I wrote out this post yesterday, but it got so long that I've decided to split it into two posts on separate topics.
1. Electoral Indicators
Allan Lichtman's System Has Been Wrong Twice Since 1984
The 13 Keys to the White House
I've previously said there's a good chance that Allan Lichtman's keys will be wrong in 2024. However I want to explain a bit more about this and how they have in fact already been wrong, although he seems loathed to admit it. Lichtman has his predictors, but even he admits they aren't the
only predictors and they're
not individually weighted. They are weighted by
stealth, eg there are two Keys for foreign policy and two Keys for performance of the economy. So he does in fact weight certain things in a certain way, but only to the extent that it provides additional keys on effectively the same overall topic. In his original
co-authored paper (November 1981) describing his system gave the following disclaimer:
“We neither claim that other parameters cannot be used for the same purpose nor suggest methods for predicting future elections.” There are two points to make here, they were clearly saying
these aren't the only indicators, and secondly it's a statement that in 1981
the system was not designed to be predictive. You would not know this now because the way he talks about his system is one of electoral forecasting, and he makes out as if that's always been the case. Further to this, if you read the Paper there are only 12 Keys described, not 13, and there are some substantial differences. I will list out the differences so you can see clearly for yourself:
12 Keys to the White House (1981):
1. Has the incumbent party been in office more than a single term?
2. Did the incumbent party gain more than 50% of the vote cast in the previous election?*
3. Was there major third party activity during the election year?
4. Was there a serious contest for the nomination of the incumbent party candidate?
5. Was the incumbent party candidate the sitting president?
6. Was the election year a time of recession or depression?
7. Was the yearly mean per capita rate of growth in real gross national product during the incumbent administration
equal to or greater than the mean rate in the previous 8 years and equal to or greater than 1%?
8. Did the incumbent president initiate major changes in national policy?
9. Was there major social unrest in the nation during the incumbent administration?
10. Was the incumbent administration tainted by major scandal?
11. Is the incumbent party candidate charismatic or a national hero?
12. Is the challenging party candidate charismatic or a national hero?
13 Keys to the White House (1988-2024):
1. Party mandate: After the midterm elections, the incumbent party holds more seats in the U.S. House of Representatives than after the previous midterm elections.
2. No primary contest: There is no serious contest for the incumbent party nomination.
3. Incumbent seeking re-election: The incumbent party candidate is the sitting president.
4. No third party: There is no significant third party or independent campaign.
5. Strong short-term economy: The economy is not in recession during the election campaign.
6. Strong long-term economy: Real per capita economic growth during the term equals or exceeds mean growth during the previous two terms.
7. Major policy change: The incumbent administration effects major changes in national policy.
8. No social unrest: There is no sustained social unrest during the term.
9. No scandal: The incumbent administration is untainted by major scandal.
10. No foreign or military failure: The incumbent administration suffers no major failure in foreign or military affairs.
11. Major foreign or military success: The incumbent administration achieves a major success in foreign or military affairs.
12. Charismatic incumbent: The incumbent party candidate is charismatic or a national hero.
13. Uncharismatic challenger: The challenging party candidate is not charismatic or a national hero.
In fairness the wording of the Keys has not changed since 1988, and possibly 1984 (the earliest version of
his book I was able to find for reference is from 1990).
“Prediction rates” (from the 2024 Edition aka
correlative ):
Key 1 (Party mandate) 66%
Key 2 (Nomination contest) 87%
Key 3 (Incumbency) 68%
Key 4 (Third party) 68%
Key 5 (Short-term economy) 82%
Key 6 (Long-term economy) 68%
Key 7 (Policy change) 68%
Key 8 (Social unrest) 66%
Key 9 (Scandal) 63 64%
Key 10 (Foreign/military failure) 68%
Key 11 (Foreign/military success) 71%
Key 12 (Incumbent charisma) 55%
Key 13 (Challenger charisma) 71%
In 1981 he did not award a single Key for foreign/military policy, later he added
two keys for that. The single-term key has been removed entirely, so effectively an incumbent seeking re-election originally was awarded three Keys automatically, now it's only two. Long-term economic growth has been re-defined. So much for his claims that his system has correctly predicted every US Presidential election since 1984 unchanged. The other thing I would note here is that the Keys are only
correlations, not
causations. There are more predictors that have even greater correlation and therefore greater predictive power than almost any one of the Keys does individually.
Performance of the S&P 500 in the final 3 months leading up to Election Day has a greater than 80% correlation with predicting the winner. That's significantly higher than most of the Lichtman Keys, one wonders why he doesn't include it. Key 12 is so boarder-line in correlation that one wonders why he thinks it has any true predictive power.
What other predictors/indicators are left out? Well things that affect an individual electoral cycle would be one. If an event happens that no one could have predicted, the Keys don't move. The Covid Pandemic, the Florida butterfly ballots as examples. Another is thing I consider that he doesn't is how united the opposition Party is behind their pick. Trump did not contend the Primaries, he took part in no Republican 2024 Primary debates whatsoever and he still won the nomination. Lichtman's system does not award a Key for that. In this election we have the unprecedented situation where the candidate selected by the party base in the primaries has been replaced after the primary elections. My view is that the Primary Process matters for
both parties as electoral outcome indicators, and usually it's about how badly your own candidate gets damaged by your own party - think Clinton 2016. Perhaps most importantly of all is
other historical trends with equal or greater correlation to election outcomes that he does not consider. I am also of the view that 40-50 year trends have more predictive power than 230 year trends that “go back to Federation”.
Lichtman has a good system, but not an infallible one. I'm far more in agreement with his way of looking at US Presidential Elections compared to over-glorified pollsters like Nate Silver. We shall see the precise effect of the “incumbency key” in November. Harris only has one viable path to victory and that is improving the performance of the current White House Administration. If she goes over to Israel and negotiates a ceasefire between Hamas and Isreal
right now, then she will win in November.
The other area we agree on, but differ is on the electoral history. He thinks the systems holds true back to Federation - no national electoral predictive system can possibly last that long. I'm sorry, but that's bullshit. History does however matter a LOT and I would look at the past 40-50 years of electoral history, not the past 230 years. Yes
some electoral indicators may go back that far, but not everything. I also think perceptions of the economy matter more than whether you're in a technical recession or not. High inflation, a cost of living crisis, and a large spike in unemployment are all bad indicators for the incumbent party - maybe not to the same extent as one of the Lichtman keys individually, but put together they would be.
Now as for his system predicting a Trump victory in 2016:
bullshit it did!
Let's first take a look at what his
2016 edition book says:
“The keys to the White House focus on national concerns such as economic performance, policy initiatives, social unrest, presidential scandal, and successes and failures in foreign affairs. Thus, they predict only the national popular vote and not the vote within individual states. Indeed, no system could have predicted the 537 vote margin for George W. Bush in Florida that decided the 2000 election. In three elections since 1860, where the popular vote diverged from the electoral college tally—1876 (when Democrat Samuel J. Tilden won the popular vote, but lost in the electoral college to Republican Rutherford B. Hayes), 1888, and 2000—the keys accurately predicted the popular vote winner. Based on the historical odds since 1860, the chances are better than twelve to one that the popular and electoral college vote will converge in any given election. However, these odds presume continuity over time in the relationship between popular and electoral college votes. Some analysts have suggested, however, that this relationship may have changed given the sharp division in America between Republican ‘red states’ and Democratic ‘blue states.’” (p.xi, emphasis added). I should note that this is exactly the same in the
2020 edition and in the
2024 edition as well - it hasn't changed.
It only predicts the
popular vote, not the Electoral College. He kept saying this over and over again in 2016. His initial prediction is
here and there are many interviews with him adding caveats including that he's only predicting the popular vote not Electoral College such as
here. Furthermore he incorrectly awarded the third party Key to Trump. He repeatedly said that the Third Party Key only falls if the Third Party Candidate is anticipated to get 5% of the vote or more - on their own. You don't add up two third-party candidates together. Well prospectively it may have looked like Johnson would get 5% of the vote, but retrospectively looking back - he did not! He only got
3.28% of the vote, and even adding Stein's 1.07% does not add up to 5% either. But you need not believe me, let's read his book for ourselves:
“For upcoming elections, any candidate who appears likely to win 5 percent or more of the popular vote is a ‘major’ third-party contender. If, however, a third-party candidacy clearly resulted from a split within the challenging party, it would not topple key 4. Although no such candidacy has emerged in past elections, some candidates, such as George Wallace in 1968, may actually have drawn more votes from the challenging than from the incumbent party.
“Key 4 is not turned against the party in power when several of the perennial third parties together have garnered more than 5 percent of the vote. This occurred in the 1904, 1908, and 1912 elections in which the Socialist party of Eugene Debs took most of the third-party vote. Only in 1912 did Debs reach the 5 percent mark, but Roosevelt’s candidacy turned key 4 against the incumbent Republicans irrespective of Debs’s performance” (Lichtman 2016 p.31).
By his own rules that Key was incorrectly awarded. Removing that Key would mean that the incumbent White House Party was predicted to win the popular vote, and they did. So it really makes no sense that he continues to say that his system predicted a Trump victory in 2016 when in fact retrospectively the Third Party Key would not have been awarded to the challenger.
The Keys are predictors but they're not infallible.
S&P 500 Currently Predicts a Trump Victory
The final three months of the electoral period as
defined by Sam Stovall is taken to mean July 31 through to October 31 (or as he says
here in 2012 “the three calendar months leading up to the presidential election.”) That means the entire early August market slide downwards counts, and the S&P needs to recover to where it was on 31 July 2024 to predict an incumbent victory.
![[Image: GPiyXT6.png]](https://i.imgur.com/GPiyXT6.png)
There's plenty of time for it to recover and there are other more sophisticated market-indictor predictive systems, but this indicator alone has a greater than 80% correlation with predicting the Presidential outcome (Stovall said 88% in 2012 which is 3 presidential elections back).
Some Other Historical Trends
1. VPs rarely get elected to President. Going back 230 years 6 out of 19 attempts were successful, therefore the predictive power according to Lichtman's thesis would be 68%. That's pretty much the same power as most of his Keys.
2. A sitting President withdrawing from the race. A sitting President withdrawing mid-Campaign (in fact well more than 2/3rd of the way into it) is unprecedented, however an incumbent President not running and the WH party putting a different Candidate is not unprecedented. 7 sitting first term Presidents have, for various reasons, chosen not to contest re-election for a second term where they were eligible. Two out of the Seven replacement incumbent WH candidates were successful. That's a predictive power of 71% which is higher than no less than
nine of the Lichtman Keys, on par with two of the Lichtman Keys, and lower in predictive power than just two Lichtman Keys (Keys 2 and 5).
3. How united the challenging party is behind their candidate. There's many ways to measure this, one would be considering whether the Primary process was a bitter contest for the challenging party, another would be discontent within the Party for their candidate. While there are never-Trump Republicans, Trump basically represents the Republican base and in a similar way Harris represents the Democrat base.
4. How well liked the Candidate is by Independent voters. This is getting more important now than it was 30+ years ago as more US voters register as Independent. Lichtman's only way to measure this is by the strength of the Independent Candidate, I think you can look at the opinion polls on this question it's one of the few things the opinion polling is helpful with. Anyway it's good news for the WH Party as Harris is presently
way ahead of Trump on this one.
So the final word on all of this is that at the moment the vast majority of the data to-date suggests a Republican Party victory in November. Not because of the candidate's quality, but because of multiple other data points. The moment everything changed was when the Dems rolled Biden. Coronating Harris was a political gamble with absolutely no historical data to back-up the plan. At that point the the stock markets (although with much less predictive power than the final 3 calendar month performance data) was
predicting a Biden victory with 59% confidence (a higher predictive power than Lichtman Key 12!) Biden was ahead and likely to win in November as I kept saying, now we have even greater data showing that rolling him has the dire outcome of returning Trump to power from the political grave - but as I have said all along, this election is not about Trump it's about the performance of the Biden-Harris Administration.
![[Image: giphy.gif]](https://media0.giphy.com/media/v1.Y2lkPTc5MGI3NjExZXpteW5yM2Z4Y2djY211eXp5aXQybzkwcTczN3htdzd6cHF2Nmp6bSZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw/mpfMDb6MB6EWQ/giphy.gif)
