Talk:Cook Partisan Voting Index

New CPVI released

I see there's been a lot of editing activity recently, but I still see some old info (like the recently-reverted edit about 2008 and 2012) and old sources. I'm wondering which parts are old and which parts have been updated to the most recent, and if there's a fast way to import the data from the Cook Political Report directly - specifically, from the source that is currently source 4, which is the most recent, from only 5 days ago as of today. Blippy1998 (talk) 18:43, 12 April 2017 (UTC)

Good news: that covers the PVI's for individual House districts. Bad news: it does not seem to cover the overall PVI's for the 50 states, which are the numbers that are used for Senate – that source still needs to be updated for 2016... --IJBall (contribs • talk) 19:02, 12 April 2017 (UTC)

CPVI Calculation Section

I had a little trouble understanding the explanation of the calculation. I wonder if I may have interpreted something. It reads:

"PVIs are calculated by comparing the district's average Democratic or Republican Party's share of the two-party presidential vote in the past two presidential elections to the nation's average share of the same. The national average for 2004 and 2008 was 51.2% Democratic to 48.8% Republican.[1] For example, in Alaska's at-large congressional district, the Republican candidate won 63% and 61% of the two-party share in the 2004 and 2008 presidential elections, respectively. Comparing the average of these two district results (62%) against the average national share (50%), this district has voted 12 percentage points more Republican than the country as a whole, or R+12."

Where does "50%" come from if the average national share was 51% for Democrats and 48% for republicans? Furthermore it is stated that the scale compares the districts average "Democratic or Republican Party's share" of the vote, but which party specifically is used for comparison in any given calculation? Is it set up such that it's always R or D + a number? Can there be minus? I think the explanation could be clearer or else I am simply not understanding. LostRoss (talk) 21:49, 28 February 2018 (UTC)

IMHO, there is clearly an error in the quoted text. --LondonYoung (talk) 23:54, 1 March 2018 (UTC)

Cook PVI District Index for Pennsylvania does not match the cited source?

I'm not a regular editor, sorry if I'm doing this wrong, but I was looking at the PVI in Pennsylvania and it does not match up with the Cook Political Report's Partisan Voting Index: Districts of the 115th Congress that it is citing (https://web.archive.org/web/20170607150217/http://cookpolitical.com/file/Arranged_by_State_District.pdf). I think this might be due to redistricting - Pennsylvania's Supreme Court ordered that the electoral map be redrawn in February and there is a new map. From what I read (https://fivethirtyeight.com/features/pennsylvanias-new-map-helps-democrats-but-its-not-a-democratic-gerrymander/) the numbers in the article are correct for the updated map - for example in the new map districts 1/2/3 should be slightly Republican leaning, heavily Democratic leaning, and heavily Democratic leaning, which the Wikipedia article shows - but the document that the article cites still has them as D+31/D+40/R+11. Maybe there is an updated version of the PVI that it should be citing instead? I'm not sure where it would be though because from what I see on the Cook website (http://cookpolitical.com/index.php/pvi-map-and-district-list) the numbers do not match up to the numbers listed in the article. I'm not 100% sure if that is the case, but either way the page's index information doesn't appear to match the source that it is citing. --NotAWikipediaEditor 17:08, 6 March 2018 (EST)

Penn. was, IIRC, one of those states forced to do an out-of-cycle redistricting by the courts. That is why the numbers don't match. They probably haven't been updated yet because I'm not sure it is clear if the current map will end up being "final" or not... --IJBall (contribs • talk) 23:00, 6 March 2018 (UTC)

WTF happened to this article???

Where is the district by district listing? It has been my go to source for checking candidate viability for years. This is EXACTLY why I continue to refuse to donate to Wikipedia - to this day overzealous mods & contributors continue to wreck articles and destroy hours of people's work in the name of their idea of what constitutes proper citation or worthiness. 172.113.104.20 (talk) 00:14, 8 September 2019 (UTC)

Are you referring to the collapsed table? You can expand it by clicking the "[show]" button in the top right of the table. It's common to collapse massive tables by default to reduce the amount of scrolling needed for people who aren't interested in the table. --Posted by Pikamander2 (Talk) at 01:48, 8 September 2019 (UTC)

Extremes section

FTR, I'm not sure I agree with many of the changes today. Collapsing the tables is neither here nor there IMO, but the 'Extremes' section should not have been removed – it's a straight WP:CALC section. Perhaps that can be made clearer at the top of that section, but it does not need to be removed. --IJBall (contribs • talk) 02:50, 8 September 2019 (UTC)

Semi-protected edit request on 16 November 2019

Update Kentucky Governor to Democrat 71.254.15.47 (talk) 00:16, 16 November 2019 (UTC)

 Not done: please provide reliable sources that support the change you want to be made. NiciVampireHeart 03:40, 16 November 2019 (UTC)

Semi-protected edit request on 16 November 2019

Update Kentucky Governor to Democrat

https://www.cnn.com/2019/11/14/politics/kentucky-governor-recanvas-begins/index.html 71.254.15.47 (talk) 11:05, 16 November 2019 (UTC)

 Not done for now: We should wait until the result is certified and the new governor is inaugurated. Sceptre (talk) 12:46, 16 November 2019 (UTC)

State Table Should Have a District Breakdown

The state table shows how many D and R House members the state has. To supplement it, I think it should have a column showing how many D+, R+, and Even districts the state has. That way we can tell if a party won a significant amount of House districts that lean the other way without having to check the rating and representative from each district. In addition, the state table should have a total row at the bottom to show how many D and R governors, senators, and representatives there are, along with how many D+, Even, and R+ districts there are. These changes could be done without making the table too big, and they summarize information from elsewhere in the article and the article about the 2018 House elections, so there shouldn't be any objections that I am proposing inaccurate information or original research. EvanJ35 (talk) 15:05, 7 December 2019 (UTC)

The 'By State' List/Section Is Problematic

The PVI 'By state' list and section at this article is problematic. The Cook Political Report itself does not publish this info, and does not even seem to track it – their focus is exclusively on U.S. House districts. I've just spent several minutes looking, and I cannot find a legitimate secondary source that publishes this data (some articles look at state PVI on an election-by-election basis, but do not use the "standard" PVI method of combining the PVI of the previous two elections (e.g. 2012 & 2016) for their "by state" PVI figures) – the only sources that do publish "by state" PVIs are effectively election junkie blogs and forums (but blogs and forums are WP:NOTRS).

If an actual bona fide WP:RS cannot be found for this data (and, at this point, I think even a source for "by state" PVIs just for the 2016 election would be acceptable), I propose that the 'By state' section/list should be removed from this article as a classic case of WP:OR. --IJBall (contribs • talk) 15:04, 22 May 2020 (UTC)

It's literally sourced at the very top of the list https://cookpolitical.com/state-pvis Andibrema (talk) 09:16, 23 November 2020 (UTC)

Understand, that wasn't added until this edit on Oct. 31. Notice that my post above was well before that citation was added/updated. --IJBall (contribs • talk) 16:24, 23 November 2020 (UTC)

Extremes section, again

So, again, somebody has objected to the 'Extremes' section here, and again they seem unwilling to discuss it on this Talk page. Thus, I am posting here in case other editors actually want to discuss this issue.

FTR, I don't think the entire section should be removed – but I do think it should be substantially trimmed back from where it is now, and that its current length is WP:UNDUE.

But, as a general concept, reporting "extremes" comes straight from reading the table (which is sourced), is basically a simple WP:CALC-type interpretation, and in the case of "most D+ R-held district", etc. will in fact have some secondary coverage, at least on the general topic if not on the specifics.

Anyone else case to chime in?... --IJBall (contribs • talk) 01:24, 18 January 2021 (UTC)

@IJBall: The extremes section is definitely too long. It's okay to say West Virginia is the most Republican state with a Democratic Senator, but not that Joe Manchin is the Senator specifically, he's a moderate, his political history in his state, etc. A lot of what we could include in the Extremes section can already be inferred in the table. Here's what I would trim it down to:

Extended content

The most Democratic congressional district in the country is New York's 15th, located in the Bronx, with a PVI of D+44. The most Republican district is Texas's 13th, based in the Texas Panhandle, at R+33. In terms of states as a whole, Wyoming is the most Republican at R+25, and Hawaii is the most Democratic at D+18. In the Senate, the most Republican-leaning state to have a Democratic senator is West Virginia (R+19 PVI), represented by Joe Manchin. The least Democratic-leaning states to have two Democratic senators are Arizona (R+5 PVI), represented by Kyrsten Sinema and Mark Kelly, and Georgia (R+5 PVI), represented by Jon Ossoff and Raphael Warnock. The most Democratic-leaning state to have a Republican senator is Maine (D+3 PVI), with Susan Collins. The least Republican-leaning state to have two Republican senators is Florida (R+2 PVI), represented by Marco Rubio and Rick Scott. Four states with a Republican-leaning PVI (Kansas, Kentucky, Louisiana, and North Carolina) have Democratic governors, while three Democratic-leaning states (Maryland, Massachusetts, and Vermont) have Republican governors. These governors are all seen as centrists who attracted voters from the opposite party. The most Republican-leaning state with a Democratic governor is Kentucky (PVI R+15), with Andy Beshear, and the most Democratic-leaning state to have a Republican governor is Vermont (PVI D+15), with Phil Scott. Two of the three states with even PVI ratings, Pennsylvania and Wisconsin, have Democratic governors (Tom Wolf and Tony Evers, respectively). In the House, the most Democratic-leaning congressional district represented by a Republican is Florida’s 26th; with a PVI of D+6, the district is represented by Carlos Gimenez, only six Republicans to represent a Democratic-leaning House district. The most Republican-leaning congressional district represented by a Democrat is Georgia’s 7th with a PVI of R+9, it is represented by Carolyn Bourdeaux. Following the 2020 elections, there were 26 Republican-leaning House districts represented by Democrats. The only Republican leaning state to have more Democratic House members than Republican House members is Arizona (PVI R+5) with 5 Democrats and 4 Republicans. The most Democratic leaning state to have more Republican House members than Democratic House members is Wisconsin (PVI EVEN), with 5 Republicans and 3 Democrats. The most Democratic district relative to its state is Tennessee's 9th, being D+28 in an R+14 state (a 42-point difference). The most Republican relative to its state is Illinois's 15th, being R+21 in a D+7 state (a 28-point difference). The most Republican states to have a Democratic Representative are Alabama, Kentucky, and Tennessee; while the most Democratic states to have a Republican Representative are California, Maryland, and New York. Of the 428 Congressional districts that are in states with more than one district, 104 lean to one party while their state leans to the other.

Wikinights (talk) 04:10, 6 February 2021 (UTC)

My $0.02? I would also completely cut the Governors from this section. PVI was originally created to track federal races, and there is much greater correlation between PVI and the results of federal races like House and Senate. Governors races are local affairs, and states have shown a much greater willingness to go "against partisan lean" – e.g. even some of the most D+ PVI states like MA, MD, and VT have Republican governors, etc. IOW, PVI is much less relevant when talking about Governors races. So the 'Extremes' section should really only talk about House and Senate seats IMO... Also, I would leave details, like Texas-13 being "R+33", in. --IJBall (contribs • talk) 04:17, 6 February 2021 (UTC)

My bad, I didn't intend to remove the R+33 part, only the Texas Panhandle part. Edited the prior collapsed section. Wikinights (talk) 19:03, 6 February 2021 (UTC)

I have now trimmed this section back some, cutting the Governors paragraph, and cutting some really WP:OR-y stuff from the third paragraph. As of now, the third paragraph is still problematic, IMO – it would be better if that paragraph was citing some secondary sources. --IJBall (contribs • talk) 18:37, 26 February 2021 (UTC)

The 'By state' Table

Three of the four columns should probably be removed, in my opinion. Cook PVI is strictly a function of presidential vote and including House delegations, etc. serves little purpose here beyond confusing readers. I understand why analysis of PVI in a broader electoral context is interesting but it would fit better in a blog. — Preceding unsigned comment added by Dkjeiau373671 (talk • contribs) 04:18, 24 February 2021 (UTC)

It doesn't make sense to do that – it makes sense to either cut the entire section, or to leave the table as is. --IJBall (contribs • talk) 05:10, 24 February 2021 (UTC)

It isn't helpful to say, "doesn't make sense," and simply suggest your preference instead. To your idea of cutting the entire section, it's absurd to cut state PVI from the PVI Wikipedia article. It's relevant because PVI is the subject here and Cook tracks it. Sure, it isn't as commonly used as district PVI but it is on their web site. However, as I said, the section can be improved by removing extraneous non-PVI data. Disagreement is great but let's either justify our opinions or move on. — Preceding unsigned comment added by Dkjeiau373671 (talk • contribs) 08:10, 24 February 2021 (UTC)

OK, have it your way – I oppose your proposed change. --IJBall (contribs • talk) 13:19, 24 February 2021 (UTC)

Footnote Error About the Definition of PVI

The footnote was written by D. Byler (2016), not anybody here, but I still want to point out that it's wrong. As an example, it says that if a Democrat wins by 5 points, he or she would be expect to win a D+2 state by 7 points. In reality, D+2 means 2 points more Democratic and 2 points less Republican, so you have to double the 2 before adding it to the 5. If anybody doesn't believe me, the data proves my point. Take Hawaii, which is D+15. In 2020, Biden won by 4.45 and won Hawaii by 29.46 for a difference of 25.01. In 2016, Clinton won by 2.10 and won Hawaii by 32.18 for a difference of 30.08. The mean of 25.01 and 30.08 is 27.545. Doubling 15 from D+15 to make 30 is much closer to 27.545 than 15 is to 27.545. EvanJ35 (talk) 13:12, 17 April 2021 (UTC)

@EvanJ35: PVI is based on raw percentages, not margins. Refer to Cook Partisan Voting Index#Calculation and format. Wikinights (talk) 16:15, 17 April 2021 (UTC)

New map

Someone removed my map, I think it was in good faith because at least for me, if I try to see the full image it still shows the oldest version, but I want to say that what is correct (it may obviously have some mistakes in a few districts, but it is correct in general) --Ngfsmg (talk) 23:29, 17 April 2021 (UTC)

There was at least one mistake that I already corrected, with UT-3 that was shown as if R+20, but it's not exactly my fault because it's like that on the table in this article, that table must be corrected --Ngfsmg (talk) 23:40, 17 April 2021 (UTC)

The same user now decided to remove my map again with no justification... If there is some valid reason for that, it should be explained here or when making the change, I don't want to get into an edit war --Ngfsmg (talk) 15:28, 18 April 2021 (UTC)

Can the "Extremes and trends" section be updated to a list format?

Hi there, I just edited the section § Extremes and trends in which I added Governorship data, which I now know shouldn't be there, as per previous discussions in this Talk page.

But I had also updated the information into a list type format, making it much easier to follow (see this revision). So I just don't see any drawbacks of updating it to a list format. But that has been reverted by IJBall, with the edit summary "...and don't convert to list-format..." in the subsequent edit. Can you please give your opinions in this matter. Thanks. Cheers! CX Zoom (talk) 20:56, 29 April 2021 (UTC)

I am against converting this section into list-format. It should not be the focus of the article, and putting it in list-format will draw undue attention to it... The question is not between what we have now and "expanding" the current section – the question is actually whether the entire 'Extremes and trends' is too WP:UNDUE and should be removed, or whether it should remain roughly as is. --IJBall (contribs • talk) 21:02, 29 April 2021 (UTC)

New PVI for redistricted districts

Can I add a list of new PVI for redistricted states such as Oregon and Texas? Or should I wait until the whole redistricting process ended?Hong Kong Ian (talk) 10:07, 26 October 2021 (UTC)

It's best to wait until the entire redistricting process is ended before updating the PVIs. Secondly, it will have to be sourced to something. Hopefully the Cook Political Report will publish the update themselves, but if they don't it will need to come from somewhere. Finally, the new figures should simply be put into the original table – we do not need a second table for this. --IJBall (contribs • talk) 12:15, 26 October 2021 (UTC)

I think there should be both sets of PVIs until the next Congress is seated (or at least elected next year). Until then, all existing members will be from the previous district lines, and the PVI to party-membership thing will be completely messed up. Let's also remember that the new districts don't even necessarily remain numbered similarly. The new PVIs are going to be valuable and important information, but you can't jettison the existing PVIs until they are no longer relevant. Gabrielthursday (talk) 17:01, 26 October 2021 (UTC)

In that case, the best suggestion may be to wait until after the 2022 elections to update the table. The figures for the redistricted districts can be mentioned in prose in the meantime, citing the relevant source. --IJBall (contribs • talk) 17:03, 26 October 2021 (UTC)

Perhaps, but I don't see the harm in having both figures available. The new PVIs will also be relevant, since primaries and elections will be fought within them for most of 2022. Right now, the house PVIs are collapsed/hidden in the article, so there isn't much of a downside, imho. Gabrielthursday (talk) 20:32, 26 October 2021 (UTC)