[EL] Did Alan Lichtman's 13 Keys Actually Predict That Clinton Would Win?
Pildes, Rick
pildesr at mercury.law.nyu.edu
Thu Nov 10 13:10:19 PST 2016
I want to run this question by the list serv, to help me figure out something I’m uncertain about:
As many of you know, Alan Lichtman developed his “13 Keys to the White House,” which had accurately predicted in advance every presidential election since 1984. Based on those keys, he predicted a Trump victory in both May and at the end of October. Since the election, he has been celebrated for having been one of the few who called it correctly.
But I have gone back to look at Alan’s prediction and I am puzzled a bit. The way his test works is that if 6 or more of the 13 keys go against the party in power (if the answer is “false”), then that party loses. If fewer than 6 do so, then the party in power wins.
In Alan’s final prediction for this year, made in the Washington Post, he says that 5 of the keys definitely pointed against Clinton. But the 6th and deadly final key, according to Alan, was that Alan predicted Gary Johnson would get more than 5% of the vote, which is what triggers this particular key. Alan explains how on Oct. 28th he arrived at this:
In his highest polling, Gary Johnson is at about 12 to 14 percent. My rule is that you cut it in half. That would mean that he gets six to seven, and that would be the sixth and final key against the Democrats.<https://www.washingtonpost.com/news/the-fix/wp/2016/09/23/trump-is-headed-for-a-win-says-professor-whos-predicted-30-years-of-presidential-outcomes-correctly/>
Johnson only ended up with 3% of the vote. So if Lichtman’s key based on the assumption that the 3rd party candidate will actually get 5% or more of the vote, then this key actually cut in favor of the Democrats. And since it was “the final key,” his system perhaps actually predicted a Clinton victory.
Of course, the prediction must be made in advance of the actual 3rd party vote. But then the question becomes whether what matters about this key is the anticipated vote for its own sake or for its sake as the best proxy for the actual vote. The answer is the latter. When Lichtman created the model based on past presidential elections running back to 1860, he used actual 3rd party vote totals, not the “anticipated” vote X number of days out before the election.
So on Alan’s self-application of his test in that interview, it looks like only an insufficient 5 keys pointed to a Clinton win. Indeed, Alan himself acknowledged this before the election, when he said if Gary Johnson slipped below five percent of the popular vote, that “would shift the prediction.”<https://www.washingtonpost.com/news/the-fix/wp/2016/10/28/professor-whos-predicted-30-years-of-presidential-elections-correctly-is-doubling-down-on-a-trump-win/?tid=a_inl> So on Alan’s own account, his keys actually predicted a Clinton victory – he just mistakenly overestimed by 100% Johnson’s vote share (and even if you add other 3rd party candidates, the total is still below 5%).
But now I want to speculate about whether there is a way to salvage Alan’s test. He does not count his 2nd key against Clinton – that key is “no serious contest for the incubment party nomination.” I gather this factor is based on how many delegates the winner gets at the party convention and Clinton got more than enough to have this key cut for her, not against. Even in May, Alan thought this key cut in favor of Clinton because her nomination was never seriously in doubt; by the fall, he had dropped any reference to it. But I have not read his book, which much more fully elaborates how the keys are to be applied.
But the bottom line, for now, is that Alan said his 13 Keys would predict a Clinton victory if Johnson (or 3rd parties in total) got less than 5% of the vote. Neither Johnson alone or 3rd parties together got 5%. So even though Alan’s test is being celebrated for having gotten this right, his test actually failed.
Even if the test failed this time, it has still been remarkably accurate. And since I have not read Alan’s book, I might be missing something about the test, which is why I am interested in whether I am missing something.
Richard H. Pildes
Sudler Family Professor of Constitutional Law
NYU School of Law
40 Washington Square South, NY, NY 10012
212 998-6377
From: law-election-bounces at department-lists.uci.edu [mailto:law-election-bounces at department-lists.uci.edu] On Behalf Of Douglas Carver
Sent: Thursday, November 10, 2016 12:51 PM
To: Sean Parnell
Cc: law-election at department-lists.uci.edu
Subject: Re: [EL] if national popular vote plan had passed, Hillary would be the winner
But I think it is safe to say that the US is unique in that the president/head of state/strongest party/chief executive/whatever you want to call where the true power lies can be selected without a plurality of votes.
On Thu, Nov 10, 2016 at 10:33 AM, Sean Parnell <sparnell at philanthropyroundtable.org<mailto:sparnell at philanthropyroundtable.org>> wrote:
One line from the piece jumped out at me:
“But if the candidate who got fewer votes wins the White House for the second time in five elections, it could put a new spotlight on the peculiar way that America picks its presidents — one not shared by any other democracy.”
This is technically true, but I’m not sure it’s quite as powerful an argument as suggested. In the U.K., for example, the Prime Minister (roughly comparable to our President) isn’t elected directly by the citizens of that nation, instead it’s effectively chosen by members of the largest party in parliament. A parliament filled with members who do not, it is worth noting, have constituency sizes that are equal (or thereabouts) – Isle of Wight has 118,00 or so people vs. 22,000 for Na h-Eileanan an Iar (see: http://www.ons.gov.uk/peoplepopulationandcommunity/elections/electoralregistration/bulletins/electoralstatisticsforuk/2014-05-01). And of course we don’t have a Queen in the U.S. (OK, maybe Queen Bey) to ask an MP to become PM and form a government.
The point is, there are plenty of forms of democratic governance in which the public at large does not directly select the chief executive/head of state. So while it may indeed be “peculiar” to use the electoral college, it’s not at all peculiar to not have direct election of chief executive/head of state.
Sean Parnell
Vice President for Public Policy, The Philanthropy Roundtable
1120 20th Street NW, Suite 550 South
Washington, DC 20036
(202) 600-7883<tel:%28202%29%20600-7883> (direct)
(571) 289-1374<tel:%28571%29%20289-1374> (mobile)
sparnell at philanthropyroundtable.org<mailto:sparnell at philanthropyroundtable.org>
From: law-election-bounces at department-lists.uci.edu<mailto:law-election-bounces at department-lists.uci.edu> [mailto:law-election-bounces at department-lists.uci.edu<mailto:law-election-bounces at department-lists.uci.edu>] On Behalf Of zacharyr46 at gmail.com<mailto:zacharyr46 at gmail.com>
Sent: Thursday, November 10, 2016 8:58 AM
To: Lorraine Minnite <lminnite at gmail.com<mailto:lminnite at gmail.com>>
Cc: law-election at department-lists.uci.edu<mailto:law-election at department-lists.uci.edu>
Subject: Re: [EL] if national popular vote plan had passed, Hillary would be the winner
Totally agree with Prof. Minnite. Here's my piece on this for those interested:
http://www.nbcnews.com/politics/elections/electoral-collage-lesson-more-voters-chose-hillary-clinton-trump-will-n681701
Sent from my iPhone
On Nov 10, 2016, at 7:52 AM, Lorraine Minnite <lminnite at gmail.com<mailto:lminnite at gmail.com>> wrote:
Brad is right that there are different values expressed in the original constitutional design of our electoral system and the means by which a president is chosen. But those values always were and continue to be contested. We are all know the many ways in which our system is not robustly democratic; for example, the more democratic direct representation of the House of Representatives stands in contrast to the original indirect election of Senators, and less democratic representation in the Senate of the states. The Electoral College falls into the 'less democratic' of our political institutions.
I'd like to go back to the assertion that, "Even in 2000 and 2016, the results will be close enough that one can't really know what would happen in a system in which each candidate would have very different incentives on how and where to campaign." I think this too easily brushes aside the critique of the Electoral College from the standpoint of a robust democratic ideal. Brad suggests campaigning would have been different if the national popular vote plan had been in place in 2016, and that this might have produced a different outcome, I guess with Donald Trump winning a plurality of the votes. I don't find the critique credible. For example, I find it hard to believe that either candidate would simply have concentrated their efforts in the states where they knew they had strong support in order to boost their numbers (i.e., Clinton spending all of her time in California, New York, and New Jersey, or Trump spending all of his time in Mississippi or Oklahoma).
Moreover, the impact on campaign strategy misses the larger point that we now again, only 16 years into in the 21st century will have twice installed presidents who lost the popular vote. I find that shocking and very disconcerning.
On 11/9/16, 11:21 AM, Smith, Brad wrote:
This is horrendously wrong.
Actually, there was a tremendous amount of voter suppression in 1876. The troops simply couldn't be everywhere, and were badly undermanned. The situation was so bad that President Grant asked Congress to authorize martial law in the South, in order to protect black voters from the Klan and other violence. Congress refused to pass the measure (it had passed a similar measure in 1871). The Red Shirts and the White League were other major Democratic paramilitary groups. In South Carolina, Ben Tillman, primary sponsor of the Tillman Act, was a member of the Sweetwater Club, which assaulted blacks attempting to vote with regularity.
The election of 1876 was quite probably worse for violence against black voters than the election of 1888, because by 1888 southern whites could largely claim "mission accomplished" when it came to vote suppression.
Bradley A. Smith
Josiah H. Blackmore II/Shirley M. Nault
Professor of Law
Capital University Law School
303 E. Broad St.
Columbus, OH 43215
614.236.6317<tel:614.236.6317>
http://law.capital.edu/faculty/bios/bsmith.aspx
________________________________
From: Richard Winger [richardwinger at yahoo.com<mailto:richardwinger at yahoo.com>]
Sent: Wednesday, November 09, 2016 11:05 AM
To: Smith, Brad; Election Law Listserv
Subject: Re: [EL] if national popular vote plan had passed, Hillary would be the winner
There was no suppression of black votes in 1876, because the federal troops were still occupying the south. That is why Mississippi's legislature sent two black US Senators to Washington, in the 1870's.
Richard Winger 415-922-9779<tel:415-922-9779> PO Box 470296, San Francisco Ca 94147
________________________________
From: "Smith, Brad" <BSmith at law.capital.edu><mailto:BSmith at law.capital.edu>
To: Richard Winger <richardwinger at yahoo.com><mailto:richardwinger at yahoo.com>; Election Law Listserv <law-election at uci.edu><mailto:law-election at uci.edu>
Sent: Wednesday, November 9, 2016 5:27 AM
Subject: RE: [EL] if national popular vote plan had passed, Hillary would be the winner
Richard,
There is pretty little reason to include 1824, when not every state even counted popular vote and the campaign was entirely different. In 1876 and 1888 the Republicans would have won the popular vote except for massive suppression of black votes and Republican votes more generally by the Democrats in the deep south. In each of those elections, the electoral college actually helped to make sure that the candidate actually favored by a majority of the populace actually won the election, by isolating the Democratic vote suppression and fraud.
Even in 2000 and 2016, the results will be close enough that one can't really know what would happen in a system in which each candidate would have very different incentives on how and where to campaign.
All of this points up that our electoral structure reflects values other than raw popular vote totals. At the same time, the popular vote usually carries the electoral college, and the system is designed to assure that no one without substantial and widespread popular support can be elected.
Bradley A. Smith
Josiah H. Blackmore II/Shirley M. Nault
Professor of Law
Capital University Law School
303 E. Broad St.
Columbus, OH 43215
614.236.6317<tel:614.236.6317>
http://law.capital.edu/faculty/bios/bsmith.aspx
________________________________
From: law-election-bounces at department-lists.uci.edu<mailto:law-election-bounces at department-lists.uci.edu> [law-election-bounces at department-lists.uci.edu<mailto:law-election-bounces at department-lists.uci.edu>] on behalf of Richard Winger [richardwinger at yahoo.com<mailto:richardwinger at yahoo.com>]
Sent: Wednesday, November 09, 2016 8:17 AM
To: Election Law Listserv
Subject: [EL] if national popular vote plan had passed, Hillary would be the winner
With the greatest number of uncounted votes in California, Oregon, and Washington, by far, states that are very strong for Clinton, it is clear to me that she will have approximately 1,000,000 more popular votes than Donald Trump.
The Democratic Party has been the victim of the electoral college five times now: 1824, 1876, 1888, 2000, and 2016
Democrats should have been concentrating on passing the national popular vote plan instead of focusing on campaign finance reform. Clinton's side spent far more money than Trump's side. We should get over the idea that voters always vote for the candidate with the most spending.
Another reform Democrats should have been working for is instant runoff voting. Yet just a few weeks ago Jerry Brown vetoed the California bill to expand instant runoff voting.
Richard Winger 415-922-9779<tel:415-922-9779> PO Box 470296, San Francisco Ca 94147
_______________________________________________
Law-election mailing list
Law-election at department-lists.uci.edu<mailto:Law-election at department-lists.uci.edu>
http://department-lists.uci.edu/mailman/listinfo/law-election
_______________________________________________
Law-election mailing list
Law-election at department-lists.uci.edu<mailto:Law-election at department-lists.uci.edu>
http://department-lists.uci.edu/mailman/listinfo/law-election
_______________________________________________
Law-election mailing list
Law-election at department-lists.uci.edu<mailto:Law-election at department-lists.uci.edu>
http://department-lists.uci.edu/mailman/listinfo/law-election
--
Dilexi iustitiam et odivi iniquitatem, propterea morior in exilio.
(I have loved justice and hated iniquity, therefore I die in exile.)
-- the last words of Saint Pope Gregory VII (d. 1085)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://webshare.law.ucla.edu/Listservs/law-election/attachments/20161110/c8545e48/attachment.html>
View list directory