Swing State Project

« Quick Poll Roundup: AZ, IA, PA | Main | MN: Slight Kerry Lead »

Saturday, May 29, 2004

All State Voting Trends

Posted by Chris Bowers

On Friday, on a day off, I spent literally the entire day producing this chart that shows the partisan index in every state since 1976 (after clicking the link, you will need to scroll down the page to see the chart). I think it reveals at least two particularly interesting pieces of information.

First, as a commenter in the thread noticed, there in fact does appear to be an ���Emerging Democratic Majority.��� Over the past seven election cycles, states with a combined electoral total of 233 are trending Democrat, while states with a combined electoral total of 188 are trending Republican. It would appear that good times are ahead.

Second, Perot clearly did not cost Bush the 1992 election. The partisan index measures the degree to which a state favors a party relative to the way the rest of the nation favors that party. This being the case, it would follow that if more typically GOP partisans had indeed swung to Perot than had typically Democratic partisans, the 1992 partisan index would reveal and anomalous pro-DNC swing due to a temporarily eroded Republican base.

However, only a handful of states that Clinton won show such trends. Perot definitely seems to have caused Bush to lose Georgia, as the usually double-digit pro-GOP partisan index in that state cratered at +5.0 GOP in 1992. The same goes for Nevada, which relatively favored the GOP by 13.2 in 1988 and 7.5 in 1996, but only by 2.9 in 1992. I���ll grant that without Perot, Bush probably wins both states.

Looking at the chart, however, only Colorado, Kentucky, Maine, Montana, New Hampshire and Tennessee are other possible states that Perot swung to Clinton. Still, even if Bush had won all of these states as well as Georgia and Nevada, Clinton swould have won the Electoral College 315-223. Further, there is no conclusive evidence that Perot actually cost Bush any of these other six states:

While Colorado had a PI of GOP +9.9 in 1996, far different from 1992���s PI of GOP +1.4, in 1988 the PI in Colorado was only GOP +0.1. It would appear that Colorado didn���t like Bush I either time he was the GOP nominee.

While Kentucky had a PI of GOP +7.6 in 1996, far different from 1992���s GOP +2.5, in 1984 Kentucky also had a PI of GOP +2.5. In 1988, Kentucky���s index was GOP +3.9.

Maine? Don���t even think about Maine. Bush finished third in every single county in Maine in 1992. If anything, Bush cost Perot Maine in 1992.

While Montana had an index of GOP +11.4 in 1996, far different from 1992���s GOP +3.1, in 1988 Montana actually was actually DNC +1.9. Like Kentucky, their pro-GOP swing did not seem to start in full-force until the 1996 cycle.

While New Hampshire���s 1988 index of +18.4 GOP in 1988 is shockingly different than 1992���s GOP +4.3, in 1996 the partisan index in NH was DNC +1.3. By 1992, New Hampshire was trending Democratic anyway.

Tennessee definitely cratered for the GOP in 1992, but having then Senator Gore, who at the time ws still a "real Southerner" on the ticket might just as well have been the cause for the pro-DNC swing as Perot.

Of course, like I already noted, even if I am wrong about all of these states, that means Clinton would still have won 315-223. No other state shows evidence of Perot costing Bush victory. Perot did not cost Bush the 1992 election--not even close. That is one popular myth that can be put to bed.

Posted at 06:38 PM in General | Technorati

Comments

Fantastic post, Chris. Nice to see the Empirical Left drive a stake through the heart of that irritating myth.

Posted by: DavidNYC at May 29, 2004 08:34 PM | Permalink | Edit Comment | Delete Comment

People are Republican based on myths. Here are a couple of more:
1). The ONLY way to balance the Federal Budget is by a Constitutional Amendment.
2).Clinton is responsible for the WTC bombings in 1993 despite having been in office for only 3 weeks. But W. is not in any way responsible for 9/11 even though he was in office for 8 months.
3). If you wake up in the morning with a hangnail it is the fault of Bill Clinton or Big Government.
4). Cutting taxes on the wealthy will bring in more revenue to the Federal Government.
5). Iraq had Weapons of Mass Destruction even though no evidence has been found for any.
6). Iraq had something to do with 9/11 even though there is no evidence of any.

I could go on but you get my drift.

Posted by: Alan Snipes at May 30, 2004 10:25 AM | Permalink | Edit Comment | Delete Comment

Hey Chris
Really, really informative. Thanks for taking the time to research and prepare the chart.You know, as a suggestion, it would be interesting seeing how the states changed their trend during and after the civil rights movement. While I realize it's probably off topic it would sure go a long way in explaining the "Republican" majority. It's just a thought. Thanks for your site. I look forward to your e-mail every day.

Posted by: Joan at May 30, 2004 11:56 AM | Permalink | Edit Comment | Delete Comment

Does the Rasmussen Poll have any validity? It seems without fail that Kerry will lead for 3 days only, and then is tied or behind. Plus,
bush's approval rating seems inflated. Is my suspicion on target - I know this guy works for fox.

Posted by: Kath at May 30, 2004 06:39 PM | Permalink | Edit Comment | Delete Comment

I agree, Rasmussen seems rather suspicious. What is his methodology? Two interesting things about Rasmussen, he's consistently shown the Dems ahead in the Congress vote and he's published many polls showing Kerry ahead in battleground states. And, Rasmussen is suppossedly conservative, so this bodes well for both Kerry and the Dems looking to take back Congress. That might be the sleeper story this fall, that the Dems are poised to retake Congress after a decade in the wilderness.

Posted by: Rock_nj at May 30, 2004 10:40 PM | Permalink | Edit Comment | Delete Comment

What I liked most about your breakdown is that it shows pretty conclusively that the cockpit of Presidential elections - where things are most volatile - tend to be along the Mississippi. Lots of folks could mention the relative volatility of Louisiana, Arkansas (in large part due to Clinton, I suspect) and Missouri, but to see Minnesota and Iowa also on the list (along with Wisconsin) is an eye-opener. I realize a lot of factors are in play in each state, but an analysis of this region would be quite illuminating.

Yes, Oregon (the biggest surprise for me over the last year) and Pennsylvania and even Ohio to some extent (your data tells me loud and clear that if we ever got off our butts in Ohio, it would be much more often in the Dem column for Presidential elections) are interesting to look at. But the Greater Mississippi Valley is where the action is.

Posted by: Palamedes at May 31, 2004 02:10 PM | Permalink | Edit Comment | Delete Comment

Just felt like following up to the "Perot Myth". Perot quit the race the week of the Democratic convention, blamed the nation's problems on Bush, and mentioned that he was leaving because the Democrats had renewed themselves. Then all those Perotistas sat down, watched the Clinton sales job on TV, and largely went for him. Those who went for Bush, of course, were largely won over by Clinton as well, as that support was softer. Even if you think that Perot didn't bring Clinton electoral victory, I don't think this bit of theater is irrelevant to the eventual outcome.

Posted by: senator_shakedown at August 2, 2004 10:14 PM | Permalink | Edit Comment | Delete Comment

Then all those Perotistas sat down, watched the Clinton sales job on TV, and largely went for him.

The exit polls don't support that statement. They show that people went a third Bush/a third Clinton/a third home or someone else. Sure, people like to exit pollsters, but it's the best info we've got.

Posted by: DavidNYC at August 2, 2004 10:31 PM | Permalink | Edit Comment | Delete Comment

The polls I saw did support that statement, but that was a while ago and neither of us have citations. I will promise to look if you promise to as well.

Posted by: senator_shakedown at August 2, 2004 10:58 PM | Permalink | Edit Comment | Delete Comment

ELECTLINE - STATE PRESIDENTIAL REVISED LINES

To those unfamiliar with the former ElectNet.org election handicapping system, each race projects an under/over point spread between the first and second place finishers in each race. It works just like football and basketball point spreads. In a two man race, a ���-4��� , for example, means I am projecting a 52-48 final result.

The system depends very little on polls, but heavily on previous results, demographics, voter registration, and miscellaneous details relevant to the candidates, issues, and the region. Incumbency is a relatively heavy factor in U.S. House races, but less for Presidential, U.S. Senate, and gubernatorial contests.

The following describes each state���s projected result for the presidential election. The states will proceed alphabetically, and will give a projection and a short analysis.

ALABAMA

6-26: Bush by 19.

NEW: Bush by 19.5.

Bush won by 14.9 last time. Nothing much indicates that margin will shrink.

ALASKA

6-26: Bush by 28

NEW: Bush by 27

Bush won by 31 last time.. Close Senate race may help Democrat turnout.

ARIZONA

6-26: Bush by 8

NOW: Bush by 8.

Bush by 6.3 in 2000. Bush has superior area pull. With McCain working hard, most indicators show Bush pulling away.

ARKANSAS

6-26: Bush by 6.5.

NOW: Bush by 5.5.

Bush by 5.5 last time. After rightward lurch in the 1990s, the state has been returning to it���s Dem roots recently. Easy win for Sen. Blanche Lincoln (D)
may stem Bush tide.

CALIFORNIA

6-26: Kerry by 8.5.

NOW: Kerry by 9.

Gore by 11.8 in 2000. Disastrous Davis administration gave GOP a shot if Governor Arnold remains popular. Kerry hasn���t blown this open and Bush would love to force Dems to work to nail this one down.

COLORADO

6-26: Bush by 7.5

NOW: Bush by 7.

Bush by 8.4 last time. Most issues a null factor here and area pull helps incumbent. Close open Senate race may pull either way.

CONNECTICUT

6-26: Kerry by 15

NOW: Kerry by 14.5.

Joe Lieberman is off the ticket, but neighbor Kerry is on. Nobody seems to know Bush was born in New Haven. Gore won by 17.5 in 2000. In this high income state, the tax cuts are popular, but Bush���s foreign policy and social programs aren���t.

DELAWARE

6-26: Kerry by 9

NOW: Kerry by 10.

Gore won in 2000 by 13.1. However, tax cuts and growth policies popular. Regional pull a factor. Seems conceded by GOP.

DISTRICT OF COLUMBIA

6-26: Kerry by 73

NOW: Kerry by 75.

Wall to wall Dem district went for Gore by 76.2 in 2000. Nader has a better shot here than Bush.

FLORIDA

6-26:Bush by 3.5

NOW: Bush by 2.

Jeb Bush is still popular and most issues favor the incumbent. This has great symbolic worth for Dems, even though Bush no longer needs it as much.

GEORGIA

6-26:Bush by 13.5

NOW: Bush by 13.

Bush won by 11.7 in 2000. He is now stronger on all issues, especially religious/social. Easy Senate pickup won���t hurt.

HAWAII

6-26: Kerry by 16

NOW: Kerry by 16

State has rebelled against far left governance lately but is in Dem camp. Gore won by 18.2. No action here this time.

IDAHO

6-26: Bush by 40

NOW: Bush by 38.5

Bush won by 39.5 in 2000 in this most Republican state. Hasn���t gone Democrat since 1964.

ILLINOIS

6-26: Kerry by 10.5

NOW: Kerry by 12.

Gore in 2000 won by 12. Union vote still a factor here. Reagan Democrats might be more sympathetic to Bush than brahmin Kerry, but but totally disorganized GOP hasn���t even guaranteed Bush a ballot spot yet.

INDIANA

6-26: Bush by 15

NOW: Bush by 15

Bush won by 15.6 last time out. Rust Belt recession could have worked against him. Kerry has conceded state as recovery has bouyed GOP.

IOWA

6-26: Bush by 2

NOW: Bush by 1

Gore won by 0.3 in 2000. Kerry not a good fit here, but state flips between boilerplate left and hard right. If economy improves, Bush should survive.
Kerry treating state as must win.

KANSAS

6-26: Bush by 22

NOW: Bush by 21

Seems to be coming in at the same rate as 2000. Not much motivation here either way

KENTUCKY

6-26: Bush by 13

NOW: Bush by 11.5.

Dems would love to make a run here, especially with Sen. Jim Bunning in
minor trouble. Won���t happen.

LOUISIANA

6-26: Bush by 8.5

NOW: Bush by 9.

Next to Florida, the strongest D state in the South. Dems clinging to power in many state races. Hard to see Kerry doing better than Gore, though, who lost by 7.7. A must win for Bush.

MAINE

6-26: Kerry by 4

NOW: Kerry by 3.5

State lurched left in 1990s, but Dem Gov. Balducci���s numbers are collaping. However, regional draw, unions, and greens should lift Kerry. Gore won by 5. Like Nebraska, Maine has a Congrsssional Electoral system. so Bush could scrape 1 Elector even if Kerry wins a close one.

MARYLAND

6-26: Kerry by 13.5

NOW: Kerry by 12.5

State threw Dems out of Governor���s mansion two years ago, but regionalism and social issues break for Kerry. Gore won by 16.4. Kerry seems to be having some problems locking this up.

MASSACHUSETTS

6-26: Kerry by 28

NOW: Kerry by 28

Republican control of the Governor���s mansion, plus gay marriage backlash may keep final respectable. Gore won by 27.3. Home state sentiment boosts Kerry. Convention in Boston may not have helped Kerry.

MICHIGAN

6-26: Kerry by 1.5

NOW: Kerry by 3.

Battleground state went Gore by 5.1. Reagan Democrats may lean Bush on security issues, but Rust Belt pull may favor Kerry unless job picture continues to brighten. Iraq seems to be helping Kerry.

MINNESOTA

6-26: Kerry by 1

Now: Kerry by 1.5

State went Gore by 2.4 in 2000. MN not a reliable D state, but has voted with the Dems every year since 1972. Dem-Farmer-Labor knows how to win here, despite recent erosion. Hardliners will pull out close one.

MISSISSIPPI

6-26 Bush by 18.5

NOW: Bush by 21.

Went Bush by 16.9 last time. Cultural swing should boost that. Dems giving Miss absolutely nothing to vote for in the cultural war.

MISSOURI

6-26 Bush by 6

NOW: Bush by 4.5

Bush won 2000 by 3.3 Ashcroft promotion probably helped. Mizzou hasn���t been wrong on a candidate since Nixon in 1960. Probably the only
must win state for both parties.

MONTANA

6-26: Bush by 26

NOW: Bush by 26

Bush won in 2000 by 25.1. Strong GOP state.Cowboy country.

NEBRASKA

6-26: Bush by 30.5

NOW: Bush by 29.5

Reliable R state went Bush by 29 last time. Issues indicate no change. GOP has won here nine straight elections.

NEVADA

6-26: Bush by 4.5

NOW: Bush by 3

Independent state. Large Lib faction with hard left, green, and Christian Right influence. Went Bush by 3.5 in 2000. Kerry made the politically correct, if illogical, promise to stop using Nevada for centralized nuclear waste storage. This has helped him.

NEW HAMPSHIRE

6-26: Bush by 2.5

NOW: Bush by 1.

Free State Project may help Bush slightly by emphasizing low tax agenda. Kerry regional pull keeps this close. 2000 went Bush by 1.3. Another Florida?

NEW JERSEY

6-26: Kerry by 12

NOW: Kerry by 9

Went Gore by 15.8 in 2000, but Dem popularity is plunging. Regional pull plus social issues should keep,this Dem. Kerry unable to pull away and would hate to have to spend money here.

NEW MEXICO

6-26: Bush by 4

NOW: Bush by 1.5

Bush must be seen as small favorite in state which went to Gore by only 0.1. Polls (which I discount at this stage), show Kerry big. I���m cautious, but this is still in play.

NEW YORK

6-26: Kerry by 19

NOW: Kerry by 20.5

Went for Gore by 25. Not much has changed, although 9-11 made Bush a more sympathetic candidate. With easy win, Dems might not be motivated.

NORTH CAROLINA

6-26: Bush by 15.5

NOW: Bush by 12

Safely in R camp. Went Bush by 12.8 in 2000. Edwards seemed to have supplied some bump, but no real Dem play here.

NORTH DAKOTA

6-26: Bush by 29

NOW: Bush by 29

Dems win statewide fairly frequently but goes GOP nationally. Went Bush by 26.6 in 2000. If this state is below a 25 point margin for Bush, it means he���s in trouble nationally.

OHIO

6-26: Bush by 3

NOW: Bush by 3

Rust Belt recovery better pick up speed or this normally GOP state remains shaky. It���s gone with the winner since 1960 and Republican 8 of the last 12. Bush won 2000 by 3.5. Must win for Bush

OKLAHOMA

6-26: Bush by 23.5

NOW: Bush by 24.5

Despite upset win for Governor by Dems in 2002, GOP leanings remain strong. Over the border pull adds to landslide. In 2002. Margin seems to be growing regardless of any outside occurances.

OREGON

6-26: Bush by 2

NOW: Kerry by 1

Gore won 2000 by 0.4. Kerry���s tax policy and New England liberalism don���t play well here. Social issues plus Green backing of Kerry seem to have moved him ahead.

PENNSYLVANIA

6-26: Kerry by 1

NOW: Kerry by 2.5

Absolute necessity for Kerry to win to maintain viability. Bush���s ill-advised steel tariffs may have helped him here, but needs added economic good news to pull it out. Spector primary win might actually hurt Bush by depressing the base vote.

RHODE ISLAND

6-26: Kerry by 27

NOW: Kerry by 27

Everything breaks for Kerry here. Over the border pull, Gore took by 29.1, and Teddy Kennedy���s son is a congressman. Anything less than a 25 point spread means Kerry���s in trouble nationwide..

SOUTH CAROLINA

6-26: Bush by 18

NOW: Bush by 16.5

Solid south GOP. Despite outsourcing jobs controversy, other issues run right. SC-born Edwards��� pick apparently helped Kerry a little here.

SOUTH DAKOTA

6-26: Bush by 23

NOW: Bush by 20.5

Safe for Bush, but Daschle���s reelection seems to have motivated Dems. Bush by 22.8 last time. If Bush slips below 20, it may mean he���s in national trouble.

TENNESSEE

6-26: Bush by 7

NOW: Bush by 6

Bush won by 3.9 in 2000. If Gore couldn���t win here, Kerry is long-shot. Bush doesn���t want to spend money here, but may have to.

TEXAS

6-26: Bush by 23

NOW: Bush by 22.5

Texas Dems are feisty but facing extinction. Home ground fertile for Bush.
In 2000, Bush carried by 21.3. No argument here.

UTAH

6-26: Bush by 41.5

NOW: Bush by 42

All issues break for Bush, who took this state by 42.5 last time. Social issues especially should play big. The most Republican state.

VERMONT

6-26: Kerry by 8

NOW: Kerry by 9.5

Dems would have preferred Dean, but Kerry is a neighbor. Gore won by 9.9 in 2000. Dean has been behaving himself and doing his job.

VIRGINIA

6-26: Bush by 9.5

NOW: Bush by 7.5

Much anti-tax feeling here and Bush should glide, but doesn���t appear to be. Bush won 2000 by 8.0. If this gets closer, watch out.

WASHINGTON

6-26: Kerry by 1.5

NOW: Kerry by 2.5

Activist Dems would have preferred Dean. Kerry a bit too traditional a liberal here. Another must-win for Dems. They need to put this one away quickly before Bush counterattacks.

WEST VIRGINIA

6-26: Bush by 6

NOW: Bush by 4

Kerry will have to run away from his gun control votes here, although strong union ties will work for him. Bush won 2000 by 6.4. Bush can afford to lose here, but would prefer not to.

WISCONSIN

6-26: Bush by 2

NOW: Bush by 2

One state where the minors hurt the GOP last time as Gore won by 0.2.
Another must-win for Kerry. Big surge in GOP registration should turn this.

WYOMING

6-26: Bush by 42

NOW: Bush by 41

It���ll be hard for Bush to surpass 2000 win by 40.1, but he may. Cheney���s roots don���t hurt. Minors nullified each other last time with each side taking 2.3.

This projects to Bush 300 to Kerry 238. However, 94 of Bush���s electoral votes and 70 of Kerry���s are in states projected as closer than 5 points. Minor swings in fortune could elect either man.

Posted by: tfagan at August 2, 2004 10:58 PM | Permalink | Edit Comment | Delete Comment

Washington Post, November 12, 1992:

The article describes the definitive "nationwide exit poll of more than 15,000 voters conducted by Voter Research & Surveys for the major news organizations." When speaking of exit polls from 1992, this is the only one I'm really aware of that people point to as being authoritative. There might have been smaller, state-by-state exit polls, but VNS was still the big cheese back then.

"In the presidential contest, 38 percent of Perot supporters said they would have supported Clinton if Perot had not been on the ballot and 37 percent said they would have supported Bush.

An additional 6 percent of Perot voters said they would have sought another third-party candidate, while 14 percent said they would not have voted if Perot had not run."

I grant that the "equal thirds" idea is a little off the mark, but Bush pere and Clinton did receive equal support, which is the relevant point.

Posted by: DavidNYC at August 2, 2004 11:30 PM | Permalink | Edit Comment | Delete Comment

I can't remember my point anymore after that giant handicapping dump. The contest panel gives you the edge, including the traditionally contrarian Soviet judge. But my larger goal, that of making people post giant, inscrutable handicapping dumps, has been achieved.

Posted by: senator_shakedown at August 2, 2004 11:42 PM | Permalink | Edit Comment | Delete Comment