Google Ads


Site Stats

A Look at the Cook Political Report's Partisan Vote Index (PVI)

by: DavidNYC

Fri Feb 06, 2009 at 8:00 AM EST


Hardly a day - hardly a post - goes by here at the Swing State Project without a reference to the Cook Political Report's Partisan Vote Index, or PVI for short. In the wake of the 2008 elections, SSP's pres-by-CD project has spurred a lot of discussion about how the PVI is calculated and why it's calculated the way it is.

Quite a few people people had a hard time believing my explanation of the math behind the PVI. But you don't have to take my word for it - this is how the Almanac of American Politics explains things:

Cook Partisan Voting Index. Refers to the Partisan Voting Index (PVI) as used by Charlie Cook, Washington's foremost political handicapper. The PVI is designed to provide a quick overall assessment of generic partisan strength. For this volume, the PVI includes an average of the 2000 and 2004 presidential elections in the district as the partisan indicator. The PVI value is calculated by a comparison of the district average for the party nominee, compared to the 2004 national value for the party nominee. The calculations are based upon the two-party vote. The national values for 2004 are George W. Bush 51.2% and John Kerry 48.8%. The PVI value indicates a district with a partisan base above the national value for that party's 2004 presidential nominee. Thus a district with an R+15 is a district that voted 15 percentage points (as an average of its 2000 and 2004 presidential vote) higher for Bush than the national value of 51.2%. Similarly, a district with a D+15 is a district that voted 15 percentage points (as an average of its 2000 and 2004 presidential vote) higher for Kerry than the national value of 48.8%. An X +00 indicates an evenly balanced district. (Emphasis added.)

The boldface sentences confirm my understanding of how PVI works. But why should it be calculated this way? I agree with the majority sentiment that it seems to make more sense to compare 2000 district performance with 2000 nationwide performance, not 2004 nationwide performance. This isn't as big of a deal with the two Bush elections because they were both so close, but comparing Kerry's 2004 district numbers with Obama's nationwide numbers produces some pretty serious gaps. I'd be curious to know what sort of justifications or rationales anyone can come up with for the status quo.

In the meantime, some have suggested computing an "SVI" - a "Swing State Project Voting Index," comparing 2004 to 2004 and 2008 to 2008. In fact, CalifornianInTexas has already gone ahead and started calculating these numbers. For the most part, these will be more favorable to Dems, as the big Kerry minus Obama splits are removed from the equation.

So, I'm asking the community: Should we use the "SVI"? Should it be in addition to the PVI? Are there any pitfalls if we do so? Any reasons not to? Let's hear your thoughts!
DavidNYC :: A Look at the Cook Political Report's Partisan Vote Index (PVI)
Tags: , , (All Tags)
Print Friendly View Send As Email

A Look at the Cook Political Report's Partisan Vote Index (PVI) | 60 comments
I never cease to be amazed at all the work that goes on here
My opinion is that:

(1) The SVI looks to be more useful than the PVI in determining a district's lean, but
(2) The rest of the political world will be using PVIs, which could make it harder for SSP analysis to translate to outside readers...
(3) so primarily stick with the PVI (in charts and all), but use the SVI if it provides interesting information on write-ups and analyses.

Does Cook have an explanation for why he uses this methodology? Maybe there is a good reason we are not seeing?

Campaign Diaries

I want to second the first two thoughts
from LeftistAddiction.  But I might prefer a slightly more aggressive approach on SVI's incorporation, though there is probably room for the use of both.  If SVI is better, and I believe it will be, and the differences are not insignificant, it should be used.  Period.  

I love Charlie Cook, but if there's a better way, let's use it. It could be really good for this sight as well.  

[ Parent ]
Use both!
Ditto, both (1) and (2) are valid points.

However, I suggest use both metrics whenever possible, which would have the added benefit of making it clear to outside readers how much awesomer the SVI is.

We in the blogosphere have more flexibility to improve our metrics than do traditional analysis institutions. If we can get others to adopt those more-effective metrics, then that's a good thing.

[ Parent ]
Let the data speak
I'm willing to convert to using the SVI exclusively.  Leftblogistan needs to be a thought leader, not just a bunch of people typing away from their parents' basements.

But my training as an engineer says that we need to look at 2008 PVI side-by-side with 2008 SVI, understand which districts have more than a 5-point (arbitrary number) difference, and make a gut call on which data fits reality better.

I know this community is fully capable of completing such a thorough analysis.  Let the fun begin!

Bland is better
As ManFromMiddletown has repeatedly pointed out over at dKos, the better measure of partisan tendencies would be the average vote share for three down-ballot state offices like Auditor or Insurance Commissioner. Because nobody knows nor much cares who these candidates are, they get votes because of their party line, not as individuals.

Basing the PVI on the Presidential vote is quick and dirty, and standardized -- because not every state elects an Auditor or whatever.

But at the Presidential level, the candidate's personalities and characteristics -- like race in '08  -- can swing a lot of votes. And in the South particularly, the partisan trend may be Democratic at the state and local level and Repub at the national level. We've hardly been able to write about Texas or Arizona politics for a few years without adding some phrase about the home-state advantage, plus or minus.

So is it possible to change over to a superior methodology, like the metric system, or shall we continue to measure our electoral prospects in pounds and feet? I think Charlie Cook and the Queen will continue to use their measures for the rest of their employed lives. I just don't know about the rest of us.

We could make within-state PVIs using GUBERNATORIAL results
Many voters don't pay attention to downballot statewide races, but the governor's race IS one that many voters DO pay attention to.

We could calculate within-state PVIs for US House districts, State Senate districts, State House districts, counties, cities, and even individual precincts!

party: Democratic, ideology: moderate, district: CT-01

[ Parent ]
Now that we have the data
I'd really think that could be wonderful.  That way we could micro-target real targets in Dem.-leaning areas instead of assuming a whole county or area (Queens, Brooklyn) is incredibly Democratic.  A microanalysis might explain how guys like Padavan, Golden, or the Long Island crew in New York.  I'm sure other states have Republicans in the statehouse from very generally Democratic areas.  This would tell us where.

30, male, Democratic, CO-01
[ Parent ]
Give SVI a test run during 2010 election cycle
The best way to compare Cook's OPVI with Swing State's SVI is run them side by side for an election cycle.  Then everyone will see which system proves the more accurate.

Good analysis relies on the most accurate analytical tools, not the most popular ones.  So in that I do disagree with those who think Swing State analysis should be based on the industry standard rather than the most accurate in the industry.  

My 2 cents.

"My name's Dr. Multimillionaire and I kicked your ass." -- Rep. Steve Kagen D-WI to Karl Rove

Like the notion of using both
Using both allows a comparison of short and medium term trends for a district at a glance.

Using only PVI ignores two cycles of Democratic uptick.

Using only SVI presumes a one-cycle result as the normal character of a district.

If the site only uses SVI, then we are basing our read of the partisan breakdown on a single cycle in which the GOP ticket was headed by a guy their base did not much care for and did not have anywhere close to an even playing field in financial resources as the economy crumbled in spectacular fashion under a sitting president of his party.

If the site uses only PVI, we overstate GOP strength based on the Bush 2004 GOTV operation which no longer exists and ignore the shift in party ID over the last few years.

In any event, for the 2010 cycle the basic flaw with SVI and PVI is each concentrates exclusively on presidential results. Off-year cycles are very different beasts when it comes to TO, especially in states which have their gubernatorial elections in Presidential years (and the usual problem of getting out Leap Year Dems generally).

Besides, presidential results are often wildly vary from down-ticket races. More useful is how a party's candidate performs compared to party registration (where applicable).

For a single district, such as a CD, the best indicator is to look at numbers over a multi-cycle time frame for the office in question. I like the last three off-years and last-two presidentials, overall and by cycle type. But we need something more shorthand, don't we?

Either way, PVI or SVI, you get a nice take on whether party nominee is overachieving or under-performing in comparison to the top of his/her ticket.

There is no magic bullet, no algorithm which will tell us by plugging in data which district can be swung with a little effort. Like with investments, past performance is no guarantee of future performance.

Of course, my preference is an Excel sheet listing vote totals by office (prez, gov and CD) side-by-side over the last several cycles... broken down by county. Get more out of that than a PVI. But it does take up too much space, huh?

Anyway, I like the idea of using both. Putting them side by side gives us a quick take on short and medium-term trends.

Look at it this way, if the PVU and SVI are the same in a distirct, that tells you the GOP isn't losing a bit of their support come Hell or high water. Where the SVI number is more D than the PVI we see fertile ground.


Perhaps we could specify how many presidential terms of data we use
Say, a two-term PVI based on 2008 and 2004, a three-term PVI based on 2008, 2004, and 2000, and so on.

party: Democratic, ideology: moderate, district: CT-01
[ Parent ]
Cook's PVI is relative
Cook's PVI is merely a relative value even using the old data.  A D+0 seat is not a 50/50 district but is top heavy for Democrats (6-1 if my data is correct).  The new Cook numbers would mean that the divide between likely Republican and likely Democratic at the House level would come at somewhere around R+5 or R+6.  The data is useful but kind of screwy.

We saw some of this during the last election when people were salivating about R+1 or R+2 districts.  The recognition was that they leaned Democratic (which they do).

I live in a house that's over 40 years old that was constructed on somewhat marshy soil.  The floors all tilt in a mostly uniform fashion.  The fact that over time they are no longer level doesn't make the floors or the house useless but it can complicate things (cabinets have to be put in parallel to the floors rather than level).  The same thing with Cook's PVI.

I think we should use a dual system of level (SVI) and parallel (Cook PVI).  One more election like the last two (hope, hope) and dealing with the Cook PVI will be like dealing with the grade in the Grand Canyon.

R+2 need not lean Democratic
Take FL-16 for instance.

party: Democratic, ideology: moderate, district: CT-01
[ Parent ]
even that is flawed
I would also do voter registration numbers, and Democratic performances in close state races as well.

Call no man happy until he is dead-Aeschylus
Works most places
But I'd be concerned about wild optimism with registration numbers, especially in places like the Deep South.  Alot of older Dixiecrats just haven't changed voter registration.  It might be useful for some statehouse races, but I think the reflexive conservative vote no matter what the party usually will dominate even to the House Rep. level.  There also always exceptions to this, of course. AL-02, MS-01, the south Louisiana districts, probably other rural southern districts we hold too (NC-07, maybe VA-05).  All are tenuous but would not appear so on paper.

30, male, Democratic, CO-01
[ Parent ]
Also NY
Registration figures on Long Island and in the Hudson Valley of NY traditionally show about 60% to 40% Republican (two parties only) advantage while the same areas either have a Democratic PVI or a mild Republican lean.  Tradition keeps registration longer than voting.  Changes in registration seem more meaningful.
[ Parent ]
Also, new registrations.
One thing that convinced me of a good Democratic cycle this past year was the number new registrations that were Democratic.  Especially in these "Dixiecrat"-heavy areas, new voters are much less likely to be the super-conservative "registers D but votes R for national races" kind.

At least I think.  I have no sources to back that up.  Can anyone speak on that?

party: Democratic, ideology: moderate, district: CT-01

[ Parent ]
Is it even necessary to compare to the national performance?
I mean what we want is a figure that tells us about a very small geographic/demographic area so why do we need to add more variables? Particularly when many states aren't contested by either candidate - surely tv ads concentrated in swing states like Ohio and Florida are going to add artifacts to the PVI/SVI that isn't there in say Idaho or Massachusetts. But maybe I have completely the wrong end of the stick. I'm prepared to be enlightened! :)
That's why I suggested above, gubernatorial-based PVI
Compare it to the governor's race, especially if it occurs on off-years relative to presidential years.  Such as the midterm years, or the odd-number years.

party: Democratic, ideology: moderate, district: CT-01
[ Parent ]
That probably is a good suggestion.
Though in some governor races such as CA-Gov 2006, those numbers can be a bit flawed. In this case, there was a moderate Republican incumbent against a weak Democratic challenger. There were many districts that are strongly Democratic on the federal and state levels that went for Ahnold: CA-01, CA-05, CA-10, CA-14, CA-15, CA-16, CA-17, CA-23, CA-27, CA-29, CA-30, CA-36, CA-47, CA-51, and CA-53. A lot of Democrats crossed over to vote for Ahnold while voting Democratic elsewhere on the ballot.

My blog
Twitter
Scribd
28, New Democrat, Female, TX-03 (hometown CA-26)
[ Parent ]
PVI and SVI = Center-Right Nation
I don't think Cook's PVI or this proposed SVI would be as useful a measurement as they could be. A district with a 2004 Cook PVI of D+1 or D+2 district is actually a Republican district, since Bush won by just over 2% nationally. You're going to see a lot more of those with Obama's numbers because his margin of victory was much larger.

The proposed "SVI," comparing numbers within one election, doesn't solve this problem either. If Obama's national win was just over 7%, every district with an SVI of up to R+7 is actually a Democratic district. But people are going to consider an R+7 district to be solidly Republican, even thought it barely went to Obama. I realize the point is to compare the district to the national average, but I'm not sure how useful that information is when predicting who's going to win in a district.  Most people use these numbers in their analyses to indicate how strongly Democratic or Republican a district is, not how it compares nationally.

The only way I can see to avoid this is to have the letter ("R" or "D") always indicate which party won the district and the number indicate by how much. So a district with a PVI of D+2 went for Obama by 2% (51%-49%). A district with R+6 went for McCain by 6%. With the SVI (or Cook's PVI not including 2004), these districts would be R+5 and R+13, respectively, and we would be a center-right nation.

Another option would be to average the House, Senate, and Presidential numbers, so we don't rely exclusively on the Presidential race to define a district's lean.

So why not call a district that Bush won 51-49 and McCain won 51-49 simply R+2?
Bush wins a district 51-49 in 2004 and Obama wins that same district 51-49 in 2008 so that is a dead even partisan district regardless of the national share. And so on...
[ Parent ]
Depends on how many elections you factor in
The main part of the debate here is whether to consider just one election or the past two. If you average the past two elections, it would be R+0 or D+0.  If you follow SSP's proposed SVI method, it would be an R+2 in 2004 but become a D+2 in 2008.

Maybe we simply need more numbers in the metric to make it useful. Another detail that should be accounted for is the shift from one election to another, instead of just averaging those two elections. Using the above example, the SVI would be R+5 (because the Republican did 5% better than the nation as a whole, since Obama still won it by only 2% and won nationally by 7%), and it shifted 2% to the Democrats since the last election.  That would leave us with R+5,D+2.

If you go by the winner's margin of victory, it would be D+2,D+2, since Obama won the district by 2 points and it shifted 2 points to the Dems from the last election.

That kind of measurement tells me exactly what I want to know about the district: that it favors Democrats (something an R+5 rating would NOT suggest), and that it's trending Democratic (something neither the PVI or SVI account for).

[ Parent ]
I guess it comes down to personal choice
And what in particular you are looking for.
[ Parent ]
Common Misunderstanding
R+7 does not mean the Republican would win by 7 points, but rather that the Republican vote would be 7 points higher than average.  So given that Obama won roughly 53-46, a 50-50 district would be R+3 or R+4.  an R+7 district would be one that McCain won by roughly 53-46.

28, Unenrolled, MA-08
[ Parent ]
What do you mean by a "Democratic district"?
Sure, in today's more polarized world of political alignments, it may be useful to sort districts like this, but what if you get an absolute tsunami of an election--say, one where the presidential result is like 61-39?

If the Democrat won the presidency by 61-39 and the Democrat won the house seat 51-49, that's an R+10 district right there.

party: Democratic, ideology: moderate, district: CT-01

[ Parent ]
Rely on the SVI
Include the PVI in brackets afterward, but SVI just seems to make more sense.

Hell, imagine how little the system would have worked after huge landslides like Reagan's in 1984 and Nixon's in 1972.

That said, if you have any contact details for Charlie Cook or one of his confederates, I'd be interested to hear a defence of his method.

what would nate silver do?
i think whatever the methodology, it should include '08  numbers and drop out '00 numbers, otherwise it's increasingly out of date.

i guess i can see the appeal for cook and maybe others in that the '00 and '04 national elections were close to 50-50, therefore you almost removed a complex variable and you can just say if they voted 60% for GOP in '00 and '04 then they are GOP +20.  if you have a national election 0f 53-46 then the calculation becomes harder, but that's no excuse for sloppiness.

but again what does nate think about this?

wrong
If they voted 60% for the GOP in both elections, they would be R+17 or something. It's not the margin of victory but the % more Republican (or Democratic) than the national average.

That's exactly the problem. It tell us something that is not very useful, really: how MUCH MORE partisan the district is, not simply how partisan the district is.

George Bush won D+2 districts, and Obama won R+7 districts, yet no one would consider an R+7 district to be a complete tossup.  These are very misleading metrics.  

[ Parent ]
Or maybe it isn't as simple as you want it to be, get over it.
It is only misleading if you want it to be.  Of course you think R+7 means wildly conservative!  However, since it's rating the district compared to the nation as a whole.  Now, if the nation as a whole is D+7, then yes.  An R+7 is a toss up.  You have to take that extra step in your thought process or of course it won't make sense.  

This isn't new, we had to do the same thing with Cook's PVI.  

[ Parent ]
I don't see the problem in trying to simply the method

[ Parent ]
Well, PVI / SVI are supposed to come out to D+0 and R+0
If you want to take SVI and put it up against the national performance AGAIN (Twice?) so that R+7 districts become D+0 districts... that just doesn't make any sense.  
[ Parent ]
I realise that
It just seems silly to call a district Republican leaning when Obama won it by a significant margin all because of an almost arbitary national number.
[ Parent ]
R+7 does not mean Republican leaning...
It just means Republicans performed 7 points better there than they did nationwide the past two cycles.  
[ Parent ]
Exactly, and that's useless information.
Most people use the PVI as a factor in predicting performance of one party or another. Knowing how the district compares to the rest of the nation is not as useful as simply knowing how partisan the district is.

I find myself always doing the match with the PVIs to determine whether it's really D-leaning or R-leaning. Why not just present that information from the start?

[ Parent ]
No
It's not useless.
The fact that some people use the measure incorrectly hardly makes it useless.

22, Democrat, AZ-01
Peace. Love. Gabby.
[ Parent ]
It's not useless
What we care about is how this district would go in a 50-50 country. We shouldn't presume that a district is leaning Democratic when Obama won it by 5 after Bush won it by 5 in 2004.

Politics and Other Random Topics

24, Male, Democrat, NM-01, Chairman of the Atheist Caucus, and Majority Leader of the "Going to Hell" caucus!

[ Parent ]
If you don't like the PVI/SVI
Don't use it, just look at how Obama and McCain did in each district. The point of this number is not to say how the district went, if you want to know that just look at the horse-race numbers from the district. This is meant to compare the numbers to the rest of the country, and it is important to know how Democratic or Republican a district tends to be compared to the rest of the country. Let's put it this way, this year OH-12 went 53-46 for Obama (a 7 point win), you might think this makes is a Democratic-leaning district, but in 2004 Bush won the district 51-49. Or Obama won NE-02 by a point this time around, yet in 2004, Bush won it by 22 points. Should we really make a presumption that NE-02 is normally a tilting Democratic toss-up district when only four years ago it went over 20 points for Bush?

Why do you think that districts or states exist in a vacuum?

Politics and Other Random Topics

24, Male, Democrat, NM-01, Chairman of the Atheist Caucus, and Majority Leader of the "Going to Hell" caucus!

[ Parent ]
See my comment upstream
a 50-50 district is closer to R+3

28, Unenrolled, MA-08
[ Parent ]
No, you're naming margin, and the idea of a voting index is deviation.
We need to beat 50% to win an election (in most cases).  So a 60-40 district is ten points more Republican than a 50-50 district--this is ONE TYPE of deviation number.  The margin is 20 points, but that doesn't mean much.

If we can convince just a sixth of the current voters who normally vote Republican (a tenth of the total voting population) to vote Democratic, that takes 10% off the Repubilcan's share and puts 10% in the Democrat's share, giving us 50-50.  So it would seem like we should focus on the 10% deviation number.

But on the other hand, if all current voters continue voting the same way, we'd need 20% more of the voting population voting for the Democrat in order to pull even, which seems to mean we should focus on the margin number.

And other solutions just take parts of both of these factors.  Change some existing voters' opinions, and bring in some new voters.

So, in summary, perhaps we should look at both margin and deviation, for this type of deviation number.  But this type of deviation is always half of the margin.  So in summary, this is just the data that we get from merely looking at the raw result data, who got how much percentage of the vote in each election.

Now, the OTHER TYPE of deviation number is the sort of partisan voting index that we're talking about here.  It's not a comparison to the "50%+1" rule of thumb; it's a comparison to the rest of the country or the rest of some other larger grouping.  This is the idea that the when the tide rises, things in the water rise correspondingly.

Using our example, we had a 60R-40D district.  What does this mean?  We need another type of deviation number in order to figure out a sort of "baseline" performance of a generic D and a generic R for a given race.

One way of doing this is to compare the result to up-ballot or down-ballot results.  For example, how did the House candidate do relative to the presidential candidate?  That's (basically, with some exceptions) how our presidential-baseline PVI of today works.  For example, if the Republican president won 55-45, but the House Republican candidate won 60-40, then the district is 5 points more Republican than the nation, so a PVI number using on that election would be R+5.  (Use more than one election and average out the PVIs to get potentially more accurate results.)  But this idea isn't perfect--as people have noted, some districts are much more open to voting one party or the other at the local level than at the national level.  Additionally, this all depends on the candidates as well--a Democrat with a very conservative profile and reputation can far outperform an inherently strong Republican lean, and (especially, say, with an incompetent Republican candidate) might even win.

Ideally, we would try to collect/compile statistics on the following:
* baseline D vote
* baseline R vote
* how the swing votes usually split, between otherwise equal-quality candidates
And we need this data specific to the level of election we're talking about.  It's gonna be different for presidential, senate, house, governor, state senate, state house, etc. races.

party: Democratic, ideology: moderate, district: CT-01

[ Parent ]
SVI
It's definitely a good idea to take the lead in developing a more rational version of PVI. Look at 538- Nate didn't get so much play by tagging along with the half-assed analyses already in use. However, if you are going to take the lead, it makes sense to try to do this carefully. Presidential votes will be way off sometimes because of home-state advantage or other special circumstances. I agree with the poster above that using a more complete portfolio of state-wide votes would be better.
I agree. We have to stick to our guns here.
Comparing Kerry's performance in District's to Obama's national performance makes absolutely no sense.  Comparing Kerry's performance in districts to his performance nationally while comparing Obama's performance in districts to his performance nationally fits like a key in a lock.  We should use SVI primarily here unless a rare situation warrants using PVI, in order to show our seriousness that we have a more compelling, rational rating system.  
[ Parent ]
Do SVI
It makes a lot more sense.  Although, us also using PVI only is giving it undeserved credit.  If we all decide it is stupid and is worth making a new system over, then why present the other?
Sense is that, over the long haul, Cook has it right
It's tempting to almost use 2000 no matter what as one baseline considering the calculations since the election was as close as it was.  I think we'll see funny things with the blow-out of 2008, but 2008 was aberrationally good for us.  1984 and 1988 were the last times the electorate was so tilted and that, of course, was for the GOP.  Before that, we go all the way back to 1964 for us and 1972 for them.  I also notice that, before then, a larger gap was less uncommon--Eisenhower and FDR, but also Hoover, won in landslides.  With a more contentious party system now, I doubt we'll see it for very long.  Obama MAY be able to improve for 2012 if he does really well, but this may be a once in a half-generation sorta thing.  PVI tells us which districts flow with the tides and which are stubbornly resistant.  SVI would also be a good microtargeting tool, but something tells me that it would just be an under/overperformance scale and wouldn't be as cross-comparable.

30, male, Democratic, CO-01
Not really true
Obama won by about seven points. Clinton won by five and a half in 1992 and eight and a half in 1996.

1980 also saw a Reagan win by over eight points.

Obama got a higher percentage of the popular vote than any of these candidates, but there's just no evidence that close elections are a permanent factor of the current partisan system.

[ Parent ]
True
I went by raw total of the winner, not margin.  That should have worked in a two-party system, but it was an error (see Ross Perot--though I think most of his support would not have gone to Clinton).  Also, I was just arguing for highest vote percentage, ignoring that others were close, but not quite there.

30, male, Democratic, CO-01
[ Parent ]
2000 and 2004 were aberrations
In both 2000 and 2004 the losing candidate received more than 250 electoral votes.  The only other time that happened was in 1916.  In electoral vote terms the election of 2008 was comparable to Bill Clinton's two wins in 1992 and 1996 and less than the Republican wins of 1980,1984, and 1988.

I would argue that 2000 and 2004 were the aberrations and not 2008.  PVI could be a great example of a model that so tightly fits a small sample that it has limited predictive value.

[ Parent ]
More Fundamentally,
More fundamentally, Cook PVI does NOT do what it purports to do.  It is supposed to reflect how much more R or D a district is over a national average of 0.  But, by comparing 2000 congressional district results to 2004 national averages, it screws things up.  If you calculate the Cook PVI for the nation as a whole, it was R+0.8 before the most recent election.  After this election, it is D+2.5.  By definition, the national PVI should be 0.  But, because the Cook formula is flawed, it is not.  I prefer using a better formula.
[ Parent ]
Seems right
That formula actually looks pretty good.  If you did it for one cycle only, it would reflect only under or overperformance for a particular candidate, which is not exactly what we want.  It could become too much of a kitchen sink formula, but why not three cycles?

30, male, Democratic, CO-01
[ Parent ]
No, the national PVI is not...
by construction 0. Well, in the sense that if you take the entire country and use the percentages of the national popular vote, then yes, since that's defined to be 0.

As soon as we break it up into districts, however, then taking a straight average of district PVIs won't yield 0, since not all districts had the same number of votes cast in them.

Montana, which cast 470,000 votes for President this year, obviously doesn't have the same effect on the national popular vote percentages than Loretta Sanchez's CA-47, which cast 130,000.

You'd have to take a weighted average based on number of votes cast for President, which does result in a national PVI of 0.

[ Parent ]
I was talking
I was talking about using the national popular vote, not averaging the PVIs of all the congressional districts.

Bush won 49.74% of the two-party popular vote in 2000.  He won 51.24% of the two-party popular vote in 2004.

Cook's PVI formula, as David has noted, compares results for the last two elections in a particular district against only the most recent national resluts.  So, using the Cook PVI formula for the USA as a whole, you get:

((49.74-51.24)+(51.24-51.24))/2 = -0.8, which is a D+0.8 under Cook's PVI.

Using the most recent election results, Obama got 53.68% of the two-party popular vote in 2008 and Kerry got 48.76% of the two-party popular vote in 2004.  the Cook PVI for the USA as whole becomes:

((48.76-53.68)+(53.68-53.68))/2 = -2.5, which is a R+2.5 PVI for the USA as a whole under the Cook PVI formula.  Basically, if Cook continues to use the same formula for 2009, then every congressional district will be about 2.5 more R than true neutral.

[ Parent ]
Keep It Simple Stu...
KISS.  I guess I don't see the value of seeing how much "more republican" or "more demcoratic" a district is than a national average.  the closest we can figure how democratic a place is IN AN ABSOLUTE SENSE is of more value.

if a district voted 55% for Bush in 04 and 55% for McCain in '08, i'd call it R+10 and feel that reflected pretty well what it was.  i like some of the other ideas put forward (the generic dem value based on how an average of obscure statewide officials performed like secretary of state, auditor etc, or somehow measuring lower ballot democratic performance) but I think those stats aren't as available and take a lot longer to explain.

i think the biggest arguments against the old ('00 and '04) PVIs is the demographic changes in these districts.  i read recently that if the electorate was the same as it was in 1992, McCain would have won easily, but because of larger numbers of young people and people of color it was a whole different story.  these trends are going to continue and old numbers will miss the story.  

I hope I'm not too late
to the discussion. On the one hand, I like having PVI (or something like it) because it lets you sum up a district in one number. (And I'd be inclined to just keep using PVI rather than our own conconction on the front page, just to maintain compatibility with the rest of the pundit-sphere... as a hyperbolic comparison, I'd hate to switch SSP to Esperanto because we've all decided, correctly, that it's a more logical language than English.)

But on the other hand, I'm getting kind of tired of PVI, as it's just one dimension out of many in describing a district, and one we shouldn't fetishize more than we should. So, for me at least, the discussion of whether a district should be an R+13 or an R+15, depending on what baseline we use, is more of a distraction than anything.

Here's a case in point. Think of all the different districts that clock in at (old PVI of) R+3. (I guess I randomly grabbed that number because I've been thinking about NY-20.) These districts have little in common. Of these 14 districts, I see them falling into at least six different categories, each of which tells its own story that's very different in its level of openness to downticket Dems:

CA-11, CA-45, FL-08, FL-24: Sunbelt districts that don't have a Democratic history because they really didn't exist until a few decades ago; they're composed of new transplants in suburban/exurban settings. Trending Democratic, and also increasingly willing to vote Dem downticket, although often in reaction to terrible GOP incumbents.

IL-06, MN-02: Midwestern middle-class suburban districts that are increasingly willing to vote Dem at the top of the ticket but are still unwilling to part with conservative GOPers downticket.

NY-20, NY-26: Northeastern suburban/rural districts where there's still a Rockefeller Republican tradition, esp. downballot, but a favorable overall trend toward Dems.

NC-02, NC-07, NC-08: Lowland southern rural districts where there's enough of a tradition of Yellow Dog Democratic voting plus a sizable African-American minority that Democrats can thrive downballot.

OH-03, PA-04: Rust Belt districts that mix urban and rural components; ancestrally Democratic but trending away from us as unionists die off, but still amenable to pro-labor socially conservative Dems.

TN-04: Appalachian rural district with a history of voting Democratic downballot (and upticket too until lately), but trending away from us fast at all levels.

So, I was thinking if we really want to go large, and contribute something to the broader blogosphere, that goes beyond a purportedly more accurate version of PVI, that really affects the larger conversation about what a district can and can't support, maybe we should try categorizing districts in terms of 20 or 25 typologies. Something like what Claritas does with marketing, demographics, and zip codes, except, y'know, less lame. For instance, I'm sure we can think of other districts that fit easily into each of the six categories I have above. Sounds like a lot of guesswork initially, but if we find variables that truly work, we could actually do some regression analysis and make sure that it meets SSP's usual data-driven standards.

Sounds like this means we need to differentiate between different levels of government
In order to isolate the time-varying nature of PVIs.

party: Democratic, ideology: moderate, district: CT-01
[ Parent ]
Esperanto allusion dead on and love the Claritas idea
We are arguing every nuance of a shorthand. And isn't shorthand by its very design incapable of incorporating every nuance?

And, if SSP switches to a measure and terminology which only has meaning insides its own domain, are we in danger of letting our heads disappear into our own body cavities? Need to speak in a language understood by others or we might as well be posting in Basque.

Use both. Still spreads the idea of SVI model and retaining PVI for comparison makes accessible to broader world of numbers-crunching junkies.

Of course, PVI has limitations, as does SVI. Each is based on a single variable for one office. For a fuller understanding of a district, one needs to incorporate many other quantitative and qualitative inputs. That does not take away from the utility of the at-a-glance number. It just means it is not the greatest thing since sliced bread.

Cook came up with PVI because party registration numbers alone do not correspond accurately with actual voting performance of districts. Chose to do multi-cycle to avoid buying in to much on one-off election cycle fluctuations. Of course, this means it will lag when one party is trending higher and another lower. It will self-correct if trends continue on their past track, but a cycle late.

The problem with SVI is its assumption that a one-off cycle performance indicates the true relative character of a district.

Which is why I like having both side by side. Can look at them and use our guts to make our guess which way it will move in light of recent events.

It will be especially useful looking at the difference between PVI and SVI in districts which are undergoing rapid demographic shifts... like increasing Hispanic and first-generation immigrant registration in exurban areas or the flight of the young from rapidly aging rural areas or states like Florida, where the immigration pattern has shifted from north-to-south to a majority south-to-south pattern

Though I point out, neither is that good for an off-year cycle. Very different beasts than presidential years. The utility of PVI and SVI for the off-off-year gubernatorial and special elections this year is minimal. But it still a starting point for reading the trends.

What I love in your post is the Claritas proposal. It would be really ground-breaking work. My head hurts thinking of the math and trying to remember how to build a model from the quantitative political analysis and Prob and Stats classes in school... but it is a damned fine idea.

I would say the model has to be built differently for off-cycles and prez or gubernatorial years, depending on the state and office in question. The TO fluctuation is so striking it has to be built into the model.

DO IT! DO IT! DO IT!

...so I don't have to do it myself.

Of course, in any case... SVI, PVI or a Political Claritas... we are using static models to try and forecast a dynamic process. Love the stats but in the end the gut must rule. And, beyond that, the character and personalities of the candidates themselves will trump everything else in the end, assuming each has adequate resources.

[ Parent ]
perhaps we should make some list of case study evaluations
against which any given system is evaluated.

For example, one such case might be Barrow's Georgia district, which I've heard flips between Republican-leaning and Democratic-leaning depending on whether there's a presidential race to increase turnout of Blacks.

Another such test may be ancestrally D/R areas that trend R/D at the national level.

party: Democratic, ideology: moderate, district: CT-01

[ Parent ]
Oklahoma
would be a good state to use to test model under duress. Dems have nearly 2-1 edge in registration but one hell of a strong GOP performance in federal elections.

On the other side of the coin, coastal Florida seats with high GOP reg numbers (but drawn from a very country club and old school GOP set of voters) would serve nicely.

[ Parent ]
What we need: a "baseline" for comparison.
The question: How do we get it?

What we need is a way to estimate a sort of generic D vs. R matchup for each district (or state, or state senate/house district, or county, or town, or even precinct).  Then we can figure out how well our candidate did, versus their candidate.

Ideally, there should be a way of finding this out.  For example, we could ask everyone in the district who they'd vote for, generic Democrat or generic Republican, for a given position, and then ask them whether they were certain of their choice.  If certain, we'd put them in the "solid D" or "solid R" bin, and if uncertain, we'd still get an idea of how big the swing block is, and whether they lean D or R.

Now, even if we do an abstract version of this, via polling, we (or anyone else) would still need to put in a ton of groundwork into gathering this data.

If we can approximate this data via some other measure, that would be quite useful.  For example, we could try to average out the performance of candidates to that seat in the past, but seeing as we are trying to gauge exactly that, this would be a wolf-guarding-the-sheep situation.  Another method is to compare to results up and down the ballot.  While this can be skewed by differing perceptions of the parties at different levels of government (such as Democrats being popular locally in heavily conservative areas), this is nevertheless the basic idea behind Cook's Partisan Voting Index, on which our "SSP PVI" is based.

Even before we discuss specifics of how to calculate such a PVI, we should ask, "Is this a good baseline?"  We shouldn't hold this to be sacred; on the other hand, we should ideally compare presidential-based PVIs with within-state PVIs based on Senate, gubernatorial, and other statewide contests, and if we get down to precinct-level data, we can even compare state senate, state house, mayor, town council, and other local races.  The presidential, gubernatorial, and senate results, however, are perhaps the easiest to sort out of all these data sets.

However, they are also the highest-profile results, easily influenced by personalities and unique characteristics of individual candidates.  Perhaps lower-profile statewide offices may be useful...

And don't forget that people undervote--not all presidential, gubernatorial, senate, or other statewide office votes will include votes for Representative, State Senator/Representative, etc..

party: Democratic, ideology: moderate, district: CT-01

hey, this is why Cook came up with PVI
to find a better approximation of average performance than party reg.

Like the idea of a multi-cycle average of performance by office. Think that is useful data as a starting point for core strength.

But then we are back to the problem at the root of the argument over whether to use SVI or PVI.

How much does the past dictate the future? Do the earlier cycles understate one side's current or projected future strength by overvaluing votes cast by a coalition which has since crumbled?

What I really like in your post, though, is the suggestion of comparing relative performances up and down the ballot.

I think there is real utility in seeing, say, a Democrat for a House seat performs five points better or worse on average over a five-election period than than the Dem running for Governor or President. Or whatever the situation. We get to see if and where the numbers diverge and start to analyze why.

Could base it off the top of the ticket and do as a plus/minus. Of course, the office used as the baseline would have to change, There will be cycle when there is no gubernatorial but there is a Senate race and vice versa. And times when the CD is the top.

We see a Democrat running for governor of Bugtussle got 52 percent in an electoral district. The average over the last four cycles was, say, 55. So the guys is minus 3. We can then break down the lower offices by how much they under- or over-perform by that particular cycle and however-many cycles. And we could also compare their averages for theit office itself over a multi-cycle period.

Could even break it out by which office is the top of the ticket to see difference when state or federal is the top of the ticket and so on.

All could provide intriguing data. And it would be useful in researching past performance of lower-tier officeholders as they start moving up the ladder. The folks who exceed the average for their top-of-ticket and their own office would be ones we mark as comers for the future. The ones who do not we might need to consider as lame horses not worthy of promotion.

But your point about the personality of the upper-ticket candidates is valid.

Where this model breaks down is the quality of their opposition. If we see a Congressman who has good numbers but has run against underfunded token challengers, the numbers will make this person look like a stronger candidate for statewide office than he really is. And one who has survived tough races will look relatively weaker.

OK, so never mind using it for weeding out the farm system.

Maybe we need a model based on votes earned compared with some net campaign spending formula.

The problem is, like with mutual funds, past performance is no guarantee of future returns.

Our models are static and we are trying to use them to forecast a dynamic process. There is no model we can build which will replace the gut and the changing of the tides.

This does not mean we should not try, but, in the end, the quantitative models can't replace qualitative judgment.


[ Parent ]
Just found this via dailyKos
I think the SVI is better.  SVI will average to 0, and PVI will not.  

But it could be even better.  One thing is to include the native son effect for POTUS and VPOTUS.  That would make it a more accurate gauge of true feeling, and it would be relatively easy to do.

Given that the data is already entered, we could also look at trends over time.

How do you include the native son effect, though?
Do you simply mark the data as such with a footnote?  Or do you go back to previous data without it?

party: Democratic, ideology: moderate, district: CT-01
[ Parent ]
A Look at the Cook Political Report's Partisan Vote Index (PVI) | 60 comments

Copyright 2003-2010 Swing State Project LLC

Primary Sponsor

You're not running for second place. Is your website? See why Campaign Engine is ranked #1 in software and support among Progressive-only Internet firms. http://www.mediamezcla.com/

Menu

Make a New Account

Username:

Password:



Forget your username or password?

About the Site

Publisher: DavidNYC

Senior Editor: James L.

Senior Editor: Crisitunity

Contributing Editor: Jeffmd

Search the SSP archives:

SSP Resources

Presidential Vote by CD

SSP Document Collection

Race Tracker Wiki

Dave's Redistricting App

Blogroll

Daily Kos Elections
DemConWatch
Five Thirty Eight
Governing
Hotline On Call
Political Wire
Pollster.com
Prop*50
Public Policy Polling
The Thicket
Powered by: SoapBlox