Our friend Dan Drasin, still crunching industriously at the 2006 election numbers, sounds an alarm for Democratic hopes of--KEEPING THE HOUSE

By Dan Drasin

[DWT friend Dan Drasin, a longtime computer technologist and software developer--as well as political junkie--from Baltimore, reported recently here on his first findings from crunching the 2006 election numbers, which suggested that Rahm Emanuel's election strategy came closer to losing than to winning the House of Representatives for the Democrats.

[We asked him what he has since gleaned from his data-mining. What follows is the first part of a two-part report, pinpointing the danger for Democrats hoping to hold onto the House in succeeding elections. In the second part he will suggest what the party will have to do to maintain its House majority.]

Now that the House of Representatives is safely in Democratic hands, the common wisdom is that it will remain so for a while. After all, the Republican wave of 1994 kept the GOP in power for 12 years.

However, as I was perusing the Partisan Voter Index (PVI) numbers, I noticed some disturbing details that make me wonder if this is a foregone conclusion. Through a combination of geographic accidents, luck, and straight-out gerrymandering, the GOP seems still to have an enormous built-in advantage in being able to maximize the impact of their voters on the House makeup.

Let me start by listing a single set of numbers that should scare the Democratic faithful. This year, the Democrats enjoyed a tremendous tailwind: They dominated the popular vote by 57.7% to 41.8%--a margin of 15.9%(!), a dominance that is unheard of in recent times. They picked up a lot of seats, but their final margin in terms of House control was only 7.4% (via a pickup of 30 seats). So a natural fear should be that if Democratic House control requires this kind of dominance, victory in the next couple cycles may be hard to repeat.

The conventional wisdom is that the difference between popular-vote percentage and House seats is mostly a product of the incumbency advantage that the Republicans enjoyed before this election. And since the Democrats now have it, not only is incumbency no longer a hurdle, but it will even help keep them in power.

But wait. In 1994, the Republic wave consisted of a 54-vote gain by the GOP. This led to a margin of control in the House of 5.7%. And to get this, they won the popular vote by 49.9% to 44.4%, a margin of only 5.5%! Get it? The GOP overcame incumbency (and an even higher amount of it at 54 votes) with only a 5.7% advantage in the popular vote. The Dems needed 15.9%--three times as much.

OK, I admit, this is just one set of numbers, and there are many other factors at play here--retirements, unopposed candidates, recruiting, scandals, etc. But it's enough to get me scratching my head and wondering why (and sleeping a little less soundly). So my first question is: Is this effect real? Or is it simply an aberration? My next question is: If this effect is real, why does it exist? And finally, what can be done about it?

IS THE REPUBLICAN ADVANTAGE REAL?

On the first question, it seems pretty clear that this effect is real. All we have to do is look at the House election results since 1994. In 2004, 2002, and 2000, the GOP control of the House exceeded their popular-vote margin consistently. (In 2004 it was by over 4% over the popular vote margin.) And in 1996 and 1998, the Dems actually won the popular vote in the House (by 0.3% and 0.9% respectively) but remained in the minority (by 4.8% and 2.8% respectively).

So in every year since 1994, the Republicans have outperformed their part of the popular vote. It's hard to see how this is all explained by incumbency alone when you have the 1994 result to compare with 2006. (You can see a visual of this.)

So if, at least for the sake of argument, there is a GOP bias in House control, why does it exist? Clearly this particular bias hasn't always been the case. If you look at House results before 1994, you see a long-term bias in favor of the Democrats even when the Republicans were incumbents. I stumbled across an answer to this question as i was doing research on the PVI.

WAIT! WHAT THE HECK IS THE PVI?

Now before I jump into discussion of the PVIs, let me first explain what the PVI is and what it isn't. (You can find a good detailed explanation here, or a simpler one here. Readers will have to forgive some simplifications I'm making in my explanation.)

The PVI is a measure of the relative partisanship of each House district. Relative to what? Relative to the rest of the country. So a PVI of +2R means that a Republican candidate in that district is likely to get 2% more of the popular vote than the Republican share of the vote nationwide. So if in a particular election year the Republicans get 49% of the vote, then this candidate would (all other things being equal) get 51%.

What it does not mean is that in any given year the Republican candidate can expect to win by 2%. It just means that this district leans a little more Republican than the country as a whole. But in a year when the Democrats win by 10%, this Republican candidate would still lose by 8%. (Of course, these are just averages and don't factor in the particular candidate, issues, or incumbency . . . )

CRUNCHING THOSE PVIs--UH-OH!

Now, here's what I did. I took the PVI numbers (which i got from a blog) and sorted them from "most Republican-leaning" to "most Democratic-leaning." The results were immediately revealing.

The first thing that stood out was that of the 435 districts, there were 232 that had a positive Republican PVI and 190 that had a Democratic PVI (with 218 needed to control the House).

That doesn't mean that the GOP will win 232 districts every year, but because of the way PVI is calculated (from the last two presidential cycles--in our case, cycles where the popular vote was very evenly split). It means that the Democratic votes in those elections were concentrated in a smaller number of districts than the GOP's. Or in other words, in a year where the popular vote in the House is close, absent other considerations, the Republicans have a 40-plus seat advantage!

The other side of this coin reveals itself when you look at the distribution of the PVI. The Democrats have a lot of heavily Democratic districts, while the Republicans are spread out more efficiently. This means the PVIs for Republicans aren't as high, but there are more districts where they have an advantage.

For example, there are 21 districts that have a PVI of +30D; there are no +30R districts. By the same token, there are 141 districts with a PVI between 0R and +10R, while there are only 80 similar districts for Democrats. Here and here are two charts to summarize--and illustrate how the Republican House results outperform their share of the popular vote consistently.

BUT HERE'S THE REAL KICKER . . .

Obvious targets for pickups are those districts where the incumbent's party is opposite the PVI rating. (This is House Campaigning 101. Just look at the Dem and GOP targets in 2006.) The problem is that when you look at the current House makeup, there are 46 Democrats representing districts with a Republican PVI and only 13 Republicans in Democratic PVI-land. On top of that, 21 of the 46 Democrats are in districts that are PVI +5R or greater.

Gulp!

That means that in the next election cycle, there are going to be a lot of Democratic seats with big targets on their backs. No doubt, incumbency will help many Dems retain their seats, but with the GOP only needing to pick up 16 seats, Democratic control is by no means assured.

If this is the problem, what can be done? In the next post, I'll go over some strategies and tactics that will help mitigate the built-in advantages that the Republicans currently enjoy in the House.