The Thirteen Keys to the Presidency

November 1, 2004 6:37 AM

Professor Allan Lichtman has predicted the results of the past four elections correctly using a system known as The Thirteen Keys to the Presidency.

The Keys predict election results by assessing the performance and strength of the party holding the White House. The thirteen points take into account all the factors that decide elections from the obvious (how the economy is doing) to the more subtle (whether the party in power has achieved major policy change). If eight or more of the keys favour the candidate of the incumbent party, he wins. Any fewer, he loses.

Eighteen months ago Lichtman forecast that Bush would retain the presidency. But the Republican Party now has seven keys turned against it for 2004, one more than the fatal six negative keys.
posted by tapeguy (31 comments total)

i get connection refused on the last link.
posted by Stynxno at 6:56 AM on November 1, 2004

I'm in the former Communist Soviet Union.
The last ink fails to connect to me!
posted by jmccorm at 7:02 AM on November 1, 2004

The dirty little secret of political science is that its theories, particularly highly-specific ones like this, serve only to explain past behavior, not predict future behavior.
posted by waldo at 7:05 AM on November 1, 2004

Yeah, I can't reach the last link either. What's the homepage for the 13 keys site? Google can't seem to dig it up for me.
posted by mathowie at 7:10 AM on November 1, 2004

From the review on the second linked (Amazon) page:

Looking ahead to 2000 has to bring a smile to Democrats. Using the keys, it is hard to envision a scenario other than a Gore victory in November. Whether that's good news or bad news is up to the reader.

Ho ho.
posted by i_cola at 7:15 AM on November 1, 2004

You all are thinking so two dimensional. Have you tried the other internets? I found it on this one.
Praise be to the Internets!
posted by jmccorm at 7:15 AM on November 1, 2004

waldo: The ideas being discussed here were put forward by a historian, and not a political scientist. His thirteen keys are not a theory, either. He has not put forward hypotheses and tested them, to my knowledge.
posted by raysmj at 7:18 AM on November 1, 2004

Of course everybody could wait till tomorrow or next month to find out the actual election results and never mind putting belief in football games, keys or other superstitions.
posted by substrate at 7:21 AM on November 1, 2004

the FPP reads like a synopsis of a dragonball z episode
posted by luckyclone at 7:27 AM on November 1, 2004

Thanks, jmccorm. Note that these are still pretty subjective, for the most part. E.g.:

6. Is the election year a time of recession or depression? No. (Yes. Kerry.) [Bushites will say the recession is over.]

....

9. Was the incumbent administration tainted by major scandal? No. (No. Bush) [Bush-opponents will say that there have been several major scandals; the only argument is whether they've "tainted" the administration. From my perspective, yes; from an opinion polling persepctive, I'm not sure.]

11. Did it achieve a major success in foreign or military policy? Yes. (No. Kerry.) ["Policy" is the key word, here. The initial combat phase in Iraq is clearly perceived as a success by most Americans -- though it's becoming clear in retrospect that it was a "success" much as Pearl Harbor was a "success" for the Japanese: They failed both to destroy our fuel reserves and our most valuable naval assets, the carrier force. And most glaringly, not because they failed to understand the important of either. Similarly, thanks to a combination of slavish adherence to Transformationist dogma and military willingness to compartmentalize and succeed at the proximate, tactical mission in spite of bad planning and strategic idiocy, we failed to enforce a rule of law after deposing the enforcers of law, thus seeding chaos in Iraq; and we failed to keep weapons out of the hands of insurgents.]

12. Is the incumbent party candidate charismatic or a national hero? Yes. (No. Kerry.) [Clearly, Bush is charismatic, to a large group of people.]
....

posted by lodurr at 7:31 AM on November 1, 2004

I have a prediction:
If Bush gets more votes, Bush will win.
If Kerry gets more votes, Bush will win.
posted by Outlawyr at 7:33 AM on November 1, 2004

If Bush gets more votes, Bush will win.
If Kerry gets more votes, Bush will win.

If only.
posted by Eamon at 7:38 AM on November 1, 2004

Someone's gonna get more votes, and I predict the one that does, will win. Unless...
posted by kamylyon at 7:46 AM on November 1, 2004

For the next 36 hours or so, we're stuck with these little parlor games, aren't we.
posted by chicobangs at 7:52 AM on November 1, 2004

[obligatory: "what do you mean, 36 hours?"]
posted by lodurr at 8:06 AM on November 1, 2004

Has anyone tried shaking a magic eight ball yet? I heard those are pretty accurate.
posted by drezdn at 8:11 AM on November 1, 2004

when the polls close, lodurr.
posted by amberglow at 8:14 AM on November 1, 2004

amberglow: I think lodurr is making a little joke about the 2000 election and the general chances of shenanigans.

Btw, this guy's theories seem to make about as much sense as that Redskins loss thing from yesterday. Honestly, every one of these is crazy subjective.
posted by malphigian at 8:19 AM on November 1, 2004

never mind putting belief in football games, keys or other superstitions.

Redskins game: Kerry
Weekly Reader poll: Bush
Higher skirts: Kerry
Halloween Masks: Bush
Fuller Head of Hair: Kerry
Taller Candidate: Kerry

Anything I'm forgetting?
posted by Secret Life of Gravy at 8:33 AM on November 1, 2004

Don't forget the Family Circle first lady cookie recipe challenge, which went to Laura Bush.

All that stuff is silly though, I'm going for reading entrails -- anyone want to donate some?
posted by malphigian at 8:50 AM on November 1, 2004

Anything I'm forgetting?

Nickelodeon Poll of Children: Kerry
No sitting senator has ever beaten an incumbent president in the popular vote: Bush
No incumbent president having lost the popular vote has ever been re-elected: Kerry

I have to say that I like the "keys to the presidency" model much better than the econometric models-- not because they seem to favor my partisan biases, right now, but because the "keys" seem more focused on understanding the political landscape, rather than simply looking at economic indicators.
posted by deanc at 8:56 AM on November 1, 2004

The analysis in the final link is subjective and lame. Even so, I do believe Kerry will win in a landslide. Not because of this, though.
posted by Slagman at 9:15 AM on November 1, 2004

Secret Life of Gravy - Kerry will win because he has more royal blood than Bush.
posted by junkbox at 9:28 AM on November 1, 2004

Skirts are higher?

Slagman: "Landslide" is probably too strong; but I also don't think it will be nearly as close as it looks to Reuters.
posted by lodurr at 9:34 AM on November 1, 2004

6. Is the election year a time of recession or depression? No. (Yes. Kerry.)

Oops. We're not currently in a "recession" or "depression."

10. Did the incumbent administration suffer a major setback in foreign or military policy? No. (Yes. Kerry.)

Frankly, only partisans have a strong sense of this. I don't think swing voters nessarily agree that there was a "major setback."

11. Did it achieve a major success in foreign or military policy? Yes. (No. Kerry.)

This one is probably closer to true, but Afghanistan and the invasion, not occupation, of Iraq could both be seen as major military successes. Again, debatable. Even the author of the column recognizes that many will disagree with him, and goes so far as to spell out his reasoning for answering 10 and 11 as he did.

The Keys are a useful tool for framing the dicussion, but hardly constitute a rigorous prediction model for Presidential elections.
posted by mikewas at 9:54 AM on November 1, 2004

Oh come one there are only 2 possible mutually exclusive outcomes

1. Bush wins
2. Kerry wins

The prof has got a phat chance of winning ..I invoke the spirit of Nostradamus over us !
posted by elpapacito at 10:04 AM on November 1, 2004

Voters who live near walmart prefer bush
Voters who live near starbucks prefer kerry
posted by grateful at 11:48 AM on November 1, 2004

I wonder what prognosticators think the chances of this nov. 3rd thing are?
posted by amberglow at 12:06 PM on November 1, 2004

I live between a Walmart and a Starbucks! I don't know who to choose!

Ah well, voting's not cool anyway. I think I'll stay home and get trashed on Listerine.
posted by billybunny at 2:29 PM on November 1, 2004

mathowie: My apologies. You can't find the Keys to the White House site I used to run because I let all my sites go dark.

Keys to the White House, as of late 2001. As you can see, I've been invested in this prediction system for some time; in fact, I published a prediction online for the 1992 election. The system, as interpreted separately by me and Dr. Lichtman, was correct in 1988, 1992, and 1996. (Lichtman also predicted 1984, but it wasn't published so far as I know.) It predicted a win for Al Gore in 2000, but because Lichtman had to design the system to account for the "stolen" elections of the 19th century, he chose to weight it by whether it predicted the popular vote winner, arguably an error. The system was borne out, but the prediction was, you know, wrong!

On my discussion site for the Keys, I published concurrent predictions based on each Key as it fell, as well as some analysis of individual Keys and their inter-relationship or predictive power.

I'm sorry to say that the Christian Schock analysis is erroneous (actually, I think he's using an earlier formulation of one key). Note that Lichtman's own prediction using his system sets just five keys, at most, against the Republicans (do a find on "shrum"). I agree more with Lichtman's analysis than Schock's.

The Republican (incumbent) party has all four of the first keys, the "political" keys. They won a mandate in the midterm elections; there was no nomination contest; the incumbent is running; and there is no third party candidate from the incumbent faction. It is important to note that the Keys were derived from an historical database of elections going back to the Civil War, and no incumbent party has lost the White House when all four political keys were in its favor. It's not clear that social discontent, measured by our era of poisoned political discourse, is up to the level of previous eras of unrest (demonstrations, riots, assassinations), but I'll allow that key provisionally. The GDP growth key was lost from the beginning simply because Clinton's eight years were so go-go. Thus at most six keys are turned against, seven in favor. Without the unrest key, though, the incumbent party has eight, which favors victory for the incumbent party. With the policy key, the GOP has nine. Personally, I favor giving them the unrest key and holding back the policy key. Still, that leaves them with a victory under the Keys system.

Note that I would much prefer that the Keys system I've touted for many years be wrong and Kerry win, but I have to call the keys the way I see them.

mikewas, I agree that we're not in recession and don't tip that key. But 9/11 was the worst attack on American soil since Pearl Harbor; by definition, it is a foreign policy or military failure. Iraq may be a quagmire or may not, but we only need one major failure to tip the key. The same goes for the success key, and Afghanistan fits the bill even if Iraq doesn't.

lodurr, malphigian: Lichtman knows and agrees that the criteria are subjective. He was attempting to design a system that was more robust than the mathematical and econometric models such as Ray Fair's, which predict the win based on stock market prices or unemployment figures. There are more of them online these days and they disagree wildly. Most of them were wrong about 2000. Subjectivity is a hazard of the system as presently designed, but then, the entire purpose may not be objective to begin with.

raysmj, chicobangs has it right. This doesn't rise to the level of theory; I have always said it is more of a parlor game. For Lichtman, the purpose is to test the predictive power of certain lemmas and see what they teach us. One of the most important things we can learn from the Keys system is that the thirteen testable propositions, which were mathetmatically ranked with scores of others by Lichtman and a Russian scientist skilled in predicting volcanic eruptions, are almost universally about the performance of the incumbent party. Often a poli-sci truism is that incumbent candidates -- like George Bush -- are subject to a referendum on their performance, but not same-party candidates such as Al Gore. That's not true, it turns out. The system also teaches us that what turns an election generally takes place long before the campaign -- fundamentals such as the economy and war. The zigzags and poll movements are noise and little more. It's disheartening for challenging parties to think that the only way they can improve their chances is to choose a candidate of unusual charisma (I still think Clinton had it, but Lichtman disagrees; the last he pegs with the key is Reagan) or national hero status (e.g. Ike), but there are opportunities to chip away at the keys outside the election cycle. It's more important to realize that the electorate is forgiving and unforgiving in different ways, and the best way to win the White House is to do well by the country while you have it.

By the way, Dr. Lichtman was the expert hired by the US Commission on Voting Rights to examine the Florida debacle; today he's on the board of VoteWatch.
posted by dhartung at 8:32 PM on November 1, 2004

dhartung - It seems to me as if there's a whole lot of subjectivity involved here - and with both assessments.

The determination of the "keys" seems a bit shaky to me.

Of course, in a few hours......
posted by troutfishing at 11:18 AM on November 2, 2004