the like) had a bigger incumbency advantage than legislators, disproving that theory as well.

Instead, the incumbency advantage may actually be a case of perception creating reality.

Stephen Levitt, famous for Freakonomics, and Catherine Wolfram from Harvard did a complex econometric study into the causes of incumbent advantage, rather than simply measuring the advantage as earlier studies had done. They found that the source may actually be incumbents getting better at scaring away high-quality opponents.

Rather than anything based in reality, the advantage is that people thinking about running for public office over estimate the ability of the incumbent to retain office. High-profile candidates decline to run, fundraising goes uncollected by default, volunteers sit unmotivated on the sidelines and the incumbent gets back in because no one bothered to run hard against him.

Ken Carty from UBC found similar evidence in Canada. Where they are not intimidated, "local party organizations of non-winning candidates are in the position of being able to realize potentially significant electoral returns through the mobilization of additional personnel or financial resources in their constituency campaigns."

Political parties should severely rethink their slavish devotion to protecting embarrassing MPs.

Rather than an unstoppable vote-getting machine, embarrassing incumbents may only be embarrassments.

And where that embarrassment begets high-profile and highly-motivated challengers willing to take on a gaffe-prone incumbent, political parties may experience additional losses as a reward for their misplaced loyalty.